FAIFUL (36 inches tall).

00 Urnln
A Step-by-Step Guide
Malcolm J. Tibbetts
Linden Publishing
Fresno
The of Segmented Woodturning
Step-by-Step Guide
by
Malcolm J.Tibbetts
All rights reserved. No part of this book may be reproduced or transmitted in any form or by
any means, electronic or mechanical, including photocopying, recording, or by any
information storage and retrieval system, without written permission from the publisher.
2003 Malcolm 1.Tibbetts
579864
ISBN 10, 0-941936-86-4
ISBN '3' 978-0-941936-86-6
Pntodin!h:na
Library of Congress Cataloging-in-Publication data
Tibbetts, Malcolm, 1949-
The art of segmented wood turning ; a step-by-step guide I by Malcolm Tibbetts.
p. cm.
Includes bibliographical references and index.
ISBN 0-941936-86-4 (pbk., alk.paper)
1. Turning. 2. Woodwork. I. Title.
TT203.T53 2004
684'08--dc22
Linden Publishing Inc.
2006 S. Mary
Fresno, CA
V .lindenpub.com
800-345-4447
TabeoICon!ents
1. Introduction 9 10. Stave Construction 89
,+ Types of Construction 1
Calculating Compound Miter Angles 89
Building a Staved Vessel 90
3

Tools 13
Mortise-and-Tenon Joint 96
4

Safety 15
11. Lamination Trickery 101
AAW Lathe Safety Guidelines 16
Building Zigzag Rings 101
Respiratory Protection 17
Multi-Generation lamination 104
5 Understanding Wood 19
Two Ways to Build Diamonds 110
Moisture Content '0
Laminated Segments and Staves 11'
Testing for Moisture Content '0
A Bowl from a Board 118
Wood Grain Orientation 21
12. Production Turning 123
Choosing Wood 2
Preparing Yow Wood
T. NscellaneousTcks
^
'
4
Oval Bowls 129
6. Glue ,6
Mitering complete Rings
'
3'
Clamping '7
Installing Round Designs
'
35
Glue Safety '9
14. Building an Icosahedron
'
39
7 Techniques and Tricks 31
Sphere Turning Techniques 143
Cutting Segments 3'
Transforming Spheres into Vessels '48
Disc Sanding 38
A Truncated Icosahedron '49
Gluing Segments Together 47
Portholes and Large Turings
Preparing Rings for Stacking 49
15
153
Uses of MDF 54
Building a Porhole-Style Ring '53
Working with Veneer 56
Large Diameter Ring Techniques '59
Building Sunrise Segments
Splitting Rings into Multiple Rings 57
,6,
Another Way to Insert Diamonds 16,
Vessel Bases 58
Improving Efficiency 6,
More Large Vessel Techniques ,63
8. The Design Process 63
16. Segmented Ribbons 167
Design Tips 63
'
7

Final Thoughts
'
75
Feature Rings
65
Appendix
Creating a Blueprint 67
Mathematical Formulas Concepts '76
A Sample Drawing 70
Calculating Compound Miters '76
9

John's Tng
73 Construction of Platonic Solids 176
Indian Blanket Feature Ring 7
3
Chart of Compound Miter Angles 176
A Cutting Ust 80 Description of Turnings 178
Joining Two Halves 84 Bibliography .80
The Finishing Process 85 About the Author ,8,
Index 18,
*1
ALtO!OAx0N
|rCD5tall
1.
Introduction
From the time I built my first birdhouse at age
five in my granddad's shop. I have always had a
fascination with wood. More than thirty years
ago, after moving into an empty house, with my
wife's encouragement. I purchased a table saw
instead of inexpensive, affordable furniture.
Following twenty years of building furniture for
our home in Lake Tahoe, I discovered the
magical capabilities of the lathe. As Jlook back
at my past furniture designs. it strikes me that I
have always experimented with designs using
contrasting wood color combinations. Given my
prior woodworking experiences and my
ownership of the necessary tools. it was only
natural that I gravitated towards segmented
wood turning.
Segmented woodturning, or as it has been
called, polychromatic wood turning, has been
around for a very long time. Many years ago
someone probably had the need for a bowl that
was larger than their largest piece of wood and
thought, I'll just glue two pieces together."
From that moment in time, the "art" of
segmented turning has continued to evolve.
Long before I glued together my first ring of
segments. many other woodturners pursued
this art form with great success, providing
inspiration to those of us that followed their
lead. In the 1970S and 1980S, Emmett Brown and
Cyril Brown, with their publication of
Polychromatic Assembly for Woodturning,
inspired thousands of turners to experiment.
Dale Nish. in 1980, with his publication of
Artistic Woodturning, sent a signal to the
woodturning world, that it was OK to glue wood
together, and woodturners continued to
enthusiastically experiment. Ray AUen, Giles
Gilson, Bud Latven, Lincoln Seitzman, and Mike
Shuler are just a few of the very talented artists
that have also inspired me. These artists opened
doors for aU of us, but they, like all of us, had to
9
start with their own very first segmented
project. If you are looking for a new challenge,
then it is my hope that these pages will provide
you with the techniques and inspiration needed
to create your own first segmented turning. If
you have already started down this path, then I
hope some of my ideas will renew your interest
and encourage you to develop new challenges.
On these pages, I have attempted to share my
experiences.l have described the techniques
that have worked for me and in some cases,
those that have not worked. I want to make it
clear, that I do not offer the methods described
on these pages as the only way. I know from
casual contact with other turners, that everyone
develops their own way. I should also point out
that not all of these techniques are my own
invention. Woodturners are ver sharing and
many techniques have been offered to the
public in workshops, symposiums, and
magazine articles; I have tried to give credit
where it is due.
Segmented woodturning is much more than
gluing together pieces of wood into a bowl. It is
about creating exciting objects that challenge
you and stimulate the interest and enjoyment
of others. It need not be a paint-by-the-numbers
type of craft -it can be a real opportunity to
express oneself with very few limitations. This
activity requires extremely precise
woodworking skills and often requires a little
inventiveness. As a segmented turner, you are
hee from the boundaries of a single block of
wood. You have unlimited choices regarding
size, shape, and color combinations. The
possibilities are endless, which is why it
continues to keep my interest. AUc!6m0N
(photo 1-01) is a perfect example of a form that
would be almost impossible to create from a
single block of wood.
1O THE AT OF SEGMENTED WOODTUNIG
1-02 MOBIU SONATA, 20 inches wide.
I have written this text for the turner who has
at least some woodworking experience. You do
not need to be an engineer or a mathematician.
nor do you have to be a master woodworker, but
familiarity with woodworking machinery and
at least basic lathe-turning skills are cerainly
necessary. If you possess these minimal skills
and are ready to take your woodturning in a
new direction. then this book is for you. It is my
sincere hope that I have conveyed the right mix
of advice and inspiration and that I have done
justice to this fascinating subject.
May all your glue joints remain perect and may
your need for challenge be completely satisfied.
Sincerely.
Malcolm 1.Tibbetts
2. Types of
Construction
The three most common methods of segmented
construction are:
the stacking of segmented rings,
assembling staves,
the simple lamination of wood layers.
Precise joinery and accurate glue assembly
techniques are necessary for any style to be
successful You might be wondering. "What's the
difference between segments and staves?"
Segments are miter-cut hom a board with the
wood grain oriented horizontally and then the
individual pieces are glued together, end grain to
end grain. to form a circle or ring of wood. The
individual segments can be different species of
wood and they can be different lengths, and their
angled miter cuts can even vary, however the
OITkC1
201 Segments and staves are the two most
common components.
11
total of aU their miter angles must add up to 360.
to form a complete circle. The ring of segments
on the right side of photo 2-01 shows segments of
two different lengths in the same ring.
The wood grain of staves, on the other hand, Is
usuaUy oriented verticaUy like the staves of a
wine barrel. Staves can be short or long. wide or
narrow, and just as with horizontal segments,
they do not have to all be equal, but their side
angles must also add up to 360. Segments are
normally produced with simple miter cuts,
whereas staves are often cut wcompound
miter cuts. Compound miters created the vessel
under construction in the center of photo 2-01. If I
had used simple miters to create the shape, a
much thicker board would have been necessary.
So there you have it. segments and staves. These
two elements compose the majority of
12 THE ART OF SEGMENTED WOOIORNNG
segmented work. There are endless ways to
assemble and embellish these elements, and
many will be described on the following pages.
Creating turnings from a simple laminated
block of wood layers is also occasionally used as
a method of construction. Most woodworkers
already know how to make simple laminations.
so I will not be focusing on this technique.
People often confuse segmented work with
inlaid work. The public seems to be more
familiar with the term "inlaid." There is a big
difference. Inlaying is the process of creating a
recessed cavity for the insertion of another
piece of wood. There are opporunities for this
technique in segmented work, although most
assembly involves layering thicker pieces of
wood together, thereby creating the same
design on the inside as well as the outside of
the turning. For example. suppose you wanted a
diamond shape of light-colored wood
surrounded by darker wood. To inlay, you would
cut a recess into the surface of the darker wood
(by hand or router) and then glue a thin but
accurately shaped piece of the Iightcolored
wood into the recess. To laminate the diamond
shape into a turning requires a different
approach. You start with a diamond-shaped
piece of wood the same thickness as the rest of
your turning components and then glue the
swrounding dark wood to the diamond's four
sides. using the disc sander to create tight glue
lines. This assembly is then machined into a
segment to be built into a ring of segments. The
diamond shape will be visible on the inside of
your turning as well as on the outside. Making
diamond shapes is detailed in Chapter 1J.
The technique known as "open segmented" has
gained popularity urecent years. This style of
construction creates air gaps between
individual segments during assembly. I have
little experience with this technique and will
not be offering advce on this subject. William
Smith has written a very detailed book,
Segmented Turning (Schiffer Publishing). which
I recommend to those who wish to investigate
this style of segmentation.
2-02 EMT VASE (18 inches tall).
3-01 These are a few of my small hand tools.
3
Tools
Segmented wodtuming requires an extensive
woodshop of tol5. The major tools that you need
are a table saw and a lathe. The size of your saw
and lathe wdictate the size limitations on your
work. A lo-inch table saw and a mid-sized lathe
are adequate for most designs. As for myself, I
have created many very large turnings that
would have been very difficult or impossible on a
smaller lathe. in addition, I mmuch of my own
lumber from very large planks and a lo-inch
table saw would be undersized.
Of course, the more complex your turnings
become. the more extensive your tool
requirements will become and to produce
quality work. you also need:
a Iz-inch disc sander (or larger),
a surface planer, and
a jointer.
Moreover, if you can afford it and have the
space, then these come in very handy:
a lo-inch miter chop saw,
a 14-inch band saw (with riser block),
a driU press. and
a drum-style thickness sander_
1
In addition to the jointer/planer. a drum
sander is invaluable for dimensioning thin
material. It is a lUXUry for any home shop. to
be sure, but I now wonder how I ever got along
without one. If a name-brand tool is not in
your budget. there are ways to construct a
homemade drum sander. which you might
want to investigate. Before gluing up multi
layered laminations that need to be extremely
consistent. I run my strips through the drum
sander. This not only provides the needed
thickness accuracy. but the smooth surfaces
also result In cleaner glue lines.
14 THE ART Of SEGMENTED WOODTORNlNG
3-02 CRtBBG BORD (16 inches diameter).
After acquiring tools for more than 30 years for
my own small shop, it is so !that I hardly have
room to move around freely. Mobile bases under
the band saw. drill press, jointerlplaner, disc
sander, and drum sander allow me to mdze
the use of my 250 square feet of floor space. It is
small and crowded, but I am U8to have l
In addition to large power tools, you will need a
variety of smaller power tools. Again. it depends
upon just what kind of work you are planning
to attempt, but I would say that for one reason
or another, I often need:
a 3/S-inch drill motor.
palm sander,
a router,
a hand-held power-plane,
a jig saw,
a belt sander.
Other small hand tools that are necessary:
lots of several types of clamps,
calipers for measuring,
a compass for drawing circles.
a good quality steel straight edge,
a small, bright flashlight,
a trustworthy moisture meter, and
the standard small hand tools
(screwdrivers, hammers, etc.).
Photo3-01 shows a few of my own small tools.
The assembly of segmented work requires good
eyesight along with very good lighting. When
you need sunglasses myour shop, then you
probably have just about enough light. use a
combination of incandescent bulbs and halogen
fixtures. Make your work area as bright as
reasonably possible; a magnifying light can be
very useful during small-component assembly.
WWWW@@@@1P WWW@ 1
4
Safety
There are literally hundreds of ways for shop
accidents to occw. Woodworking clubs are full
of members who have had close calls 1 serious
accidents. S,how do you avoid becoming an
accident statistic? BE CEFU This probably
sounds Simplistic, but I truly believe that the
most effective safety tool that you possess is the
six inches between your ears. Think. think.
think, and when you find yourself not thinking,
then take a break or maybe end your session. I
am not a safety expert. but I have been around
the construction trades most of my life and
most of the injuries that I have investigated
were due either to poor judgment, lack of
concentration, fatigue, failure to use personal
protection equipment, or more commonly, a
combination of all four. Rarely is it the
q-O: Personal protection equipment is
essential. Turners should seriously consider the
use of an air helmet system -your lungs will
thank yu.
equipment's fault. Some of my own personal
protection equipment is shown in photo Q~OJ.
This equipment is designed to protect your
eyes, face, ears, lungs, and in the case of the
anti-shock gloves, your hands (I suffer from
non-woodtuming related carpal tunnel
syndrome). Use this type equipment uyou have
it and acquire it if you lack it. Pau helmet
system is a big investment, but is invaluable at
times. His far superior to a cloth-type
respirator.
Some of the work required to produce
segmented pieces can be rather monotonous.
How do you keep your concentration while
cuttIng hundreds (or thousands) of segments? I
break up the work.l cut segments for a while, I
sand for a white, I glue up a few components,
1 THE ART OF SEGMENTED WOODTURNING
then I go back and cut more segments. I never
spend hours trying to focus on just the cutting
of hundreds of segments. I keep mixing up the
workload. This helps keep my mind focused on
the job at hand. it keeps my mind in the game.
As a segmented tuer, you have the dangers of
many power tools in addition to the dangers of
the lathe. Take the time to devise whatever kind
of jig is necessary in order to avoid putting your
hands near moving saw blades, drill bits,
sanding discs. etc. A very general piece of
advice: if you have to cut or drill or sand a very
small piece of material, then attach it to a larger
piece that can be safely held. It is really quite
simple if you stop and think about it: if your
hands are never very dose to a moving saw
blade. then it is difficult to cut your finger. In
addition, instead of looking for your safety
glasses when you think you need them. just
make it a habit to always put on eye protecton
upon entering your shop.
For many years. I managed outdoor operations
at a major ski resort and as such, I was
responsible for the safety of hundreds of
employees. A few additional things that I
learned about safety are:
Hyou allow a dangerous condition to exist.
you will eventually have an accident;
a messy work place is a dangerous work
place;
the one time you neglect to put on your
safety glasses is the time something flies into
your eye;
when you rush a job and take shortcuts,
accidents are more likely to occur, and
a dull tool is a dangerous tool
Of course, in addition to all the dangers of
operating table saws. miter saws. and other
woodworking machinery, as a woodturner you
face the dangers of the lathe. The American
Association of Woodtumers (the AAW) has
published a list of safety tips that do a very
good job of covering lathe safety. The AAW, with
over 1O,OO0 members worldwide. is a great
international organization dechcated to the
advancement of woodturning, check them out
at N .woodturner.org. With their permiSSion
and encouragement. 1 have included their lathe
safety guidelines.
AW Lathe Safety Guidelines
1. Safe, effective use of the wood lathe reqUires
study and knowledge of procedures for using
this tool. Read and thoroughly understand the
label waring on the lathe and in the
owner/operator's manual.
2.Always wear safety goggles or safety glasses
that include side protection and a full-face
shield when needed. Wood dust can be harmful
to your respiratory system. Use a dust mask or
helmet and proper ventilation (dust collection
system) in dusty conditions. Wear hearing
protection during extended period of operation
3. Tie back long hair. Do not wear gloves. loose
clothing, jewelry, or any dangling Objects that
may catch in rotating parts or accessories.
4. Check the owner/operator's manual for
proper speed recommendation. Use slower
speeds for larger diameter or rough pieces, and
increased speed for smaller diameters and
pieces that are balanced. Hthe lathe is shakng
or vibrating. lower the speed. If the work piece
vibrates, always stop the machine to check the
reason.
_.Make cerain that the belt guard or cover is in
place. Check that all clamping devices, such as
on the tailstock and tool rest. are tight.
.Rotate yow work piece by hand to make swe it
clears the tool rest and bed before turning the
lathe ON. Be sure that the work piece Dfreely
and is firmly mounted. His always safest to tum
the lathe OFF before adjustmg the tool rest.
7. ExerCise caution when using stock with
cracks, splits. checks. bark. knots, irregular
shapes, Or protubrances.
8. Hold ting tools securely on the tool .
and hold the tool in a controlled but
comfortable manner. Always Ua sloer
s
when st uthe work pieCe is bC
This helps avoid the possibility of an
unbalanced piece jumping out of the lathe and
striking the operator.
9. When running a lathe in reverse, it is possible
for a chuck or faceplate to unscrew unless it is
securely tightened on the lathe spindle.
10. Know your capabilities and limits. /
experienced woodturner may be capable of
techniques and procedures not recommended
for beginning turners.
11. When using a faceplate, be certain the work
piece is soHdly mounted. When ting
between centers, be certain the work piece is
secure.
12. Always remove the tool rest before sanding
or polishing operations.
13. Do not overreach, keep proper footing and
balance at all times.
14_ Keep the lathe in good repaiI. Check for
damaged parts, a1ignment, binding of moving
parts and other conditions that may affect its
operation.
15. Keep tools sharp and clean for better and
safer perormance. Do not force a dull tool. Do
not use a tool for a purpose not intended. Keep
tools out of the reach of children.
16. Consider your work environment. Do not use
the lathe in damp or wet locations. Do not use
in presence of flammable liquids or gases. Keep
work area well lit.
17- Stay alert. Watch what you are doing and use
common sense. Do not operate tool when you
are tired or under the influence of drugs or
alcohol.
18. Guard against electric shock. Inspect electric
cords for damage. Avoid the use of extension
cords.
19 Remove chuck keys and adjusting wrenches.
Form a habit of checking for these before
SWitching on the lathe.
20. Never leave the lathe running unattended.
Tu power off. Do not leave the lathe until it
comes to a complete stop.
SA
17
Respiratory Protection
One m1biggest hazards in any shop,
mentioned in the AAW guidelines, is the nasty
waste substance produced -the dust! It gets
down your shirt and into your shoes and before
you know it, there is a trail of it throughout
your home. However that is not the worst: it
also coats the inside of your lungs. Once you
destroy lung tissue, the body can never
regenerate it, the damage is permanent.
ObViously, anything that makes you cough and
choke cannot be good for you. While some wood
species seem to be more harmful and irritating
than others, no dust is good dust. So how do we
protect ourselves? By all means necessary! In
my own shop I operate a 3HP dust collection
system, a ceiling-mounted dust filtration device,
and a 2o-inch fan mounted directly though the
exterior wall right next to my lathe. When this
fails to keep the air clean, I use an air helmet
system or cloth-type respiIator. My favorite
time of the year is the summer when I can
operate my large wall fan continuously, with
the windows and door open, without worrying
about heat loss.
Even with these precautions I know that I
have inhaled unhealthy levels of wood dust. His
so easy to be lazy and not take the time to put
on the protection, especially if you know that it
is only going to be dusty for a few seconds.
Unfortunately, those few seconds add up and
may result in health problems in the future. In
the heat of battle. in the excitement of creating
a new turing. and in the rush to complete the
project, it is so easy to allow safety to take a
back seat. No turning is worth a fmger or
permanent damage to your lungs.
Take your time and work safely!
1b H! PkT Ol bECNENTEDVOODTUIN!NC
5
-
0
CARETD VASE (12 inches diameter) -caretta, purpleheart. gabon ebony.
UNDERSTANDING WOOD
-O1 Because of its stability. mesquite is a
favorite wood for segmented work.
5
Understanding Wood
It is very imporant to understand certain basic
properies regarding your raw material
wood. A very good book on the subject is R.
Bruce Hoadley's Understandin9 Wod. I highly
recommend it to those wishing an in-depth
edUcation on the subject. Wood is a wonderful
substance. but it is very diferent from
materials such as plastic and metal. Having
grown as part of a tree. it has unique
characteristics and no two species are the same.
For that matter, no two boards from the same
tree are the same. Hyou have been a
Woodworker for any length of time, you have
probably discovered some of the difficulties
associated with wood. The most troublesome
characteristic is it moves, its actual dimensions
change. It expands with high humidity and it
shrinks as it dries and it does not do this
conSistently. Each species acts and reacts a little
diferently. "So what?" you might ask. Well. I can
teU you frst hand that if you ignore this
characteristic. then you Wlikely experience
failures i some of your glue joints. The two
biggest reasons for glue jOint failure are:
high wood moisture content, and
incompatible wood grain orientatiOn.
20 THE ArOF SEGMENTED WOODTURN!NG
Moisture Content
Hoadley defines "moistwe content" as "the
weight of water in the cell walls and cavities of
wood expressed as a percentage of oven-dry
weight." In other words, if a not-so-dry piece of
wood weighs 1.1 pounds, and this same piece of
wood weighs only 1 pound after it has been
baked in an oven until it is completely dry, then
its moisture content would be 10%. If the wood
that you use is not at least as dry as its
environment. then it will continue to lose
volume as it continues drying. Two pieces of
wood on either side of a glue joint with
different moisture contents wlprobably move
differently. Glue is unable to restrict this
movement and a gap or unevenness may
appear between the two pieces of wood. It is
common to experience some movement even
when your wood is very dry. Different species
can react differently to changes in their
surrounding environment and even with the
best of sealing finishes on your turgs, high
or low humidity will eventually affect the
wood. By using dry wood, you will dramatically
improve your chances of producing stable
turings that will stay together. "How dry?" you
ask. I would recommend never using anything
above 10% moisture content and I would only
use wood at 10% dall the wood were
conSistently the same. In general, wood in the
8% (or drier) range is much safer. Anything
higher and you Wbe taking a risk -it is just
not worth it! Do not trust your wood, just
because the supplier tells you that it is "kiln
dried." The wood may have been 1% moistUre
content when it came out of the kiln, but if it
has been improperly stored in a moist
environment, then it may no longer be at that
percentage. Photo j-o1 displays a recent
pwchase of mesquite that was imported from
Argentina. This is terrifc material for
segmented work, because H is known for its
stability, that is, its lack of movement after
drying. It was described as being "kiln dried"
and it probably was, but unfortunately. H came
by ship and was stored fOT several months in a
warehouse in Seattle before being purchased. I
was very happy with my pwchase and not
terribly surprised when I measured the
moisture content. It ranged dramatically from
about 9% up to 14% from board to board. By the
way. the main wood in the vessel on the front
cover is Argentinean mesquite from that pallet.
How you store your own wood inventory is
important, especially if you live in a humid
climate. Your storage area obviously has to be
dry. I live in the Sierra Nevada Mountains of
Califoria (Lake Tahoe) and for about the
year, we enjoy a very dry climate. Then winter
comes. It can be dry one day and snowing the
next. Untill built a heated drying shed with a
dehumidifier, my wood inventory was always i
a state of change with the seasons.l call this
storage device a drying shed, not a kiln. His just a
warm, dry, low humidity, insulated, large, storage
closet. During the summer months. unless I am
in a hwr to use particulcu bocuds, I do not
operate the heater. The outside of the shed is
painted black and solar heating keeps the inside
quite warm.. but dwng the Ver months. a
small electric heater keeps the shed a toasty goO
and the dehumidifier runs on a regular basis.
Wood that is drier than its environment will
quickly acquire moisture Himproperly stored,
which is why you must test your wood.
Testing for
Moisture Content
How do you measure the wood's moisture
content? One method involves repeatedly oven
drying a sample and then weighing it until it
no longer loses weight. I do not have much
.
experience with this technique and frankly, It
seems like quite a lengthy process. On the other
hand, there are several manufacturers of very
easy to use electronic moistwe meters. These
devices are able to deteDe the amount of
water content in the wood by measuring
electrical resistance between two
sharp
probes
that you insert into the wood. Wet wood is a
much better conductor of electricity than d
wood. and these meters detect this difference.
Be aware, wood does not dry uniformly The
outside can be considerably drier than
the
inside and exposed end grain walWaYS

"

NDlN

OO """""

9
5-02 Wood must be tested for moisture content.
measure drier than the rest of the board.
Actually, testing the end of a board is a waste of
time. Hyou are ready to start your project, and
you know the required width of a strip of wood,
then rip-cut a strip from some of your boards
and test the interior of the wood. Place the
probes in line with the grain as shown in photo
5-02 and check several locations.
What do you do if your wood is not dry enough?
Imagine -you are all ready to star a new
project, you have done the design work and
at the last minute, you discover that some of
your wood measures 14%. Do not be tempted to
proceed, you will most likely regret it later. Your
only choice at that point is to continue drying
the wood. Do not lear this lesson the hard way!
The important messages are;
test your wood, and
take the necessary steps to ensure that your
wood stays dry.
Importance of
Wood Grain Orientation
The second and probably most common reason
for glue joint failure is incompatible wood grain
orientation. What do I mean? The important
thing to remember is almost all wood
movement is perpendicular to the direction of
the wood grain. A board moves Sideways. not
lengthways, therefore segments move mostly
up and down and in and out; they move very
little from end to end.
This predictable movement is certainly a
disadvantage and a nuisance. but you can
design your turnings to allow all the wood to
move together by orienting as much B possible
all your individual pieces of wood so that the
grain is positioned the same directIOn. Da
simple vessel composed of stacked segmented
rings, this is easily accomplished. By glumg 8
the segments end to end. the wood gram W
be horizontal. As you stack the nngs together.
all the wood gram contmues to remam
honzontaL Hthe wood moves. It all moves
22 POF bICNN1IDYODURNINC
together, without exerting opposing pressure
on the glue lines Ais well and the vessel
should lIVe happily for a very long time_
Complex turnings, containing elements that are
not all horizontal. present a more difficult
challenge: you must minimize opposing forces
during the deSIgn phase of your project. That
does not mean that you cannot position two
pieces of wood perpendicular to each other. D
small dimensions, this can be done with
confidence. The total movement within very
small pieces of wood is usually not a problem. H
we go back to that simple vessel with all its
segments positioned horizontally and we now
decide to place a thin piece of contrasting wood
vertically between each segment, then this is
OK dthe horizontal segments are not too tall
(thick). I would recommend not exceeding 3/4
inch, and that might be pushing it. I have
constructed many such Dngs with short
vertical pieces glued between horizontal
segment ends and those Dgs have
remained sound for many years.
However, sometimes designs require a taller
vertical element next to a horizontal element.
This i potential trouble! Hmay require extra
work in the milling of your wood, but instead of
glumg a thin strip of wood with its grain
oriented vertically, use a short, wide piece of
V,thereby keeping all the grain in the same
orientation. Photo -u]shows an example of
this; in this turing, large segments with
portholes make up the outer apron of the table.
The 1/4-inch wide strips of walnut positioned
between the large segments of myrtlewood are
orientated with their grain horizontal just as
the large segments. Many people do not notice
this type of attention and care. but your glue
lines will thank you.
People with backgrounds in woodworking are
sometimes surprised at the extensive use of
end.grainto-end-grain glue jOints in segmented
woodturing. After all. everyone knows that
end-grain glue jOints are not nearly as strong as
side-grain glue joints. While that is certainly
true. woodtumings are subjected to few
exteral forces. it Mthe interal forces caused
by possible wood movement that need
most
of
the attention. Besides, the weaker end-grain
joints are usually reinforced by the
overlapp
ing
side-to-side grain joints that are present
between rings. By overlapping the alignment of
vertical glue lines (like laying bricks), you are
essentially creating a series of mortise-and
tenon type joints, which are very strong and
efective.
Choosing Wood
Choosing wood for a tug can be 1
Because most of the time, relatively small
quantities are required, we have the
opportunity to work with exotics that
otherwise might be too expensive. There are so
many types of wood from which to choose.
William A. Lincoln's book, World Woods in Color
(Linden Publishing), displays more than 250
color photos of different woods_ It is a great
reference tool for woodworkers. My own
personal experiences with the woods of the
world cODSIS!Oa much shorter D,I h2vJIS|60
below those woods with which I have had
enough experience to form an opinion. These
are woods that I have cut, sanded, glued and
finished during the past 30 years. As you will
see, I have categorized these woods into three
designations -limited use, OK, and prsonal
faorites. These are simply my personal
preferences and opinions.
Limited-use woods include alder, ash. aspen,
birch.lacewood. madrone, mahogany,
maple
(soft), oak (red white), padauk. poplar,
redwood, teak, and zebrawood. These vju
do not make the grade. they are either too 8H,
too oily, too open-grained. or too dull in
appearance. although hom time to time, in the
right circumstances, do use some of them.
OK woods include |(spalted), birCh
(spatted), cherr, ironwood.limba (light
and
dark), osage-orange, walnut, and wenge.
These
woods are OK, but for one reason or
another
.
they do not inspire the same enthusiasm
as the
ones listed next. I do use some of
these
qUite
a
bit. but usually not as a stand-alone
leatwed
wood.
Walnut and wenge prOVlCk W nice
lowrr
UNDERSTANDING WOD
23
5-03
As much as possible, wood grain direction should be consistent.
cost alterative when a large volume of dark
wood is required.
Personal favorites include apple, blackwood,
b!ooowood, bacote, bubinga, carob, ebony
(Gabon), ebony (Macassar), holly, jarrah, maple
(bird's-eye), maple (hard), maple (curly), maple
(spalted), mesquite, mountain mahogany,
myrtlewood, persimmon, pink ivory,
purpleheart, rosewoods (pau ferro, East Indian,
Honduras, Brazilian, coco bolo, flamewood.
tulipwoodJ, yellowheart, and ziricote. These are
faVOrites for several reasons: they have great
color and/Or figure character, and except for the
rosewoods they all glue well and most of them
are tight-grained, which means they polish
nicely.
When designing a turning, you are faced with
many chOices of wood. What goes with what?
What combinations do not work well together?
Over the years I have probably combined most
of the woods listed above. The easiest and less
troublesome combinations involve woods that
are similar in density -hard maple next to
most any of the exotics as an example_ I have to
admit,l am partial to tight-grained, dense,
hardwoods_ They cut and sand uniformly and
usually take a nice finish. Soft woods and open
grained woods. in general. just do not project
the same type of look. You can certainly
combine soft woods with hard woods, such as
holly and ebony. However, it requires a little
more diligence to achieve a smooth surface, it
requires high-speed shear scraping, and you
need to minimize coarse-grit sanding. Oily
woods present their own special problems. They
do not glue as dependably and they often
shorten the lifespan of your sandpaper by
quickly clogging the grit. The best advice that I
can give you regarding oily woods is to glue
them quickly. Do not allow the segments to Sit
around overght and possibly release any
extractives onto the glumg surfaces. 1also
caution against creatmg rings that consist of
24
TE AT OF SEGMENTED WOODTURNING
only oily woods such as cocobolo and teak.
Gluing a wood like cocobolo to itself is much
more risky than gluing cocobolo to a non-oily
wood.l have heard people suggest that cleaning
oily wood glue surfaces with substances such as
acetone or lacquer thinner just before gluing
Wensure success. but I have never found it
necessary as long as the gluing is done soon
after the cutting and sanding.
Some other combinations that can cause you
problems are combining a wood like padauk
with a wood such as holly -a bad idea. The
colorful padauk dust is very effective as a
staining agent and often makes a mess of
adjacent light-colored wood, especially a white,
soft wood like holly. You might be able to
overcome this problem on the outside of your
turning, but to maintain white holly on the
inside is just about impossible. If you desire a
red/white color contrast, a combination such as
bloodwood and maple is a better choice.
Combining contrasting wood colors is a big part
of segmented turning, but the sad truth is, all
woods darken with age. Exposure to sunlight
accelerates this darkening process. Woods that
are dark to begin with Wbecome darker and
eventually Wstart to look alike. Therefore, it is
important to choose wood combinations that
Wretain their contrast over time. Obviously,
white and black combinations will provide the
most dramatic contrast and last a1most forever.
Other combinations will be less dramatic and
not Ulong-lasting. P an example, you might
decide to combine a red wood such as
bloodwood with purpleheart. When freshly
machined they will look very nice together;
however, within a few years (or less), from
across a room, they wlappear to be the same
color and the effect that you desired will have
been lost. The lesson here i, select woods that
will provide long-lasting contrast if that is the
efect you desire. Study your choices and try to
visualize your turgs as they will appear after
years of inevitable darkening.
As woodworkers, it is easy to get excited about
spectacular grain in a board; however, as
segmented woodtuers we have the ability to
create our own spectacular grain and figure
effects. We do not have to purchase the biggest,
most expensive boards at the store V order to
create beauty. We can utilize small boards, cut
offs, and scraps at conSiderably less expense. I
have constructed many turnings from
numerous Similarly colored woods in order to
achieve an overall effect. In those cases, I was
not as concerned with contrast as I Y with
the look. An example of this is shown in the
photo of TIkIf7PPL(photo 5-0
4).
In this turning, I attempted to duplicate the
shades of an actual apple by randomly using
bloodwood, pink iVOry, and tulipwood. This
apple (a gif to my schoolteacher wife) is also a
box with an outside profle that has been
flattened in places to imitate an actual apple.
Approach your wood selection decisions like a
painter choosing colors from his palette, while
keeping in mind the individual characterist
of each wood species.
In case you have not figured it out yet, I love
wood. All woods are good for something. Even
the sawdust we produce has value to someone.
sometimes feel like a kid at Christmas when I
surface-plane rough stock and discover the
hidden beauty that lies in wait. I observe gallery
visitors approaching my turnings with a sense
of awe and with an almost uncontrollable
deSire to touch the sutfaces. Beautiful WD
that effect on people. I love the opportunities to
share my passion with this amazing material.
Perhaps this is why I became a segmented
woodturner instead of an ironworker.
Preparing Your Wood
Before starting the process of cutting segments,
you should accurately dimension aU of u
wood, or at the least, athe wood that will
be
needed for a particular phase of your
projec
t
This is important for several reasons: it allO
you to confirm the quality and the
quantity
of
your useable material and it saves
you
trom
having to interrupt the construction
process
U
order to prepare more material. By
preparing
8
your
wood at the same time, you
WIU create
5-04
1tkF`!7PPL(4 |Dtt> tall) -think of wood LOlOI> as paint
consistent dimensions, which will make your
assembly easier and more accurate. You might
be thinking that you are only going to be using
small pieces so why bother truing-up the whole
board. Believe it, the straightness and flatness of
your boards affects the accuracy of your miter
cuts. It I5 Important to take the time using
whatever technique works for you to flatten
and then accurately dimension your lumber
ahead of time. Using a jointer followed by a
planer and sometimes even using a drum
sander is my usual routine. Otherwise, during
the cuthng and gluing phases of your project.
you will have to deal with problems such as a
board that does not sit right on your saw. Think
about the unpleasantness of trying to make
accurate miter cuts across a board that wobbles
on your saw, or think about gluing segments
into a ring that do not sit flush with one
another because their heights differ. At a
minimum, it is a hassle, but it can also easily
lead to inaccuracy. It is very difficult to build
precise segmented rings from crooked boards of
varying thickness.
26
THE ArOF bFCMFNTD VODTUkN!NC
6-01 Different glues are needed for different situcons
6.
Gl ue
Without the availability of effective, long
lastmg, dependable glues, segmented turing
could not exist. There are many different types
and brands. 1 am often asked, "So, what is the
best one?" lf 1 could have only one type of glue
in my shop, then the answer would Dpretty
simple - original Titebond (and the company
did not pay me to say that). It is versatile and
has always performed wel for me, as wel as
many other turners that I kow. However, I also
use many other tyes of glue for a variety f
tasks. Here is a list of what Iuse, how I use It.
and a few other thoughts;
Titebond Original
I use Titebond Orginal (P, or polyvinyl actate)
glue on most gof my ring assemblies and ng
stacking glue jobs. I also use dto attach t
gs
to faceplate-mounted waste blocks and to glue
together design elements. I can flatten and stade
freshly glued ring assemblies in as little Ua
hour after gluing, but I am usually nat in that
much of a hwry. While a glue joint may b strong
enough to t, un-cured glue squeeze-out on the
sidelmyur tings can be messy to deal
with Hgets all over your gouge and then,
when sning yow gouge, it glazes over your
grinding wheel. Ti difficult to avoid without
waiting for haws or oveght. This glue's quick
curing time is a big advantage most of the time,
except for those big. time-consuming glue jobs.
Titebond Extend
Titebond extend gives me about 30% to 40%
more working time depending upon
temperature. I use it just as I use the anginal
when I need more working time. I have not
noticed any difference in performance
compared to the original.
Epox
Epoxy makes a nice paste when mixed with
sawdust to fill voids. Epoxy works well for small
inlaid work, but for most large glue jobs, in my
opinion, it is a little too brittle, not allowing the
wod to move at Even extremely dry wood
Wsometimes ty to move a little due to
moisture content changes. I have seen epoxy
fail because it had no give.
Cyanoacrylate (CA or super glue)
Super glue is great for quickly attaching small
components to waste blocks. I use Hto fill small
cracks and voids (in the wood. not between
segments). I do not use it for permanent
assembly of segments or for any large
attachments to waste blocks. It's an effective
stiffener when soaked into soft, punky wood.
Polyurethane
Polyurethane glue sometimes is used for oily
wood segment assembly. Uprovides longer
working time than Titebond Extend. It is good
glue, but messy and difficult to clean off your
hands. Rubber gloves are an absolute necessity,
but I still manage to get some on me and it
takes days for the cured glue to wear off. Good
damping pressure is essential and 24 hours of
cure time are recommended. One oddity about
GWE
this type of glue i that it is moisture activated,
meaning, that it needs a little moistwe to cure.
Super-dry wood should be lightly moistened
with a misting spray bttle before commencing
the glue job.
Hot Melt (sticks)
I use hot melt often for temporary attachment
of rings to waste blocks. It is easily removed
with a utility knife and/or chisel. 1t will not
hold items that protrude very far from the
waste block (vibration), but I have used it to
temprarily attach short rings more than 40
inches in diameter.
Spray Adhesive
I only use spray glue for the temporary
attachment of sandpaper to sanding blocks. A
light spray on the back of the papr, without
spraying the sanding block, allows you to easily
remove and change the paper.
Urea Resin
Sometimes called plastic resin, I use urea glue
for large lamination jobs. Hprovides about 30
minutes of working time. His a little messy and
inconvenient (you have to mix it), but it has
never failed to do its job.
Turning a vessel that is composed of hundreds
(or thousands) of pieces of wood means that
you are also turning that many glue lines. Glue
is much tougher on your lathe tools than green
wet wood, it blunts your cutting edges quickly.
Think of turning a circumference with 24
segments with your lathe speed at 1000 rpm.
That accounts for 24,000 tiny attacks on your
nicely sharpened edge every minute. You W
find that you need to re-sharpen much more
frequently when turning glued assemblies.
When purchasing turning gouges, look for the
hardest steel available.
Clamping
I always t to apply some pressure to freshly
glued joints in order to force all the excess glue
from my joints. How does one gauge the
reqUired pressure and how much should be
applied? His probably an acquired sense, but
28
THE AOF bGMNIDWOODRNING
6-02 You can never hav enough clamps.
my experiences tell me that many people try to
apply too much force with their damps. I
generally never apply more damping pressure
than I could I!Iwere applying the pressure with
just my hands or weight (1 weigh over 2OO
pounds). The manufactures of Titebond suggest
as much pressure as 200 pounds per square
inch; I am sure that in general, Iapply far less.
Remember, most of the tme the total surface
area is quite small. Getting the fit right ufar
more important and effective than cranking
down on your damps. 1tamount of glue that
you apply is imporant. too much glue only
requires more pressure to force the excess from
the joint. Ideally, when using glue such as
Titebond, only a small bead msqueeze-out
should occur. A thin coating of glue on both
surfaces ufar more effective compared to an
excess amount of glue on just one surface.
Because of all the necessary cutting I generate
hundreds of little pieces mscrap wood. I keep a
stockpile of them available with which to
spread glue.
There are many styles ofdamps available and I
use dtfferent types depending upon the task:
pipe damps for heavy duty lMtions;
.bar clamps for smaller laminations;
quick clamps (a small bar damp with a pistol
grip tightening handle);
spring damps of various sizes for small parts
and sometimes for joining rings;
rubber bands for small rings and irregularly
shaped small parts;
hose clamps for larger ring assemblies (they
can be joined to form any diameter), and
strap damps for the really big ring
assemblies.
Some of the most used clamps in my shop are
rubber bands. Passortment of sizes Cbe
cheaply acquired from any office supply store
and by doubling or tripling a single band,
a
variety of pressures is possible. Just when
I
think that I have finally accumulated enough
damps, a project comes along that requireS
even more clamps. I do not know a w
worker
who has too many clamps (photo 6'02)
GLUE 29
6-03 Gluing cauls enable pressure to be applied in the right di rection.
Pa segmented woodturner you will spend far
more time cutting, sanding, and gluing than
you will turning. Meeting the challenges of
creative damping is crucial to your success.
Generally, the goal is to apply pressure at a
perpendicular angle to the glue line. Sometimes
it is necessary to first attach gluing blocks
(cauls) onto your components temporarily, thus
creating the ability to apply clamping pressure
in the desired direction. An example of using
gluing cauls is shown in photo 6.03. This
particular glue job required two applied cauls
(MF pieces) in order to apply the needed
perpendicular pressure. Other times, you may
find that it is just easier hand-holding your
pieces for a few minutes. Regardless of
technique, at least some momentary pressure is
necessary. The components being glued in
photo 6-03 formed a feature ring for the cover
vessel, HRCKN5PIRI13.
Glue Safety
Most people never think about glue dangers,
but they should. Many of the synthetic glues
such as epoxy, polyurethane, resorcinol,
cyanoacrylate, and urea resin produce fumes
that are quite harmful to your respirator
system. Some of these glues also produce
harmful dust when sanded. I know of
woodturners who have had particularly bad
experiences with cyanoacrylates (super glues)j
these glues are handy, but you have to use
extreme caution. Sticking your fingers together
is minor compared to the damage that is
possible to your respiratory tract. Take the time
to turn on a fan, open a window, or step outside
-whatever it takes to protect yourself.
While the PYAs (Titebond and others) are quite
safe to use, they can eventually irritate your
skin with prolonged exposure. When I am
gluing many ring assemblies, I try to wash my
hands frequently and I usually apply a hand
cream to help protect my skin. Another glue
danger that few people ever think about
concerns the sharp edges that are often formed
by cured PYA glue squeeze-out. I have cut
myself on glue edges more than once. so be
aware.
30
JHE PTOF bECNENTFD VOODTUIN!NG
(OO LOAND WHI1 1POT
FcHNtUF8,RtcK8,ANDbHOkI-L 31
7
Techni ques, Tri cks, and Shortcuts
This is the meat potatoes chapter of the
1x k. I vdiscuss basic construction
techniques: how to cut, sand, glue, and assemble
components into forms. Several jig designs are
also offered -few things ensure successful
w working joinery more than good jigs.
Cutting Segments
Using a Miter Saw
Some of mypast works have contained more
than 6,000 pieces of wood, hence the term,
segmented woodturning. Some turers cut their
segments using a table saw and a few even use
a hand saw (with lot of sanding). but I do 95%
of mycutting on a compound sliding miter saw.
Jis faster, less tiresome. and safer. With good
equipment and technique, I believe it is more
accurate. Here is mymethod for accomplishing
safe, accurate cuts using a miter saw:
Start by installing a disposable zercrtolerance
back fence and bed on your saw; MDF (3'4 inch
or I inch) works well. I simply place about a 2
inch wide strip on edge against the sa back
fence, with another wider strip flat on the saw
bed. I secure these using a pair of small bar
damps, one at each end, that hold the flat
bottom board tightly against the back fence.
Relocating these each time you change the
blade angle gives you a fresh saw kerf. To reduce
the chances msmall debriS from interfering,
put a small chamfer on the back top edge of the
bottom board.
I use a damping device to hold the to-be-cut
strips against the back fence. The star-up shock
of your saw can cause small boards to jump
slightly away from the fence, causing inaccurate
cuts Uthey are not secured. A quick and easy
method is to spring-clamp a round piece of
wood to the saw bed and position a wooden
wedge between this circle and your strip of
wood. Ilike to use small circles of I-inch thick
MDF cut with a hole saw. The wedge can be
quickly removed and reinstalled between cuts,
as shown in photo 7-01 (next page). Spring
clamping the wood directly to the bed also
works, but it is tiring and not as fast to change.
\ the photo, note that I am gang-cutting two
strips at the same time.
Make sure your strips of wood are flat and
have square, parallel edges. It's difficult to make
accurate miter cuts from a warped board.
Use a stop block made from hard material,
something that will not dent easily. Bevel or
chamfer the bottom edge of your stop block to
prevent sawdust from interfering.
Make sure your blade is sharp. Even carbide
saw blades need sharpening after a few
thousand cuts.
Use a hold-down device to preent any
movement of the cut-off segment during
cutting. I use simple wooden L-shaped pieces of
wood, as shown in photo 7-01. On large
segments, there is no hold-down device as
effective as your fingers. I have never seen a
manufacturer's guideline for "how close Dtoo
dose" regarding fingers near the saw blade.
When cutting small segments (under 2inches
in length), using a handheld device is prudent.
You have to make the choice regarding when
not to use your fingers. If the segment is small
enough to make you pause and think about it.
then you should be using a device. In the set-up
shown in the photo, the segments are being cut
on the left of the blade. If I were left-handed, I
probably would reverse the set-up.
32
JH 1OFb5CMTDVCOD|UkINC
Hold-in
7-01 This miter-saw setup is how I (ut most of my segments. Note the hold-down deVice
The miter saw blade has 1chances to cut
the wood: on the down push, and on the up
$g. Segment-ends that move before the
blade ucompletely retracted are liable to
receive unVted scoring. Segments can move
as result ma Nrped wood strip. from debris
under the strip, or possibly from unintentional
side pressure by your hold-dL device. The
blade should Dsmoothly retracted fom the
wood without additional contact. The fact that
the saw blade has a chance to contact the
segment-end after it has been cut is the one
disadvantage of using a miter saw.
A dust extraction system connected to your
saw will greatly reduce the dust interference
problem and improve the air that you breathe.
Regularly check the 90 vertical accuracy of
your blade. One way to do this is to cut six
rectangular segments (no miter) about V inches
long, fom a strip of scrap about Vinches wide
and 3'4-inch thick. Cut these with the strip on
edge against the back fence. Lay these pieces
end-to-end on a smooth surface and then tum
over every other piece upSide-down. Use a bar
clamp to apply pressure from end to end, and
observe the alignment of the segments. |your
vertical blade alignment was off, the pieces of
wood will not fonn a straight line adjust
your saw accordingly. )this is unclear, visualize
a series of j cuts through a board. If the
orientation of the cuts remained unchanged,
the pieces would still fit together in a straight
line, because all the cuts would be parallel to
one another. By reversing every other piece, a
10 angle results at each joint. The goal is to
adjust your saw's verical angle so that any
error is not detectable.
To set your stop block to a specified segment
length, make a first cut at the end of your board,
then flip )over and use a caliper to make a
pendl mark indicating the desired length. Line
up the pencil mark with the edge of the kerf on
your saw bed. Make another cut and measure
the segment length, and adjust the stop block as
TECHNIQUES, klCK$,ANObHOk-'UT ]]
MD|bec
]O2 Using a cross-cut sled on the table saw Is an alternative method for cutting segments
necessary. If the segments are large, I usually
cut my first segment a smidgen too long and
then adjust Hneeded, i order not to waste
wood. A calculator usually displays at least five
or more decimal places. Cutting segments to the
nearest hundredth (plus or minus a few) is
usually dose enough. Calipers are much more
precise than a ruler for making small
measurements, however, expensive machinist
calipers are not necessary.
Once you have everything adjusted properly,
it is just a matter of cutting and flipping over
the strips and cutting again. Before each cut,
make sure your pointed board ends are resting
against the stop. It only takes one short
segment to cause a ring inaccuracy.
Using a Sled on a Table Saw
Before the availability of the modem sliding
compound-miter saw, the table saw was the
natural choice for cutting segments and uis still
a very accurate method. The goal Uto make
repetitve cuts that are consistent in length and
angle. Using the small, 6-inch wide miter gauge
that came with your table saw Ua poor method;
a sliding sled wlimprove the accuracy of
crosscutting on most table saws. Some table
saws are equipped with sliding tables; mine is
not, so the next best thing (and maybe even
better thing) Uto build a sled. A sled Unothmg
more than a sliding surface that transports jD
board through the saw blade. The simplest
design is just a piece of plywood with an
attached miter board that is guided on one Side
by your table saw fence -not a very reliable
design My exprience in cutting segments on a
table saw is limited, but if this were my primary
tool I would want a device that qUickly and
safely produced consistent cuts. Achieving thlS
requires more than just a sliding piece of
plywood I built the sled in photo y-u2usmg ]/q-
inch MF, but other materials such as Baltic
birch plywood could certamly be used. I have
painted the various components of the sled to
3
4
THPRTOPbFCMFNTDVOODHkNINC
better convey the design and construction. The
picture i taken from the front of the saw; the
operator stands on the side at the top of the
photo. The exact dimensions are not important
long as you understand how the device
works. To build thIs sled:
Start with a pIece of 3'4-inch MDF or plywood
big enough to span your table top from near the
left side (as you stand at the saw) to several
inches past the right-hand miter slot. The front
to-back dimension should be at least 20 inches. I
have shown this bottom layer painted white in
photo 7-02.
Mill two runners that barely fit into the width
of your miter slots and plane their height to
slightly less than the depth of the slot. Using
scres and a little glue, attach one runner
parallel to one end of the sled about 3 inches
from the edge. This would be the runner on the
left side of the photo; the ends of the runners
are painted red. With the saw blade lowered
below the table, check to see if the sled W
slide freely with just this first runner. Use a
sanding block to loosen the fit Unecessary.
Place a couple layers of veneer into the
bottom of the other miter slot, stick a piece of
double-sided tape onto the top of the other
runner, and trim the tape to the width of the
runner. Then peel off the tape backing. and
place the runner in the slot with the tape facing
upward. The veneers will elevate the runner
slightly above the table swface.
Carefully position the MOF with the attached
runner into its miter slot and then lower it onto
the other runner. Tap the top several times to
make a tight connection. Secwe the second
er by drilling, countersinking, and
installing three or four screws. There is no need
to remove the tape.
Place the sled back into the miter slots and
check the fit. Once again, use a sanding block to
adjust as necessary, creating a snug but smooth
sliding 1\Grasp the outside comers mthe sled
and check to see if any slop exists. A snug fit is
essential -an occasional spray of Teflon wl
keep things sliding.
Before
sliding the sled into the blade,
install
a
bridge across its leading edge. This will stabilize
the platform by keeping it rigid after you ct
partially through the platform. I glued several
layers of 3'4-inch MOF to create the bridge,
painted blue in the photo. In addition to
providing rigidity. the weight of the MDF
dampens vibration, thus improving the quality
of the cut.
Slide the sled into the blade to create a kerf a
little more than halfway towards the back edge
One problem with many sled designs is the
retrieval of small cut-off segments. One very
dangerous option is to retrieve the segment by
reaching over the blade before retracting the
sled. Pbetter solution is to use a hold-down
clamp on the cut-off so you can retrieve the
segment after sliding the sled back from the
blade. Without holding down the segment. on
most sleds you would risk the loose segment
making unwanted contact with the blade. On
the other hand, a hold-down adds another step
to the operation and it leads to unnecessarily
transporting the segment back alongside the
blade. Holding down very small segments C
also be difficult. Efficient cutting of hundreds of
segments demands a simpler design and
proedure. Ideally. segments should be easily
retrieved, the length stop-block should be easily
adjusted, and fingers should never b placed in
jeopardy. I think these requirements have
been
addressed with this particular design:
Instead of screwing and gluing a miter
fence
board directly onto the main sled platform.
install another layer of 3'4-incn MDF
across
the
platform at the approximate angle that you
want to cut segments. In this case, my goal is to
cut 11.250 angles on my segment -ends in order to
produce 16-segment rings. In photos 7-02 and 7-
03, the second layer of MDF is painted
green
and a strip of maple sits upon it ready
to be
C
A close look at photo 7-03 shows that the
green
MDF platform that holds the maple is
notche
d
back on the right side of the blade
and
:s
replaced by a small tapered piece m
MDf
(painted red). This tapered ramp
catches
the
cut
of and gravity prevents further
contact
WIth
the blade while the operator pulls back the sled
At no time during the cut or the retrieval do
fingers need to be near the blade.
For a miter fence. use a straight piece of stable
hardwood (I used oak). The angle of the oak
fence, painted yellow in both photos, is critical
and requires a little trialand-error to get
perfect. Using a protractor, position the miter
fence board as close as you can to 11.25. Using
two large screws, attach it about 3 inches to 4
inches back from the leading edge of the second
(green) layer of MDF, creating a shelf to hold
strips of wood. The shelf should be as wide as
any boards that you intend to cut. Secure the
end nearest the blade (1.03) with a zero
tolerance screw. that is, one in a tight hole
preventing side to side movement. Secure the
other end with a screw that passes through a
small slot in the board. This slot will allow
slight angle adjustments to create a perfect
11.25. In photo 7-02, the sloHnstalled screw is
the one with the large washer. The bottom
leading edge of the oak should be chamfered to
prevent sawdust from interfering with the
position of the to-be-cut wood strips.
During cutting, the saw blade ties to push the
board away. To prevent slippage, a thin strip of
80-grit sandpaper should be adhered to the
leading edge of the fence. When cutting sharper
angles, install a holddon clamp to eliminate
board slippage.
easily adjustable stop block can be created
by attaching an assembly of MDF to the table
saw rip fence. The exact design of the stop block
depends upon your saw's rip fence. My
configuration (painted blue) simply damps onto
the fence with two small bar clamps, as shown
in photo 7-02. Instead of using MF as the
actual stop-block material, attach a more
durable piece of hardwood to use as the contact
surface. After building the stop-block assembly.
slide it along the fence into the blade, trimming
the hardwood portion parallel to the fence and
blade. In photo 7-03, the piece of maple is
positioned against the stop block (painted red).
When using the sled, position the stop block on
the fence in a convenient location between
yowself and the blade. The stop block does not
FCHNlUFS,RICKS, ANDbHOkTLUW
7-03 Notice the red ramp that keeps the cut
segments away from the blade.
move with the segment, it only allows
consistent positioning of the board prior to each
successive cut. With a stop such as this, simple
adjustment of the table saw's fence easily alters
the segment lengths.
Now it is simply a matter of cutting a few
segments and adjusting the angle as necessary.
To start this process, I cut eight segments (haH a
ring). I cut the segments from 3-inch wide
material because inaccwate miter cuts would
be more obvious than if I had used narrower
material. Eight segments snugly held in a half
round shape against a straight edge will quickly
display most errors. Depending upon the natwe
of the inaccuracy, adjust the miter board and
cut 16 new segments. Secure these with a
rubber band or hose clamp and once again
check the seams. Continue this procedure until
a ring of 16 segments can be created and held
up to a light without shOWing any detectable
faulty seams. This may take several attempts
and requires patience. The slightest adjustment
affects a11 32 angles that make up the nng. A
tenth of a degree change in the cutting angle
adds up to a total change of 3f. Adjusting the
miter board angle is really splitting hairs.
Install several additional screws to prevent
future movement of the miter fnce once you
have achleved the perfect set of cuts.
This device is now ready to cut segments at
j JHPKOF bCMFN1DVODFNINC
7-04
There are two basic ways to cut a board into segments.
11.2S. but what about the need to build a 24-
segment ring or a 12-segment ring? You can
either install a different fence to cut different
angles, or build an entirely new sled for each
different angle. This is too much hassle for me.
that's why I prefer to cut segments using a
miter saw,
There are several very good commercially
available mitering devices. I have no
meaningful experience with them. but my
guess is they are not designed to quickly cut
hundreds of small segments, They appear to be
very good at cutting miter angles on a larger
scale. and easily adjustable to different angles.
For cutting hundreds (or thousands) of small
pieces I think a home-made device is superior.
and it sure is less expensive,
Two Methods
ofCutting Segments
A long time ago, someone figured out that if
you cut a segment from the end of a board and
then flipped the board upside-down. you could
cut another identical segment witil just one cut
(and so on and so on). This (ommon method
saves time and conserves wood. However, the
appearance of the resulting segments will differ
from one edge of your board to the other edge
and, dependmg upon the grain of the wood. the
look can be quite different. Even if the grain was
almost Identical in appearance, there would
most likely be a difference i the way that light
reflects from the two different surfaces. Most of
the time I do not worry about this lack of
consistency. nor do I often desire a perfectly
matched set of segments. However. there are
times when the distraction of inconsistency is
unwanted. For example, a feature ring with an
intricate design surrounded by a light-colored
wood could lose some of its effectiveness if the
light -colored wood appeared as different
shades, When a consistent look is the goal. then
segments must be cut another way. Instead of
flipping the board over to cut another segment.
it is flipped over to only cut as shor a segment
as possible (which is discarded) and then
flipped back to the original position to cut
another full-sized segment. By cutting in this
manner. all the segments will have the same
grain orientation and will be consistent m
appearance. Ualso pays to label the segments to
maintain the same order during gluing, This
tec.hnique requires more work and it uses mote
wood, which is why I seldom resort to it. but
sometimes the design demands the sacrifce.
With some pre-planning. the short segment
pieces can be used to form a different ring and
thereby aoid being discarded. A simple
illustration of the two methods is shown in
photo y-ug,I used a red markmg pen to
represent the saw kerfs of the two methods,
With this particular angle, one board produced
eight segments, while the other board produced
only six,
Gang-Cutting
A typical turing can require cutting hundreds
of individual pieces of wood. Using a miter saw,
CQone of those cuts requires placing and
securing the wood, turing on the saw blade.
lowering the blade. raising the blade, and then
removing the segment. With a little practice,
this can be done quickly. However, if you can
stack two or three layers of wood and cut them
at the same time, then you will significantly
reduce the number of operations and save
yourself considerable effort. There is little
difference in the cutting technique; the same
care must be taken to ensure that all the pieces
are against the stop block and against the fence.
When cutting more than just a couple of pieces,
it is helpful to wrap masking tape around the
bundle of strips. There is another slight
advantage to gang-cutting: the support of the
adjacent layers often reduces tear-out on the
underside of the wood. An example of gang
cutting two different strips, one holly and one
bloodwood, is shown in photo 7-01_
Cutting Thin Strips
My shop is equipped with a large, 14-inch table
saw, not the ideal tool for ripping extremely
thin strips of wood. If you have ever attempted
to push a 1/16-inch wide strip of wood between
the saw blade and rip fence, then you know
what I mean. The solution is to apply a general
woodworking technique: if the piece is too
small, attach it to something bigger. When I
need to cut thin strips, I proceed as follows:
I start by cutting a large piece MDF (at least 2
feet by 3 feet) with the fence positioned about 4
inches from the blade, but I stop cutting when
the leading edge reaches the outgOing side of
the saw table.
With the MOF clamped in place on the saw
table. it becomes a smooth, zero-tolerance
cutting surface. This is even better than a zero
tolerance table insert, because there is no insert
seam with the table surface.
I thickness-plane a piece of wood to the same
thickness as the wood from which I'll cut the
thin stnps. Using the MDF as the saw tabletop. I
rip the wood to the same 4-inch Width. This wIll
TECHNIQUES, TRICKS, ADSHORT-CuS
7-05 Cut thin strips by taping the ebony to a
wider board.
be my handle or push-board.
37
I adjust the fence away from the clamped
piece of MDF to the desired thickness of cut.
There will now be a slight gap between the
fence and the MF tabletop. The gap is no
problem, since it provides an escape for any
small debris that might otherwise interfere with
a tight fit between push-board and the fence_
Using masking tape. I secure my stnp
material to the edge of the 4-inch wide board
Usually I wrap several pieces of tape
perpendicular to the board. then run a piece of
tape down the entire length of the intended cut
If I am gomg to cut extremely thm stnps (less
than 1/8 inch), then I also place tape along the
entire length of the underside of the two pieces
of material. Photo 7-0S shows a small board of
ebony taped to a push board. bemg cut less than
liB- inch wlde. l can push this assembly through
the saw blade smoothly and achieve a clean.
accurate cut Without ever endangenng my
fmgers It is unportant to mamtam consistent.
38 THE PkQSEGMENTED WOODTlNING
7-06 Freehand disc sanding
(filst step) .
7-07 Freehand disc sanding
(second step) .
7-08 Freehand disc sanding
(wrong second step).
steady pressure against the fence to ensure a
smooth and parallel cut, but because of the size
of the push board, this is safely done.
I repeat the procedure as necessary. It is time
consuming and it uses a lot of masking tape,
but it works extremely well and is worth the
trouble: uniform, cleanly cut strips down to 1/16
inch thick, even less! Hthey are for a
lamination, I cut them just slightly oversized
and, using a transport tray. run them through
the thickness sander to erase any blade
markings while guaranteeing consistent
thickness.
Disc Sanding
No matter how precisely you cut segments,
professional results usually require disc sanding
the segment-ends. When joining dark wood to
dark wood, because minuscule imperfections in
the glue lines can be difficult to see, sometimes
it is possible to skip some sanding, but when
gluing light-colored wod to light-colored wood,
sanding is the only way to achieve the best
looking joints. A few turners disagree, believing
that acceptable joints w possible directly from
the saw blade. Many years ago Ray Allen, a well
known segmented turer, convinced me that
disc sanding was worth the effort. At a
presentation in Provo, Utah, he passed around
two samples of glue lines between curly maple
segments. One joint had been sanded before
gluing, the other had not. It was easy to see the
difference. The disc sander is the only practical
method that I know of for producing the best
possible joints, especially when joining light
colored woods. So how is it done?
First, you need a good sander: one that runs true
with minimum wobble, and whose sanding
table stays locked at 90 to the disc. Not aU
sanders are built alike.
The sanding disc has to be sharp, not wor-out
or gummed-up with wood extractives, pitch, and
oils. Dulsandpaper produces poor glue surfaces
and unnecessarily heats the wood, possibly
causing small heat checks in the end grain.
To achieve not only good glue lines but also
round rings, an accurate sanding jig is required.
If perfectly round rings are not critical, then
freehand sanding can be effective, but it has to
be done a certain way.
Freehand Sanding
and the Rub Joint
I use both freehand and sanding-jig techniques,
depending upon the size of the ring and the
importance of its roundness. If I only need a
small ring of conSistently colored wod, such as
a vessel bottom ring of ebony, I Woften not
use a sanding jig, however, I will still sand the
segment-ends. Before sandmg any segments, I
d-tit my nng together to check the fit. I can
qUickly determine it my angles are right-on or if
they are tight to the inside or outside of the
ring. I then freehand sand the segments
accordingly: if the fit was tight to the inside. I
apply a little more pressure to the inside of the
segmentends. and vise-versa. This is pure
guesswork. the size of the segments and the
type of wood both affect the pressure required.
Generally. very light pressure with very little
wood removal is required; I do not like heating
up the wood. I am mostly interested in
removing only the saw-blade markings.
Applying uniform pressure from end to end
takes a little practice. One exercise is to make
pencil marks on a segment-end before sanding.
sand just a little bit. and check to see if the
pencil marks are still barely visible. Repeat this
exercise until you can regularly achieve
uniform removal of wood.
The goal of disc sanding is to achieve perfect
glue lines. both vertically and horizontally
along the miter seam. When doing any
freehand sanding it is important to hold all the
segments in the same relative position as they
contact the spinning disc. I like to place the
outside (the longer side) of every segment
towards the outside of the disc; photo 7-06
shows this orientation. After sanding the first
end, to sand the second end. the segment
should be flipped upside-down and end-over
end. as shown in photo 7-07. not rotated end-to
end as shown in photo 7-08. By turning the
segment upSide-down you effectively erase any
vertical angle error. because any inaccuracy in
the O position of your sanding table is offset.
Imagine that your sanding table surface is at
j to your disc. By flipping the segments
during sanding. each glue joint will mate one
go.So angle to one 8g.So angle. resulting in a
perfectly tight vertical glue line. If your goD
sanding table setting is off just a tiny bit and
you do not flip your segments but instead rotate
them. then even if your miters form a perfect
360
"
, the segments will form a dish shape under
clamping pressure.
When freehand sanding and gluing together
pairs of segments. do not sand all the segment-
FCHNIUF5.R|CK5,AND bHOkT-LUT5 39
7-09 The rub-joint gluing technique starts with
pairs.
ends initially; sand only the ends that are to be
glued first. The sequence goes like this:
Layout your segments in pairs. then sand only
the opposing ends between each pair.
When making contact with the disc. use caution
to keep the segments flush against the disc.
Slide the segment a short distance (an inch or so
depending upon the segment size) along the
disc, being careful to maintain equal pressure
against the disc from end to end. Sliding the
segment reduces heat build-up and produces a
better surface. A smooth {almost slippery}
sanding table is essential. Hyour sanding table
is at Oto the disc and your miter blade was at
Oto the saw bed. then very little pressure
should be reqUired to erase all the saw-blade
markings. Good cuts leave hardly any
imperfections on the segment-ends.
Referring to photo 7-09. first only sand the
suraces of ends 1 and 2,jand 6. 9 and 10. and 13
and 14 (the red numbers}. lf you were to sand
the ends at the same time. then you would nsk
accidentally smearing some glue onto a sanded
surface before its assembly time.
Glue each of these pairs together using the rub
joint technique.
"What's a rub Jomt?" you ask. ThIS technique
40 THE Al OF bFCMNTD VOOD1URNINC
7+10 Rubber bands are optional. but recommended for large pairs.
forms good, tight joints with only momentary
hand pressure. You simply apply a little PVA
glue to one of the segment.ends and rb the
two ends together, checking to make sure glue
coats both surfaces. Then you place the pair of
segments on a flat surface covered with wax
paper or other non-stick surface, and keeping
the segments flat, aggreSSively rub the two back
and forth a few more times, keeping pressure
on the joint, until you start to feel slight
resistance. indicating the glue has begun to set.
Make sure the two outside corners line up. then
simply let the glue cure with no clamping
pressure. Tumers and joiners have used this
technique successfully for many years.
However, I do like to apply a little sustained
pressure whenever possible, either with rubber
bands around the segments as shown in photo
7-10, or with a damp on large segments. Most of
the time this probably is not necessary, but it
can't hurt and it might result in a tighter joint.
Hyou attempt to apply rubber bands, you will
quickly discover that you need to adjust their
tension in order to apply uniform pressure
across the glue surface. Experiment first with a
pair munglued segments by pulling them
apar while confined by a rubber band. The goal
is to adjust the band so that it requires equal
force to separate either end of the glue joint. It
only takes a few tries to get the right feel for
how to apply the bands. Rub joints will set
without the rubber bands, but I usually do use
them on large segments.
Sometimes, after only a few seconds, you
cannot readjust the alignment of two segments,
so Uis important to pay attention and align
them accurately the first time. Soon after gluing
segments together. 1 wipe the squeeze-out off
the joint using a piece of scrap wood: this saves
the sandpaper on my disc sander and
accelerates the glue curing. Rub joints and
freehand disc sanding go hand-in-hand.
After the glue has cured for a 15 minute
minimum on the first four segment pairs, disc
sand the top and bottom surfaces (the glue
lines) so that the segment pairs sit flush on a
smooth surface.
Now sand and glue end numbers 3 and q
together, and n and 12 together (the blue
numbers). During sanding, remember to orient
the segments in the same pOSition against the
sanding disc. Think of the pair of glued
segments as a single segment, which essentially
is what they have become.
You have transformed eight segments into two
half-sections. In a ring with more segments.
continue until two halves exist. You may have to
join just one segment at some point if your
halves have an uneven number of segments.
7-11 Half-rings can be trued up using the disc
sander.
Hyour ring is large it is a good idea to
occasionally dry-fit the sections together, to
check the overall ft. I usually dry-fit the final
four quarter-sections to see if they are going to
form two similar halves. I adjust the miters
slightly by disc sanding to achieve two
matching diameters before the final assembly
of the half-sections. This ensures that the two
ha-sections will align with each other with
minimal modification, thus resulting in a
relatively round circle.
The final gluing of the two halves requires one
last sanding of the ends, followed by an
examination of the final joints with a bright
light behind the jOints. This final sanding job
can be a little tricky depending upon the overall
half-ring size, especially if the diameter of the
half-sections is dose to the diameter of your
sanding disc. When sanding large half-rings,
you will be sanding one end against the side of
the disc that is rotating up with the other end
on the down side, as shown in photo 7-11. The
rotation of the disc as indicated is counter
clockwise, meaning the disc will try to pick up
the right-hand end of the half-ring. This
requires a firm hand on the up side to keep the
half-ring in contact with the table surface so u
won't be thrown.
Because the travel speed of the disc surface is
faster towards the outside diameter compared to
TECHNIQUES, TRICKS, AND SHOR-CuS
4'
7-12 This ri ng is too large to sand both ends at
one time.
the inside diameter, more wood is sometimes
removed towards the outside edge of the half
sections, causing a slight gap towards the
outside of the glue joint. This is why it is
important to check the fit with a bright light. Ha
gap does exist after very light sanding, then you
have two remedies: play around with the disc
sander as you sand only one end at a time (this
is how I usually do it), or hand sand the ends
using sandpaper adhered to a smooth, flat
surface such as a piece of MOF Hand-sanding is
easiest when the sandpaper is stationary,
allowing you to rub the half-sections across it.
Once a good final fit is achieved, then it is simply
a matter of gluing together the two halves, using
hose clamps or rubber bands for pressure.
Extremely large half-rings, those larger than the
diameter of the sanding disc, have to be sanded
one end at a time. Photo 7-12 shows such an
example. The disc is 20 inches in diameter and
the half-ring is 26 inches in diameter. P
accurate surface can still be sanded, but it can
require a number of trial-and-error attempts.
Sand a little and check the fit repeatedly until
no visible light comes through the joints.
Freehand-sanded rub joints should only be used
on small rings or when perfect roundness not
critical The method IS qUick and easy and
produces very good jomts
42 THE ART OF SEGMENTED WOODTURNING
ThiS is a typical disc sanding jig.
Building and Using
a Sanding Jig
There are many designs that demand completed
rings be round, not just almost round. H
segments are freehand-sanded, there is a good
chance that the joints could look fine but the
ring could be slightly ovaL Freehand sanding
takes a little practice whereas the use of a jig is
almost foolproof. When turning an oval ring on
the lathe, several problems can occur. Valuable
shape options can be lost because wall
thickness will have been lost while rounding
the oval shape. and design elements may not
remain consistent in appearance around the
ring. A sanding jig results in extremely
consistent segment angles and lengths
producing round rings. So, what is a sanding jig
and how do you build one? Photo y-1shows
one of my jigs. His important to understand
that a sanding jig can only be as accurate as
your disc sander. If your sanding table is not
firmly anchored with its miter slot parallel to
the disc, then no matter how precisely the jig u
built, it will not perform as deSigned. Check the
parallel alignment of your table slot and adjust
as necessary, and
check this alignment
regularly (photo 7-1).
)your miter slot is
not parallel, then yow
segments will be
different lengths
depending upon
where on the disc
they were sanded.
Consistent lengths are
as important as
consistent angles.
To build sanding
j
ig,
For a platform, cut a
piece of MDF or Baltic
birch plywood
approximately the
same size as your
sanding table_
Mill a miterslot runner that snugly fits
widthwise, and when placed into the slot is
slightly below the table surface.
Shim the runner with veneer scraps so that it
is slightly above the table surface. Stick on a
piece of doublesided tape and trim to the width
of the runner.
Carefully position the jig platform on the
double-sided tape with one edge of the
platform held against the sanding disc. Press
the platform onto the tape to create a solid
bond.
Remove the platform to drill. countersink, and
install five screws to secure the runner. The tape
can stay in place.
Check the fit of the runner in the sander
slot
and adjust using a sanding block as necessary.
While holding the MDF firmly in place. tum on
the sander and slide the platform back and
forth. You should have a snug fit with the diSC.
but the disc's rotation should not be restncted
after a few seconds of sanding action.
Cover the runner WIth masklOg tape and dnll
a small hole near the center of Its outsld edge
7-14
The sanding table alignment is important.
so you can hang the device. Spray several coats
of vanish to protect the MDF. Strive for a smooth
fnish that will reduce friction with the sanding
ta ble and allow segments to slide easily. Be sure
to coat both sides equally with fnish, so the
MDF does not warp.
1o make a fence for the sanding stop-block,
Da straight -grained piece of hardwood the
same length as the width of the platform and
about 2 inches to 3 inches wide; in photo 7-13 you
can see a piece of purpleheart attached to the left
side of the platform. Put a small bevel along the
bottom long edge that faces to the right, then use
screws and glue to install the piece across the
left-hand end of the platform. Strive for O

alignment with the disc. This assumes that your

disc rotation (as you face it) is counter-clockwise.
If your disc rotates the other direction. then
position the fence at the other end.
The next step is the sanding stop-block, a
piece of stable straight-grained hardwood (I
used mesquite) with an angle at one end. other
woods are certainly adequate -I chose
mesquite because it is durable and
dimensionally stable. The goal is to create a
perfect angle between the end of the stop block
and the sanding disc. The width of the stop
block depends upon the average size of your
segments. Uyou need to sand large segments,
1ECHNIQUFS,1k!CK5,AD SHORT-CUTS
9j
then the stop block should be as wide as the
segments. The stop block in photo 7-13 was
created to sand segments with an :1.2j

angle.
The stop block angle needs to be 22.j
=
because it
represents the overall angle of the segments
(both ends). The space between the stop block
and the disc represents the desired segment
angle and length. Notice in photo 7-13 that a
hold-down clamp is pOSitioned to secure the
stop block to the jig platform. On this stop
block, one end was cut at 22.j

and the other

end was left long so an adjustable limit-stop
could be installed. This adjustable limit-stop
controls the amount of possible wood removal,
creating consistent lengths .
To create the limit stop, I cut a 3/8-inch wide
slot about 4 inches long down the center of the
stop block at the opposite end from the angle
cut. The actual stop can be most anything; I
used a turned circle of hardwood with a 3/8-
inch hole in the center. A 3/8-inch carriage bolt
was used to install the round limit stop on the
underside of the sanding stop-block through
the slot. The length and position of the slot
should allow the angled end to be adjustable
from zero to about 4 inches away from the disc,
or whatever distance you predict will match
your longest segments.
Here is how the jig work:
Before turning on the sander, place one
segment-end against the sanding disc and the
other end against the angled end of the sanding
stop-block, but do not bottom out the inside
corer of the segment against the stationary
fence. In other words, adjust the position of the
sanding stop-block so that a slight gap exists
between the stationary fence and the inside
corner of the segment.
Secure the sanding stop-block with the hold
down clamp and tighten the carriage bolt with
the limit-stop held against the side of the
platform.
Sand one end of all your identical segments by
sliding the segments along the end of the stop
block until they bottom out against the
stationary fence. Try to alter your sanding
THE ART OF SEGMENTED WOODTURNING
7-15 The stop block angle must be perfect.
position by sliding the jig along the disc. I the
segments are small, it's prudent to use a
retrieval tool (photo 7-13).
With the hold-down clamp still iocked, loosen
the underside limit stop and place a thin shim
between the limit stop and the side of the
sanding sled. For a shim, I usually use a thick
piece of card stock paper or a piece of thick
veneer. Retighten the limit stop while holding
the shim in place.
Now,loosen the stop-block clamp (the hold
down clamp seen in photo 7-13), remove the
shim, reposition the sanding stop-block with
adjustable limit-stop against the platform side,
and then re-damp with the hold-down. It is a
lot simpler than it sounds. The thickness of the
shim dictates the amount of adjustment
towards the sanding disc, and equals the
amount of wood that will be sanded off the
second end. By adjusting the sanding stop-block
in this manner, you will avoid having to fiddle
with the adjustment. You Wbe able to quickly
adjust for the sanding of the opposite ends, and
avoid sanding more than necessar.
With the sanding stop-block repositioned,
sand
the other ends of the segments.
By now, you should have a good idea how the
sanding jig works, but before using it, the
sanding stop-block angle must be perfected. If
you were incredibly lucky when you cut this
angle, it will be exactly 22.5

. More than likely,

the angle will need a little tweakig and this
can require a lengthy process of trial-and-error.
To check and adjust this angle:
Cut and sand enough test segments to form a
half-ring. Cut fairly wide segments, since they
will show inaccuracy more clearly than small
ones. In photo 7-13 and 7-15. the holly segment
has a slash of red marker on it. This makes it
easy to keep track of tops and bottoms. You can
also see in photo 7-13 the small hook tool I use
to retrieve segments away from the moving
disc. One of these is simple to make from a
sharpened bent nail and a scrap of wood.
After sanding the segments assemble them
against a straight edge and determine whether
the angles are too acute or too obtuse. Hopefully,
the needed correction is small and here i where
things can get a little confusing. HgU test half
ring is less than lS0D, which means the segments
are tight towards the inside, then each segment
needs to be slightly more acute (or shorter on
the inside). The angle between the stop block
and the disc needs to be opened slightly, by
placing a few layers of masking tape on the
fence at the end closest to the disc. Photo 7-13
shows how the blue tape shim affects the angle
btween the stop block and the disc. The
addition of the tape has changed the angle of
the stop block so that the next test segments
Wbe slightly sharper in angle. If the test half
ring was more than ISO", then place the tape at
the other end of the fence guide.
With the masking tape adjustment in place.
adjust the sanding stop-block slightly closer to
the disc, re-sand the segments, and check the
ring accuracy again. Repeat the procedwe
unUl
a perfect set of segments can be created. It can
take a while, but hme spent perfectmg the JIg
wlll save much more time later.
Once the test half-ring is perfect. sand enough
segments to form a complete ring. Secure the
segments in a hose clamp (or rubber bands) and
inspect the fit against a bright light. There
should be no light visible through the seams. If
necessary. tweak the angle one more time.
To finahze the angle at the end of the stop
block. remove the tape shims and place one of
the perfect segments against the fence. As
shown in photo 7-15. carefully position the stop
block against the segment-end and make light
contact with the sanding disc to create the
required perfect 22.5 angle.
This perfect angle will not remain perfect if any
of the components becomes altered. most likely
the aHgnment of the sanding table with the
disc. If your rings start to show inaccuracy,
check the miter slot alignment first. These
instructions have resulted in a sanding stop
block for 16-segment rings. Other angles can be
created using the same procedures. and one jig
body can be used with an assortment of
sanding stop-blocks.
WhiJe I have used sanding jigs for many years, I
learned the technique of using masking tape to
adjust the sanding block fom Curt Theobald, a
tuer from Wyoming, at an AAW symposium
in Pasadena. California His trick can save you
hours of adjustment fiddling and an accurate
sanding jig can save you hours of ring assembly
time. Frequently slide the jig to utilize the
width of your disc. especially when sanding
many oily segments. Reverse direction on a
sanding disc is a handy featurej another set-up
can be installed at the other end of the jig, thus
increasing the life of the sandpaper. Do not use
the jig to correct grossly inaccurate miter cuts. A
saw blade removes wood much more efficiently
than sandpaper, so take the time to adjust your
saw as accurately as possible to minimize
sanding. The gluing technique for assembling
rings is discussed later in this chapter.
CHN!QU5,R!CK5, ANDbHOkT-LUT5 45
7-16 Accurate. tight gl ue lines are achieved by
disc sanding.
J
Z 4

5
^
b

10
1
2 l
1
3 i
14
1
1
5
7-'7 These are the individual steps to create
this feature ri ng component.
Disc Sander Fitting
Feature Ring Components
Feature ring components often require detailed
planning. They can be a puzzle, requiring
thought to determine the best sequence of
steps. As designs become more complicated. the
order in which you add pieces becomes cntical
It pays to write down your plan of attack before
starting. To illustrate. I will step through
building the feature-ring segment 10 photo 7-16.
46 THE ART Of SEGMENTED WOODTURNING
thIS is not a very comphcated design compared
to some -as shown in photo 7-17. there are 16
steps. Lcourse, the actual number of segments
reqUIred depends upon the diameter of the ring.
When makng this type of feature ring, you
would proceed by making all the necessary
segments at the same time. For this exercise I
am going to focus on the construction of the
design in a single segment.
Step 1 Rip dark wood strips about 1/8 inch wide
and light colored strips about 1/4 inch wide. The
length of the strips depends upon how many
segments are needed. Laminate one strip of
each color together.
Step 2 Clean up the laminated strip and
crosscut pieces 1/2 inch long. The number of
pieces is twice the number of segments.
Step . Cut short pieces (about 1/2 inch long) of
the dark wood and glue them to one end of the
short laminated pieces. Cutting them a little
longer provides room for alignment error. Make
sure the ends extend slightly beyond the width
of the first pieces so that they can be sanded to
fit. Spring clamps are particularly useful for this
type of clamping.
Step 4 Use a disc sander to true up the top and
bottom of the assembly by sanding off the extra
material. creating a nice smooth surface for the
next layers.
Step 5-6 Continue this process of gluing on
pieces and sanding the surfaces for the next
pieces, H shown in photo 7-17. Whenever
sanding off excess wood, be careful to keep
things square and try to minimize the removal
of material from the previous piece.
Step 7 Instead of adding one piece, glue on two
pieces as shown.
Steps 8-1 Continue the process of sanding and
adding U shown.
Step 15 Miter a center piece from dark wood,
the exact angle depending upon the number of
ring components. Turn one of the completed
sections upside-down, and glue the three pieces
together. Placing a mitered piece in the center is
optional; if a parallel pIece were used. then all
7-18 A bright l igDt can help expose i mperect
joints.
of the required angle would have to be inserted
between the completed segments.
Step 16 Sand the top and bottom surfaces.
To assemble these segments into a ring,
position another spline piece between them.
The second spline can be either mitered or
straight-sided, depending upon the design.
Assembling rings is discussed in much more
detail on the next few pages.
The key to successfully building such a ring is
very precise disc sanding. It takes a little
practice, so be patient. It is a good idea to
build
a few extra components. so you have the luxury
of discarding the flawed ones. Countless designs
can be built using this technique. The size of
this example is fairly small, which allowed for a
few glue lines of opposing grain orientation. U
the same design were built in a larger size, then
a few of the vertical components should b cut
with hOrizontal grain in order to avoid conflicts
in wood movement. This particular
feature
ring
was used in the creation of TANo WF
WoD.
a
turning in photo 13-16 (page 135).
7-19 Two opposite small spacers will absorb
tiny misalignments.
Gluing Segments Together
The Hal-ring Method
With the near-perfect results of a sanding jig,
segments do not have to b glued in pairs as
described in the rub joint technique. A faster
way, the haf-ring method, can be used. It is
similar to gluing of the segments together at
one time, with one big diference: two joints are
not glued initially. After jig-sanding all the
segment-ends, the ring should b dry-fit and
checked for accuracy against a bright light, as
shown in photo 7-18. I use a ceiling-mounted
300-watt bulb to check dry-Oped rings.
Aer confirming that the ft is near perfect,
attach a small piece of spacer wood to the
center area of t segment -ends using a little
dab of glue or double-sided tape (photo 7-19).
During the glue up these joints do not receive
glue. The spacers act as a fulcrum so that any
error will be taken up by the space they provide.
Glue and damp as Hit was a completed ring,
except the spacer joints remain dry. This u
shown in photo y-2o.After the glue has cured.
the two halves can be taken apart, touched up
on the disc sander as necessary, then glued
together to 1a complete ring. The key to
success with this technique iS,the fit of all the
pieces has to be near perect, which Iequires
using a precise sandingjig.1f your angles
produce two slightly oval halves. then the
chances of success are much less: the pressure
TECHNIQUES, TRICKS, AND SHORT-CUTS
47
7-20 Spacers help ensure two perfect half rings.
of the hose clamps wltr to form a circle and
wllikely separate a seam somewhere within
the ring. If I am gluing together lots of rings in
this manner, then I usually gang-cut a stack of
small l/8-inch dowels into short pieces, then
attach them using just a dab of glue. The
roundness of the dowels creates a very effective
fulcrum.
The All-At-One-Time Method
The all-at-one-time method is just what Hsounds
like. His si to the halfring method except
there is no room for error, and errors wl
occasionaly OC.Every joint uglued and
clamped at one tme. Hucertainly faster than
other tedmiques, but it is also more risky. To be
successful, t conditions should apply: the dry
fit of the ring should be perfect to the naked eye,
and there should not be any light-colored-to
light-colored seams. The frst condition is self
explanatory. The second condition (light to light)
should also Deasy to understand: dark woods
hide seam imperfections much more effectively
than light-colored wos. When gluing an entire
ring at once, it only takes one segment slightly
out of place to spoil the ring. If the segments fit
perfectly when dry-damped, then they wlnot
fit perlectly when glued unless all the pieces are
in the same position as they were during the test
fit. Improving your chances of success requires a
generous application of glue to lengthen working
time, careful attention to the aligrunent of all the
48 THE ART OF SEGMENTED WODTURNING
7-21 Gluing a full ring at one time requires
pe|Jetl miters and extreme care.
segment comers, uniform clamping pressure all
around the ring, and accomplishing the job as
quickly as possible. When applying pressure
around a ring of freshly glued and slippery
segments, the miter angles naturally try to force
all the segments away from the center, but the
clamping device prevents that from happening.
However, the hose damps or rubber bands
cannot pull a segment to the outside if the glue
joint were to set prematurely, which would
prevent the segment fom sliding. A segment
that is not squeezed to the outside (a
misalignment of outside comers) will likely
cause an unacceptable seam somewhere in the
ring. Photo 7-21 shows an all-at-once glue jab. I
usually limit the use CUtechnique to small
dark rings with narrow segments.
Rubber bands are my clamping chOice when
clamping relatively small diameters; they apply
pressure more uniformly than hose clamps and
they are quick and easy to apply. Additional
clamping pressure i Simply a matter of applying
more bands. A wide variety of rubber band sizes
can be found at any ofce supply store; I keep
several sizes on hand. When clamping very small
rings (under 4-inch diameter), you also have the
option Odoubling or tripling a large band in
order to produce the needed pressure. I have the
advantage of large hands and I can stretch bands
and place them around rings up to about 7 inches
In diameter. Lrger rmgs require a rubber band
7-22
A rubber band gluing jig Is just a D|0
with small holes and nails.
gluing device. This is nothing more than a
of MF drlled with concentric rings of small
holes that accept nails. To use this jig. Ipl \
paper over the board and stick nails through U
paper into a circle of holes slghtly larger
the ring to be glued. Before applying glue to U
segments, I position a rubber band around the
nails. Then after applying glue to all the segment
seams and placing them within the stetche
band, [ carefully remove the Hto entrap the
ring. To apply additional bands, I keep two of the
nails in place (between the band and the
segments) and use them as an anchor while
stretching more bands around the ring. I usual
apply at least four bands, depending upon the
ring diameter and the size of the bands. After
applying the bands, I inspect and adjust all the
segment corners, I check the evenness of the
segment surfaces by using a piece of scrap W
to remove glue squeeze-out. and I flip the ring
over to inspect the other side. Once I approve
the
fit, I hang the ring on a horizontal pipe covered
with masking tape, which allows glue squeeze
out on both ring surfaces to cure. When working
on a big project, I sometimes have
dozens
of
D
hanging on pipes awaiting the next step. The
rubber band clamping method works wJ when
gluing rings up to about ] inches in
diamet
er.
After that, the choice becomes hose
damps.
Photo 7-22 shows a sample of this
gluing
Jig.
The
use of the rubber hand jig is not
limited
to
aJ-at
-
TECHNIQUES, TRICKS, AND SHORT-CUTS 49
7-23
A disc sander can be used to flatten one side of a ring.
one-time gluing jobs; I frequently use spacers
and glue half-rings as previously describd.
Note: Rubber bands should not be used with
polyurethane glue, because it expands as it cures
and requires more restrainingjorce.
1have found that when using Titebond, I have
enough time to glue up to a mBum of 36
large segments at one time (provided the
temprature is cool). lf the temperature is
N then 36 is too many. When I have too
many pieces within a ring to glue at one time,
then I glue pairs together, thereby reducing by
half, the number glue joints that I have to deal
with in one operation. As a rule, I want the first
joint that receives glue to remain slippery until I
damp around the Whole assembly and check all
the comer alignments. For the novice, let me
stress: do not attempt to dose poor glue joints
with extreme clamping pressure.
Preparing Rings
for Stacking
Segmented rings always need flattening before
attachment to other rings. Here is how I usually
do it. When I need to glue a small ring (under 14
inches diameter) onto a vessel, lusually use my
20-inch disc sander to smooth one side. My
sander has a vertically adjustable sanding table
that can b quickly cranked out of the way,
allowing access to the entire disc. I simply
handhold the ring against the disc and make a
few swirling motions. Bo.grit sandpaper qUlckly
flattens and smooths the surface, thereby
creating one side of a tight glue line between
layers. If the ring is thin and fragile, making it
difficult and/or dangerous to hold against the
disc, then I make a temporary handle by hot
gluing It to an MOF backng plate, as shown in
photo 7-23.
50
THE ART OF SEGMENTED WOODTURNING
7-2
4
The 1/2-inch bowl gouge i s by far my most-
used cutting tool.
7-26 The flatness of this surface needs
improvement
There are times when the disc sanding option is
not available to you. For example. as you build a
vessel by stacking rings, you should flatten the
uppermost ring surface on the lathe, because
this Wensure that the seams remain parallel
to each other. Lathe-flattening is very accurate,
a little more time-consuming. and requires a
little practice. Whether you are flattening a
permanently glued surface. a top rim, or a
temporarily mounted rig. the technique is the
same. Here uhow I do this:
7-25 Checking for flatness requires a good
quality straight edge ..
7-27 This diamond-point tool is versatile,
quickly resharpened, and does a great job on
NU|
First, using a I/2-inch bowl gouge, I tum away
the
outside and inside comers of the
segme
nts.
Thi s results in a round shape, which is simply
easier to modify (photo 7-24).
Next, I use my bowl gouge to smooth
the
glue
surface.
This step removes any glue
squeez
e.out
and
eliminates
the ridges between
segmen
ts
The exact type of cut that you use is a
pers
onal
preference.
just be careful to remove
a
U:D
Itn8
!
amount of
material. A light touch
of your
hng
tt
ECHNIQUtS, 1RICKS, AND SHORT-CUTS 51
7-28 Using a flat sanding block is the final step prior to gluing the next ring.
while the ring is spinning wdetect most
imperfections.
Now stop the lathe and visually (and by
fingertip) inspect the surface. If the surface is
smooth, I place a steel straight edge across the
surface from side to side. Smoothness
flatness are not the same thing. By holding a
small bright flashlight under the straight edge, I
can inspct the surface for flatness. To achieve a
tight glue line between layers, the two surfaces
have to be perfectly flat.
eDepending upon the results of my
inspection. I shave a little more wood off the
highest area. Photos 7-25 and 7-26 show the
difference between acceptable and not quite
dose enough. The straight edge has to be
extremely straight and you have to take a very
dose look at the contact between it and the
ring surface. My goal is to achieve a perfectly
consistent light transfer from the underside.
eAt this stage, to make tiny fn adjustments, I
often switch to a small diamond-pointed
scraper in an angled shear-scraping poSition to
perfect the surface (photo 7-27). small tool.
with its slightly Oc cutting edge, allows me
to remove very tiny shavings. Hmay take three
L more attempts to get it right. When lathe
flattening an extra wide segment ring (2 inches
to 4 inches from outside to inside), the process
usually requires even more attempts.
I cannot overly stess the imporance of
accurate ring flattening. I am talking about
smidgens (the smallest difference detectable by
the naked eye). In photos 7-25 and 7-26, the
surface of the ebony ring is ready, while the
holly ring needs just a little more shaved off
towards the inside. Hyou make the shghtest
compromise, you risk contact between two
imperfect surfaces, and then the error could
become doubled and very unacceptable The
52 THE ART OF SEGMENTED WOODTURNING
7-29
Using a clamp against the ceifing is one way to join rings.
surface imperfections;
before sanding, the
surface should be as
good
as you could get it. When
touching and removing
the sanding block from
the spinning ring, avoid
rounding the edges of the
ring surface. When I first
started using the
technique, I was so
concerned with pOSSibly
rounding an edge that I
would start and stop the
lathe before applying or
removing the sanding
block. I no longer stop
and start, but I am
careful to maintain the
parallel alignment of
sandpaper to ring SID
faces. Replace the sand
paper frequently, it is a
tool and like your gouge
it needs to be sharp.
lathe cutting tool and the technique that you
use is not nearly as imporant as the result. Use
what you are comfortable and familiar with.
Just be sure to closely inspect the surface and
avoid any compromises.
The final step in preparing the surface for
gluing is to hold a sanding block (photo 7-28)
across the ring from one side to the opposite
side, with the lathe turning at a moderate
speed. I use 80-grit paper, attached with spray
adhesive to either 3/4-inch or I-inch MOF. When
flattening a large-diameter ring, I like to use l
inch MOF with a piece of sandpaper attached to
each end. Having several different sized blocks
on hand u a good practice -you would not
want to try sanding a 2-inch diameter ring with
a 24-inch sanding blok. If you spray a light
coating of adhesive on the back of the
sandpaper, it wl hold, while being easily
removed later for replacement. If you have done
a good job with your lathe tools, then just a few
seconds of sanding should eliminate any
remaining Imperfections on the gluing surface.
Do not rely on the sanding block to correct large
Clamping Techniques
Over the years I have devised many ways to
glue and clamp rings together. Extreme
pressure is not necessary, but the accurate
centering of the rings to each other, and the
vertical alignment of the seams, are important.
To glue an un-mounted ring onto a mounted
ring, the easiest method is to simply place the
un-mounted ring on a table surface with wax
paper and a rubber router pad under it, apply
glue to both sides of the jOint, and place the
mounted ring (or turning) on top, as shown in
photo 7-29. Moderate clamping pressure can be
applied using a push style clamp or by just
stacking some weight. Betore applying
glue, l
always dry fit and determine the rotational
orientation of the two rings. A pair ot divider
s
Whelp determine the center location between
segment-ends. I make a pair of pencil marks
once I finalize the deSired position, or I apply a
piece of masking tape cut at the seam
Accur
ate
centering of the two rings can be achieved by
TECHNIQUES, TRICKS. AD SHORT-CUTS
5
3
7-30
Using your lathe's tailstock works well for centering and clamping.
measuring the distance between the two
outside diameters during the dry ft. Adjust the
position until the distance is equal on all sides
and make a note of this measurement. Now
during the gluing you have to align the pencil
marks and check the outside measurement in a
couple of locations. Hthe difference in
diameters is small, the centerng can usually be
done by eye-balling and/or using your finger
tips to gauge the ring positions. Use plenty of
glue, which will give you a little more adjusting
time. Apply moderate damping pressure and
make sure the two surfaces do not slide out of
poSition. Hthey do slide, adjust the clamp angle
until they stay in place. The amount of pressure
need not be great, between 1OO and 300 pounds
depending upon the total surface area, enough
1Oforce the excess glue from the seam and
produce tight contact between the surfaces.
I use various techniques for gluing layers
together. The previous example involved joining
an un-mounted (multi-Sided) ring to a vessel
under construction. Another method involves
using the lathe Qyow damping device to join a
mounted ring. Mer flattening the ring. using
either the disc sander or the lathe. it is dry fit.
aligned, marked, and glued H shown in photo
7-,0. If I am joining a larger ring to a smaller ring.
I usually tum the mating surfaces close to the
same diameters, which simplifies the centering
and reduces the ring width that needs flattening.
The technique of using the lathe tailstock as a
clamping/centering device work eremely well
Htes up the lathe for a few minutes, but usually
I have something else to work on while I wait for
the glue to partially cure. The ring to be added
does not have to be lathe-tured round in order
to use this tcque. Any ring that has been
center-mounted flattened on a round circle
of MF w a center-drilled hole can be pressed
onto a lathe-mounted ring by positoning the
tailstock center in the center hole.
Before applying glue. always perform a dry fit,
determine the rotational alignment. and make a
couple of pencil marks. Apply glue to both
surfaces and then bring the tailstock into place
to hold one ring against the other. Apply slIght
pressure while twisting the two rings back and
54
THE ART OF SEGMENTED WOODTURNING
7
-
3
'
Unmounted thin rings can be joined with spring clamps.
forth. This spreads the glue evenly, same as a
rub joint. Align the two pencil marks and apply
additional pressure using the tailstock
tightening-handle. Photo 7-30 shows the two
pendl marks on the bubinga and holly. Uyou
are in a hurry and the temperature is warm, you
can remove the assembly from the lathe after
about ten minutes. This gluing technique works
well on small- to medium-sized rings.
Segmented forms can contain dozens of layers.
It is not necessary to join those layers one at a
time. To speed up the process, consider joining
two rings to form a thicker ring. After using
either the disc sander or the lathe to flatten the
mating surfaces, rings can be damped together
with spring damps as shown in photo 7-31.
Uses ofMDF
(Medium Density Fiberboard)
The design of small turngs usually calls for
small, short rings; to speed up the process of
consolidating dozens of small rings, I he
devised a technique using MDF circles.
I create
MDF discs using large hole saws on a drill press,
because the hole saw cuts a round disc with a
centered 1/4-inch hole. By enlarging the
center
hole to 5/16 inch, it accommodates Sl16-inch
bolts or all-thread, which when fed
through
the
holes, automatically centers the two
rings.
By
centering and attaching rings onto
these
MDF
circles,l can bolt them together without being
concerned about the centering, thus allowing
my attention to focus on rotational
alignm
ent.
Using a single carriage bolt, a couple
of
was
he
I>.
and a wing nut, I can join flattened
rings
together very quickly and accurately
By
glUI
ng
732 MDF circles are very handy as temporary
waste blocks.
hole-sawn circles onto larger circles of MOF, I
can mount, disc-sand, and glue together larger
pairs of rings very easily. Tassortment of MDF
circles is shown in photo 7-32. By using only
small beads of hot-melt glue, rings can be
removed easily from these circles and thus
many repeat uses are possible. In photo 7-33, I
am joining a single ring onto a stack of
previously glued rings. On the underside of the
stack, another 3-inch diameter MDF circle acts
as a means to lathe-mount the assembly using
a four-jaw chuck. In the photo, the underside
MDF circle is clamped in a vise, which makes it
possible to twist the upper MDF circle using a
large pair of groove-joint pliers. After the glue
job shown has cured, the assembly will be
lathe-mounted. the smaller-end MDF circle will
be removed. and the next glue surface will be
lathe-flattened. Then another ring that has
been disc-sander flattened will be joined, just B
shown in photo 7-33. This technique is a big
timesaver when faced with consolidating many
rings. More examples of this technique are
shown in Chapter 12.
A few tips about hole-sawing MDF circles:
Use a relatively slow drill-press speed.
TECHNIQUES. TRICKS, AD SHORT-CUTS 55
7-33 A centered bolt and nut works well for
clamping small rings.
Orient the hole saw edge slightly outside the
edge of the MDF board so that it creates an
escape for the sawdust. This is essential,
otherwise. you will generate too much heat.
Set the stop on your drill press quill so that
you cut almost all the way through the
material, but not quite.
Flip the board over to complete the cut. which
makes it easy to remove the disc from the hole
saw. With a little practce. the cut disc can be
retrieved without stopping the drill.
Cutting all the way through from one SIde
traps the circle deep inside the hole saw and
requires tedious removal
I use a lot of MDF in both 3/4-inch and 1-inch
thickness. His relatively inexpensive and
extremely versatile. However. it is also Dof
chemical adhesives that produce very
unpleasant sawdust. A good dust mask and a
dust extraction system are essential. MDF glues
and holds screws well. making it ideal for jig
building as well as for lathe-mounted waste
blocks. If you are not familiar with MOF. think
of super-smooth particlebard made from W
dust instead of from wood dups. l should
56
THE ART OF SEGMENTED WOODTURNING
7

34 Veneers can be used as a design element.

7-35 Veneer strips stacked between MOF cauls
can be quickly cut into segments.
7-36 Using veneer segments (as opposed to one
piece) maintains consistent grain alignment.
caution you however: MDF should not be used
as a waste block for large items unless Ris
penetrated by numerous mounting screws. For
example, do not glue a circle of MDF to a large
turning and then mount it in a four-jaw
chuck,
because the MDF is liable to delaminate
without reinforcing screws. Number 14 pan
head screws in various lengths are usuaUy my
choice for attaching faceplates.
Working with Veneer
Veneer commonly provides a contrasting
narrow band of color. Photo 7-34 shows an
example af veneer near the top rim of a vessel.
Adding an occasional thin, contrasting band can
dramatically improve a turning's appearance.
Even though veneer is quite thin, it has grain
direction and should be oriented as you would
orient thicker pieces of wood. Hyou desire to
place a harizontal layer of veneer between two
thicker rings of wood, then it should be mitered
together the same as the other segments to
maintain wtiform grain direction. Uwould
certainly be a lot easier to glue one large piece
of veneer, thereby saving time and hassle. and
with a very small diameter, this is probably not
a problem. With larger diameters. and therefore
longer segments, it would be risky. Also, with
just one piece of light-colored veneer, the
sides with exposed end grain would appear
darker than the other two sides. Mostly, I use
maple and holly veneer for light
contrast,
and
ebon-x or walnut for dark contrast. I
typically
keep a supply of veneer strips stacked
betwee
n
two pieces of MDF or particleboard and tightly
secured with masking tape (photo 7-35). When I
need a ring of veneer segments, J gang-cut them
on the miter saw. The saw blade will sometimes
fracture a few pieces, so Tis a good idea
to
always cut a few extra.
To assemble veneer segments into a ring, use
masking tape, not glue. Position the
segm
ents
on a smooth MDF surface and apply a
small
piece of tape to hold each joint tightly
togeth
er
Masking tape does not have much stretch, but
by holding down one end of the tape
on
one
segment and pulling Honto the
other.
shght
pressure can be applied between the veneer
segment-ends. One at a tme, tape the segments
together until two separate halves exist. The fit
of the two halves Wsometimes require a little
touch-up on the disc sander and then they too
can be taped together to form a complete ring.
In photo 7-36, I show an eight-piece veneer ring
in the process of being taped together. The
veneer segments were gang-cut On the miter
saw at 22.50 to form the complete circle. The four
segments at the bottom of photo 7-36 have been
taped together end-to-end to form half of the
ring. After joining the other four segments, the
ft of the two halves will be checked and
adjusted as necessary.
Fi, consistent clamping pressure is required to
glue veneer rings onto another surface. Place a
Upiece Orubber (a TOuter pad works well) on
a smooth surface. Put a piece of wax paper on the
rubber pad and place the veneer ring on the wax
paper, taped side down. Spread glue onto both
surfaces and clamp the joint with slightly more
Unormal pressure. I use my workbench with
a push clamp from the ceiling, but many other
clamping techniques are possible. His important
that before applying glue you check the evenness
Oathe veneer segments, because an
overlapped piece Oveneer V Othe joint
The rubber pad ensures DOdistribution of
Oping pressure to aOthe veneer segments.
Hone of the veneer pieces were just slightly
thicker or U er, then Vout the rubber pad a
poor glue line could result. The rubber pad takes
up any differences. To prepare the veneer layer
for the net ring in the turning, use a lathe tool
to cut away the overhanging excess veneer and
lightly touch up with a sanding block.
How to Make Ebon-x
Black veneer is expensive and not commonly
available, that's why I create my own by dyeing
walnut veneer, otherwise known B ebon-x. This
is the only time I ever alter the natural colors of
any of my wood. It's a simple process:
Cut walnut veneer into pieces that will fit
into a large baking dish (9 inch by 1 inch by 2
ic works). If you require larger pieces, then
use a larger container.
TECHNIQUES, TRICKS, AND SHORT-CUTS 57
7
-37 Making two rings from one can save time
Make a solution of apple Cider vinegar and
iron filings. I shred a pad of rusted steel wool
and let it soak in the vinegar for a few days,
shaking it occasionally.
Stack the veneer in the baking dish, one piece at
a time, while pouring the solution onto the
sheets, making sure that asurfaces receive a
coating.
Top off the pan with any remaining solution,
cover with a piece of plastic wrap and soak the
veneer for a couple of days.
Remove the veneer one sheet at a time and
stack the pieces neatly with a white paper towel
sandwiched between each sheet. Place a flat
board onto the top of the stack and weight it, or
clamp it together.
Allow the wet veneers to thoroughly d,
which can take several weeks.
Splitting Rings
into Multiple Rings
This opportunity frequently presents itself
especially in small turnings. Imagme a vessel
S
8
THE ART OF SEGMENTED WOODTURNING
7-38 Small plugs can be used i n vessel bases.
design that calls for many short (less than 1/2-
inch tall) rings. By first constructing I-inch tall
rings and then splitting each into two rings, you
V reduce ring construction time (photo 7-37).
To split small rings on the lathe.
Disc-sand both horizontal surfaces, then use
hot-melt glue to center-mount a round piece of
MDF on each surface. The MDF circles should be
small enough to ft in a four-jaw chuck, or
should have a center-drilled hole that fits a
screw chuck.
Mount the assembly and twn the outside
profile round. Use a pg tool to separate the
nng into two halves. Sometimes, when faced
with the splitting of fragile rings, instead of
completely parting the rings, I only cut about
halfway through with the parting tool and use a
band saw to complete the cut. This aVOids the risk
of catching a separated half-ring. Cutting roud
objects on a band saw is not a recommended
practce; hLer. it can be safely done using the
handles that the MDF dIdes provide. In photo 7-
/.
a 12-inch ring of ebony segments is being
parted. In this situation I do not want to cut all
the way through the ring for fear of losing the
tailstock portion Therefore, I only part halfway
through and then use the bad saw or handsaw
to complete the separation
7
-3
9 Make tapered openings for plugs.
The result is two mounted and centered rings.
These rings are available fr use as any other
rings. Using them on opposite sides of a feature
ring is one option Look for these opportunits
and save yourself ring assembly time ad effort
In chapter 12, another method of ring splitting
using a table saw is demonstrated.
Vessel Bases
Tapered Plugs
There are three basic ways to construct the
bottom of a vessel: for small bases Iuse a
solid
piece of wood upon which layers of rings are
attached, for slightly larger bases I like a ring of
segments surrounding a tapered plug, and for
large bases, a floating disc-shaped plug wOtKs
best. The goal is to create an assembly that w
not become over-stressed due to inevitable
wood movement.
Many segmented turners build all of their
vessels by staring with a solid piece of wod,
and you might be thinkig, "Why not?" When
constructing small turings with small bases,
there is probably little reason not to start in
that manner. However, I prefer the tapered
plug
technique.
Havoids the exposure
of end
graHI
.
which a solid piece of wood displays
, and I |
us
thInk it looks better. Remember, changes in
moisture content V result Umovement
perpendicular to the grain direction (Chapter 5).
A small plug inside a wide ring of segments is
going to move less than a large and wide solid
piece of wood. In addition, if you use a wide
solid piece of wood alone, then the two sides
displaying end-grain V often finish in a
darker color compared to the other two sides.
Another consideration is the type of wood you
select for the vessel bottom: you want very
stable WStudy the grain patterns and t to
select wood that is quarter-sawn. Examine the
ends ma few boards and if possible use
something that displays growth rings parallel
to the edge of the board, as opposed to growth
rings parallel to the surface of the board.
Generally, this type of grain orientation is more
stable and the piece of wood will move less
with humidity changes.
Accurately fitting a plug into a ring of segments
requires a little patience. Start by using calipers
to check the diameter and enlarge the inside of
the ring dose to the desired plug diameter with
a slight taper, then turn a slightly tapered plug
a little larger than the ring's inside diameter.
Photo 7-38 shows a small tapered holly plug
ready to be glued into a base ring.
Now, remount the base ring on the lathe and
carefully remove material, checking the fit
frequently. As you get closer to a good fit. you
also need to create matching taper angles so
that the fit looks good on both the inside and
outside of the vessel. To do this, make sure that
the bottom or narrowest part of the fit is tight
first and then take very small cuts until the
top of the fit comes together. If you first
concentrate on the outside fit, then you have
no way of determining the accuracy of the fit
on the other side.
Another technique, once you are very close to a
good fit, involves insering the plug into the
ring while the lathe is spinning. This will
immediately indicate the location of any
tightness by slightly burning the plug. You
should use this procedure very carefully
because the plug can be grabbed and torn from
TECHNIQUES, TRICKS, AD SHORT-Curs
19
your hands. Keep a firm grip and insert the plug
with extreme gentleness. Also, try not to create
dark burn marks, which will harm the glue
joint. If a dark burn does occur, stop working on
the ring and remount the plug to remove the
bur. In general, l try to avoid using this
technique because burished wood does not
glue reliably.
When doing the final ftting of a plug such as
this, I like to use a small diamond-pointed
scraper, which I grind from 3/S-inch square
turg steel. Ua shear scraping position, it
allows me to take very small cuts and gives me
the necessary control (photo 7-39). Remember:
finalize the plug shape first, and then shape
the hole in the ring by very gradually
enlarging the bottom of the hole until the top
seam comes together.
Floating Discs
Uthe past, about the only glueline stability
problems that I have experienced have been in
the base ring area. Over time, a few large sized
plugs have split or separated fom their
surrounding ring. Even though the plugs were
relatively small (under 4 inches), the wood grain
was restricted from movement and the
resulting stress caused an unsightly defect.
While this has not been a common occurrence,
just once is one too many. I happened to
mention this problem to a good friend, Bruce
Friederich, who is a very accomplished 1er in
Auburn, Califoria. Most of Bruce's work is with
solid wood, not segmented, but he got me
thinking about using a different technique. He
casually suggested, "Why don't you just make
the base float?" Cabinckers have addressed
predictable wood movement problems for
centuries by using frameand-panel
construction, so why not use the same
principles in the base of a vessel? This has since
become my preferred method of constructing
large vessel bases.
This type of base plug is actually easier to build
than a precisely fitted plug The procedure is as
follows:
First. build the base ring. Lathe-mount It and
60 THE ART OF SEGMENTED WOODTURNING
7-40 large bases require a floating plug.
7-4'
Glue only where the arrows point.
tum it round both inside and outside.
Tum a recessed groove about 1/4 inch deep
and 1/4 inch wide into the inside of the ring.
Photo 7-40 shows an ebony ring with this type
of groove.
Tua round. 1/4-inch thick disc of wood to fit
inside this groove (photo 7-40). Use quarter
sawn material if possible. The fit should not be
tight. about a Ilt6-inch gap should exist
between the disc and the side of the groove. The
disc shown in photo 7-40 will be flipped upside
down when it is inserted into the groove. Using
only two small dabs of glue at each end-grain
side of the plug, secure the disc into the center
of the groove, making sure an equal gap exist
on each side (photo 7-41). In the two photos I
have indicated the gluing locations. Prior to
gluing the disc into place, I also apply a coat of
paste wax to its edge. This seals the wood and
helps keep glue from sticking to the edge dwing
the next glue job.
Now, tum the base ring and the outside 1/2
inch of the disc fIat and smooth, and tum the
center portion of the disc slightly concave so
that it does not interfere with the fit of the next
ring. Be gentle, because the disc is only held in
place by two small dabs of glue.
To ensure that the disc remains fee to move,
before gluing on the next ring apply a little
paste wax around its outside surface. If you
look
closely at the disc in photo 7-41. you can see
the
discoloration of wax around the circumfer
ence .
Use a Q-tip to apply the wax and be
careful
to
aVOid smearing any wax onto the outside nng.
Prepare the next ring. Its inside
diamet
er
should be smaller than the diameter of the disc,
thereby hidmg the open seam. In this example.
the next ring is shown in the background of
photo 7-4
The next step is to glue the adjoining ring
into place. Use a minimum amount of glue and
spread it very thin near the expansion gap.
The goal is to prevent glue squeeze-out from
entering the gap and restricting the disc's
movement. The only glue that should contact
the disc should be the two dabs that hold it in
place.
When using this technique, the final
dimensions and shape of the vessel profile and
the position of the disc must be well planned.
During the final shaping of the vessel's
exterior, you certainly would not want to cut
though to the inside disc groove. This type of
plug also requires that the shape of the vessel's
bottom contain a depression, you cannot tum
the base flat.
This method allows the plug to shrink and
SJ,and it allows the vessel walls to move
independently -a simple but very effective
solution to a very annoying problem. You never
know what might result fom communication
with fellow woodtumers, thanks to Bruce
Friederich for the suggestion.
Improving Eficiency
like many activities, experience usually results
in easier and qUicker techniques. I can look back
at some of my earliest work and remember how
painfully slo the forms took shape. Getting
faster requires analyzing the steps that W
be required and looking for opportunities to
eliminate or consolidate some of them. A few
examples:
Dimension all your boards at the same time.
When possible, stack strips of wood and gang
cut segments.
Glue-up thick rings that can be split into
multiple rings.
Work on more than one section at a time (the
TECHNIQUES. TRICKS, AD SHOR-CUS 61
top, the middle, the base, the feature ring).
I you are making a time-consuming.
complicated element that requires lots of
machinery set-up time, consider making more
than one. Look for opportunities to incorporate
the element into other turings.
Make multiples of the same turing, thereby
increasing efficiency (see Chapter 12, Production
Turning).
Look beyond your current project. Think abut
how your current activities might affect your
future turgs.
62 THE ART OF SEGMENTED WOODTRNING
8-01 TPPLODBOW (10 inches dia.) -si mpl e forms stili require careful pl anni ng.
8-02 FOR THE Boy XDG (26 Inches tal l)-
extensive di sc sandmg created the feature ri ngs.
8-03 MYRn MOONS (35 inches tall)
-let
your
i magination go wild.
THE DEIGN PROCESS
8.
The Design
Process
OK, you have all your tools sharp. you have
cleaned your shop, and you are ready to create
your frst segmented turning. What do you
make? The choices are endless. that's one of the
great joys of segmented work You can create
just about any shape that you desire using most
any type of wod that you have available. Of
course, as shapes and design components
become more complex, then the design process
becomes more complex. The simplest designs
are bricklaid open bowl shapes, but even with
simple shapes and designs, accurate assembly is
important. A look at photo 801 shows a lidded
open bowl. relatively easy to turn but still
requiring careful constuction and alignment of
the segment ring components.
Closed forms (vessels with small top openings)
are a popular choice. This shape is actually
much easier to turn as a segmented piece
compared to turning the shape from a solid
block of wood. The reason: you can create the
shape as two separate open turnings and then
join the sections together. There is no struggling
with turng blind or with the diffculty of
removing wood though a small opening. FOR T
Boy KING, an example of a closed turing that
was constructed in sections, is shown in photo
802.
While the vessel is certainly the most common
type of form created on the lathe. you are not
limited to vessel forms. Quite the contrar.
sculpturaJ forms ofen stand out. and receive
attention and interest from the viewing public.
Youhave endless choices of shape and color and
you have the ability to combine those shapes to
create anything that you desire. An example of
a sculptural form, MYRL MOONS, is shown in
photo 8-03. Be creative and feel free to take a
chance. let your imagination go wild.
Design Tips
Regardless of what type of form you choose to
create. the shape is what matters most.
Interesting exotic wods and patterns wlnot
erase the pitalls of an unpleasing shape. I
recommend that you first focus on the shape
aboe all else. Pretend that you are making a
solid. blackcolored form. Ask yourself, "Can the
shape be improed?". Look at the lines that you
design and decide it they flow smoothly. There
are many different opinions regarding good
form. but I think most respected turers would
agree that the profile lines at your shape should
usually change direction as gradually as
possible. Think of a piece of string. freely
suspended between two points, compared to the
same piece of string with a weight attached
somewhere along its length. Gravity torms a
much more pleasing shape with the uneighted
string. Hyou vary the separation and/or
elevation ofthe two ends of the string, the shape
changes. but it generally remains pleaSing. Most
pleasing forms can be created by combining
sections of different gravity formed curves. To
emphasize thiS. let me say that I am a big fan of
the Spanish architect Antonio Gaudi. who
designed many fanciful buildings in Barcelona
i the 1890S and early 19005. One of his design
techniques was to create an upside.down model
of his buildings by hanging stringmounted
elements above a large mirror. By viewing the
mirror image. he could see his creation right side
up. An example of his geniUS is on display at the
Sagrada Familia cathedraJ in Barcelona. If you
appreciate whimsical. innovati ve. intncate
architectural design. then check out the work of
Antonio Gaudi at any public library or bokstore.
His work might even inspire you to go bldly
where no other turner has gone before.
When creating vessel shapes. try to aVOid a
64 THE AR Of SEGMENTED WOOOTRNING
clunky look. Shive for a vessel base that is no
larger than needed for stability. A small base
will often give life to your shape. One trick,
which I picked up from the late Gene Pozsesi, is
to create a small air gap at the edge of your
vessel's base, which will give the illusion that
the vessel is floating. In other words, hide the
contact point between the vessel's base and the
surface upon which it sits. Another tip that I
picked up a long time ago fom the well-known
turner David Ellsworth: look at your design
upside down, it should remain just as pleasing.
Vessel wall thickness has an effect and projects
a feeling. There is no formula for proper wall
thickness, although in general, the smaller the
turning, the thinner the wall. His also
important that the wall thickness be consistent
fom top to bottom. The wod is going to move;
inconsistent Wthicknesses could cause
uneven movement and result in unwanted
stress. If you think of wood movement as a
percentage of its dimension, then it is logical
that the smaller the dimension, the smaller the
total movement. I generally strive for about 1/8
inch wall thickness on small tungs, up to
about 3/8 inch on large pieces.
Good form is perhaps the most diffcult aspect
of wood turning to achieve and understand. I
know that I continue to learn with every piece
that I make. A Supreme Court justice once said
he had difficulty describing porography, but
he knew it when he saw it. Good design is also
difficult to describe, but quickly and easily
recognized. I suggest studying the works of the
masters and not just of other woodturners. Visit
a museums, look at works done in ceramic and
glass. Study the forms that natwe creates,
because naturally occurring shapes are among
the very best. Do not be fooled into thinking
that a spectacular piece of wood or a
particularly complex segmented design will
overcome an unpleasing shape.
Non-segmented woodturners study their next
piece of wod before deciding what shape to
turn and they make their decisions on the
characteristics of that piece of wood. Segmented
turers can go though a similar process. Instead
of studying a block of wood, I often study my
wood inventory and then design based on what
I have available. Uyou do not own much of an
inventory, then you need to analyze what is
available at your supplier. You do not want to
near the completion of a turning only to
discover that you do not have enough wood of a
particular type. While you might be able to
acquire another board. its grain and coloring
might not match the rest of yow turning and
you will regret your lack of planning. Hreal1y
pays to plan ahead.
One easily made mistake is using your boards
as you need them during the constructon of a
turning. Uyour boards were not sequentially
cut from the same log, then their coloring
probably will differ. If you use of one board
before cutting into the next board, then you |
the risk of unwanted color breaks in your
turning. For a more consistent look, it is better
to use a little of each board as you build your
rings, then the color differences will be random
and the effect will be a blending ofthe colors.
It is very common to incorporate several
different types or colors of wood into a
segmented ting. That is another reason this
style of turing is so much fun. The colors catch
the eye and dearly separate this style from solid
wood forms. His very easy to overdo it, and I
have certainly been guilty of the over-use of
multiple wood types. Using many different
types of wood is not always a bad thing. one
just has to be aware that it can be overdone
and
that too many colors can result in a busy
turning. Sometimes, less is better.
As you design your turnings, you must make
many other decisions. One choice is the
number
of segments to include in each ring of the
construction. This is a key decision that can
have a dramatic effect on the appearance of
your turning. There are no minimums or
maximums. it is up to you. The advantage of
fewer segments is, you will not have to cut
and
glue as many individual pieces of wod.
However, if you choose too few. your turning
may appear clunky or heavy and it may not
project the desired effect to the viewer. As the
number of segments per ring becomes fewer
the sharpness of the ring corner joints
increases, and you will have to contend with
end-grain chip-out during the turning process.
Conversely, the more segments per ring, the
more gentle the corers and as a result, the
shaping on the lathe will go more smoothly.
As a general practice, I avoid constructing rings
with fewer than 10 segments, and 12 to 24
segments per ring is more the norm. On large
diameters (20 plus inches), 24 to 60 segments
should be considered. Even on large turnings,
segments that are longer than 5 inches usually
appear too big. Try to visualize the appearance
before deciding upon the number of segments.
Before deciding, calculate the segment lengths
for both the largest and the smallest rings.
I like to maintain the same number of segments
per ring throughout the turing. It is cerainly
possible to vary the number of segments per
ring, however, 1would recommend that you do
it in a way that is not immediately apparent to
the viewer.
Regarding staved construction, you do not have
the same segment corner issues while turning
the prOfile, but you need to select a number of
staves that creates a nice proportion to the
overall size of your design. A column created
fom only three or four staves is pOSSible, but
you might as well turn a solid block of wood. I
like to construct staved vessel forms with a
minimum of eight staves. A big concern with
stave construction is, "How do you add any
layers above or below the staved portion of the
form?" The wood grain in the staves is vertical:
"Do we have to maintain the same grain
orientation in the other adjacent components?"
In general the answer is yes, although Ihave to
admit that I did not always do this in some of
my early turnings and many of them have
survived for many years, but a few have not.
The success or failure of perpendicular-grain
glue lines mostly depends upon the length of
the joint. Short intersections of opposing grain
direction are much more likely to succeed.
However, when possible, avoid these types of
JOints. In Chapter 1L,1 describe, a step-by-step
THE DESIGN PROCESS
8-04 Des1gns can be created by brick-laying
different colored woods.
construction of a staved vessel and I offer a
solution to the grain orientation challenge.
Feature Rings
It is common for segmented vessels to contain a
feature ring within their design -a decoration
that adds interest for the viewer. The number of
design elements should correspond to the
number of segments that comprise the other
rings in the turning. For example, if I decided to
construct a turning with eight Indian blanket
designs positioned around it, then I would
normally build the other rings with eight,
sixteen, or twenty-four segments. This would
allow me to align all the rings in a balanced
configuration. I do not want to create a
configuration that displays dramatic differences
in the alignment of the verical glue joints
between layers. which would occur if all the
rings were not multiples of each other. In most
designs, I think a lack of symmetry between the
vertical glue lines creates an unpleasmg effect,
although there are exceptions such as when
using randomness as a design element.
What are the choices regarding feature nng
designs? Here again, there are endless choices
and you have the opportUnity to create
something that no one else has ever done. On
the next few pages I offer a few posslbultIes
66
THE ATOf SEGMENTED WODTURNING
8-05 Large diamonds require the use of a disc
sander.
8-06 A router created the opening for thi s small
diamond shape.
8 07 ZIGZAG Box (3 Inches di ameter) -just one
of many feature ng de$igns.
southwest Indian pottery and basketry have
had a big influence on segmented woodturng
designs. Their geometric designs lend
themselve
s very nicely to the assembly
proce
ss
of segmented work, which is why so many
segmented
turners have focused on these
types
of designs. southwest designs also seem to D
quite popular with the buying public. but you
are certainly not limited to them. Quite the
contrary. because there is so much Indian-style
work being done, I encourage you to explore
other possibilities.
Brick-laid designs appear relatively Simple, but
accurate construction and alignment of the
rings is essential and in a large design,
rotational alignment is very difficult to
maintain. Complete vessels can be constructed
in this manner, so the intricacy of your designs
is completely your decision. This type of
construction is very stable because Bthe wood
grain is conSistently horizontal. Successful
alignment of the segments does require round
rings. Rings that are just slightly oval will cause
unequal segment lengths and the result Wbe
inconsistency. In Chapter 12, a production DO1
six similar brick-laid vessels is presented. In
addition to brick laying with equal-sized
segments, unequal segment lengths also can be
used. To avoid confusion during construction.
you definitely need a detailed sketch fom
which to work.
Diamonds created with a disc sander are
another choice and they are actually quite
simple to make. Photo 8-05 shows a rather large
diamond design (about 2-1/2 inches wide). If U
look very closely, the order of assembly can D
seen. Photo 8-06 shows a router-created
diamond deSign; this is done very differently,
requiring precise woodworking skills. The
diamond in photo 8-06 is actually
less
than
1/2
inch wide. See Chapter 11 for detailed
instruction regarding both of these techniques.
Zigzag patterns are time-consuming. but they
impress vieWers. The sample in
photo
8-01
u8
small, 3 inch lidded box. The lid
has
been
removed to show the zigzag on the
insid
e
Zigzags
can be built using multiple
laye
rs. a
nd
8-08 Indian blanket designs are easy once you
know the trick.
they can be flat or steep depending upon the
miter angles. Detailed. step-by-step instruction
is also presented in Chapter II.
There are countless variations of Indian blanket
designs (photo 8-08). These designs are easily
achieved by executing numerous woodworking
steps as described in Chapter 9.
HI have a signature element. it is probably
portholes or segments with windows. They
provide an opportunity for incorporating
endless designs into your turnings. The sample
shown in photo 8-09 is quite small and displays
small. I-inch cabochons of macassar ebony. I
have also used stone cabochons in some of my
porthole deSigns. Larger windows can be used
to display marquetry designs; the technique is
covered in detail in Chapter 15. The curved
pieces of maple in the photo result from
exposing a layer within the segments. another
technique covered in Chapter 13-
Feature rings are usually the focus of a turning.
They grab the viewer's attention and they
provide many opportunities for expressing your
creativity. These few examples barely scratch
the surface of possible designs. Uyou need
more inspiration or additional ideas. investigate
the public domain designs of the past. Look at
ancient Egyptian art. Indian art. and computer
clip art; look at product packaging. and look at
the patters that Mother Nature creates. Design
TE DESIGN PROESS
8-09 Porthole designs require careful planning.
ideas surround our lives, though perhaps most
satisfing are the original designs that we as
artists invent.
Creating a Blueprint
In all but the Simplest of designs. you must have
a plan. Ucan be a hand-drawn sketch or a
detailed computer drawing. but YOU HVE TO
HVE A PL. Before creating that plan. let us
reiew a few basic geometric principles:
ea circle contains 360 degrees.
ethe circumference of a circle is equal to its
diameter times pi.
epi is equal to 3.1416.
ethe diameter of a circle is equal to its
circumference divided by pi.
Also.
ecircumference divided by number of segments
equals segment length.
esegment length times number of segments
equals circumference which. when diVlded by
pi. equals diameter.
This is most of the math that you need to know.
with the exception of compound mlters, whlch I
discuss in Chapter 10. With a small calculator
68
THE ART Of SEGMENTED WOOOTURNING
,
8-10 The terms can become confusing.
you can determine everything you need by
doing simple multiplication and division. lust to
avoid any possible confusion regarding
segments, let me define and illustrate a few
terms using photo 8-10:
esegment length: length of the long side (the
outside corner to corner).
esegment width: width of board from which
segments aIe cut.
-segment height: thickness of board fom
which segments are cut.
_segment angle: the angle of cut. usually 3600
divided by the number of segments divided by 2
(2 angles per segment).
Here are a few additional bits of information to
consider as you go about the design process. If
you design a turning with rings 3/4 inch tall
and you use boards that are 3/4 inch thick, then
your total completed height wiH be less than
designed. While stacking the rings, you wiIl lose
some height due to machining and sanding the
horizontal surfaces between rings. I normally
draw my designs to account for a 1116-inch loss
Dheight on each ring (less if the turning 1
quite small). Coincidently, the thickness
dimension of many commercially available
hardwoods is 13/16 inch. In effect, the waste
factor has been figured in for you. This may
seem inSignificant, but with a tall turning, it
can add up and distort the designed profile.
Ucalculating segment lengths, when you
divide
circumference by number of segment
s,
your given measurement is an arc length (a
curved measurement) of the circumference
curve. It's not quite the same thing as the length
of a segment, which should be thought of as a
chord length (a straight measurement). This
difference can usually be ignored. Urarely
makes a difference because a little extra
thickness is generally added to the required
diameters and to the width of the segments
during the design process.If you consistently
ignore this difference throughout the entire
shape, everything should hne up just fine. I only
mention it for the math whiz who might
question my calculation method.
Using Graph Paper
Computers are powerful tools, but they are not
necessary for creating quality designs. A pencil
with a good eraser, some graph paper, and a
ruler are all you need. First, decide on the
overall size of your turning. it is bigger than
your graph paper sheet, then you can tape
multiple sheets together or you can work at a
smaller scale. I personally like working with
actual-size drawings. Since you do not have the
ability to stretch or shrink the pencil drawig of
your shape, as you can in a computer program,
you might want to first calculate the total
height of all your components (rings, feature
ring, etc). As an example, imagine a vessel with
these components:
-a bottom ring 1/2 inch high, .50;
-three main body rings each 3/4 inch high,
2.25;
- a feature ring 2 inch high, 2.00;
_two more body rings at 3/4 inch high.
1.50;
_a top ring that is 1/4 inch high, .25.
-Vessel height ' 6.50 inches
By calculating the height first, you have avoided
having to alter your component heights after
you have determined the overall shape. If you
prefer a vessel taller or shorter. then it Ma
matter of adding or deleting rings. or of
changing ring heights. Next, you need to
decide
on the approximate diameter and shape for
THE DESIGN POCESS
69
~ @_ I> ~

l..
=
* y : )
!
7. 7
8-11 Usi ng graph paper makes drawing a design much easier
your vessel. Mark on your graph paper the base
and top elevations. Also, make a few reference
marks so you know the location and diameter of
your feature ring. Uis likely to be the largest
diameter ring, usually positioned at a point of
profle transition. Then draw a few curved lines
representing the vessel profle (photo 8-11). For
now, focus on just one side of the vessel
centerline. Play around with the profile until
you aIe happy with the shape. Transfer the
vessel profile to the other side by making a few
corresponding pencil marks from the centerline,
or by folding the paper down the centerline.
Study the shape once again and repeat the
process as many times as necessary. Now,
thicken your outline to represent the wall
thickness. Next, overlay onto the vessel profile a
senes of rectangles that represent the side
views of the individual rings. To prOvide a little
room for error, draw these rectangles slightly
wider than the diameter of the vessel.
Measuring the length of these rectangles W
give you the individual diameters of each ring.
Next, overlay another series of smaller
rectangles starting on the outside and going to
the inside, to encompass the vessel wall
thickness. Again, give yourself a little extra
wood for errors. These smaller rectangles
represent the individual segment Widths in
each ring or layer, that is. the width of the board
from which you cut the segments. Now, you
need is the individual length of the segments in
each ring. To determme the segment lengths.
multiply the ring diameters by pi (3-1416) and
then divide by the number of segments per
ring. Photo 8-11 shows a Simple draWing using
7
0
THE AR OF SEGMENTED WOODTlRNING
the previously listed ring heights. Though
nothing fancy. it provides all the information
needed.
If you are familiar with the computer
spreadsheet program Excel and you are
designing a large turning with many layers.
then it is very simple to create a cutting list. The
program formulas can do the math for you.
otherwise. just grab a calculator and manually
do the math for each ring.
Using a CD Program
Instead of using graph paper. the same design
can be quickly developed using a computer
drawing program. Using a computer doesn't
mean your vessel design need be any different.
it's just faster and more accurate. Using a
computer also allows easily changing your
design unti you are completely satisfied. Years
ago when I first became interested in
segmented work, I Wfortunate to already
possess some Ldrawing experience, so it was
just natural that I started creating my
blueprints on a computer. I still use a very old
handicapped without the advantages
of
computeraided design.
Specialized Computer Programs
Within the past few years several computer
savvy woodturners have created commercially
available programs that make the chore of
designing a segmented turning very easy.
H
would be unfair of me to endorse one program
over another. I really have not had the necessary
experience with them to judge their differece
s
Quite frankly. because of myoId Mac computer
and Lprogram, Ido not need additional
computer help. l will say that the limited
experience that I have had with these new
programs is very positive. and I definitely think
they are great tools for woodturners. The
programs are well worth the investment. They
have their limitations -they are not CAD
progams. meaning they will not allow you to
draw out complex design elements, but Wa
minimum of learning time, one can quickly
produce perfect lines and useful cutting lists.
MaCintosh Classic computer with an ancient
A Sample Drawig
program called Clans L.People are amazed
The most effective way to explain the complete
when they see this old relic still operating, but it
design and construction process is to document
works great and has athe power and features
the creation of a vessel This is probably the
that I need; 1 wl mourn its eventual demise. On
next best thing to being in the shop with me. A
a computer, you can draw a profile and
friend, John, who provided me with some
manipulate the shape untl you are satisfied -
preferences. commissioned the vessel that I V
you do not have to keep erasing pendl lines to
make. He desired a 1O.inch to 12.inch diameter
start over. In addition, the computer is very
Indian pot shape, he liked the look of what I C
effective at smoothing your lines and
an Indian Blanket feature ring, and he really
transforming them into pleasing curves. The
liked bubinga as the main wood. John also
benefits of a computer-generated drawing are
wanted something that would be stable on a
two fold: it allows easy development of the
coffee table in case it was bumped into from
tuing's shape, and it provides you with all the
time to time. He liked one of my vessels that
dimensions you need for constructing and
was on display at a local g
a
llery, but he
wanted
assembling your piece.
it wider and shorter. What you Wsee is
what
When using a computer program. the steps are
he gets. Since John had a hand in the
design
not much different from paper and pencil.
proess and since he is paying for it. I will refr
Shape comes first, followed by an overlay of
to this piece as ]OH,s TRNIN. Based on my
rectangles in order to calculate dimensions. The
conversation with John, I have drawn a rough
computer is a major time saver and because it is
sketch of the shape (pho 8-12).
so flexible, it enables you to focus on the form
At this point I )mthe components,
but I
have
Cyour turing. Saved fles can be recalled,
not dedded the exact dimensions. I
begm b
modified, and then reused. I wou
l
d be seve
.
,
_
e
_
I
Y ___
th
_
i
g a
b
out this p
cl design
and
s
a
y.
THE DESIGN PROCESS
8-12 Computer drawing programs are very handy. but not necessary.
"Where do I want to start?" The most complex
component is the feature ring wUthe Indian
blanket-style geometric design. The exact final
dimensions of the vessel are not critcal, so why
not construct the feature ring first? In the next
chapter I will build the feature ring. finalize the
vessel design and dimensions. and build this
tung.
EBONY
HOllY
EBONY
71
72
TH AR OF bFCN5NTD VCODTUkN!NC
9-00 JOHN'S TURNING (12-112 inches diameter)-a lot of work, but worth the effort.
9 01 ThiS IS the feature nng for JOH'S TuING.
9
John
'
s Turning
The feature ring in JOHN'S TuRNING is the focal
point of the vesseL Indian blanket designs are
very popular with the buying public and they
are not nearly as difficult to construct as most
people would guess. They do, however, require
many accurately executed steps. While the
feature ring may contain many individual
pieces afwood, it's not necessary to individually
cut every piece of wood. Follow along and you
will see what mean. Photo 901 shows the
completed feature ring that will be incorporated
into /oHlIs TURNING,
Building an Indian Blanket
Feature Ring
From an instructional point of view, it might
seem strange to start out with the most
complex step in this vessel's construction, but
once you understand this process, the
remainder of the project will seem simple. I
chose bloodwood and holly as the two
contrasting colors in this design element. This
combination creates a dramatic contrast.
However. it is a difficult combination for several
reasons. The woods are of diferent densities,
which presents a minor turning difficulty. Holly
is so white in color that glue lines are difficult
to hide. Bloodwood can stain the whiteness of
the holly. and bloodwood is an oily, difficult
wood to sand that gums up sandpaper very
quickly. Having said all that. I still like the
combination for its look, so it was worth the
trouble. Here is my step-by-step procedure for
constructing an Indian blanket feature ring:
First. I planed the bloodwood and holly boards
to the same thickness of .9 inch. Using a table
saw. I ripped strips 1.3 inch wide from each type
of wood and ripped thin strips about 1/8 inch
thick from the same boards. Ripping thin strips
JOHN'S TURNING
9-02 Accurate dimensioning of boards is
essential.
9-03 This is the arrangement of components
prior to gluing.
7
3
is discussed in Chapter 7; a few of these stips
can be seen in photo 9-02. One advantage of
segmented work is, we do not require perfect
boards. Defects such as the ones shown in the
holly strips are easily worked around. discarded.
and not a problem.
Using a miter saw set at 45

, I cut polygon
shaped pieces from the 1.3 inch wide strips. The
point-to-point length of the holly pieces was not
critical, though I did need about an inch of
length on the short side. I cut the bloodwood
pieces with their short sides less than a quarter
inch long. I have arranged a few of these
components in photo 9-03 to show you how I
intended to glue them together.
The thin lIS-inch wide strips were passed
through a drum sander. erasing all saw blade
marks and ensuring consistent thickness. I
banded them together with masking tape. then
gang cut them (at 45") to lengths equal to the
d1agonal Side of the polygons. I needed one
14
THE A OF SEGMENTED WOODTRNING
9-04 Anchor the f|1st piece in the glui ng jig_
9-05 Prior to pplying glue. a dr fit |salways
good idea.
9-06 Accurate clamping |scr|ticaIto success
piece of each type of wood for each side of the
bloodwood pieces; therefore, I cut 56 small
strips
of each color. plus a few extra in case I
discovered a defective piece while gluing.
A few
of these thin. short strips are also shown in
photo 9-03.
In this feature ring 1 decided to place 10 design
elements around the vessel, as shown in photo
9-01. Each piece of bloodwood will produce one
half of a completed element, so I needed at least
20 pieces of bloodwood. This will become much
clearer in the next few steps. I always make a few
extra elements. s I wlbe able to reject a few
after completion. With this in mind, I cut 28
pieces of bloodwood and 28 pieces of holly. Then I
cut four of the holly pieces in half so that half
Qecescould be QLsiti1ned at eachend of the
strips that would be glued together later. This
was enough pieces for Qcomplete designs.
Since I had 28 pieces of bloodwood. I glued
together tour strips consisting ot seven
bloodwood pieces each. The gluing of these
strips was perhaps the trickiest step in the
construction. I used a simple gluing tray (photo
9-04) consisting ot two strips of I-inch MF
glued together to form a straight right-angle
channel. I covered the jig with blue masking
tape to allow removal of the strips after the glue
had set for 30 to 40 minutes. Titebond was my
glue choice. It was important that the bttom
shelf of the gluing jig be slightly narrower than
the polygons, so it would not interfere with the
damps. This meant the tray bottom was slightly
less than 1.3 inch. Photo 9-04 shows the gluing
tray is slightly narrower than the holly strips.
To glue all these components together I used a
(-clamp to firmly anchor a half-piece of the
holly at the left end of the jig. with its long
side
against the backstop (photo 9-04). Then I
gathered two thin bloodwood strips. two thin
holly strips, one bloodwood polygon, and one
holly polygon. Before applying glue I made
sure
that the surfaces were dean and free of any
debris. and did a dry fit as shown in photo 9-oS
I then clamped the piece of holly with a spring
clamp and positioned it with a slight gap
between the short side of the bloodwood
piece
and the back of the tray
ThIS gap would be closed
dunng the actual gluing
step that follows.
I placed these pieces on
the gluing tray with glue
spread on all mating
surfaces. With the spring
clamp holding the holly
piece in place I added a
quick clamp to force the
bloodwood piece against
the back, which in turn
forced the holly piece to
slide slightly to the right
while being resisted by
the spring clamp. Done
properly, there should be
JOHN

5UkNtN
7S
enough pressure on q
d I d 9 07 Here are the components of one strip all glued an c ampe
the glue joints to ensure
-
tight seams. If I have
forced the piece of holly too far away, thereby
causing a loose glue joint, I release the
bloodwood clamp and force the holly piece back
to the left, then attempt it again. Udoes require
a little practice. Accurately gluing these pieces is
critical to the creation of the completed design.
The location of the center piece of bloodwood is
especially important. It must be conSistently
and firmly seated against the back fence of the
jig. Hthere is any gap, the components that are
later cut will not be consistent in length.
This process was repeated, going fom left to
right, until all the pieces had been glued into
four separate strips. One of the glued strips is
shown inphoto 9-07.
These four strps were quite fragile, so I
cautiously used a chisel to pry them from the
gluing tray. Using wax paper instead of masking
tape on the gluing jig would have allowed easier
removal of the strips, but it would not have
provided enough resistance during the gluing
process, the holly piece might have slid too easily
away and the joints would not have been tight
Before the next cutting step. these strips were
cleaned up. You can do it with a little handwork
followed by light passes through a planer using a
9-08 To rip-cut thi n strips, attach the board to a
wider push board.
backing board. Once prepared, the strips could b
smoothly rip-cut on the table saw. Photo 9-08
shows ripping one of the cleaned-up stips
attached to a wider push board of maple.
These glued strips were cut mto four lIS-Inch
thick strips using the techmque described in
Chapter 7. The exact thIckness is not cnhcal,
but they must all be the same and It is
extremely important to keep each set together
Photo 9-09 shows one set of four stnps In the
16 THE ART OF SEGMENTED WOODTURNLNG
g-Cg I this example, !ourstrips were cut from
each assembly,
g-1O To create the desired design, each strip i s
flipped over.
g-11 A_Iu|_tray is used to assemble the thin
S1r|p5
order they were cut. The gaps
represen
t
the
saw
kerfs.
Flipping these strips over and reassembling
them creates one-half of the design. The result p
flipping
the strips u shown in photo
9-10.
The
flip seems to confuse some people: each
individual strip is flipped, not the stack
as a
W\
Before gluing the strips together I passed
them
through the drum sander to erase all saw blade
marks and to ensure tight glue lines. I was
especially concerned with the appearance of the
holly-to-holly glue lines. It was important to
drum-sand off equal amounts from each of the
strips, to accurately maintain the design a
consistent thickness.
Centering all the strip elements with respect
to each other would be difficult to do one at a
time, but because of the way they were created
it was quite easy. By accurately aligning the
ends of each strip with one another, all the
elements within the strip automatically lined
up with each other. Even if one of the
components was misaligned during the first
gluing operation, everything must still line up
now. You might have one unusable
component, but the rest will be unaffected.
Photo 9-11 shows the ends of one grouping,
glued together and clamped using the same
gluing tray as the first glue job. I did have to
rip the gluing tray down to accommodate the
narrower assembly. In addition to the four thin
strips I have added a thicker strip of hoUy.
which will become the top and bottom
border
around the blanket design. Glue obscures the
view, but the four thin strips near the end of
the gluing tray are perfectly aligned with eac
other. It the ends of these thin strips are not
accurately aligned, then every design element
will also be out of alignment.
Using a combination of a hand
plane,
the
0
sander,
and the table saw, I cleaned
up
these
four new strips.
Using a miter saw at go", I separated the strips
into
individual pieces. These cuts
did
not
hav
e
to be exact. I just eyeballed the
center
bte
n
the deSigns. Each piece re resents one half of 8
design element Photo 9-12 shows the miter saw
separatmg the components.
These half-pieces were carefully paired before
being glued together. If two halves did not quite
match perfectly, I put one aside and tried
another. By trial and error and a little
adjustment using the disc sander, I managed to
match the half-pieces; a few unglued pairs
are shown Dphoto 9-13. Inaccurate alignment
of the two halves is easily noticed, so I took
extreme care while gluing.
Even though I only needed 10 completed
elements, I glued all the halves together because
a few were not likely to pass muster and at this
point it is far too late to go back and make
more. Photo 9-14 shows two halves clamped
together to form one complete design.
I cleaned up the completed deSigns on the disc
sander and inspected them once more. By
checking both sides, I was able to select the 10
best. These elements now needed to be miter
cut in order to form the 3600 ring: 10 pieces, 20
angles, therefore, 18 per end.
Centering the bloodwood between the miter
cuts is important. By placing one side of a
square even with the end of the longest
bloodwood piece, I made a pencil mark on each
segment. These marks were then aligned with a
mark on my saw bed and one end was miter-cut
off all the segments (photo 9-15). Then, using a
stop block carefully positioned to center the
bloodwood shapes, I cut the other ends. I
intentionally cut these segments a little long,
which allowed me to check the fit and to check
the centering of the bloodwood within the
holly. I readjusted my miter saw stop block to
remove a tiny bit of length from the segments
then re-cut whichever end appeared longer,
thereby improving the centering of the
bloodwood. I repeated this process once again D
order to get every component as centered as
possible, while maintaining equal segment
lengths.
The width of the segments (fom outside to
inside) left little room for error, the ring had to
be perfectly round. In cases such as this, a
JOHN'S TURNING 1
9-'2 A miter saw is used to cut individual half
components.
9-'3 The half-components must match
perfectly.
9-'4 Precise allgnment during gIu|ng IS
eSsentialM
78
TE ART OF $EGMENfEO WOOOlRNING
9-'5 The miter saw is USed to create segments
from the glued assemblies.
g-1b Splines are cut with the same horizontal
grain orientation.
9-
'7
Good planng ensures a Successful glue
Job.
sanding jig is necessary. Sanding jig
construction and use is discussed in
Chapter
7. 1
lightly sanded the ends, assembled the ring
using a hose clamp, and checked the fit
agains
t
a bright light. The joints have to be pedect to
the naked eye when dry-clamped and
held up
to a bright light.
My design called for a vertical piece of ebony
positioned between each element.
Perhaps
it
would have been easier to cut short pieces of
ebony from a long strip, but that would have
positioned the ebony grain perpendicular
to the
grain of the blanket segments. Instead, I cut
them using the miter saw at gog, orienting the
grain horizontally to match the holly and
bloodwood (photo 9-16). Notice the shape of the
segment hold-down device. Another example of
maintaining consistent grain orientation
shown in photo 5-03. In general, try to avoid
any glue lines with perpendicular grain longer
than about 3/4 inch. I decided that 1/4-inch Wdt
pieces of ebony would look right. These parallel
sided pieces of ebony had no effect on the fit of
the miter joints. Instead of using the disc
sander, I just made a few light swirling passes
by hand on a piece of So-grit paper adhered to <
flat piece of MF before the gluing step.
Gluing the fnal assembly was the easiest step. I
decided to glue the entire ring at one time. I
wanted to accomplish this quickly so I gathered
up everything I needed before starting (photo
9-17). During gluing, the key things to pay
attention to are; all surfaces clean, all surfaces
receive glue, gcorners meet evenly, and
segments sit flush with one another. A dean
gluing environment is important, so I usually Ut
my shop vac to blow off shavings and dust
from
my arms and clothing, which minimizes
the
risk
of debris falling into the glue joints. I applied
glue quickly and liberally to both ends of each
segment and to both sides of each piece of ebony
His critical that qthe outside comers
line
up
just as they did during the dry-fit
inspecti
on.
It is
also
important that the tops and
bottom
s of
a
the parts remain flush with one another
. A
few
light taps with a hm er before
fnal
dam
p
tightening uusually sufficient. Iran
a |QM
around the top suface to make swe
every
thlflg
JOHN'S TURNING 79
4.11"
.90"+- 3.78"
-I
12 8.00"
12.00"
12.31"
12.63"
8
7
6
;"-12.M"
5

+ .

1
2.50"
1.50"
.1.50"
'*
12.75"
12.50"
11.00"
8.83"
9-18 Afer building the feature ring. final dimensions can be determined.
was flush. Wiping excess glue off the surface
with a small scrap of wood is another way to
double-check the evenness of the surface.
The technique for constructing this feature ring
(photo 9-01) can be easily altered to create bigger,
smaller, simpler, or more complex Indian blanket
designs. His easy way to make very complex
looking designs, containing hundreds of pieces of
wood that do not require individual handling.
This feature ring contains 590 individual pieces
of wood, but as you Lnow see, I did not
individually cut and glue 590 pieces.
Now, back to JOHN
'
S TRNING_ The finished
feature ring measured 12.8 inches in diameter.
Using that infonnation, I adjusted the original
turning design and created my final blueprint
(photo 9.18). During the construction
deScription that follows, I make many
references to the various numbered rings layers;
these numbers appear on the left side of the
drawing_ The original drawing was done actual
size but here has been reduced to fit the page
size. When creating such a drawing, first focus
on the shape_ Once you are satisfied, add the
rectangles that represent the rings and
segments. The little bit of extra rectangle length
that you provide here will give you latitude as
you construct and shape the vessel profile.
Using the measurements fom this blueprint, I
created a cutting list (next page) that would be
ver handy in the shop while constructing the
ring layers. I created this list on a computer
using the spreadsheet program Excel, but a
small calculator cerainly would have worked.
The list contains all the infonnation I needed to
construct the turg. The highlighted segment
length column is probably the most critical and
requires the most attention during the cutting,
If using a calculator, just multiply the ring
diameters by pi (316) and then divide by the
number of segments_ The figures in the board
length column are simply segment lengths
times number of segments. No allowance is
added for saw kerf loss because the angle of the
miter cut usually offsets it. Having this checklist
in the shop was invaluable.
80 THE POF SEGMENTED WOODTURNING
Cutting List for fohn
'
s Trning
Ring Ring Miter Number of Width of
Height of length of Wood
Boo,d
Numbr Diameter' Angle Segments Segments
Segments' Segments' Typ
"
ngth"
"
3
5 66 '8' 1 , 50 040

78 ebny
,8
2
633 ,8
'
1 , 50
0.1] , 99 holly
2L
3 883 9
'
20 2.50
0.80 , 39 bublnga
28
4 11 00 9
'
20 2.1] 080
l
73 bubinga
35
3
12 So 9
'
20 , 69 0.80 1 96 bubinga
39
6 1275 9
'
20 , 00 0.13 2 00 holly
40
7 12 7S 9
'
20 J L 0.19 2 00 ebony 40
LPKt 1288 ,8' 1 0.88 , 63
nt, nt'
nt,
9 12 63 9
'
ZL 1.29 0.19 , 98 ebony
40
'0 12 31 9
'
L 1.29 0.13 193 holly
1
9

" 00 9
'
20
5+
0.80

1.89

bubinga 38
J 8.00 9
'
xL 2.63 0.80 1.26 bubinga 25
'3 4
"
J

-
1 1.00 0.13 1.29 holly
'3
'
4 3-]8 ,8' '0 0.90 0-40 1.19 ebny 1
all figures represent Inches
" rough estimate b4sed on the length of segments times the number of segments
The featurering construction was the most
difficult part of this project. The rest was simply
a matter of building and stacking segment
rings. Because of the shape of the vessel,
internal turing would be difficult if the shape
were created one layer at a time starting at the
base. For this reason, and because working in
two directions is more effiCient, I built the top
and bottom separately and then joined the two
halves. started by constructing the bttom and
top rings of ebony. Using the cutting list, I cut
the ebony segments using the miter saw (the
table saw would have worked as well).
The top and bottom ebony rings were small and
absolute roundness was not critical. Therefore I
sanded the segment-ends freehand, without a
sanding jig. Gluing all the segments at once was
a little risky. I preferred either of the other two
methods, rub joint or half-ring. I glued these
rings using the rub-joint technique of joining
pairs. Photo Q-1Qshows the second step of
glumg together the top and bottom ebony rings.
The base ring needed a plug in the bottom of
the vessel. I chose a piece of holly and used the
technique described in Chapter 7 to create a
good fit between the plug and ebony ring. Plugs
require VQlittle clamping pressure to be held
in place -a spare faceplate or similar weight U
more than enough
Aer the plug-joint had cured, l tured the
sutface that would be the bottom of the vessel
flat and smooth. I then removed this base r
with its plug and permanently glued it
(Titebond) to another faceplate and waste block
that had been tured to the same diameter,
making it easy to center. Perhaps this all
sounds
a little complicated, but it is really quite simple
and quickly accomplished. Photo 9-20 shows
the top and bottom ebony rings ready for
attachment to additional vessel rings.
I flattened the ebony rings so they were ready
for the attachment of the adjoining holly rings.
Even though the two finished holly rings
needed to be only ,12jinch tall, I glued
them
U
from .37s-inch
thick material. This was
easier
to
handle and easy to tum down to the final
thickness after attachment to the ebony
nngs
I
temporarily attached the holly rings to
small
waste blocks that would act as handles, and
flattened them on one side using the disc
sander. Using a clamp from the ceiling, I then
glued them to the ebony rings.
The order of constructing the remaining rings
made little difference. However, to save wood u
is best to rip the widest segment material first.
Aer I cut the wide segments, was able to re
rip the unused portion of those strips to the
next widest dimension. If I had started with the
narrowest strips of wood first, then I would
have been unable to use leftovers for the wider
segments. I continued to build the bubinga
rings in the order of width, from widest to
narrowest. My bubinga strips were all cut from
the same wide board, thus ensuring similar
coloring. If several boards of varying color had
been used, then would have mixed up the
anangement of segments by randomly cutting
strips from different boards.
Before completing the construction of the
remaining bubinga rings, I prepared the base of
the turing for its first bubinga ring. The holly
ring was turned down to its designed thickness
(.125 inch) and I finalized the gluing surface
using a sanding block_
Ring #3 was attached with hot-melt onto a
faceplate-mounted centering board. Using
concentric pencil circles, the bubinga ring was
centered by eyeballing its outside edges. If you
distrust your eyeballing capabilities, then use a
caliper to center the ring as shown in photo 9-21.
When joining the holly ring to the first bubinga
ring, only the portion of the bubinga ring that
contacted the holly needed flattening. The
portion of bubinga that extended beyond the
holly was turned down below the glue surface.
thus allowing me to focus on the glue joint. In
photo 9-22 (next page), ring #3 has been
flattened and cleaned with canned air. The
surface can be cleaned with compressed air if
you have it, or with the exhaust port of a shop
vacuum. For small dusting jobs, canned air is
effective and convenient. You can sometimes
fnd three-packs, reasonably priced, at the large
warehouse-type stores.
JOHN
'
S TURNING
9-19 Construction starts with both top and
bottom ebony rings.
9-20 Top and bottom rings are secured to
faceplates@
9-21 Precise centering O1 rings !5 lmportant
82
THE ART OF SEGMENTED WOODTURNlNG
9-22 Clean surfaces help ensure good glue
joints.
LICC this flattening 1OccSs was completed I
glued the two rings together using the lathe as
my clamp and centering device. My tailstock
live-center just happens to fit very nicely into
the ttueaded recess of my faceplates. Photo 7-30
in Chapter 7 shows the procedure.
I added rings #4 and #5 just as ring #
3
had
been prepared and glued. I also added rings
#12 and then #11 to the top of the vessel. While
preparing the glue surfaces I also did a little
rough turning on the inside and outside of the
vessel. I only removed wood that I was certain
would have to be removed later, as shown in
photo 9-23. I maintained a majority of the
possible wall thickness until most of the vessel
had been constructed. Retaining maximum
wall thickness gave me options during the
final shaping.
There were several ways to put together rings
#6 through #10. There are many steps involved
and it can get confusing. Here is how I did it.
Since rings #6 and #10 are thin layers of holly
about the same diameter, I cut, sanded, and
glued together (using the half-ring method) one
thicker hoUy ring. I also did the same for ebony
rings #7 and #9 I then lightly attached the
ebony ring to a round piece of particleboard
usmg a few beads of hot melt glue. The
particleboard provided a handle for holding the
nng agamst my disc sander to flatten and
9-23 Initially, keep the vessel walls at maximum
thickness.
smooth one side. I could have done this on te
lathe, but my sander is big enough and I was
not concerned about turning the ebony ring
round at this stage.
I flattened the feature ring using the disc
sander and then glued the ebony ring to it
using many spring clamps (this is shown in
Chapter 7. photo 7-29). I flattened the holly r
on the disc sander same as the ebony ring. Then
I glued it to ring #5. clamping with a pipe clap
that pushes from my shop ceiling. I have many
different length pipes and I use this technique
quite often, because it does not tie up my lathe.
This holly ring is fairly sturdy. but just as with
thinner and weaker rings I positioned Hon a
thin layer of rubber mat that was covered with
wax paper. The rubber router pad supports the
ring very evenly, helping to distribute the
clamping pressure.
In photo 9-24, the bottom of the turning with
the holly ring has been lathe-mounted. The
exposed side of the holly has been flattened
and the upper portion of the turning
has
bee
n
glued
onto the holly ring using the
tailsto
ck d
a clamp_
After the glue cured, I parted the holly ring V
half as shown in photo 9-25. I then turned
down the holly rings (rings #6 #10) to a
thickness
of about 1/8 inch and
flatten
ed
the
gluing surfaces. These two halves of the
9-24
Planning ahead can save assembly steps.
9-26 Here is another example of two-from-one
ring usage.
turning were then set aside.
I mounted (again using hot-melt) the feature
ring with the ebony ring to my centering
faceplate. I flattened the exposed surface of the
ebony ring, and as you can see in photo 9-26, I
glued the upper portion of the turning to the
ebony ring.
I parted the ebony ring into two rings as I had
done with the holly ring. and turned down the
two ebony rings to a height of approximately
3116 inch. When the vessel is shaped, the angle
of exposure will make this 3116 inch ring appear
a httle thicker, closely matching the 1I4-inch
Wide vertical pieces of ebony in the feature nng.
JOHN
'
S TURNING
9-25 Splitting a ring results in a two-for-one
benefit.
9-27 Masking tape can be being used as a
positioning aid.
In photo 9-27. I have removed the featwe ring
from the centering plate and I have carefully
centered and glued to it to the base of the
turning using the ceiling damp system. At this
point I have assembled the entire tng into
two halves and both pieces are still mounted
and centered on their original faceplates.
Now it is a matter of turing the outside
profiles to their final shape. A frequent visu
check of the outside profe is necessaz dunng
the outside shaping -the line has to be Just
right. The outside profile can be turned with the
two halves temporarily joined using a httle
double-sided tape plus support from the
tailstock. In thiS case I simply held the two
8
4 THE ART OF SEGMENTED WOODTURNING
9-28 Finish-sand the inside prior to joining the
two halves.
9-29 This is the last sanding step prior to gluing
the ha Ives together.
9-30 Take steps to minimize glue squeeze-out
contamination
halves together frequently, to examine the
profile as I finalized the shape. I carefully
matched the inside and outside diameters of
the mating surfaces and turned the insides of
the two halves to their final wall thickness. U
this particular vessel, 1 was striving for a
consistent wall of 3116 inch. If the vessel had
been much smaller, my thickness goal would
have been closer to 1/8 inch. While it certainly
would have been possible to tum the wall down
thinner, based on my cOnversation with its
owner 1 wanted the vessel to be able to survive
a fall to the floor.
The vessel's top opening was too small for my
hand, so it only made sense to sand the inside
as much as possible before gluing the two
halves together. Both insides were power
sanded to 400-grit (photo 9-28) and both
mating surfaces received one last touch of the
sanding block (photo 9-29). I also checked the
matching diameters one last time.
I did not want the difficult task of cleaning up
any glue squeeze-out that might run d0the
inside surface. I used masking tape to create a
barrier, as shown in photo 9-30.
Joining Two Halves
The big moment had finally come. I glued the
two halves together as shown in photo 9-31. I
predetermined the rotational alignment and
positioned a piece of masking tape, which I
then cut at the glue joint line. The tape gives me
quick rotational alignment, allowing me to
focus on the critical centering of the two pieces.
Usually the diameters are not a 100% match,
creating a tiny ridge. Using my fingerailS, I can
judge the consistency of this tiny ridge all
around the circumference and make slight
adjustments as necessar. The clamping
pressure in this case is simply a few
faceplate
s,
about 15 pounds. The total surface area ofthe
glue joint is quite small and does not require a
lot of pressure. l did not use the ceiling clamp
because I did not want the distraction of having
to adjust the angle of pressure, which otheJ
se
might force the upper portion to slip
sideway
s
Stacked weights provide enough centered
force
9-31 Concentric alignment of the two halves is
critical.
I try to time this step so the glue can cure
overght, for two reasons: I want a strong glue
joint prior to contiuing. and I want any glue
squeeze-out on the inside to be dry so that
when the vessel is spun, the squeeze-out does
not smear.
The next moring I mounted the vessel using
the top faceplate (photo 9-32). The glue job had
perfectly centered the two halves, making it
possible to tum the vessel from either faceplate.
This made it easy to finish the base shaping
without the interference of the lathe headstock
-not a big deal, but having the vessel
positioned this way did help.
After using a t/2-inch bowl gouge and a t-inch
shear scraper, I made a few passes with a
hand-held cabinet scraper (photo 9-33). My
goal with these tools was to create a surface
that did not require coarse sanding. I wanted
to start sanding with lSo-grit, because WIth
J
OHN'S TURNING
8S

.

I

9-32 A perfect glue job al l ows lathe mounting
from either end.
9-33 Before sanding, a 1ne burr on a cabinet
scraper can improve the surface qual ity
80-grit, the combination of soft holly and hard
ebony would be easy to over-sand. creatmg an
uneven surface.
The Finishing Process
Afer completing the base I re-mounted the
vessel using the base faceplate, removed the
top faceplate. and turned off the waste block. l
finish-turned the upper half of the vessel and
power-sanded its entire surface to -loo-gnt. I
cleaned up the inside glue
J
oint with a hook-
86 TE AT OF SEGMENTED WOODTURNING
9-34
My wife's small hands are invaluable.
9-35 This is an effective method of reverse
mounting the vessel.
shaped Stewart scraping tool. All surfaces were
then cleaned with a tack cloth and made ready
for the first of two coats of sanding sealer. The
top opening, as I knew it would be, was far too
small for my hand to enter. Photo 9-M reveals
one of my secret tools -a willing,
enthusiastic wife with small hands. Without
Tere's assistance, 1 would have been fumbhng
with a rag on a stick for who knows how long
His just as important to protect the inside as
the outside. The finish is a barrier ag8
moisture. To prevent unequal wood
movement, both sides need the same
treatment. If a vessel opening is too small for
anyone's hand, then the next best thing is to
pour oil in, swirl it around, drain it back out,
and wipe it dry as best you can with a rag
attached to a dowel.
After drying for 24 hours, the vessel was sanded
again with 400-grit and rubbed with super-fine
steel wool, preparing it for the second coat of
sanding sealer. The final fnish was four coats,
applied one each day, of a satin tung
oil/urethane product made by General Finishes
(the green can). Before applying each finish
coat, I rubbed d0the surface with steel wool
and cleaned it with a tack cloth. After the final
coat, [ buffed the surface with a poliShing
compound using a bufng disc mounted in a
hand drill. did the fnishing and final
buffing with the vessel Smounted on the
base fceplate.
I prefer non-glossy finishes. I want the wood to
look and feel like wood. not plastic. although. I
have to say, I have seen some very professional
glossy finishes on turings. Sometimes I have
put a shine on a piece by applying a final coat
of buffed , but it is not my most-used finish.
A satin sheen is just my personal preference.
We are almost done. With a vessel this size. I
partially part off the waste block and then
finish the job with the band saw. If you have
ever cut something round on a band saw, then
you know it can be tricky. The saw teeth wtry
to rotate the turning in your hands. Be careful if
you attempt this -keep your hands clear and
maintain a firm grip as you cut slowly. [fthis
procedure makes you nervous, then do not try
it. There are other ways to remove the base
block. A handsaw while the piece is still on the
lathe is one way. Reverse-mounting the vessel
and turng off the block is another My most
common method of reverse-mountmg is to cut
a recessed groove into a mounted piece of MDF
to snugly fit the vessel top. and then attach a
donut-shaped ring of MF using sl16-inch all
thread rods. Thin layers of foam rubber and
pieces of paper towel protect the vessel finish,
as shown 1photo 935.
I'm not sure why, but one of the first things
people do after they pick up a woodturning is
look at the bottom, so I try to give them a little
something to look at. To create a professional
looking bottom, I reverse-mount the vessel,
turn, sand, and sign the bottom before
applying finish. If there is room, I usually
record the wood species used. My choice for
any writing work is a simple wood buring
tool. It may seem like an insignificant detail,
but people will look at your Signature with
great interest, so it pays to design and execute
it with great care. Iusually try to include a
white piece of wood in the base, just to have a
good surface upon which to burn (photo 9-36).
Even if you are not selling to the public, your
family and heirs will appreciate the signature.
If your design requires dark-colored wood on
the bottom, then use a silver or white fine
tipped permanent marking pen. Do not remove
the vessel from the reverse-mounting device
until you have finished signing the bottom. If
you remove it before signing, and happen to
mess up with the wood burner, then you will
have to go through the mounting process
again. If it is still mounted, it is relatively
simple to turn off the mistake and redo the
signing. Hurray, we are done with this one!
In the world of segmented turning, this vessel
(photo 9-00, page 72) is a very typical design.
You might be thinking, "This is too much work,
I think I'll stick to big blocks of wood." Do not
be discouraged that easily. While it is a lot of
work, it can be very satisfying. If you give it a
chance, many of the techniques that I have
described will become almost second nature
and you will quickly learn methods that
significantly shorten construction time.
JOHN'S TURNING
"J -t
'3
9-36
Always sign your work.
88
THE ART OF SEGMENTED WOODTURNING
10-00 LH55Ptf5 (tallest is 5-1/2 inches) are examples of forms built using small staves.
1001 THk WfDDING Ot1 (10 inches tall). hollow turnmgs constructed of staves
10.
Stave
Construction
While stacking rings is the most common way
to construct a segmented tung, there is
another technique: staves. Wooden barrels were
built from staves hundreds of years ago. The
wne industry still favors stave-constructed oak
barrels for wine aging. There are two basic
types of stave-coFcted fanTIs: one is built
using simple miters, the other is constructed
fom staves with angled miters. KO as
compound miters. Simple miters are certainly
easier to construct, they are just tall segents
with the wood grain oriented vertically.
However, they do not provide much
opportunity for creating a non-parallel shape,
unless the boards are quite thick. The chess
pieces i photo 1o-ooare miniature examples
of stave constructon Simple-mitered staves
comprise the portions of the chess pieces with
the dark veneer splines. Because of the small
size (the king is j.jinches tall), there was no
reason to use compound-mitered staves. In case
you were wondering, the grain orientation of all
the sections is vertical.
Compound miters make possible a wide range
of vessel wall angles. Shapes can range from a
very shallow dish to a vertical profile. By
adjusting two angles. the saw blade angle and
the miter gauge angle, we can achieve a variety
of shapes. Calculating compound miter angles is
more complicated than simple miters, it is not
just a matter of dividing 360" by twice the
number of segments. There are two key
variables: the number of staves, and the slope
angle of the desired form. The number of staves
is simply your decision based upon stave width
and vessel Circumference. In other words,
desired diameter multiplied by pi and divided
ST CONSRUCON 89
by desired stave width, equals number of staves.
The other variable, slope angle, is measured
between the side of the form and the flat
surface upon which it sits (assuming the form is
cone-shaped with the small end at the base). A
tall vessel would likely have a slope of more
than 60", whereas a shallow platter would
probably have a slope of less than 25
"
. The slope
is completely up toyau to determine, use a
protractor and draw a few lines to help you
decide.
Calculating Compound
Miter Angles
Do not worry, you do not have to go back to
_mhigh school geometry boos. Uthe
appendix of this book I have provided a chart
listing the angles most commonly needed
Hoever, if you need to detennine a set of
angles not listed in the provided chart. and you
are not afaid of a little mathematics, then you
will need a calculator with trig functions and
the following formulas:
(If you have no interest in calculating compound
miter angles, then skip ahead afew pages.
Miter Angle (MA) " inverse tan (1 {casS
lan[360.. (2N)]))
Blade Angle (BA) = inverse |(cosMA. tanS)
MA is the miter angle.
S is the slope of the vessel (measured from
horizontal to side),
SA is the saw blade bevel angle,
N is the number of staves.
go THE AT OF SEGMENTED WOOoRNING
-
~

M 3
. ..
dY
m D

Q
Wey
8 ry
m N
WT4 W
.
e c
m sr
10,02 Staved vessels also require careful
planning.
Trust me, this is easier than it looks. You do not
have to understand cosines and tangents, you
only need to know how to push the right
buttons on a calculator that has those functions.
Make sure the calculator is in degree mode, not
radian moe. Start by first selecting two
numbers, the number of staves and the slope of
the vessel. I will take you through the process
using 75" as the slope angle and 12 as the
number of staves. The width of the staves has
no bearing upon the angles, His only varied to
change the Circumference (and diameter) of the
form. Because the blade angle (BA) formula
needs the miter angle (M), I must calculate it
first. Below I have insered 75 as the slope and
12 as the number of staves.
Miter Angle (M) = inverse tan (1 7 [COS75"
tan(360' " (2'>2))])
Using my calculator, I next determine the cosine
of 75" to be .258819, I have done the math (360 ^
(2.12)) which equals IS", and I h calculated
the tangent of IS" as .2679491. Therefore:
Miter Angle (Ml ' inverse tan h (.258819
.2679491))
continue by perfoOg the multiplication
Miter Angle (M) ' inverse tan (1
.0693503)
and the division as shown above.
Miter Angle (M) = inverse tan 14.419548
To conver the inverse tangent to degrees of
angle, use the calculator once more. You might
have an inverse button (IN) or, as on my
calculator, hit shift,tan to display the inverse:
Miter Angle (Ml ' 86.032872" or 86.03"
Now with the miter angle (M) K0,I can
find the blade angle (BA). Below, I have put the
miter angle and the slope angle in the formula
Blade Angle (BA) : inverse tan(cos86.032872"
tan7S')
The calculator provides these figures:
Blade Angle (BA) inverse ta(.06g1B41
17320soB)
The multiplication results:
Blade Angle (BA) = inverse tan .2581985
The calculator converts the inverse of the blade
angle tangent to the blade angle in degrees:
Blade Angle (BA) = 14.4n49S' or 14.48'
The next step is to adjust your saw to these
angles (blade angle 14.48", and miter angle
86.03"). I hope this is a little clearer than mud; it
can certainly be confusing, especially if you do
not use it very often. That is the convenience of
chars, but if you ever have to calculate angles
for a staveconstructed form not listed. now
yu
know where to find the prOure.
Building A Staved Vessel
Staves cut with the grain positioned
vertically
present another problem Hyou desire 10 attach
additional layers. Consistent grain orientation
becomes difficult to maintain. I wlshow you
how I deal with that problem as Ibuild a stavr
constructed vessel
The first step, just as WIth the previous project
(JOHN'S TRNING), is to develop a design. I looked
over my wood inventory and decided to build
three smaU vessels using narrow, 2-inch wide
curly maple boards, cut-oft 1 purchased from a
guitar blank supplier. Guitar suppliers secure
some of the finest wo s available they sell
some terrific looking scraps at reasonable
prices. Based on mywood selection, Icreated a
simple drag (photo 10-0:). I decided to build
the stave portion of my three vessels using
eight sides, with a slope angle of 75 (or 1jfrom
vertical), and to use solid pieces of maple burl
for the top shoulder section of the vessels. I
needed wedding gifts for two of my nieces, so
these vases were just the ticket.
Referring to the table of compound miter angles
(page 176), I find that I need to cut mystaves
witha saw blade angle of 21.69" and a miter
gauge angle of 83.88. I wish it were that simple.
Unless your equipment is a lot more high-tech
Umine, setting up these angles requires trial
and error. That's why I decided to build three
similar vessels at the same time. After the
lengthy process of adjusting my saw, Iwant to
make more than one turning. The boards were
narow and about 22 inches long. I first
machined them flat and straight using a
jOinter/planer, then crosscut them into 7-inch
long rectangles. To cut these rectangles into
staves, 1 used a shop-built sliding table saw sled
th securely held the wood for consistently
accurate cuts. Photo 10-03 shows my sled device.
It is made from I-inch thick MOF with two
runners on the bottom that fit quite snugly into
mytable saw miter gauge slots. This eliminates
any side-to-side sloppiness in the travel as it is
pushed into the saw blade. At the top of the
photo, you will notice a bridge of I-inch MDF
spanning the two sides. It stabilizes the two
sides of the sled by securely connecting them
across the saw blade kerf. At the back of the sled
is an adjustable 2-inch thick lamination of MDF
that acts as a miter gauge.
This sled usimilar to the miter-cutting sled
described in Chapter 7, with adjustabUtty for
different angles and without the tapered exit
ramp. Before setting up this sled to cut the
ST CONSUCTION
91
10-03 A table saw sled is used to cut compound
miters.
10-04 After cutting one side, set up the sled for
the second sides.
compound miters, I cut t-inch thick MDF into
rectangles that matched my curly maple pieces,
for makng test cuts before sawing the god
stuff. Using a protractor, J positioned the MDF
miter gauge as closely as I could to 83.88",
realistically, to within 1" of that angle. Next. I
adjusted my saw blade as close as possible to
the desired 21.69. For safety and accuracy's
sake. I positioned a hold-down clamp on the
sled to flImly secure the WIf you look to the
right side of the hold-down clamp. you can see a
piece of MDF attached as backstop. thereby
ensuring identically dImensIoned pIeces. The
inch thickess of the sled allows for easy. secure
screw-attachment of components such as stop
blocks holddown danlps
92 THE ART OF SEGMEN1D WOOOTURNING
10-05 Checking the dr-fit is ver i mportant.
I first cut eight test pieces on one side. Then I
crosscut a piece of MF to create an angled stop
block. which I attached to the other side of the
blade. In photo 10-04. you can see this stop
block with the hold-down damp mounted on
top of it. You will also notice a piece of tape
attached to the piece of wood which is about to
be cut. Because I have used this sled for many
other projects. the kerf has become much wider
than the blade. allowing narrow cut-ofts to fall
into the ker and jam. By holding the end of the
tape in one hand as I pushed the sled with the
other, I could retrieve the cutoft before it caused
a problem.
After cutting the other sides of the test pieces, I
laid them out on a flat surrace and taped them
together much like the maple pieces in photo
10-05. Irolled the form into a cone shape and
checked the accuracy of the angles. After the
first test cuts I was only off a tiny bit. The angles
were a little tight on the inside, causing a slight
gap between the untaped outside edges. This
meant I needed a tiny bit more angle on one of
my settings. It Uimportant to note that any
time you vary from a given chart angle, you will
also alter the slope of the form. I was not
worried about achieving an exact 75 slope,
because 1 had enough wothickness for
turmng the desired profle. Therefore. I
proceeded by adjusting only one angle until the
fit was right. Since I had screwed down the MF
miter gauge, it was much simpler to adjust the
blade angle. After looking at the first test
assembly, I estimated that I was oft a total of
approxmately 2 or less. By dividing 2 by 16
(the number of angles). I deteted that I only
had to increase the blade angle by about .12,
which is not very much. I knew that mysaw
blade angle adjusting handle moves the blade
1.50 per revolution, so had to turn the handle
less than 1/8 of a revolution. This sounds like
splitting hairs. but that is how I decided my
next move.
Youmight be wondering how I estmated that
my frst set of cuts was off by approxmately 2-.
Since the circumference of any ring contains
360, ua circumference were 360 inches. then
each degree would span 1 mOLikewise, if a
Circumference were 36 inches, then 1 would
equal 1lto inch (J6 inches divided by 360). The
vessel under construction had a crcumferenc
e
of approxmately 16 inches. therefore each
degree equaled approximately .04 inch (16
inches divided by 360.). Iestimated that
the total
width of gap on the outside of the fust set of test
pieces was less than .1 inch, therefore I needed to
increase the total of Bthe angles by
approximately 2- (.1 inch diVIded by 04 Inch)
ThIS type of estimating IS rough, but it is better
than haphazardly adjustmg the saw blade.
I made the 1/8 of a turn adjustment and cut
eight more test pieces, taped them together, and
checked for accuracy once again. I was lucky,
they ft perfectly; usually at least one more set
oftest pieces is required. I cut the first eight
staes and checked the fit once again, using the
tapetogether technique shown in photo 1005.
Afer conDng the settings I cut the other 16
staves, gving me enough to glue together three
different cone-shaped forms. No matter how
many attempts it takes, the fit has to be near
perfect and if the form is to be glued together
all at one time, then the fit has be absolutely
perfect, at least to the naked eye.
Before eiuing these staves together I needed to
create a laminated spline, to glue between the
staves. This was a five-piece lamination, as
shown in photo 10-06, consisting of t pieces
of maple veneer, two pieces of ebonized walnut
veneer, and one piece of l'8inch thick ebony. To
either side of the lamination components you
can see a piece of 3'4-inch MDF coered with
blue masking tape. These are cauls, to Dplaced
on the outside of the laminations to provide
uniform damping pressure. I glued the layers of
wood together. cleaned them up. and cut them
to length. These splines had parallel sides, so
they had no effect on the fit of the staves.
I then laid out the assembly of staves and
splines flat (outside up) and applied tape to
each seam. I paid special attention to make sure
the tape was well adhered to the splines, so it
would keep them in place during the gluing and
clamping. When applying the tape, do not
squeeze the pieces together tightly. The tape
should not restrict the staves from easily
folding in to form a circle. The tape is only an
aid in the assembly process, not a clamp. After
applying the tape, I used another board on top
ofthe assembly to turn uoutside down as
shown in photo 10-7.
Gluing together compound miters is more
df cult than gluing simple miters. Because of
the angle of the outside profiles, hose clamps
may slip and not provide the needed pressure.
STAV CONSTUCTION
'oo6 The splines are a fivetayer lamination.
1O07 The assembly is now ready for glue.
93
Depending upon the acuteness of the outside
slope, rubber bands might not stay in place.
Gluing these forms together presents challenges
for which I wlsuggest a few solutions.
My preferred method of clamping this type of
cone-shaped form requires constructing a
customized gluing jig. To build this jig, I band
saw fow circles of 3'4inch MDF and mounted
one to a faceplate. I then secured the other three
to the mounted circle with a single centered
screw, as shown in photo 10o8 (next page). The
circles were then all turned to the same
diameter (this was not necessary. but it looked
better than rough bandsaw cuts). Before taking
the circles apart I drilled three evenly spaced
holes near the outslde edge, to accommodate
sh6-inch allthread rods used to damp the
94
THE ART OF SEGMENTED WOODTURNING
10-08 This will become a gluing ji g for all-at-one
time stave gluing.
10-09 A parting tool is used to cut donut-shaped
rings for the gluing jig.
10-10 ThiS style of glUing JIg reqUIres perfect
miter angles.
form. Ialso made reference marks in order to
maintain
the original alignment of the
outsid
e
holes while gluing and damping. This will
make more sense in another couple of
photos
I took the assembly of circles apart and
using
the center holes I mounted each one onto a
screw chuck as shown in photo 10-0g.
Using a parting tool, I then cut diferent
sized
donut -shaped rings from the three circles. Their
inside diameters were cut to roughly match
three different outside diameters of the cone
shaped form, and the cuts were done at an
angle to closely conform to the slope of the
staves. I did a dry-ft of the jig surrounding the
cone shape and positioned pieces of all-thread
with nuts and washers at each layer. Aer
removing the jig I applied smooth, slippery duct
tape to the inside edges of the donut rings, to
reduce fiction against the staves and to allO
easier removal later. Then I laid the staves back
out with their open inside seams facing up and
visually checked once again for any sawdust
that might interfere. Because I knew that the
tightening all the nuts would take some time, I
waited until the next moring when the
temperature was cooler, to have a little more
working time with the glue.
To assemble the cone I applied glue (TitebondJ
generously to all the glue surfaces, positioned
the staved cone on the mounted faceplate,
installed the three donuts with the all-thread,
and quickly threaded wing nuts onto the all
thread on the back side of the face plate. By
tightening the nuts that I had positioned at
each donut, I was able to squeeze the donuts
towards the faceplate. Photo 10-10 shows the
glued assembly positioned upside-down to
allow glue squeeze-out to drain from the
insid
e.
The staves must fit periectly for this technique
to succeed. Building the jig is time consuming.
but it does a great job of applying
pressure
in
the needed directions and it results Da
perfectly round form, attached and
center
ed
on
a mounted faceplate. ready for turing.
For demonstration purposes I glued up
anoth
er
of my three cones using another method.
Similar to gluing flat segments. glued
toge
thll
pairs of staves until two halves existed. Photo
10-11 shows qUick clamps applying pressure
between two staves with a spline in between.
Because of the angles and the widths of the
staves, the two outside edges lined up opposite
each other and allowed for this type of
damping. If that had not been the case, I would
have used numerous rubber bands.
photo 10-12, because the outside profie of the
form was not very acute, I was able to use
rubber bands to clamp the quarter-sections
together. The bands stayed in place without
slipping, allowing me to apply many around the
shape. This is tricky, the bands must b
stetced towards the inside ofthe form in
order to provide uniform pressure on both sides
of the glue line, otherise the rubber-band
pressure will try to pull open the outside of the
seam. Play around with this dry before gluing
and you will see how to adjust the tension.
Now that there are t halves, it is just a matter
of tuing up the mating surfaces to achieve a
perfect fit between the two halves. My 2o-inch
disc sander makes short work of this chore
(photo 10-13). Sandpaper mounted on a flat
surface is another option, it just takes a little
more muscle. A little rough-sanding with a belt
sander can reduce the amount of hand-sanding.
Once you achieve a perfect fit between the two
halves, then it is a simple matter of gluing and
damping them together with rubber bands.
The first two forms, after removing the donut
style clamping device, were center-mounted
and ready for turng. I glued the third form to
a faceplate-mounted waste block. Next I
flattened the base of all three turngs to
prepare them for a couple layers of veneer. This
presented a compromise in grain orientation:
the grain of the staves uvertical, while the
grain ofthe veneer is horizontal. Normally I
would not consider such an anangement,
however in this case, because the overall
dimensions were qUite small, the risks were
acceptable. To glue on the two layers of veneer, I
used my ceiling clamp system as shown in
photo 10-14 (next page). I glued both layers of
veneer at the same time, with just a slight
offset ot their respective grain directions. To
ST CONSTUCON
10-11 This is another method of joi ni ng staves.
10-12 Rubber bands can provide plenty of
clamping pressure.
10-13 A disc sander makes short work of
flattening the halt-cones
96
THE ART OF SEGMENTED WOODTRNING
10-14 Veneer layers are added, with the damps
pressed against the ceiling
ensure consistent damping pressure, the
veneers were positioned on a thin, wax-paper
covered rubber router pad.
For the base rings, I assembled rings of ebony
from segments with their grain oriented
vertcally, same as the staves. Instead of cutting
these miters on the miter saw, I rip-cut strips on
the table saw at 22.S
G
each side, and then used
the miter saw (at goO) to cut these strips into
segments. I attached the resulting rings to a
waste block with hot-melt, and prepared them
for joining to the vessel base. Because of their
small size I could have built these base rings of
ebony with horizontal grain, but I wanted
vertical grain in an upper ring. so it was easier
to build both rings the same way. By having
vertical grain I was able to create a stronger
joint, as described next.
Mortise-and-Tenon Joint
Between Layers
I wanted to improve the strength of the joint. so
instead of glumg two flat surfaces to each other
(the ebony end-grain and the maple veneer
side-grain), 1 decided to join them using a half
mortise-and-tenon type of joint. This provided a
small side-grain-to-sidegrain connection
within the seam. Photo 1o-1shows the vessel
base with its half-mortise cut, and the ebony
10-15 A mortise and tenon joint creates a more
reliable glue jOtnt.
base ring with its tenon cut. The easiest
sequence that I have found for accurately
tung this type of connecton i:
1. Determine the center of the wall thickness by
turing at least one of the profles (inside or
outSide) dose to its final shape.
2.Use a sharp diamond-pointed scraper to
carefully tum a mortise in the base of the
vessel, as shown in photo 1o-1.Turn this
mortise diameter close to the intended
centerline of the vessel wall thickness. In
photo 1o-1,I have intentionally left more
wood to the outside of the vessel until the
assembly is complete, when it wbe tue
down to the final profle.
3 Next, tum the tenon or male portion of the
connection as shown in photo 1o-1.Because
of the darkness of the ebony, you must look
closely. Temporarily attach the ring of ebony
to a circle of MDF and then screw it onto a
screw chuck or grab it in a four-jaw chUck.
Using calipers that were set to the outside
measureme
nt of the mortise, turn
the
teno
n
diameter
to a very close match, but
not
all
the way. Also. make sure the tenon is
sligh
tly
longer than the depth of the
mortise.
4 Remove
the ebony tenon piece fom
the
screw
chuck and re-mount the
vessel
onto
10-16 A tapered plug completes the vessel base.
the lathe. Check the ),it should still be a
little too tight. Using the pointed scraper,
remove a smidgen of material fom the side
of the mortise and check the ft again.
Continue removing tiny shavings and
checking the fit until it's snug.
j.Check the horizontal gap between the two
pieces. It should be slightly open because the
tenon has bottomed out in the mortise.
Remove tiny amounts from the bottom of
the mortise and check frequently until the
gap becomes tight.
6. There is no way that I know of for measuring
the accuracy of the right-angle side cuts that
you need to make. They are so small you
should be able to eyeball the surfaces and
achieve a good fit.
7. One way to check for tight spots is to insert
the tenon into the mortise while the lathe is
turing, but burishing the wood will seal
wood pores, which will jeopardize the gluing.
I suggest minimizing this type of fit check.
It takes a little time to fit up this type of
connection, but the improvement in joint
stability is well worh the effort. EVen though
the side-grain-to-side-grain portion of the joint
is quite small, it adds a lot of strength.
STAVI CONSl UcoN
97
10-'7 In lieu of a steady rest. braces can provide
stability.
Back to Vessel Constrction
The next step was to insert a plug into the
ebony base ring. I used a piece of curly maple, as
shown in photo 10-16. When fitting up this kind
of plug, remember to shape the plug first, and
then cut the recess in the base. It is a lot easier
than the other way around.
I tured the outside profles pretty close to their
final shape and sawed the forms off their
faceplates. then re-mounted them with their
bases glued to the same waste blocks. I tured a
shallow recess (about 1116 inch) into the
removed faceplates and used these precisely
dimensioned recesses to fit the base ebony
rings, thereby keeping everything nicely
centered. I did not turn the base ebony rings
down to their final diameter, in order to provide
more stability and strength for attaChing to the
waste blocks. Because the base diameters were
small, pOSSibly not prOViding enough strength
during the interior turing, I decided to install
braces on the outside of the vessels (photo
10-17). I did not want to risk losing a piece off its
waste block. A steady rest could have been used
instead of the braces. The insides were then
tum down to a wall thickness of just over 1/8
inch. I figured that later. after shear-scraping
and sanding the outside. the final wall
thicknesses would be pretty close to 1/8 inch I
98 THE AT OF SEGMENTED WODTURNING
10-18 Veneer layers are added to the vessel top.
10-19 The upper section of the vessel gets
prepared.
10-20 The tallstock effectively clamps these
sectIons together
-MMM
sanded the inside and then cut a half-mortise
into the top, to accommodate the gluing of
another ebony ring.
For the second ebony nng I glued together a 3'4-
inch tall ring with vertical grain, mounted it on
a piece of MOF, and machined it to fit the
mortise that you can see at the top of the staves
in photo 10-17. After gluing the ebony ring onto
the staves I pared it off, leaving enough ebony
still attached to the MOF to provide another
ebony ring for one of the other vessels. I shaped
these ebony rings with a mortise, just as had
shaped the top of the staves, in order to accept
the next component. The third vessel was a
little different: I used East Indian rosewood for
the center of the splines and also for the center
and top rings, instead of ebony.
Ithen started constructing the upper portion
of the vessel. For the top rims, I created small
ebony rings with horizontal grain and
mounted them to center -drilled discs of MDF.
After I flattened the ebony I glued on two
layers of veneer, and cleaned them up as
shown in photo 10-18.
1 had a few nice pieces of maple burl that I had
decided to use for the shoulder area of two of
the vessels. For the vessel with the rosewood. I
used a piece of vertical-grained spalted maple.
These burl pieces. with their grain oriented in
many directions, were center-drilled and
mounted onto a screw chuck (photo 10-19). I
turned a laugh outside shape and created a flat
surface in order to glue on the top ebony ring.
the one with the veneers.
unscrewed the shoulder piece (the burl) fom
the screw chuck and screwed the top ebony ring
back onto the chuck. This allowed me to
pOSition the tailstock to act as a damp and
centering device for gluing on the bwl piece. as
shown in photo 10-20.
I rough-shaped the inside of the burl piece and
cut a half-tenon to ft the top of the ebony ring
that was attached to the top of the staves.
After
achieving a snug fit. I put the two pieces
together WIthout glue. ThiS assembly was
lathe
mounted and I used the tailstock to hold the top
onto the lower half I removed the braces and
final-shaped the outside. Ater taking the
assembly apart, reinstalled the top onto the
screw chuck, turned the inside to the firushed
@thlckness, power-sanded. and then
protected it from glue squeeze-out by applying
tape (photo 10-21). Note the tenon on the burl
], which will fit the ebony ring.
I glued the two halves together in an upsIde
down position to prevent glue from running
down the inside of the staves. The next day, I
tured off the top MDF disc and shear-scraped
and sanded the outside. After removing the
interior tape I cleaned up the inside seam,
wiped the vessel clean with a tack cloth. and
applied sanding sealer. If you look closely at the
waste block in photo 10-22, you'll see that I had
to remove part of it in order to accomplish the
final turing and sanding near the base. With
hindsight, l could have saved myself some
touble by laminating another layer onto the
waste block before gluing on the vessel, thus
allowing much easier access to the lower
section.
After another coat of sanding sealer and four
more coats of satin finishing oil, I took the
vessels off the waste blocks, reverse-mounted
them, and prepared the bases for signature and
finishing. The finished vessels are shown in
photo 10-01 (page 88).
One other stave-gluing method that I did not
show during the constuction of these three
vessels is the use of gluing blocks an the outside
surfaces of the staves. The donut compression
ring technique does not work very well if the
vessel slope is flatter than about 60 and as the
slape approaches 45, it does not work at 8The
ratio of vertical pressure compared to horizontal
pressure creates an inefective clamp. My
solution to clamping a flatter cone shape Bat
one time is to temporarily attach glue blacks and
use hose damps. Photo 10-23 shows this type of
arrangement. The glue blacks prevent the hose
clamps from sliding on the angled surface.
STAVE CONSUCON 99
10-21 Tape can be used to prevent glue squeeze
aut contaminatIon.
10-22 Generally. | like to apply finish while the
vessel is still mounted.
10-23 Glue blocks can also be used to assemble
staves.
100 THE ART OF SEGMENTED WOODTURNING
11-00
/DH R5 1It0P (6-112 inches tall) -an example of alternating stave orientation.
11.
lamination
Tri ckery
Mother Natwe does a fne job of creating layers
qcolor in the woods that she grows and solid
wo woodtumers experience the joy of
exposing those layers to the world. Different
shapes and profile angles expose different wood
and create a wide array of images. Laminated
wood is the same, completely different images
appear depending upon the angle of exposure.
In this chapter, I offer a few of the many design
effects that are possible with layers of wood.
Building Zigzag Rings
Zigzag patters require precse ftting and gluing.
You can use them as a stand-alone featwe rings,
or as a complementary ring alongSide a feature
ring, almost any size or shape is possible. The
first step is to decide how many zigs and how
many zags you want in your ring. that is, how
many points upward and downward To caku1ate
Uyou need to know the circumference of the
ring and the characteristic of the adjoining
rings, that is, how many segments or feature ring
elements. Let's assume a diameter of 6 inches,
therefore, a circumference of 18.8S inches. In
addition, assume that the rings next to the zigzag
ring contain 12 segments, which means the
segments next to the zigzag ring are
approxtely I.S7 inches long (18.8S inches
divided by 12). With this information, you can
derde the shape of the zigzag pattern. Hcan be a
flat design with only six upward points (one for
every two segments), or it C be a sharper
pointed design with 24 upward points (two per
segment). For every upward point, there'll be two
slopes downward. This means that to create 12
upward points, 24 angled components are
required. To create a cirtrence of 18.8S
inches, each of the half-point components needs
lNAnON PICKFkY
11-01 h|Sis a typical zig-zag feature nng.
101
to be .79 inch wide (18.8S inches divided by
pieces). Another decision relates to the number O
layers to laminate in the zigzag. it is just a matter
of preparing and gluing together strips. The
technique for building the zigzag remains the
same regardless of the numbr of layers. To keep
DSdemonstation reasonably Simple, I W
build a zigzag ring with just one centered
contrasting band of wood. Photo 11-01 shows a
completed 6-inch diameter ring with 12 upward
and 12 downward points. Follow along and I w
show you how I buill this ring
I needed 24 angled pieces of wood .79 inch long;
to create a 6-inch diameter ring, I deeded to
increase the length to .84 inch (a little fudge
factor). I laminated three 30-inch strips of wood
together: two sIB-inch wide strips of
yellowhear and one liB-inch wide strip of
purple heart. The width of the outside stips is
important. If they are too narrow, there wlnot
be enough thickness to create the required
shape. To illustrate this, I glued together two
zigzag components shown in photo 11-02 (next
page). The outlined area of each segment
represents the tallest dimension possible from
W.As you can see, dI were to transform the
assembly on the right into a rectangular
segment, there would be no border around the
zigzag, and the pomts of the zigzag rmght
actually become cut off. Iconstructed the
segment on the left With a Wider outSide stnp
102 THE OF SEGMENTED WOODTURNING
11-02 Be sure that the outside stnps of wood are
WIde enough.
11-03 These were cut from a three-layer
lamination.
11-04 Carefully alI
g
n the points.
of wood. thus providing plenty of optional
matenal. Keep this in mind as you design any
zigzag configuration.
I cleaned-up my three-strip lamination to
prepare it for the miter saw, set the saw at 22.5,
and made a cut at one end. I examined the
lamination and confirmed that the center strip
of wood was parallel to the outside surfaces.
The zigzag pattern will be very dificult to
construct accurately if the center strip is out of
square. The choice of 22.5" was simply my
decision, I could have chosen 15, 45, or most
any other angle just as easily. This angle
determines the sharpness of the zigzag points.
Sharper angles produce taller patterns, flatter
angles produce shorter patters.
I set a stop block in order to make a series of
angled cuts that measured .84 inch wide. Unlike
ring segments, I did not flip the board over
during cutting so the cuts are all parallel to each
other. I maintained the order in which I cut
these angled segments in order to preserve any
existing color match when I later reassembled
the pieces. A few samples of these angled cuts
are positioned at the top of photo 11-03.
After cutting 26 pieces (I wanted an extra pair), l
flipped over every other one and paired them
together (photo U-03).
I lightly disc-sanded the opposing surfaces to
ensure nice tight glue joints and then glued
them together in pairs. Aligning the
purpleheart points on both sides of the joint
required extreme care, and a magnifying light
was especially helpful. Spring clamps were
used, as shown in photo 11-04.
Next. I used the disc sander to clean up the glue
squeeze-out on both sides of the seam, and
trimmed these pairs on the table saw to form
rectangular segments.
I ripped a narrow strip of 3/4-inch MDF and
stuck a piece of double-sided tape down
its
length. Holding the MOF strip against the side
of my table saw fence and using the fence
as a
gUide, I placed the pieces onto the tape as
shown in photo 11-05. To position the
pieces
m
a nice straight line, I held the two points

11-05 Double-sided tape can be used to secure

segments for table saw trimming.
11-07 The miter saw is used to Create 5egments
from the rectangular assemblies.
against the fence as I lowered them onto the
tape. My left thumb, hidden from view, is
pressing the MDF strip against the saw fence.
A few good whacks with a rubber mallet
ensured a good bond.
Using the table saw, I timmed the points off the
segments (photo u-06) and then adjusted the
fence to trim the other side, thus creating
rectangles.
Now it is a matter of miter-cutting these
segments to form a 12-segment ring (360
divided by 24 angles, equals 15 per end). I
adjusted the miter saw angle and locked the
stop block to cut one end of each segent. This
first cut hardly shortened the overall length of
the segments (photo 11-07). Then I barely
INATION TRICKERY
11-06 The table saw makes short work of
creating rectangular segments.
103
11-08 When gluing pairs together, focus on the
point-to-point alignments.
tapped the stop block a smidgen towards the
blade and cut the other ends. If you look at the
bottom of the segment. you will see that I left
a little bit of the angle indentation. This
proVided an instant orientation reminder
during cutting and assembly. If one segment
were to be placed upside-down Hwould stick
out like a sore thumb.
Using a sanding jig, l perfected the 15

angles
and dry-fit the ring. Once satisfied that the
angles were right on, I glued together six parrs
of these segments as shown in photo n-08. If I
had glued the ring together aat once or had
used the half-ring method, I would have not
been able to focus indiVidually on each Joint.
During all of the gluing steps, the only focus is
to align the zigzag points. The top and bottom
104 THE ART OF SEGMENTED WOODTURNING
11-09 The use of a paring tool creates a two-for
one opportunity.
surfaces can always be trued up later. Don't
imagine that the inside alignment is not as
important. A perfect outside alignment can
easily be turned away, exposing less than
perfect alignment Vthe vessel wall, if the
inside is not just as accurately aligned.
I continued the process of joining these
segments together until I had two halves. While
joining segments, I used the disc sander to
expose their mating comers and inspected the
intersections before proceeding. I did not want
to discover a defective joint later. The half
sections were touched up on the disc sander and
then joined together.
The key to successlaligning all the zigzag
points is accurate cutting, sanding. and eye
balling. A few minor errors here and there C
accumulate and cause an obvious misalignment
within the ring. I the diameter is small and the
Zigzags are few, then you Dneed to construct
your ring a little differently: instead of joining
the first pairs B described, mitering them wl
create a ring with more comers, thus conforming
more closely to the roundness of your vessel
Usually unot necessary, but it i an option.
You might have noticed while looking at the
previous photos that the constructed zigzag ring
was quite thick from outside to inside.
Normally. this would not be the case. I did it this
way to demonstrate aother little-used
technique, splitting rings concentrically as
opposed to horizontally. Photo u-og shows the
original zigzag ring parted into two rings: a
11-10 CHIL'S PLY (7 inches diameter) -an
example of multi-generational lamination.
freebie so to speak. This is not often U1,but I
show it to demonstrate -always be on the
lookout for time-saving techniques. With a
wider ring, several separate zigzag rings are
possible and the initial assembly time would
not be any longer.
Multi -Generational Lamination
Now the real fun begins. Instead of building a
simple zigzag patter of conSistently up and
down points, this technique can produce
incredibly complex patterns. You glue
laminations together, cut them apart, glue them
back together again, ad so forth. With each
cutting and gluing step, the design
becomes
more complex. The different angles
selected
for
cutting the lamination apart create different
patterns. The number of generations is
limited
only by your imagination and by your
ability
to
maintain accuracy. Clcuence Rannefeld's
book,
Laminated Designs in Wood. explores this
subject in great depth. My goal here is to
familiarize you with the technique,
so you
can
explore the endless possibHities.
Photo JO shows an eple of
multi
generational
lamination. This design
IS
Simil
ar
to a zigzag. though more complex.
This
LNATON TRICKERY lOS
11-11 It starts with cutting strips from a linear lamination.
particular dose-up photo is of the top portion of
a sculptural piece, CHID'SPLAYA.K.A. HRRY
POT R'S Toy (photo 11-10). The zigzag ring near
the top of the sphere was made from the same
lamination as the previously described zigzag
ring. The design below is more complex. you
can count eleven layers of wood mitered into
different angles. )inating eleven layers is the
easy part, creating the multiple angles is the
challenge. Designing a ring such as shown in
photo 11-10 is very time-consuming; most of the
time Iwing it without a plan and decide each
step as I go, not worrying too much about the
exact finished diameter or the exact shape of
the zigzags. if you wish to predetermine the
outcome, then draw each step on paper and
scissor-cut the paper to mimic the table saw.
Then reassemble the paper strips to display the
results of the next step. If you are a computer
whiz, you can draw designs that way. For me, it
is more Dto create the design as I cut and
glue in the shop. Here's how 1 build a ring with
multiple generations of lamination.
First a linear lamination is needed.l chose 1V
types of wood to create a nine-piece lation
about 32 inches long. The four strips on either
side of the center strip were milled and
arranged to create mirror images. Because of
the multiple steps involved, a lot of wood is
reduced to sawdust. Normally, a 32-inch long
strip of wood would produce enough segments
to create a lO-inch diameter ring. In this case,
the largest diaeter win be drastically smaller
because each step (or generation) reduces the
length of the lamination. Every strip of wood in
the lamination must be perfectly dimensioned
before gluing. This technique maes cumulative
errors a big concer. Each step depends upon
the accuracy of the previous steps. A drum
sander is particularly valuable for
dimensioning.
After cleaning up the lamination, Iprepared a
table saw sled for angled crosscutting. I
modified the same sled that I used to cut the
compound miters in Chapter 1O.The sled's
106 THE AT OF SEGMEND WOODRNING
11-12 A drum sander can ensure consistent
thicknesses and smooth surfaces
previous wide kerf was closed by gluing in an
angled piece of MDF, and I adjusted the miter
fence to about 25 (the exact angle is not critical
for this step). Photo 11-11 shows the lamination
clamped in place and ready for cutting. Notice
the width of the outside strips of myrtlewood.
Just as with simple zigzag segments. much of
the outside strip can become waste.
I installed a pointed, pivoting stop to cut 1/2-
inch wide strips. To use this type of stop, secure
one end with a screw and install another screw
to act as a stop for the stop block. A red arrow in
the photo points to the stop screw. With the
pivoting stop held against the screw and the
lamination end held against the stop block and
fence, lock the hold-down clamp into position.
Pivoting the stop out of the way makes it
possible to retrieve the cutoff as soon as it is
free; otherwise, it becomes confined between
the blade and an H ovable stop block. With
my left hand pushing the sled, 1 can grasp a
piece of applied masking tape (as shown) with
my right hand and safely retrieve the cutoff. It
is always a good idea to avoid loose pieces of
wood near a moving saw blade. If this method
of cutting makes you nervous, then a hold
down clamp on the cutoff side is an option. The
disadvantage of a hold-down clamp, in addition
to the extra step. is that it requires bringing the
cut-off back alongside the moving blade, which
can leave unwanted scoring marks on the wood.
11-13 A gl ui ng jig can help keep components
square during gluing.
The table saw with a freshly sharpened crosscut
blade does a great job of producing ultra
smooth cuts. Because the wood grain direction
is at an angle to the blade, a crosscut blade is
more effective than a rip blade. There is a
temptation to glue the pieces together directly
fom the saw, however, I know that a better glue
joint is possible by sanding the suraces. In this
case, because of the length of the strips, a disc
sander will not work. For a tray for making
several light passes through a drum sander I
used a piece of 3/4-inch MDF, covered with 80'
grit sandpaper to avoid slippage. Photo 11-12
shows the tray packed with the cutoffs. Before
sanding the pieces, I used the miter saw to
square the ends, because I did not Vt an
angled, unsupported end to go through the
drum sander. I arranged the pieces end-to-end.
with the wood grain facing into the rotation of
the drum. To prevent the spinning drum from
throwing one of the cutoffs, attached a small
Wof wood to the end of the tray and
positioned the cutoffs against this rail. I took
great care to ensure that the pieces rested flat
on the tray. Small debris can prevent a piece
from lying flat. causing an uneven surface.
Aer
sanding one side, I flipped the pieces and
lightly
sanded the other sides.
It's a challense to accurately glue all these
pieces back tosether into a flat zigzag
destgn
.
Accomplishing the glue job in one step $5
desuable but not practical. Therefore .
11-14 Maintai ning accurate point to point
alignment is critical.
assembled eight pieces at a time as shown in
photo u-13. I used a wax paper-covered MDF
gluing tray with a goO stop (the blue-taped piece
at the left end). First, I dry-ft the pieces to
ensure that all the points lined up when all the
ends were positioned against the stop. It was
important to align the points and to be sure
that the eight pieces were glued square.
I passed the glued sections of eight through the
drum sander to remove the glue squeeze-out,
and proceeded to join them into one continuous
lBaton. A few very light disc sander touch
ups created tight glue lines between the
sections of eight. Photo 11-14 shows this stage of
the project. I once more passed the completed
lamination through the drum sander, then the
edges were squared up on the table saw: the
two long edges must be parallel with the
zigzags as well as with each other. When gluing
together an arrangement such as this, it is
important that not only do the points line up.
but also that the elements form a straight line
from end to end.
Where to make the next cut? To create a
uniform, mirror-image design, the next set of
cuts must all pass through identical locatons
on opposite sides of the center strip. The
astuteness of the next cut will determine the
appearance of the next generation. By drawing
a few pencil lines directly on the lammation
and imagining every other piece flipped over, it
LINATON RICKFRY
107
11-15 A new pattern emerges as a result of
another series of cuts.
is possible to visualize the next design shape.
The sharper the angle, the more dramatic the
results. Play around with the options to decide
your next move.
I selected two glue lines on opposite sides of the
lamination, about 3 inches offset from each
other, as my alignment for the next set of cuts.
Here is where things start to get interesting.
The cuts must create perfect mirror images on
either side of the blade. Otherise, when the
pieces are flipped over and re-glued, the
alignment will not be correct. To do this, l
positioned the laminaton on the sled with the
two seams that I had selected directly centered
on the sled kerf. and I relocated the miter fence
with the hold-down damp against the
lamination. The sharpness of the angle
prompted me to attach a strip of 8o-grit
sandpaper to the fence to help eliminate
slippage. For this operation a stop block was not
used, each cut was individually eyeballed over
the kerf Achieving perect mirror images on
either side of the blade is just about impOSSible
by eyeballing, but I will be able to correct minor
differences later. For this step, my goal was to
achieve the dosest mirror images that could.
Look at photo 11-14 and visualize the lamination
cut into angled 3/4-inch wide strips; photo U-15
shows those new strips. To create matching
pairs I have turned over every other strip and
squared all the ends. An entirely different
108
THE AT OFSEGMENTED WOODTURN1NG
11-16 A simple sled enables safe and accurate
trimming of smail strips
pattern has emerged. The process of cutting and
reassembling can continue. though the
limitations of standard woodworking
equipment and my difficulty maintaining
accuracy usually make it impractical to continue
beyond one more generation. 1 decided to stop
this particular project at this stage.
The next step was to glue together the pairs
shown in photo 1115. I used the drum sander
with the same tray to sand their mating
surlaces. Because of minor misalignments (from
eyeballing the table saw cuts). } passed the
surfaces through the sander multiple times.
Before each pass I examined the mirror images
and positioned whichever side needed more
trimming. Using spring damps. l glued the pairs
to form ten pieces.
The next challenge was to transform these ten
pieces into segmented staves that would form a
ring -a tricky task because so many elements
must line up while being joined with tight
mitered seams. I cleaned them uP. examined
them. and selected the best suraces for the
outside of the ring. The accuracy of the
intersections up to this stage was acceptable.
11-17 Accuracy of both the saw cuts and the
point to point al ignments is critical.
even though a few tiny mismatches exsted.
After examining the next set of dry glue joints, I
decided to improve the alignments by making a
series of very tiny rip cuts on the sides of the
pieces. My goal was to create identical edges
that would align perfectly with each other. With
the saw fence adjusted to remove just a
smidgen. I examined each piece and selected
whichever side needed trimming. After making
ten trim cuts, Itapped the fence towards the
blade and repeated the process. I did this about
three times, trimming whichever edge seemed
longer. until all the opposing edges
would
align.
Because of the length of the pieces,l cut the
miters on the table saw instead of the miter saW.
These tall, narrow pieces needed an 18 miter
on
each side. I did not like the prospect of pushing
them through the saw blade with just the
saW
fence H a guide, because after cutting
their
to
sides the narrow bottom side would
be
utabl
e
on the table. In addition, l wanted ultra-smoth
cuts that would require little or no sanding.
A simple sled was the answer. I used two
rectangular pieces of MDF: a pIece 1
mch
thick
glued onto a slightly wider piece of
3/4
Ulch.
f
11-18 Successful gl ui ng requires pract|ce and
planning.
ten laminated pieces were slightly thinner than
1 inch, therefore the linch MDF provided an
effective stop block on the sled. This sled
(shown in photo 1116) simply provided a means
of smoothly transporting the segments (staves)
through the blade, achieving a much smoother
cut than would be possible using the table fence
alone. Before making the final cuts, I ripcut a
test strip of linch MDF 2 inches wide, 18 on
each side. 1 made crosscuts from this strip and
assembled a test ring to check the accuracy of
the blade setting.
To eliminate unwanted stave movement, l used
a toggle-style hold down clamp to secure the
pieces to the sled. The clamp also provided a
nice handle. I cut one side of all the pieces,
tapped the fence towards the blade just a
smidgen, and cut the other sides. The sled, along
with a sharp 8o-tooth crosscut blade, did a
remarkable job of producing acceptable ready
to-glue surfaces.
A dry-fit of the ten pieces confirmed the angles
were right on, but what about the alignment of
the intersections? Before dry-fittng the
pieces, Iarranged them side by side to achieve
LINATION TRICKERY
109
the best fit. I secured them with rubber bands
(photo U17) and adjusted the ends to align
everything as closely as possible. With the
cylinder of staves still secured, I sanded both
ends on the disc sander. 1 did this twice with
tiny adjustments each time, striving for the best
alignments. The alignment match between
pieces was not perfect, but they were all very
close. The cylinder end-sanding allowed aU-at
one-time gluing without the need to inspect
point intersections. with the cylinder vertical
(as in photo 1117), a few taps with a hammer
would reposition everything as it had been
during the dry-fit. Before disassembling the dry
fit, I also wrapped a few pieces of masking tape
around the assembly and cut the tape on one
seam, to further ensure that the pieces would
go back together as desired.
1 glued the assembly at one time. If I had glued
pairs together, there would have been a risk
that the final joint points would not align. By
gluing it at one time, perhaps tiny
misalignments would exist, but I knew that I
would not experience a major mismatch within
a seam.
I wanted plenty of working time so Idecided
upon Titebond Extend. lfthe temperature had
been warmer 1 probably would have used
Gorilla Glue or possibly plastic resin glue. The
pieces received a last dusting with canned air.
the hose clamps were adjusted and ready, a drill
with a socket bit for the hose clamps was
nearby, a hammer and a thin piece of scrap
wood for spreading the glue was awaiting.
Photo 11-18 shows the assembly laid out flat and
ready for glue. The masking tape holding the
staves together in alignment made the
clamping very easy. Pre-glue prep is always
smart; wasting glue working time while looking
for tools can lead to problems.
I later incorporated the created cylinder into a
vessel titled CRLIk8.as displayed in
photo 11-19 (next page). In the world of multi
generation laminations. this was a relatively
simple project. Designs that are much more
complex are pOSSible by contmumg the process
of cutting and reassembhng. Extremely precise
110
HE POF bEGMFNTED VOODTUN\NL
11-19 SCRBLED AG8 (10 inches tall) -it's not
as difficult as it appears.
milling and accurate gluing are essential to
success. A tiny compromise early in the project
easily produces more noticeable errors later.
This type of segmented woodturning is very
time-consuming and risky, things can easily go
wrong. For the experienced woodworker
looking for a Challenge, this could be just the
answer, good luckl
Two Ways to Build
Diamon
ds
Diamonds are a girl's best friend, and the shape
is also attractive on the side of a woodtuming.
They are much Simpler to create compared to
multi-generation laminations. One method of
creating diamond shapes involves
extensive
disc sanding, while the other method
relies
on
a
router. The disc sander technique is a
variation
of the feature ring building technique described
in Chapter 7: you surround a diamond-shap
ed
piece of wood with other pieces of wood
use the disc sander to create straight and
smooth gluing surfaces between the
components. Photo 11-20 shows a typical
diamond segment, a leftover from a large ring
of diamonds.
In the photo I have labeled the four sides and
marked with a pen al1 the glue lines that
required disc sanding. To build this diamond:
Using the miter saw Icut the center diamond
shape. For appearances, I wanted the grain
direction of the curly maple to match the grain
direction outside the diamond; notice that
the curly maple grain is vertical.
Two sides of the diamond (#1 and #4) were
lightly disc-sanded, then I glued on thin pieces
of purpleheart to both sides at the same time.
The purple heart pieces and all the other to-be
added pieces were cut slightly longer than
needed, which simplified the gluing steps.
I carefully sanded sides #2 and #3, and sanded
the ends of the previously glued-on purpleheart
pieces flush with the side of the
diamond.
I glued on two more strips of purpleheart to
sides #2 and #3. This process of sanding and
gluing two sides at a time continued until 3
the
components were in place. If you
study
lines in the photo you'll be able to
visuali
ze
the
order of assembly.
With the diamond shape surrounded
. it
was
then only a matter of squaring the
recta
ngle
and
mitering
two edges to create a
segme
nt
fO
ring construction.
The easiest way to make a mistake is to sand
too aggressively, thereby removing too much
M
LNATON TRICKERY
width from the previous layer. , mentioned
that the diamond shown in photo 11-:0 was a
reject. Look closely: the thin strip of maple on
side is thinner than the other similar strips,
a result of too much sanding. A very minor
defect, but because I had made a couple of
extras, I had better choices. A slight variation
of this technique, which does a better job of
hiding the last outside glue lines, is described
in Chapter 15
The second technique is trickier. Photo n-21
shows a very elongated diamond shape on the
lower portion of a vessel. These diamonds are
actually small l/2-inch squares positioned on
edge at a 45 angle. The elongated shape
resulted from shaping the vessel wall at an
acute angle. So, how did I do this using a router?
First, I built the diamonds by sandwiching a 3/8-
inch thick board of holly between two thin
layers of ebony.
I then cut small strips hom the lamination,
passed the stips through a dsander, and
glued two more Mlayers of ebony to the other
sides of the holly strips as seen in photo ~Z.
The goal was to create J/2-inch square strips. I
band-sawed the glued assembly into individual
strips, which I cleaned up and squared with a
little handwork and the drum sander. l then cut
them into about 2-1/2 inch lengths.
I used a very simple jig, shown in photo u-23
(next page), to cut V notches into the ends of
segments. The jig is nothing more a
straight edge clamped to a smooth surface. A
bar clamp holds a segment in place so that a
router bit can cut the notch. The bit has to be a
90 V shape. Hrequires a little fiddling to get
everything centered and at the right depth;
multiple passes with a stop on a plWlge-router
works best. The goal is to create two opposing
notches that will fit around the square piece of
holly/ebony. I try to cut the notches just a tiny
bit too deep, then I disc-sand to remove tiny
iOcrements from the segment ends lUltil a
perfect snug |A eXists. When hghtly sandlllg the
segment ends, address both ends of the notch.
The fit around the diamond piece has to b snug
11-20 A disc sander can create straight, tight
glue joints
11-21 An example of router-created diamond
shaped elements
11 22 To create a dl3nlOnd With a older
surround It With 3 contrJ<tlng wtd.
IJ1
112
THE ART OF SEGMENTED WOOoRNING
\\2j |||n|jcI8mpthe segment before cU\\|n_
the notch.
from end to end. with no gap between the
pointed segent ends.
Photo U-24 shoWS a couple of the diamonds
that I made for the vessel shown in photo U-21.
Essentially. I have embedded a diamond within
a segment. The elongated diamond shape on
the vessel wall is simply the result of an angled
cut throuEh the segment.
In this diamond assembly the holly and ebony
components were oriented with their end
grain perpendicular to the outside of the
vessel. I did this intentionally. i I had oriented
the wood grain parallel to the vessel wall. the
tuing would have exposed fragile short
grain at the upper pOint of the diamond. which
would easily break away and leave a defective
surface. With the end grain oriented
perpendicular. the wood fibers stayed in place.
We lear from our mistakes! The grain of
diamonds used in a vertical portion of a vessel
could instead b oriented to match the other
segment components.
11- After installing the diamond, you can treat
the segment |ike anjother.
Laminated Segments
and Staves (Five Examples)
Peasy way to produce dramatic designs is to
use laminated wood as your raw material The
three tungs displayed are examples of
laminated segment/stave trickery. The tunting
WRENCACS (photo u-2S) Vbuilt with
segments cut from a lated board with the
layers oriented verticall. The lathe shaping
exposed the layers and the pattern of oval
shapes emerged. ToPkAa (photo 16)
resulted from tUring a cyinder of laminated
staves that I offset-mounted on the lathe. I built
CUR(photo U-27) using laminated
segments
with the layers at an angle within the
segments. No curved pieces of wood were used
i any of these constructions.
To demonstrate these techniques I lted
strips of cherry and jarrah (photo u-2B) with a
layer of maple and walnut veneer in between. I
used these laminated boards to cut gglue
together five different assemblies.
Toachieve any kind of effect, I must expose the
layers
within the lamination. Sharper
outside
ring comers. which require deeper
cutting
to
create a round ring. will result in me drama
ti
C
1-2 WAlTRMELON CACUS (21 inches tall) -an
example of vertically oriented segment layers.
+1 26 TECUP 1ALlCE (4 Inches tall) -an
example of off centered laminated staves
LAMINATION TRICKERY 113
11-27 CUrES (29 Inches tall) -an example of
angled layers wlthm a laminated segment.
11q THE ART OF SEGMENTED WOOD1RNING
11
-
28 laminated strips can be
cut many different ways for
dierent effects.
11-29 trom left to right. vertical layers,
an off-centered staved cylinder, and
angled layers.
11-30 The
lamination modified
to reposition layers.
:1-j1 Three different results from the same
laminated strips.
designs. Photo u-29 shows three different rings
made fom the laminated material of photo
u-28 Tring on the left was made from eight
segments with the veneer layers vertical within
the segments. For the ring on the right, I re-cut
the laminated board to orient the veneer at a
45 angle (photo 11-30) before I cut the
segments. I built the ring in the center by
cutting eight staves from the original laminated
material, which then was offset-mounted onto
a waste block. All three rings came from the
same laminated material, but the layers are
oriented very dlfferently.
When turned, the three rings display
completely different images. Photo 11-31 shows
the transformations. The results are intriguing
even though the effort was mmor. For the
designs to emerge umformly the rings must be
11-j1 Two cylinders, constructed dierently from
the same laminated strips.
perfectly round, otherwise obvious differences
wappear. The cylinder in the center
dramatically displays the effect of non-centered
laminations.
The next two examples relate to staved
construction. Photo 11-32 shows two cylinders.
The one on the left has /staves with the
position oftwo woods alternating around the
outside (and inside). The trick is to expose the
layers by turing a ribbed surface; the actual
patter depends upon the width and depth
of
the coves. On the left side of photo 11-39. I sho
the same cylinder after a few minutes of
turning -a very dramatic effect with very
little effort.
The fifth example starts with the cylinder
on
the right of photo 11-32. I have not yet glued
the
cylinder together. but I have confirmed the
fit
My
objective was to create a vertically
ribbed surface wIth oval shapes on the ribs.
One
approach would have been to glue this
flng together and then use carving tools to
create a ribbed surface. However 1 am not
much of a carver -I prefer to do my carving
on the lathe whenever possible. After ail, what
is a lathe but a carving tool? To shape these
individual staves on the lathe, I mounted them
two at a time as shown in photo U-33. Using
double-sided tape. 1 attached two staves to a
third piece of wood. l do not own a small 4-jaw
chuck for my mini-lathe, so instead used a
cup-shaped drive at the headstock end and a
live center at the tailstock. If you look closely
at the tailstock end of the staves. you'll see
that applied an additional piece of tape
across the ends to help keep the staves
together. The middle piece of wood does two
things: its thickness determines the radius of
cut, and the live center can press on it without
forcing the two staves apart.
Before miter-cutting the staves,l had to choose
their width. I drew a sketch, which allowed 1
to predict the results of various turg
radiuses. Photo 11-34 shows a layout of
rectangles, miter angles, and likely turning
profiles. As you can see, the thickness of the
spacer wood affects the turg profile. If the
spacer utoo thick, the radius wlbe too large
(flat) and the oval pattern will not emerge. H
the spacer is too narrow, the radius will be too
small (sharp) the turing action will
remove the mitered sides, leaving nothing to
use as a glue surface.
Photo u-35 displays the results of turning the
pair of staves shown in photo 11-33. The layers
that form the oval are quite thin. any more
turning and they would disappear. At 1 inch
wide, these staves are relatively small. there
would be more leeway with a larger stave.
After turning the staves I smoothed them with
a palm sander, using a little piece of double
sided tape to hold the stave in place during
sanding.
I then covered the outside stave surfaces with
lINATION TRICKERV
"\
11-j Staves can be turned two at a time, with a
spacer i n between.
LBmHBIBG
IBVB
8[BI
H- The spacer thickness determines the
shape of the oval and the profile of the stave
11g Carefully t1rn away the outSIde layers
116 THE ART OF SEGMENTED WOOoRNING
11-37 Befre assembling the staves, attach
contrastfng splines of veneer.
11-38 Before assembling the cyIi nder, first glue
together pairs
3J jQ Completely different results have been
obtained from the same type of matenal
masking tape and attached walnut veneers to
both sides of each stave. Photo 1-37 shows the
gluing of one side. Originally I had planned to
use a single layer of veneer between each pair
of staves, but the veneer was so thin I doubled
the thickness by gluing a piece to both sides.
It's tricky to join 16 staves with rounded
profiles. To attempt it while also inserting
veneer strips would have been foolhardy. To
ensure success not only pre-glued the
veneer,
but also began by gluing together pairs as
shown in Photo 1-)8.
With the eight pairs of staves pOSitioned i a
ring formation, applied a few rubber bands
and wrapped a few revolutions of masking tape
around the cylinder. This allowed me to remove
the bands, cut the tape down one seam, and lay
the staves flat with the open inside seams face
up. I applied glue to the seams and rolled the
form back together, then aligned all the inside
comers to one another and applied many
rubber bands.
The protective masking tape on the outside U
the staves was a big help, but handwork with a
utility knife and a small carving tool was still
required to clean up the seams. An 80-grit
sanding mop sped the dean-up process. Photo
11-39 shows this example on the right, along
with the previously described example. You
might wonder where could such techniques
used? I used this last trick in the construction
of
a tuing titled FOR GADI. in photo 11-g
lamination trickery is a fascinating aspect
of
segmented turning. I have only scratched
the
surface with these five examples. For years.
turners have precisely layered and turned items
such as lamps and rolling pins. Virginia
Dotson
,
a well-known turner in Arizona. has achieved
great success turing laminated assemblies.
Combining some of these techniques
could
produce results that are even more stunning.
Give )some thought and see what
innovat
ions
you can discover.
L|NAT|ONRlCKEkY
\\
FO
QAuD)(18 Inches tal\ ) -an exampl e O!pre turned staves
117
u8 THE A CISEGMENED WOODRNING
11-42 This project requires 40 segents and 40
splines.
A Bowl from a Board
You might be thinking that segmented ting
uses a lot less wood compared to conventional
turg because there is very little hollowing
involved. You would be wrong. In conventional
segmented ring construction, many board feet
of good wood goes into the trash. Many years
ago at an AAW symposium, Mike Shuler, a well
known woodtUI er from Santa Cruz, Califoria,
demonstrated making a bowl with very little
waste. He had advanced a technique based on
construction methods used in the salad bowl
industry. Shuler also credits Dale Nish, in his
classic book Creative Woodturing, for
describing the technique of stacking angled
rings band-sawn fom boards to create a bowl
shape. This is a great example of how sharing
ideas leads to innovation and the whole feld of
woodtUI ing advances one more step. That is
not to say that we should all copy each other's
ideas, there is a big difference between
plagiarism and what Shuler describes as,
"spring-boarding from an existing concept."
With Mike Shuler's encouragement. I will
describe the basics of this ingenious technique. l
offer it so you can use the basic technique and
possibly take another step in a new direction.
Instead of cutting, gluing, and stacking
IndIvidual rings, I will glue together two large
half-nngs and then band-saw angled rings that
llA3 Assembly starts by gluing pairs together
with splines.
will be stacked into a bowl shape. I chose a
board of straight -grained purpleheart about
5-1/2 inches wide by 30 inches long. Using my
miter saw, I cut 40 segments with a 45 angle
on each side (80 times 4.5 equals 360"). I cut the
pointed ends of the segments less than 1/4 inch
wide. Before cutting the segments, I lauted
pieces of maple veneer and thin ebony together
in order to create 20 splines that I Wglue
between every other segment. For the other 20
seams, I cut the same size spline fom solid
holly. These unglued segments and splines are
shown in photo 11-g2.
Sanding segments of this size presents a
challenge because of the length of the glue lines
and the short length of the wood grain. Just a
little bit of sanding heat vwarp the surface.
Actually, just releasing wood tension by cutting
such short, wide segments can result in some
warping. His certainly tempting to glue without
sanding, but I know I can improve the glue lines,
so I sand. First, I changed the paper on my disc
sander. Sharp, fresh sandpaper will reduce time
on the disc and therefore reduce heat build-up.
A
dry-ft of the segments confrmed that my cuts
were right on and that very little sanding
would
be necessary. decided to freehand-sand half of
the seams and then glue those 20 pairs together
with laminated spline in between (photo 11-
-43) To minimize warping, I sanded the
segments in stages, just a little sanlimg tl ll\+wd
LAMINATION TRICKERY
11Q
11-q Gluing cauls apply pressure in
the right direction.
JJ-qy The two half-sections wi tt not be glued together.
by cooling time while I sanded the next. After a
couple of very light sandings. the joints were
ready for glue. I performed the bright light check
on each seam and waited for any sanding heat
to dissipate before gluing each of the 20 pairs.
Never attempt to glue wann segments. the glue
wlset too quickly and you won't have enough
aligment time. Perfect glue joints are V1, one
bad joint and the whole project is a loss. Unlike
other ring-construction where there is an
oppofty to replace a bad ring, with this
technique replacing a single defective ring is not
an option.
I continued to assemble the large ring by joining
pairs to form groups of four and then groups of
eight, with two groups of four lett alone. The
sections of eight required a gluing cul to apply
perpendicular damping pressure. I attached
triangular pieces of MDF with a bead of hot-melt
glue, as shown in photo 11-44. With smaller
segments rubber bands provide adequate
damping pressure, but with segments this large.
I chose small bar damps. Cphoto 11-44, I am
gluing the center seam, the other seams were
previously glued. Before each gluing stage
performed a dry-fit and made minor disc sanding
adjusents to ensure that the ring stayed round
It )critical that the final nng shap b perfectly
Uu,not oval. If any minor adjustments are
necessary I want to spread them across many
seams, I do not want to make a major adjustment
near the end of the construction.
Each half-ring now consisted of two sections of
eight and one section of four. Another dry-fit
followed by a few minor disc sanding
adjustments resulted in two matching halves. To
glue these (photo 11-45), 1 used the half-ring
methoo as described in Chapter 7. I only had lO
deal with three sections and two splines per
side, which made the job relatively easy. After
this step, unlike conventional segment ring
building, I did not join the two halves into a
solid ring. I smoothed the surfaces using the
disc sander and drum sander, placed the last
holly splines in between the halves, and dry
damped them together with a hose damp.
Tocreate a bowl shape from this large flat ring
reqUired cutting angled rings whose top and
bottom diameters matched the diameters of
adjacent rings. There are choices: how narrow
and at what angle should the rings be O\
Narrow cuts will result in more rings and a taller
bl. but Uthe rings are too narrow the wU
may be too thin. resulting in disaster The size
and shape of the angled rings wdetermine
and limit your bl shape options. Usmg U
120 THE ART OF SEGMENED WOUNINU
11-46 Carefully plan the location of band-saw
cuts
J1q] Cut carefully: thcwhole project depends
upon follOWing the lines.
: q8 Hcrcb thcresuItof band >mwng.
actual dimensions of my ring, I have drawn the
layout I used for this bowl in photo 11-qTo be
on the safe side, I allowed almost 18inch for
saw-kerf loss. To achieve a slight curve I drew
the rings a little wider, wUa slightly steeper
angle than necessary toward the inside (the
bottom of the bowl). This layout resulted in less
height because fewer rings could be cut -not a
big deal. the bowl is tall enough. If your goal W
maximum vessel height with minimum waste,
then you must lay out cuts that are steeper and
coser together. In any layout, make sure yow
diameters line up to provide you with adequate
wthickness. It's tempting to cut very U
rings and thus create a taller vessel, but for me
the risk is not worth the extra height. M
addition, if you cut very Urings you eliminate
room for error and yOU limit the bwl profile to
almost a straight line. Straight lines are not
necessarily bad, but I wanted a slight curve.
After deciding on my layout I used a compass to
draw concentric circles on the two half-rings
that were still clamped together. I separated the
two halves and, using a t/2-inch fne-toothed
blade, proceeded to band-saw the cuts. PI
made successive cuts, I adjusted the blade angle
to achieve the results as drawn. In photo 11-47, I
am carefully making a cut; a lack of focus here
and the whole project is lost The half-rings are
shown in photo 11-48.
The remainder of the project was
straightforward. I glued the half-rings
together
with a small piece of holly in between, built
ebony and holly rings for the top and bottom,
stacked the layers together one at a time, and
turned the bowl shape, titled Bcxxv
$wxt,
that you can see in photo 11-49.
The technique of using bandsawn angled
rings
has been around for many years, but as far as
know, with the exception of Mike Shuler,
whose
turnings are much more complex than
thlS
example, few turners have experimented
Wlth
Itor tried to take it to a new level. As I write
these words, images of combining
other
lamination tricks in combination with
thiS
technique enter my head. I hope yow
own
br:ul
l
Uspinning With ideas.
LINA1ION TRICKERV
11-49 1LACUJWUL (6-112 inchCstall) -a vjCficiCn!u>eoflumbr.
122 THE ART OF SEGMENTED WOODTURNING
12-01 QUIT 1OW (5-112 inches tall) -the result of productIon turning technIques
1 2.
Production
Turni ng
One does not normally associate segmented
tung with production turg. Most
segmented work tends to be one of a kind. built
one piece at a time. However, occasionally I do a
production run of half a dozen turings, to meet
a Christmas gift deadline or perhaps to quickly
re-stock a gallery with lower-priced items.
Such was the case with the six bowls that I
profle in this chapter. There are many
advantages to constructing multiples of the
same design, even though the work can be quite
tedious. The design for Uproduction rn is a
small open bowl shape consisting of 29 layers or
rings. In the drawing that follows, I did not
bother to number the veneer layers, but they do
account for 72 pieces of wood. I made the top
fow rings and the bottom four rings using 18

12 02 It al l starts with a drawing.

PRODUCION TURNING
12-03 Without a sketch showing the color
arrangement, gluing would be vqdifficult.
12]
segments each, while the remaining rings
consisted of 36 segments each -a total of O1
pieces of wood counting the bottom plug.
Considering I made six bowls, that adds up to
more than 5,000 small segments. Hwas a lot of
work, but not as much as you might U
Photo 12-01 displays the six completed bowls.
The first step is to create the design With a
project such as this, I needed two designs: the
bowl shape and ring dimensions (photo 12-02),
and the arrangement of the brick-laid colored
segments in each ring design (photo 12-03) I
124 THE Au OF SEGMENTED WODTlllNING
drew the shape design Just as any other. though
I dId not bother to overlay rectangles to
determine the segment widths because they
were almost all the same. The rings were 8
built hom /b- inch wide strips. except for a few
WIder rings near the top and bottom, which
determined by using my calipers on the
draWIng. Using the blueprint information I
created a cutting list with the segment length
information.
chose 36 segments per ring so I could repeat a
pattern six times around the vessel. The actual
layout of the colored segments required a little
doodling. I used the same layout for six
vessels, but substituted different woods.
creating three different combinations. Photo
w-o](previous page) displays the arrangement
of segments. To calculate the number of each
type of segment per ring, it was a matter of
counting the different colored rectangles in the
drawing and then multiplying by 6 (the layout
drawing is only 1/6 ofthe circumference). It is
very easy to make a mistake while cutting and
gluing so many different colored segments.
Instead of "measure twice, cut once, it's more
like "check three times, glue once." I made a
checklist for each of the three different designs
that included the number of segments of each
type of wood per ring. also made three
different layout drawings to guide assembly.
Even with these aids. it is easy to make an
arrangement error. You have to stay on your toes.
I constructed the top and bottom portions of the
bowls using standard techniques. building the
ebony and maple rings from 3/4-inch material
and then splitting them into two rings. thus
saving a little ring-building time. Constructing
all the multicolored rings consumes the most
time. Sanding so many individual segment-ends
was just too much trouble. so my plan was to
cut perfect miters and sand only the Iight
colored to light-colored jOints. which still
accounted for a lot of sanding. The design called
for nngs about 1/4 inch taU. My plan was to
construct nngs hom 3/4-inch material and then
split each one into two rings. thereby cutting
nng assembling time in half. Each 3/4-inch
thick nng would produce two Idenhcal l/4inch
12-04 Three sets of 3/4-inch tall rings (same
paTern, but djfferent woods).
taU rings, therefore 21 colored rings would
produce two identical bowls. I milled the boards
to a consistent thickness, ripped strips for the
widest segments first, and proceeded to glue
using Titebond, rubber bands, and the gluing jig
shown in Chapter 7 (photo 7-16). This required a
couple of days of monotonous cutting, sanding.
and gluing. I sanded only the ends of the light
colored to light-colored seams, which I rub-joint
glued separately, before using the rubber band
gluing jig (Chapter 7) to assemble the rings. I
used the half-ring method of gluing (Chapter 7)
Gang-cutting two segments at a time helped to
speed the process and I maintained ring
identity by labeling them with a marking J
Photo 12-04 shows sixty-three, 3/4-inch tall,
colored rings (twenty-one rings of each color
combination). the six bases, and three ebony
rings, which Vbecome the six top rings.
Uwas then time to split the colored tings.
thereby creatng six sets of 21 rings each. The
options were to part them on the lathe, or to
partially part them followed by band-sawmg. or
(as I decided) to cut them on the table saw. I
flattened both surfaces of the nngs on the disc
sander and smoothed the outside circumference
slightly. so the rings would easily rotate ) a
cutting jig At first glance the table-saw
technique might appear risky, but I
12Q_ Splitting the rings using a table-saw jig
creates six different sets.
accomplished it without any problems or safety
compromises. I used a jig (photo 12-05). which I
normally use to cut spline slots into picture
fame comers. Instead of sliding the jig along
the saw fence I clamped it in place. centered
and with the blade raised just high enough to
cut through the rings. Keeping my hands well
above the blade location. I lowered each ring
onto the blade and. while maintaining solid
ring contact with the side of the jig. rotated it
into the blade's rotation. Maintaining a firm
grip prevented the blade from grabbing a ring.
Tis operation reduced hours of lathe-parting
work to just minutes. while producing smooth
surfaces and consistent thicknesses. I would not
recL end this method for splitting smaller
rings because your fingers would be too dose to
the saw blade, creating a risky, unsafe situation.
The result. as shown in photo 12-06. was six
stacks of thin rings, two sets of each color
combination. One surface of each ring was
already smooth and flat from the previous disc
sanding.
Toassemble these rings into bowl shapes, I
divided each stack ohings into thirds; photo
12-02 shows the ring numbering. I attached the
#8 rings and the #14 rings onto MOF waste
blocks using a continuous bead of hot-melt
oOUcCNUkNINC
\2C Many layers await assembly.
glue. I also attached the top rings of ebony to
waste blocks so that I could add additional
colored rings to the upper portion of the bowls.
assortment of these types of MOF waste
block circles is shown in Chapter
1
(photo 1-2). 1
lathe-mounted and flattened the gluing
surfaces of these 18 mounted rings using a four
jaw chuck. Next I attached the adjoining 18
rings (# ,#13. and #20) to another set of waste
blocks, centered and attached with six short
beads of hot-melt glue. By using MOF circles
that are similar in diameter to the rings,
centering can be eyeballed. Just a quick touch
up on the disc sander and they were ready for
gluing to their mates. I used the carriage
bolt/wing nut technique as described in
Chapter to glue all these rings together. Each
joint consisted of a lathe-flattened surface on
one side of the seam and a disc sander-flattened
surface on the other side. After each jOi nt cured,
I used a utility knife to remove its waste block.
scraped the MOF circle clean of hot melt. and if
necessary. reduced its diameter to accommodate
the next ring. The advantage of the caniage bolt
technique is accurate centering. The rotational
alignment of each ring is another story. The
individual segments were quite small (average
.1 UOlongl, which made it easy to eyeball thl
segment centers of one ring to the segment
126
THE ART OF SEGMEND WODTNING
1z-O] This l a section of one bowl being
assembled, one layer at a time.
seams of thE nExt ring. However, with so many
rings it is difficult to accurately maintain the
designed patter. The tnck is to focus not only
on the adjoining nng, but also on the first nng
in thE stack. If you focus only on the adjoining
ring, then cumulative errors can produce an
obvious distortion in the pattern. Photo 12-07
shows onE of the "8 rings, lathe-mounted with
successive rings added. In the photo. the outSidE
12-08 All six bowls are in various stages of
construction.
waste block is about to be removed and the
smallest ring is about to be prepared for the
next ring.
This probably seems like a long procEss, but it
actually went quite quickly. The next photo
(12-08) shows the progress. In the upper left
corner, three of the bowls are starting to take
shape: I have glued rings #8 through #3 onto
the base assembly and have added the mid
section, rings #14 through #9. to ring #8. If you
look closely at the center row of rings, you will
see that I have mounted the top ring of ebony
with its veneer layers and maple rings to
waste
blocks. Starting with ring #21, the top portion of
the bowls is taking shape. l continued thE
process until I had joined all the rings into O
bowl forms.
Once I completed the gluing work the
vesse
ls
were
finish-turned,
sanded, and oiled. A
look
at
photo 12-09 shows one of the bowls ready for
sanding sealer, while the other bowl awruts
final turning. Note the two bottoms: the
.
unfinished
bowl still has a wide
base,
proVd
m
plenty of support during the shaping
.
whLie
I
have cut away the waste block on the
fmlsh
!RODUCION TURNING
"7
IlOg To maintain stability, shape the bottom of the bowls last.
bin order to shape its lower section.
The bowls each received six coats of finish (one
coat of sanding sealer, four coats of oil, and one
coat of buffed w) while they were stll
mounted on the base waste block. Reverse
mounting was done as shown in Chapter 9,
photo .2,so could finish and sign the
bottoms.
A big projec -maybe. It depends upon your
perspective. One thing is for sure, making six
bowls one at a time. with 900 segments each,
would have taken probably three times longer.
The basic techniques used to build these bowls
can be adapted to many other designs, although
am not sure I would enjoy constructing larger
tUrngs in this manner. Frankly, would prefer
to design and build one-of-a-knd turnings, but
this type of activity does help pay the bills.
lfyou ever decide to attempt such a project:
Check three times before you glue anything
together.
00 not allow yourself to lose focus. which
could lead to making a mistake or worse,
Jeopardizing safety.
Avoid altering your saw settings until you are
sure that you no longer need more pieces
similarly cut. During this project I messed up
the patter on one bowl by accidentally
misaJigning one ring with another. Fortunately,
I caught the mistake before adding more rings. I
could not save the ring, but I was able to cut
another 6segments because my miter saw was
stll correctly set.
Spend plenty of time planning the project,
looking for shortcuts, and inventing techniques
to accomplish the job as efciently as possible.
128 THE AR Ot SEGMENTED WODTURNING
'3 O6 OABOWL t9 (S inches tall) -reorientation of two halves creates an oval.
13
Mi scel l aneous
Tri cks
Inthis chapter I will describe a few mOle ways
to embellish a segmented turning. along witha
feWadditional methods of construction. Once
you have mastered basic construction
techniques you are probably going to want to
continue to challenge yourself. Some Othese
next techniques are risky -after many hours
lor
days) of work. disaster 1strike and all can
be lo There are many rds for
lccomplishment, while a small mistake can
wipe out a large investment Utime and wood.
:casional disappointment is part of the game.
Oval Bowls
Jshapes are a nice deviation fom the normal
round tand they not dificult to build.
ipdal lathe equipment is not necessary, but lke
0Osd vzkcarefl planning is
5al. Vy simply, the technique requires
bu a cnventional taller-than-wide bowl
,hap, C g it vertcall into two halves, then
joining the two at their top rims. Photo 13-01
dplays a tured bowl shape. The patter on the
|oopposite sides of the vessel Vfonn a
completed patter once I separate and re-joi n the
halves. Note the ebony segments that will
later become the top edge of the finished oval
bowl. I carellaid out the shape of the bowl to
duplicate one end of an oval with this particular
widthand length. I vlater carve feet from the
rm that you see protruding from the outside of
the bowl, removing most oIthe Oin the
proess. While turng the vessel it was
i
mportant to align the exterior profile on either
sqe
of the rim, so that a smooth surface could be
created after caling away most of the rim.
Consistent wall thickness 1S \3\5 important.
but even more so 1)this pro
ject. When I split
MISCELlANEOUS 1RlCKS )2g
i-o: Notice the small dri||edho|esthat can be
usedto check wall thickness.
the vessel apart any inconsistency wDvery
noticeable along the exposed edge. There are
many methods and tools for measuring wall
thickess, but in this case I can use an
unconntional method. After determining the
location Othe cut line. I simply drilled a series
of tiny holes. and used the depth gauge on my
caliper. The holes did noharm because they
disappeared when I cut the bowl in half. If you
look closely at the dark ebony segments. you
might be able to see my depthgauge holes.
After completing all the turng and sanding
that could be done. I cut of the waste blok.
drilled a few more holes at the bottom, and
placed tape over the inside of the holes. I
reverse-mounted Ubowl and rounded the
bottom using the caliper depth gauge against
the tape to check wall thickness. Once you have
the vessel apart Uis too late to adjust w
thickness. except by sanding.
In photo 13-02 {nepagel. for stability. I have
attached the bowl to a sm:ll piece of MDF with
double-sided tape. and also have applied
lllaSklllf tape to the desned cut Ime. Theedge
of the tape j\vle5 a Illee gUide for the saw
130 THE AlT OF SFCNNFDWoUkNINC
1j-Oz Mounting the bowl on a flat board
provides stabilt during cutting.
1j-o To maintain this parICular pattern. one
layer is removed from one side.
blade. Using a fine-toothed 1/2-inch blade. lsplit
the bowl into two halves. and sanded the cuts
(hoto 13.03) before removing the halves from
the NF
The designed pattern of colored segments
contamed a smgle center st1p. therefore I
needed to remove one of the half-rings from
one of the half-sectIOns. I aached one Side to
the MF shown I ghoto13-04 and cut off
one layer, then sanded the cut lme Wh
i
le the
pl was still attache.
The two ha^ebwere almost ready to Dglued
together, but hrst I had to cut notches 1 the ]I
1j-O The cuts on the half-sections can bdisc
sanded while still attached to the board
:-o Notches and rubber bands prOVide an easy
clamping method
so lcould apply rubber bands as sh1
photo 13-OS. You can see the precise alignment
of the two halves.
Using a combinatIOn of burr-style power
CalVIng tools. small drill-mounted sandtng
diSCS, a palm sander. and hand sanding, J
shaped the feet. cleaned up the glue seam, and
sanded the top edge. The result [ght13-06
page :z)1 an oval bowl tured on a
convenhonal lathe. This technique IS
probably
the Simplest example of transformmg one
lathe-formed shape into another. The
goal
was
to produce an oval shape. try to Imagine the
other shapes that are poSSible by first st.11nt
ActUmOUST8ICkS
JjC] XO 'NROUR (16-inch diameter donuts) a precise fit is required.
with a different shape. Think about the results
of cutting an object into more than two pieces.
The concept of transforming objects by cutting
and reassembling presents endless possibilities.
Stephen Hogbin, the well-known Canadian
woodturner, has created many innovative
pieces of art by realigning lathe-formed pars. If
this type of work intrigues you, then check out
Hogbin's work, it will open your eyes.
Mitering Complete Rings
XOK'NROL (photo 13-07) required assembling
tapered rings into a donut shape. The title refers
to the rocking motion that results from tilting
the piece on edge and releasing it on a smooth
surface: because of its shape and balance, it w
rock and roll (or almost three minutes. The
construction of the front donut is unusual.
I nstead of stacking rings in the conventional
sense, mitered indiVidual rings and treated
them as segments. I considered several methods
for accomplishing this task, including a jig with
a router, followed by disc sanding. !had done
that in order to build the outside donut ring in
MYR MONS (photo 8-03). A faster, more
accurate method is to miter the rings on the
drum sander.
To demonstrate this unique technique I w
build another ROCK 'NROlR style turning. My
design for this piece is two interlocking donuts.
each with a 12-inch outside diameter and 4-inch
inside diameter (ROCK

NROlR has 16-inch

diameter donuts). The goal is to fit each donut
through the hole of the other donut. l must
build one of the donuts in two halves, which
can be joined towards the end of the project. My
solution to this challenge was to miter small
rings into a donut shape. To start, I assembled
fifty-five 4-1/2 inch diameter rings from
segments that were slightly less than an lOch
tall. I dlscsanded both sides of the rings to
1
3
2
THPOtSEGMENTED WOODTUJNING
13-08 Rmgs are tapered using a drum sander and transport trays with a shim under one edge.
create smooth sUlfaces. Since they were going
to be further sanded, perfect glue surfaces were
not necessary. Then !used a band sa with a re
saw guide to split each ring, thus creating ))o
rings about -4 inch tall. Hthe rings had been
any smaller, for safety's sake I probably would
have parted them on the lathe instead of on the
band saw. Toconvert these small rings into
segments required milling each side to a
specific angle.
|photo 13-08 I have attached a 1/8inch shim
to the underside of one edge of 3/4-inch
particleboard transport trays. and have
attached the rings with their smooth sides
down using a few beads of hot-melt glue. !
passed the trays through the drum sander
enough times to erase all the band saw marks.
The shims made one side of each ring 1/8inch
thmner than the other Side. !was careful to
remove mmimum wood from the thick Side.
!removed all the rings and scraped the trays
clean of hot-melt residue. The ultimate goal was
to creat nngs whose thick and thin sides
would estabhsh the correct angle. I had no
means of duectly measunng the miter angle
associated with the rings, but I could accurately
measure the two thicknesses. I built the rings
4.5 inches in diameter in order to have a little
room for final shaping. My initial goal for the
outside diameter of the donut was 12.25 inches,
with an inside diameter of 3.75 inch. Using these
diameters, I calculated that the circumferences
were 38.5 inches and 11.785 inches respectively.
My plan was to use 98 rings to create the donut.
so dividing the circumferences by 98 meant
that the outside and inside thicknesses of each
ring needed to be .392 inch and .12 inch
respectively. This is a diference of .273 inch, so I
increased the thickness of the tray shims to .273
inch. ( glued a second shim to the first. creatlOg
a slightly oversized thickness, then I passed the
upside-down trays through the drum sander
mU the difference between the shimmed
side
and unshimmed side was .27 inch. did not
expect to achieve a set of 98 perfectly mitered
rings, but to achieve a useable set. perfec1ion
had to be the goal.
After establishing the thicker shim. remstalled
all the nngs Wth their tapered
Sides
factng
down. I could have done all the
sandmg
on
Ju"t
on( Oof each nng. but I wanted an equal
angle
on either side. After multiple passes
through
the sander. my measurements
indicated that the shims were a smidgen too
thick. I applied two layers of masking tape
under the un-shimmed side. After another pass
through the sander I removed four rings.
stacked
them. and measured their total
thickness inside and outside. Measuring four
thickness and then dividing by four is more
accurate than measuring just one. The
measurements were close enough. so I removed
all the rings from the trays. The rings, miter
sanded on both sides, are shown in photo 13-09.
The red slash marks, which I made before
removing the rings fom the trays, indicate the
thickest portion of each ring. Before sandin.g I
aligned half of the rings on the trays with a
segment seam positioned to the outside, and
the other half with a segment center to the
outside. This was important because it allowed
me to assemble the rings in a brick-laying
pattem without any loss of angle.
Assembling 98 mitered rings into the desired
donut shape was a multi-step process. Using
spring clamps I first glued together pairs. The
rings were quite fragile (.12 inch thick on the
thin side) and a little flexible, so to ensure
adequate damping pressure all around the glue
joint I used lots of damps. as shown in photo
13-10. hung the rings as shown in order to
minimize glue squeeze-out on the unglued
surfaces.
Next. very lightly disc-sanded the next set of
opposing suraces to remove any glue
contamination. and assembled groups of four
rings. By caiipering the sectional circumference
of each group of four, deterined that the
outside length was slightly short, meaning that
98 rings would produce a diameter less that
12.2jinch. The solution was what woodworkers
call a design change; instead of 98 rings, my
donut would now contain 102 rings. The
important lesson here is to measure early in the
gluing process so you are not surprised when it
utoo late to modify your plan. I would have
been out of luck had I not made a few extra
nngs at the beginning. I proceeded to glue up
MISCEUANECUS TRICKS
13-09 Accuracy of the tapers depends upon
minuscule shim 8dj|stments.
'33
13-10 Gluing the ing together starts by gluing
pairs.
13-11 This is an e"ample of creative cl3mplOg.
134
THE ART OF SEGMENED WOOD1N1NG
13-12 A large jam-chuck provides access to the
inner surface.
13-'3 An inner jam-chuck allows the outer
surface to be worked upn.
13 '4 NTEC8ATION (u-inch diameter donuts)
d the wood grain direction l> conSistent.
groups of eight rings. The groups of four were
stiff and solid so I was able to use just two
damps, one on the inside and one on the
outside. I continued to join rings in this fashion
until I had assembled four quarer-sections,
which I then joined into two half-sections as
shown in photo 13-l (previous page). Creative
clamping is often necessary: I used gluing cauls
attached with hot-melt, and rubber bands.
Now the two halves were just about ready for
turning. I touched them up on the disc sander to
create a good joint and temporarily joined them
with double-sided tape_ The remainder of this
project was relatively simple and I am not going
to outline every detail. I constructed a second
ring with the layers conventionally oriented,
then used large outside and inside jam-chucks
made from layers of MF to turn the two rings
(photos 13-12 and photo 13-13).
I split the first ring apart at the taped joint and
checked how it fit around the second ring.
Careful measuring of the corresponding
diameters and the use of a 4-inch diameter
template helped create the desired fit. Before
installing the two half-donuts through the
other donut, I applied multiple coats of oil to all
three components. Then I carefully glued the
two halves in place, did a little handwork to
dean up the tinal seam, and applied finish oil
The fnished turning, IND681ON,is shown in
photo 13-1.
Building the first donut in this project is a good
example U1unconventional problem solving:
the challenge was how to miter completed
rings. For me, one of the biggest joys of
segmented work is inventing techniques. As
your designs become more complex, that old
expression. "Necessity is the mother of
invention," certainly applies.
This technique tor mitering complete rings U
not restricted to building donut shapes. Another
example is a turning caUed GRWH, shown in
photo 13-15. In this piece the rings vary in
diameter and the miter angles gradually change
to create the desired curvature. Assembly of thiS
shape was also different: the nngs are not only
tapered, but their outSIde and inSide surf.:S
MLSCELLANEOUS TRCKS 13
5
'3-16 TANWI WOD(16 inches diameter) -inserting round elements can tricky.
13'15 Qxow(27 inches tall ) -another
example of uSing tapered complete nngs.
were also indivjdually tured and finished
before assembly. Gluing the rings together
would have unavojdably resulted in glue
squeeze-out, something did not want to have
to clean out of all those little valleys. My
solution was to attach each ring indi vidually to
its neighbor with small wood screws. Hwas a
time-consuming chore, but preferable to
dealing with glue squeeze-out.
Instaling Round Designs
As segmented turers we deal with things in
the round. but most of the components that we
assemble have straight edges. This is not always
the case. Instead of building a feature ring to
decorate a turning. objects can be modifed by
cutting openings and installing round
decorative plugs. The vessel TA1 N WI WOD
(photo 13-16) is one such example. Its top
surface contains three Indian faces that I
constructed as plugs and then inserted into
turned holes. I have experimented with router
jigs for makng round holes. but to achieve the
best results the hole needs to be slightly
tapered, not easy to rout into a cured surface.
Cuttmg the holes on the lathe IS the altertive
and it can Da tncky and time-consuming job.
136 TH AT OF SEGMENTED WOODTIlRNING
1317 Off-center turning requires a large
counterbalanced disc.
1j1y This is not a time for aggressive wood
removal.
13-18 Secure and accurate positioning is
required.
T WI WOD is appmxtely 16 inches in
diameter, and offset-mounting the top portion
was necessary to cut the openings. I used a 30-
inch diameter disk of I-inch MDF as my
mounting surface (photo 13-17). Attached to the
MDF you can see a counterweight in the form of
a heavy ginch faceplate on one side and 1!
sloped pieces of MDF on the other side. The top
portion of the turning was temporarily glued to
a 16-inch circle of MDF using hot-melt and then,
using a single centered screw, attached this
assembly to the center sloped piece of MDF (you
L see the screw hole in photo 13-17). The angle
of the sloped pieces positioned the curved
section of the tung at a right angle to the
lathe spindle. To improve the attachment of the
assembly, in addition to the center screw
applied a few beads of hotmelt glue to the
underSide seams. Vphoto 13-18 you can see the
centered screw under the right side of the tool
rest. This screw not only held the assembly in
place, but also enabled me to rotate the
assembly to the next position, Aer ftting a
plug I only had to loosen the screw, cut the hot
melt beads, and rotate to the next opening
position. Photo 13-18 shows the start of the first
hole. Photo 13-19 shows the assembly spinning
MISCElEOUS TRICKS
13-20 Multiple test-fitting attempts are usually necessary.
on the lathe while I cut opening (I trying
to not block the caera vew), and photo 1]-2O
shows the second opening ready for its plug. !
ft the plugs Udescribed in ChaQter ].As you
can imagine, a lot of work went into
constructing the ting prior to this particular
step. l really did not want to see anything go
wrong and HWa relief to successfully
complete the operation
ADeasier method of inserting plugs into a
vessel surace is described towards the end of
the next chapter.
13
7
138 THE ART OF SEGMENT1 WOOOTURNING
q Cq
LOSAOS H SA1rU1rBOXE (36 Inches tall( -can you find the Icosahedrons)
14
Bui l di ng a n
Icosahedron
Qyou are like most people. you are probably
saying to yourself, "What the heck is an
icosahedron?" Wen, believe it or not, there have
been many books written on the subject. An
icosahedron is one of fve Piatonlc solids, which
are regular threedimensional solids
constructed fom regular and identical two
dimensional polygons. The most familiar
platonic soUd ute rube, WhICh has siX
identical square sides. The icosahedron consists
of 20 identical equilateral triangles. My focus
here is to demonstrate a construction process.
At the time I did not know the form's name nor
that its orign went back to the time of the
Greek scholar Plato. but I constrcted my frst
2o-faceted sphere in 1980 as a learning toy for
my infant son, Andy. This was long before I
became interested in lathe turning. I just
thought it was a clever wodworking project I
recall seeing a plastic version in a toy store and
thinking that I could build one of wood. I later
discovered that the form is caled an
icosahedron. Somehow that toy has survived
the play of two kids and numerous family pets;
I show it to you in photo 14-01.
Years later it occurred to me that I could tum
thi s type of form into a spherical shape and use
it in my wood Bprojects. Near the front of this
book there U a photo of AuO':Gmw (photo
1.01). I attached 12 trumpet shapes to a sphere to
create this tung. The icosahedron's
configuration of 20 triangles creates 12
intersections of ytriangular points, thus the
decision to attach 12 shapes to the sphere. photo
14-02 shows a close-up interior view of this 20-
sided sphere. Another example of using the
icosahedron spherical design is TJu U
(photo 14-03); again, I attached 12 shapes to a
centrally poSitioned sphere.
BUIWING AN ICOSAHEDRON
14-01 AwO:s1OY(9 In. dia.) -my first
icosahedron.
'39
14-02 AUO: GADWi nterior view showing an
icosahedron-turned sphere.
14-03 1HJUULOanother example of an
icosahedron-turned sphere.
140 THE Al OF SEGMENTED WOODINING
1qO An icosahedron is composed of 20 mitered
triangles.
The use of this spherical design is not limited to
anchoring other elements. I have also built
vessel forms and many small boxes using this
type of constructon. COS, pictured in photo
14-04 (page 138) contains a fee-spinning
icosahedron at its center and the small boxes in
the foreground are also constructed fom
mitered triangles. lack Cox's book, Beond Basic
Turing (Linden Publishing), offers several
projects using Platonic solids. I only wish that I
had discovered this fascinating book years ago
when it w first published.
Te key to successfully constructing this type of
sphere is two fold: the miters have to be precise,
and the _Iuin_caDgin_operation must b
accurate. I have found that an all-at-one-time
gluing technique works best. To attempt gluing
20 mitered triangles together one at a time
presents too many opportunities for cumulative
errors. Gluing all 20 triangles at the same time
is no easy task either, but I have successfuJly
done it many times.
The first questions that arise are: what miter
angle to use on the triangles, and what si2e to
cut them? By trial and error, found the miter
angle b approximately 21. I later discovered
that the exact angle is 20.90S. The approximate
finished diameter of a sphere built in this
manner will be less than twice a triangle side
length. This Is a rough estimate and does not
14-06 Triangle edges (an be mitered using a
table saw jig with a hold-down damp ..
include any consideration for splines placed
between the triangles. Another factor is the
thickness of the wood. For spheres up to about 8
inches in diameter, 3'4-inch thick material is
adequate. If you attempt a larger shap, the
\atwe of the turned sphere will likely
require thicker wood.
Here is my method for building a 20-sided
sphere using mitered triangles:
The first step is to cut the triangles. I have done
this several ways and have determined that a
two-step approach works best. Set your miter
saw (or table saw) at O

and cut three triangles

fom scrap material. Place them together and
check to see dthey form a straight 180" line, and
adjust yu saw until they do. Then, cut six
triangles. clamp them into a ring, and check the
accuracy again. It helps to sand off the inside
points a little, to avoid the chaUenge of aligning
six pOints.
After you have adjusted yow saw, cut 20
triangles (photo 14-05). These are not
compound-miter cuts. but the triangle sides
must be equal in length. If the lengths are all
equal. then the angles have to be equal. Another
quick method of checking the accuracy is to
select random pairs of triangles and position
them on top of each other. Rotate them and
examine the match of the three Sides. they
should match every which way.
14-07
Masking tape holds pieces i n alignment.
cut 15 extra triangles the same size from MOF,
to use later as test pieces (photo 14-05).
The easiest and most accurate method of
mitering the triangle sides is to use a small sled
on the table saw. I have used a larger sled such
as the one for the compound miters in Chapter
1L but in this situaton the smaller sled works
better. Photo 14-06 shows my device. It is just a
piece of MDF with two angled pieces of MDF
mounted on top to create a bC`trap for the
tiangles, and a toggle-style damp to securely
lock the triangles in place. The 60

trap must be
dead-on. Cut the two angled pieces of MDF with
the same setting that you used to cut the
mangles.
The mathematicians say the miter angle is
21.905
.
, so that is the ideal blade angle to set.
Make your best guess setting this blade angle
and cut just one side of fve test pieces. Before
making all fve cuts, confirm the accuracy of the
jig trap by making a few cuts that only remove
a portion of the triangle edge and examine the
alignment of these cuts with the top edge of the
tiangle. Is the cut parallel to the side of the
tnangle? After cutting one side fom five pieces,
adjust your fence towards the blade just a
smidgen
and cut one more side fom the same
fve pieces of scrap.
Sand
the paint off the comer of the miter cuts.
ThIs Wlll ease the task of fitting them together.
BUILDING AN ICOSAHEDRON
14-08 Fold the pieces and double check the
accurac of the fit.
Place the five pieces on a flat surface and tape
them together as shown in the bottom of photo
14-07. Roll the pieces together to form a cone
shape and apply one more piece of tape to hold
them together, as shown on the left side of
photo 14-08. Closely examine the joints. Close is
not good enough, they have to be perfect. Adjust
the saw as necessary and test again until you
have achieved the proper ft. To confirm the ft
cut ten identical pieces, assemble them into a
straight line as shown in the top of photo 14-07,
and roll them into a ring as shown on the right
of photo 14-08.
You can now proceed to cut the good stuff. Cut
only one side from each tiangle, followed by a
slight adjustment of the fence towards the
blade before cutting the next side, and so forth.
The exact amount of wood removed from each
triangle side makes little difference. If the
cutting-sled trap-angle is accurate, then one cut
effectvely shortens all three sides equally. How
does one cut shorten three sides? Look at a
triangle: one cut shortens two sides and creates
a new, shorter third side. The reason for
adjusting the fence inward after each series of
cuts is just to make sure that you make a
complete cut and don't leave a flat. uncut spot
at the top of the side.
Precise assembly of 20 triangles With 30
matchmg SIdes and 12 perfectly matched
intersectIOns of 5 pomts is well beyond my
1
4
2 1POf SECMENTED WOODTURNINC
14-09 This can be folded into a sphere.
'4-10 Gluing requires quickness and maybe a
little luck.
1q:: The s|xhose damps have been remstalled
for ease of viewing
ability, am not even tempted to try it. Sure, the
2o pieces can be glued together, but tiny
mismatches at the point intersections will ruin
the appearance. By installing a spline at each
seam, !can dramatically improve my chances of
success. The splines will create openings at each
of the 12 intersections of points, thus hiding
minor misalignments. The the splines can also
dilute any inaccwacy in the final fit. Ifthere's a
tiny gap between two triangle miters, then
adding a spline should divide that gap into W
smaller gaps and reduce its likelihood of
detection. The splines do add another element
to the gluing challenge. The thickness and
choice of spline material is completely up to
you, the only requirement is that the splines
have parallel sides. U the example that I am
describing, I laminated fve narrow strips of
wod -two of wenge, two of maple, and oneal
purpleheart A single strip just as easily coud
have been used.
The biggest challenge is the actual gluig
operation. requires taping the entire assembly
together as shown in photo 14-09. This photo
shows that all the wood grain in all of the
triangles is oriented the same horizontal
direction. While the splines do present a small
compromise, there are no perpendicular
alignments. The least desirable joints are sat
an angle to each other. Later, when I insert
plugs at the intersections, the lengths of these
joints will b reduced. After taking this photo I
applied a second layer of duet tape to the
assembly to improve the holding Jwer
(photo
14-10). This assembl was very floppy and
difficult to handle. To flip it over, I placed
another board on top and flipped both layers
with the tapd pieces in between.
Now for the real |,you need to
perform a d
fit with hose clamps. I highly recommend
enlisting the help of another pair of hands. can
maneuver
small sizes without help, but
spher
es
of this size are to cumbersome to control
alon
e.
helps to place the pieces into a
circular
hollo
w
such as a rough-tued bowl during
dry_tit
ling
ad gluing - the circular rim acts as
anothe
r
pair of hands. The arrangement of triangles
determines the pas.honing of the hose camp
\Xdifferent alignments accommodate the
gI8ccmcD\of a clamp around the form's
circumference. These six circumferences avoid
thcpointed intersections and provide the most
efecive clamping pressure. I took a photo of
tbcglued form, but unfortunately. squeezed-out
foam
from Gorilla Glue obscured the view.
)v to warn you that gluing a sphere of this
size Ua very messy afair. To give you a better
view, J reinstalled the clamps around the
cleaned up sphere (photo 14-11).
Even with slowsetting glu. time is critical.
Gather all tools, clamps, rbber gloves.
spreading devics. Talk about the procedure
with your assistant. Polyurethane is the best
choice of glue because it gives you the longest
slippry working time and because it is not
necessary to meticulously coat both surfaces of
each glue joint. One surface is enough, which
saves tme. As you apply the hose clamps. it is
imprtant that the joints remain slippery so
they wslide into place. It is as important
that you tighten the damps very methodically.
as you would tighten lug nuts on a wheel. while
checkng and adjusting the alignment of all the
comp: ents. Ha component such as a spline
slips inward, use a screwdriver though one of
te openings to realign it as necessary. This is a
nerve-wracking procedure requiring quickness.
teamwork, and maybe a little luck. You might
want to consider gaining a little experience by
cnstructng a smaller sphere bfore
attempting one this size.
Aer the glue has cured the next step is to sand
the exterior glue seams and check the tightness
of Ujoints. I used a large, 7-inch hadheld disc
sander with coarse grit to dean the surface and
expose all the seams. Ha poor joint is
dered. there is little that can be done to
salage the sphere, though perhaps a section
could be transfrmed into an open bowl. Before
spnding time at the lathe. I like to confirm the
appearance of all the seams. Icosahedron is
quite a challenge and unfortunately. most
viewers have no appreciation for the
accomplishment though occasionally an
experienced woodworker comes along and
expresses his admiration.
bUI|OINC IcCsAHO8CN
'
4
3
Uyou are intrigued by Platonic and
Archimedian solids, two books, Sphericl Models
and Polyhedron Models, both by Magnus
Wenninger, offer dozens of designs. Most of
Wenninger's paper models are probably too
complex to be adapted to wood construction.
but for anyone looking for the ultimate
woodworking challenge, I recommend his
intriguing volumes. Another mathematiCian,
George Hart, |done an incredible amount of
work with spherical forms ad has an amazing
web site devoted to this subject. highly
recommend a visit to both of their web sites at
+ .georgehartrom and
http://employees.csbsju.edUlmwenninger.
Sphere Turning Techniques
I usually do the rough-twning of a sphere with
the aid of a tailstock. Unfortunately, I was still
waitng for a back-ordered tailstock assembly
for a new lathe, so Ihad to shape this sphere
another way. Normally I would use a large cup
style drive to hold one end, vOa Dvcenter
cr g another smaller cup at the other end.
By mounting the form between the cups in
a variety of positions I can the shape in
multiple directions, resulting in a ver round
sphere. without a tailstock. this is how I
accomplished transforming the icosahedron
into a round sphere.
I glued a 3-inch disc of maple to one of the flat
triangular swfaces used this to lathe
mount the sphere in a four-jaw chuck as shown
Qphoto 14-12 (next page). A fceplate-mounted
waste bloc would have also worked. Notice
that I did not use a circle of MDE Without
penetrating screws, the MDF by itself is not
strong enough.
My ita focus with my gouge N to
ete the high ridges and sJts :O
order to determine the likely finished diameter.
After achievig a fairly round form I measwed
the diameter using two squares, shown 1
photo 14-12, and then used the 7-1/4 lOch
dimension to create a template, which I lathe
cut from 3/4-inch MDE While frequently
checking the sphere with the template, I

THE AT OF SEGMENTD WOODTlU:N1NG

1-12 A glued waste block permits rough
shaping.
14-13 A thin, turned ring makes an effective
template.
\: A large Jam chuck wIth keeper rmg
secures the sphere.
proceeded to create a smooth round surface.
Turing a sphere in this manner is not as quick,
easy, or accurate as turning between centers in
multiple directions, but I managed to
accomplish the task. Photo 14-13 shows the
outward portion of the sphere with templates
positioned against its surface.
At this point I built a large jam-chuck with a
retaining ring and mounted the sphere with the
maple waste block facing outward, as shown in
photo 14-14. This jam-chuck is simply a hollow
form built from layers of NI,with its inside
diameter tured to snugly ft the outside of the
sphere. The jam-chuck only makes contact with
the sphere at its outside rim. I bolted the
retaining ring onto the chuck without the
sphere in place and turned an opening that
would secure the sphere. Then I secured the
sphere using carriage bolts, whose rounded
heads made things a little safer.
If you have a copy of David Springett's book,
Woodturing Wizardr, take a look at his use of
jam-chucks with retaining rings. He executes
some very intricate sphere modifications using
smaller versions of this type of containment
device. If you happen to see his book in a
bookstore, there is actually a photo on the front
cover of a small jam-chuck with retaining ring.
Using a section of the template, I shaped one
end of the sphere. Repeated repositioning and
turning within the jam-chuck would have
resulted in a more perfect fonn, but I chose to
first insert plugs at the intersections.
I made the plugs before cutting the plug holes.
The two-piece plugs consisted of a ring OflO
wenge segments around a center piece of solid
bird's-eye maple. I mounted the wenge rings to
smalJ NFcircles using hot-melt glue and then
attached band-sawn circles of maple to the
wenge, as shown on the left side of photo
14-15
By lathe-mounting the NFcircle in a
four-ja
w
chuck, I could first tum the maple plug, fouowed
by the wenge plug. I used a caliper to sue the
plug diameters approximately 1.9 inches and 2-4
inches respectIVely. By accurately sizing the
plugs, I only had to tum the holes to fillf the
plugs 1\,then consistency was automatic.

tricky
yO
of Inserting the plugs is accurately
g>
ltlonm
g the sphere within the jam-chuck.
This
requit
es a little trial-and-error. By cuttmg
3n
undersi
zed opening, you can compare the
spacm
g
between the splines. If they are not
equal
then reposition the sphere
accordingly
before turing the opening to the required
diameter. photo 14-16 shows that the jarrah
wod
dimension between the splines on the
right
side i less than on the left side. By placing
a dowel into the hole and giving it a leftwards
hammer-tap, the centering is improved. By
tapping, turning, and checking a few times,
equal spacing can be achieved.
After centering the opening it was a matter of
tightening the nuts one last time to firmly
secure the sphere and then turning the opening
to fit the wenge plug (see plug fitting 1 Chapter
7). Without taking the assembly off the lathe, I
glued the wenge plug into place and then
worked on something else while the glue cured.
Ten minutes is plenty of waiting time for small
plugs such as these.
The maple plug hole was then tured inside the
wenge plug and that plug was glued into place_
Before repositioning the sphere, I used a section
of my template to make sure the plugs did not
protrude, so they wouldn't interfere with the
jam-chuck fit. I also twned a tiny indentation
into the center of the maple plug and used this
to dila small centered hole using a drill press.
These small holes would later be used to attach
other spheres to the icosahedron. Photo 14-17
shows a dose-up of a few of the installed plugs,
before finish-twning the swface. I repeated the
process until al12 openings were plugged.
Using the same jam-chuck I tured the sphere,
repositioned it, twned it some more, and so
forth, and then sanded the surface. Before
sanding I enlarged the jam-chuck opening
slightly and glued a few small pieces of rubber
pad to its inside rim in order to avoid marring
the sphere. I did not use the retaining ring, but
instead, I just kept one hand in contact with the
sphere while the lathe tured and used the
other
hand to operate a 3-inch disc sander.
This particular sphere became the center of a
BUILDING AN ICOSAHEDRON
'
4
\
:: InstaH all the components of the plug
before relocating the sphere to the next location.
14-16 The opening is centered if the space
between splines is equal.
14-17 All of the openings have been plugged,
146 THE ART OF SEGMENTED WOORN1NG
14-18 PLATONIC MO!EcULE (28 inches taU) -a
small tubute to Plato
sculptural piece that I titled PLATONIC MOUCULE (m
honor of Plato), as shown in photo 14-18.
The remainder of the project consisted of
creating the smaller spheres and building the
base. The only other noteworthy technique
pertains to constructing the 3-inch spheres with
their orbit-like rings. These were first assembled
by stacking five segmented rings using the MDF
disc/carriage bolt technique that 1 described in
Chapters 7 and 12. To create the thin contrasting
stripes at assorted angles, I tued the stacks of
rings into spheres by first using a four-jaw
chuck on my large lathe and then a cup
drive/cup live center combination on my mini
lathe. Photos 14-19 and 14-20 show the two
different lathe mountings. Photo 14-19 shows
the assembly of rings mounted directly into the
four-jaw chuck without any waste block, which
was possible because most of this portion
would be turned off in the nex step. Using a
template, I rough-turned one side of each
assembly round on the large lathe and then
transferred it to the mini-lathe in order to use
its tailstock. After rounding the other end, I
used a small cup center in conjunction with the
live tailstock to tu the sphere in several
positions. as shown in photo 14-20.
Once I had achieved a fairly round form I
positioned the spheres to make a scoring cut. as
also sh in photo 14-20. I then used this
shallow parting-tool groove as my guide to saw
the spheres in half on the band saw. Photos 14-
21 and 14-22 show the sequental steps. I used
the applied tabs of duct tape as handles while
cutting and disc-sanding. Rwas a matter of
repeatedly cutting the sphere apart, disc
sanding the cuts smooth, gluing a ring of
bloodwood segments to one side. sanding the
bloodwood ring. and then joining the two
halves back together. The bloodwood segments
were treated as thick veneer. The strips that
they were cut from had been accwately
dimensioned and sanded and the segments
taped together. At the top of photo 14-21. a
bloodwood ring is being glued to one-half
of a
sphere, using piece of dense 1/8-inch thick
foal
ll
rubber between the blood wood segments 3n1
1q!
Turning a stack of rIngs into a sphere_
!q21 The sphere's diameter is mai ntained
bcause the ring thickness equals the kerf
the circle of flat MOE With the assistance of the
rubber pad, the damping pressure creates a
tight bond with the individual segments. The
piece of wax paper in the photo simply
prevented glue contact with the rubber pad. On
the right of photo 14-21 you can see the results
oftms gluing step, The next step was to sand
the bloodwood circumference to match the
canarywood, smooth the gluing surace, and
then jOin the halves back together. Photo 14-22
shows a few examples of the various steps. In
total, twelve spheres were cut in half at least
four times each and two separate gluing steps
were required to rejoin them each time. In order
to keep the spheres fairly round, the thickness
BUILDING AN lcOSAHEDRON
'
4
7
1q2O The tailstock holds a sphere as it is split
with a parti ng tl .
1q22 These spheres are in various stages of
construction.
of the inserted ring needed to closely match the
sawkerf that had been removed. By matching
these two dimensions, it was also possible to
maintain the alignment of the original
horizontal layers.
To create a smooth sanded surface I made a
small modified jam-chuck. as seen in photo
1423 (next page). This jam-cuck was simply
two layers of I-inch MDF mounted into a four
jaw chuck and turned to fit the diameter of the
spheres. Instead of turning a tight fit. I enlarged
the opening and glued three small pieces of
1/8inch dense foam rubber to the contacting
edge of the jam-chuck. This allowed me to
quickly fe-position the spheres during sanding.
148 THE AT OF SLCMLN1ID WOODURNINC
1q2j A small jam-chuck (and fingers) holds the
spheres during saIdi ng.
I used a relatively slow lathe speed (about 350
rpm) and made sure the sphere would not
become dislodged by keeping a couple of
fingers in contact with it while sanding. I was
quite amused when I viewed photo 14-23 and
noticed that my little finger had been extended
well out of the way of the four-jaw chuck. This
may appear dangerous, after all, there are two
risks: the spinning chuck, and the spinning
sanding disc. Sometimes a little danger is a
good thing, it keeps you aler. As with all
sanding. be careful not to overdo it. With such a
small item it only takes a few seconds of
sanding with each grit to erase the scratch
marks from the previous grit.
The remainder of the project required creating
the smaller spheres and the base unit. To
securely attach the spheres, I used small steel
threaded dowels. If you are not familiar with
this type of hardware, they look like double
ended wood screws With no head. One end is
screwed into one component using locking
phers, then the other component can b
screwed onto the other end. Glued wooden
dowels would have worked, but the threaded
steel dowels are superior
Transforming Spheres
into Vessels
The technique of inserting an angled ring into a
sphere is not limited to small spheres or small
rings. A few years ago, when senator/astronaut
John Glenn went up on the space shuttle. I made
a turing (photo 14-24) to commemorate the
event 1 have always been a ofthe space
program. I delivered it to a San Francisco gallery
on the day that Sen. Glenn's picture appeared on
the cover of Time magaZine and it sold
immediately. The tuing, FOR JOHN GLENN, with
its three orbits representing Sen. Glenn's space
flight, is an example of two techniques: inserting
angled rings into a sphere. and transforming a
sphere into a vessel. The rings in this turing are
not just thin layers of one wood, they are multi
laminations. To maintain the alignment of the
horizontal layers, I removed an exact thickness
of material to match the thickness of each
inserted ring. The same basic technique that was
used to make the previous small spheres was
used for this large. 14-inch sphere. After
installing the three rings, it was relativel
simple to replace the top and bottom portons
using a large jam-chuck with retaining ring as
my lathe-mounting method.
By using a jam-chuck with a retaining ring.
many modifications can be done to a sphere.
Decorating a sphere with plugs and then
transforming the sphere into something else is
another method for solving a design challenge.
The vessel. 24 HOUR PIHER (photo 14-25. page
150), is a good example of this technique. The
center portion of the vessel started out as a
sphere, which I could easily position within a
large lathe-mounted containment device. The
title refers to the four designs, which show four
different time of day scenes. The scenes were
created using marquetry techniques with
3/4-inch wood. built four large plugs, which I
then Dtinto large turned holes 10 the
sphere.
After finished the sphere modifications
using
the jam-chuck, I removed the lower
portion
of
the sphere and glued it to the base
section
Then
I removed the top portion of the sphere and
added the vessel top. Inserting plugs toa
sphere is much easier than tryin to sc:1
BUILDING AN ICOSAHEDRON '49
'4-24 FOR JOHN GLENN (14 Inches diameter) -a larger example of angled rings
them into a finalized vessel shape. Of course, to
ensure a smooth profile transition. a detailed
bluepnnt is essential.
Another example of converting a sphere mto a
vessel IS M'DNltHT SNOW (photo 14-26, page 150).
ThIs turmg also began as a sphere. allowing
me to reposition It to mstall randomly spaced
snowflakestyle plugs The snowflakes were
created usmg dlscsander techniques and then
tUrned mto round plugs. The darkness of the
ebony effectively hides the plug glue !tnes.
A Truncated Icosahedron
Constructing forms such as an Icosahedron will
challenge all of your woodworkmg skills The
icosahedron is Just one of many possible
spherical forms that await your saw blade
There are many complicated forms that use a
combination of shapes. A soccer baUIS a
common example. With Its arrangement of
pentagons (five-sided) and hexagons (slxslded)
Many years ago I did a series of turnings called
my spors bowl series Three of those turnmgs
1O H AR OF bFCMINFDWOODTURN1NG
14-25 2q HOUR PITCHER (20 i n. tall)the pictures
were inserted into a sphere as large plugs.
are shown in photo 14-27. This was both a fun
and challenging project. Athe balls were made
as bowls approximately the size of actual sports
balls. In the photo, from left to right, the titles
are, FOOTBOWL, SOCCR BOWL, and BOWL 'N BALL. I
include them here to talk about the SOCCER BoWl,
by far the most difficult piece in the series. A
soccer ball made from 20 hexagons and 12
pentagons. In the world of geometry the form is
called a truncated icosahedron, to truncate is to
cut off a section of a shape. Look at the soccer
ball and imagine the six-sided sections as
triangles with their three comers cut off by the
pentagon sections. When I made it had glued
together many icosahedrons, but this challenge
had me scratching my head for several days.
Instead of trying to assemble 32 mitered pieces
14-26 MIDNICH SNOW (17 in.) the snowflakes
were built as plugs and inserted into a sphere.
with 60 precisely aligned intersections of
points. my solution was to build the
icosahedron first. using yellowheart
triangle
s
with ebony splines. The splines created a
gap
at
each of the 12 intersections of triangular
points.
The challenge was to insert pentagon
shapes
at
these 12 locations. My solution was to
create
round plugs that contained the five-sided
shapes surrounded by matching ebony
strips.
The pentagons that you see are actually round
plugs fitted just as previously
described.
using
a
large jam-chuck with a retaining ring.l
discovered one of these plugs in my box
of
leftovers. To help you visualize the
techni
que I
show it to you in photo 14-28.
1 secured
the main sphere in a large
jam-c
huck
with a retaining ring. centered each locatlm.
BUILING AN ICOSAHEDRON
Jg-2] SPORTS BOWS -FOOTBOWL, SOCCER BOWL, and Bam NBALL (life siZe)
and twned the recess to fit a plug. The trick was
to align the five points of the pentagons with
the five splines that radiated fom each hole. H
worked and was a lot easier than tring to fit
mitered pentagons between mitered hexagons.
The top of the form (the bowl portion) was
initially part of the sphere. It was parted off,
ted inside-out, and then re-attached to the
sphere with a ring of ebony to )Jthe parting
tool kerf. Solutions to challenges do exist, they
just have to be discovered.
14-28 The SOCCER BOWL pentagons were installed
as plugs.
152 JHE ART OF SEGMENTED WOOOTURNING
1-OO RKNP!R [)omchctall)
15
Porthol es and
large Turni ngs
The vessel RXNF!Twas constructed in
three different sections, almost like three
spate turings, which I eventuB y joined
into one vessel. The middle portion is composed
Oa large ring Oportholesjle segents with
smaller
feature rings above and below. To start
the project, I constructed the portholes from
laminatons with the Indian faces sandwiched
between layers. I introduced m15turning
vportholes in 1995 at an AWsymposium at
Davis, California. Since then, this feature has
bcome one of my most popular design
elements. The technique inolves crcaJIng
dsgDs,gIuIDg!hmbetween layers,
tansforming te laminations into segments,
and turning the individual porthole
indentations before gluing the ring together. A
sVe t photo IS-01, shows the composition of
the laated segments
Building
Porthole-Style Ring
The foUowing steps outline the procedures that
I ubd to build this porthole ring:
I Tst constructed the mosaic Indian face
d<ig1using a disc sander to fit B the
components together. This was not difficult but
it M qu1te a Oecons\ing, one-step-at-time
chore. I did not follow a precise plan for each
face, but designed them as I assembled the
cmponents. I started with assortment of
3/4-inc thick scraps of wood, but ended with
about 1/2-inch thick squares because there Nt1t
many sanding steps. To help determine the
position of the faces within the portholes, 1used
a template during thei\ construction, a 4-1/2
inch
square piece of MF, with a 3-inch centered
VORT!O!$ ANQ )RG URNING5
`jO1 This is a six-layer laminated segment
containing a mosaic Indian face design.
'53
hole. During the gluing steps I freq\ently used
the template to ensure that I had properly
centered my designs and to predetermine
which portion of the design would be revealed
after t1Dng the indentaton. When creating
this type of mosaic design, creati 1e clamping
techniquesare often necessary. I used rubber
bands, an assortment of clCmps, and an
occasional gluing caul to achieve the needed
pressure alignments. I haNe built many
porthole-style rings, but this was the first to
incorporate Indian faces. My past designs had
been simple landscapes built using band-saw
marquetry techniques.
My ring design consisted of 16 portholes. I built
eight faces, each about 4-1/2 inches square, with
the intention of splitting them to create 16
designs. Aer completing the 8 faces, I squared
them and then used my disc sander to smooth
one surface of each.
For the inside (bottom) layer of each segment
lamination, I D ed two 4-1/2 inch wide boards
of mesquite about 1/2 inch thick and 48 mches
long. I glued the 8 face squares smooth side
down to one of these boards of mesqHte.I used
a ]/4inch piece of wood as a spacer to
umformly separate the face squares on the
mesquite board, creating a saw kerf location for
later cutting.
After the glue cured I passed the assembly
through my Qsander UOthe e]sed face
15
4
TE ART OF SEGMENTED WOODTURNING
1C2 These are all the components that compose the lami nated segment.
1O It pays to have lots of clamps.
suraces were smooth with a consistent
thickness of about lIz inch.
I glued the other board of mesquite onto the
sanded face surfaces, sandwiching the mosaic
faces.
After deaning up glue squeeze,out ad
squaring the lamination edges of the mesquite
boards, I split the faces by making two rip cuts
on the table saw with the lamination on edge.
This resulted in the Tstrips that you ca see
in photo 15'02. If you compare the rows of
faces, you will see minor images.
I again used the drum sander to flatten and
smooth the 16 exposed face surfaces. Their final
thickness was slightly more than 1/8 inch.
Photo 15'02 also shows the other laminaton
components: in front of the face strips you can
see square rings of purpleheart and maple
segments. I initially glued these together as
3/4'inch thick round rings, dcsanded them
smooth, then squared and split them on a table
saw. The resulting 16 thin, square rings therefore
possessed one smooth, ready-to-glue surlace.
Photo 15-02 also shows a stack of 3z maple
veneer squares, which I later glued to either side
of the purpleheart/maple layers within the
lamination. The 16 squares of jarrah, which will
become the outside (top) layers of the
lBations, are shown in the bottom right of
the photo. Before cutting the individual squares
I machined the jarrah boards to a lIz inch
thickness with a smooth drum-sanded surface
ready for gluing.
Assembling the six layers (photo 15-01) required
three steps. The first step was the previous
attachment of the faces to the mesquite boards.
The next step was to glue one layer of veneer
and the purpleheart/mapJe layers to the
faces. I
did this using at least six spring clamps per
square, as in photo 15-03. The top layer of
purpleheart/maple presses the veneer, but in
this case the segmented ring has a large
center
hole, and therefore the center of the veneer
layer had no clamping pressUIe applied.
Veneer
that is glue-moistened and not clamped will
:0
The circle of MDF is attached with hot
melt glue.
bubble and distort. To provide for stress relief
gQglue escape. I drilled a 3/4-inch centered
hole in the veneer squares prior to gluing. The
inside center portion of the veneer squares
would be cut away later. but I did not want
distortion to affect the outside glue joint.
Before adding the second layer Oveneer and
the jarrah squares, I
passed the assemblies
through the drum sander to flatten and smooth
the purpleheart/maple suraces. I reduced their
thickness to slightly less than 1/8 inch.
At this point there was no longer a reason to
keep the strips intact. I therefore crosscut the
udual sguatqsfrom the two strips. T
improved clamping access during the next
g\uIDgstep.
Tocomplete the laminaton process I glued on
the second layer of veneer and the top layer of
jarrah. After adding the faces I had marked the
bottom of the mesquite squares with a
centerline that could later use to assist
centering on the lathe.
After a little disc saDdDgand table saw work. I
h16squares ready to have prtholes turned
into them. I could have mitered the sides of the
squares before turrng the portholes, but In case
a mishap occurred, I wanted the ability to reject
VUk1OsNOJkCTUkNINCS
'55
-0 Do not turn too deep. the design is only
lIB-inch thick.
one or two. By waiting until after I had finished
the portholes, I kept my options open. I also
needed to create a spline element to position
between the finished porthole segments.
would take some tme and I did not want the
mitered segment-ends to age. The glue surface
includes numerous pieces Ooily wood end
grain, and more reliable joints resuh when there
is little delay Deen cutting and gluing.
One challenge was to maintain the centering of
the faces within the turned portholes. After
squaring the laminated pieces, I used the
previously applied centerline to hot-melt glue
3-inch MF circles to the mesqUite side of each
square. The MF circles could then bheld in a
four-jaw chuck to lathe-mount the ents
(photo 15-04). Once I had partially exposed the
faces during turing. I checked the centering
and in a few cases repositioned the MF cirde
before completing the porthole.
The actual tmg (photo 1S-OS) was the
quickest and easiest part of the ring
construction. I turned the openings with a
1/2-inch bwl gouge and a small-radius round
scraper for the inSide corer. I removed a
minimum of wood from the face layers and
attempted to create consistently dImenSIoned
portholes. Before removmg each square from
156 THE POF SEGMENTED WOODllJRNING
15-06 While the arrows poInt vertically. the wood
graIn in the splines is positioned horizontally to
match the adjoining segments.
7

~~

15-07 The outer layer must be thick enough for
the outside vessel profile.
PW%P
WV
-------
R`
w
15-08 Ifthe outer layer is too thin, inner layers
will be exposed.
the chuck, l power-sanded the openings to 400-
grit and applied a coat of sanding sealer.
Over the course of several days I repeatedly
used the MOF circles to lathe mount the
segments so could apply and buff numerous
coats of 011 finish to the porholes. I could have
done this after assembhng the nng. but it was
qUicker and easier to use the lathe's rotation.
For a spline between the segments, I created a
broken arrow design. This associated a little
symbolism with the vessel title, BROK SPIR.
Splines are not necessary, but they do nicely
fame the portholes. Since these splines are
mare than 4-12 inches long, the grain direction
had to match that of the horizontal segments. I
began by crosscutting narrow strips of maple
and mesqUite and laminating them together
with the grain direction oriented sideways, not
lengthwise.
Much like building a zigzag ring (Chapter n), I
angle-cut and reassembled the spline material
to form the broken arrow design. Instead of
gluing the angled pieces together as in a
conventional zigzag, 1 positioned a narrow piece
of mesquite between each section to separate
the maple sections (the arrow). At one end I
added a triangular paint and at the other end, I
added a couple of simple feather shapes. These
splines are shown in photo 15-06. Note that the
wood grain is either horizontal or slightly
angled, but not vertical, which would have
positoned it at a right angle to the grain of the
laJge segments. To ensure that the spline sides
were parallel I hot-glued them to an MOF
transport tray and passed them through the
drum sander .
I mentioned at the beginning of this discussion
the need to keep your options open for as long
as possible. Well, as luck would have it. while
turning/sanding one of the portholes I went
through the mosaic face layer. Instead of
making a replacement, I decided to reduce the
number of faces from the ori8ina1 16 to 14 This
of course reduced the diameter, which in tur
increased the sharpness of the outside curve.
Before making the decision to build the ring
with just 14 portholes. Idrew the curve and
overlaid the segment profile to confirm that te
turing would not remove too much
material
and expose the underlying layer. This is
important to understand: if the segment
oute
r
layer is too thin, or the curve too sharp.
then
the
second layer will be exposed at the
glue
JOlOts

couple of simple drawmgs (photos 15-07 and

15-08) show the difference between
accepu
ble
109
Atable saw sled results lOa smoother cut
and disaster. The layout U @OIO1g-Odisplays
my actual laminated porthole segment with the
predicted outside curve of the vessel. As you can
there is enough wood thickness on either
side Othe porthole. The layout in photo 15-08
shows what would happen uthe outside curve
were sharper and/or the segments were wider.
Shaping the vessel's outside surface would
expose the second layer within the lamination
and probably would ruin the appearance.
With the splines completed and the portholes
oiled, the next step was to miter the segments. I
used the table saw with a crosscut blade and a
small sled to make the 12.86w segment-end
TCcuts (360. divided by 28 angles). Several
series of test cuts using I-inch MDF were
necessary to set the correct blade angle. Photo
15-09 displays a miter cut in progress. The small
square of MDF on top of the segment provides a
contact point for the hold-down damp. After
cutting both sides I examined the centering of
each porthole within its segment, adjusted the
table fence slightly towards the blade, and
timmed a little off whichever side was longest.
You might be wondering why I used asled to
cut the miters instead of just sliding the
segments along the table and fence -the
segments were certainly large enough to handle
safely. A sled built from three layers of MDF has
weight, which dampens vibration. It also
prOVides a smooth ride over the tablesaw
PORTOLES AND LARGE TURNINGS
'5-10 Gluing on the splines saves time later
insert. which otherwise could cause tiny
inconsistencies. The result is a smoother
crosscut on the end-grain of the segment.
I dry-fit the Qsegments and confrmed the
accuracy of the fit. To protect the porthole
surfaces, before any gluing I covered the
openings with maskng tape.
15
7
Instead of an aU-at-one-time glue job, I first
glued one spline to one side Ceach segment
(photo 15-10). This reduced the total number of
components, simplifying the job. If you look
very closely, you will notce that the spline is
not as thick as the segment. I have aligned it
with the inside of the segment, not the outside.
During the lathe-turing the segment wat
the corer intersections Wbe tured away. so
there was no need to build the splines as thick
as the segments.
Th type of ring must be constructed perfectly
round because just a slight Oashape will
result in portholes that are not athe same
depth. For that reason l like to add a matching
sized waste block when I glue the ring together.
Th allows me to confirm the roundness of the
ring and to tweak the roundness slightly. u
necessary. before the glue sets. If the mlters are
precise, very little tweaking should be
necessary. I once again dry-assembled the ling
of 14 pieces with attached sphnes. and tured a
piece of 3/4 Inch MDF to m.tch Its outside
diameter
158 THE AR OF SEGMENTED WOOOTURNING
'5-11 Be sure to have everything at hand before
speading any glue.
15'3 A sharp cutting tool is essential.
I gathered everything that was needed (photo
15-11) and glued the ring together. To increase the
open time, I used polyurethane (Gorilla) glue. In
the photo, notice the water bottle: I sprayed a
hght mist oer the Jomts before gluing.
Polyurethane glue is moistureactivated and sets
better with the presence of a little moisture.
Four hose damps provided c1ampmg pressure
and a pipe clamp from the ceiling held the waste
block } place (photo 1-1z). took special care to
align all the segment mtersections and to center
the WLwaste block
'5-12 A waste block can be attached at the same
time that the ring is glued.
Aer allowing the glue to cure overnight, I VeIy
carefully turned the ring on the lathe (photo
15-13). The porthole openings meant that my
sharp bowl gouge was cutting Qmost of the
time. To minimize chip-out at the porthole
edges I took very light cuts at the highest
reasonable lathe speed. The ring's inside surface
much easier to tur, although care had to
be taken to maintain sufficient wthickness
behind the face designs. Far too much work had
gone into the project to get careless at this stage
-I checked frequently with a double-ended
wall thickness caliper. Because of the porthole
segments, this portion of the vessel had to have
thicker walls than the top and bottom porions.
was unavoidable and not obvious from the
outSide; on the inside, the portholes do slightly
interrupt the vessel curvature.
With the porthole ring done, I finalized my
blueprint for the remainder of the vessel. l
divided the construction into three
diferent
sections: the porthole ring with its two
adjoining feature rings, the bottom
section.
and
the upper section. I independently built each
section and then combined them into the fm
al
vessel form. am not going to descnbe mQ
little detat regardmg the rest of the project
because have discussed nngbwldmg
techniques
in previous chapters, but will
detad
a few noteworthy techniques.
1j-1g The center porion of the vessel is
composed of eleven layers.
Large Diameter
Ring Techniques
)rOkenSpirits is about 22 inches in diameter. I
built many larger vessels, but this is
certainly large enough to demonstrate a few
techniques. After completing the porthole ring I
built the two smaller feature rings, which I
positoned above and below the porthole ring. l
uthe disc sander extensively to fit the
individual elements of these two rings. In
additon, my design called for thin layers of
wood above and below these two smaller
feature rings. If you take a close look at the
mddle portion of the vessel, you can count five
layers above and below the porthole ring. The
technique used to join all these large rings was
slightly different than for smaller rings. Photo
15-14 displays the eleven layers, from left to
right they are mapie, jarrah, feature ring, jarrah,
maple,
prthole ring, maple, jarrah, feature ring,
jarrah,
and maple.
Here is how I assembled these layers together:
I had
already attached the porthole ring to a
facepla
te during the previous glue job. Instead
of gluing together the four U maple and
janah rings, I created two thicker rings of each
PORTHOLES AND LRGE TURNINGS 15
9
'5-1
5
A sabre saw can be used to spht a large
ring.
type of wood, which 1 could split into thinner
rings. Photo 1S-1S shows one of the maple rings
center-mounted on a circle of MF with a
continuous bead of hot-melt glue. I have turned
the outside of the ring round and flattened its
surface for gluing. I did not turn the interior
curve of the ring, but left it alone with its
segment corners. To split the ring, I used a thin
parting tool from the outside surface. I did not
cut all the way through the ring because it was
too fagile -a slight bump could have broken
it_ Hyou look closely at the inside corners of the
ring, you'll see where the parting tool barely
broke through. A little uncut wood between the
segment corners still holds the two ring-halves
together. The small break-through slot created
by the parting tool provided an entry point for a
sabre saw blade, making it easy to separate the
ring. then put aside the cut-off thin ring with
its one ready-to-glue surface.
flattened the lathe-mounted half of the maple
ring, then hot-glued a ring of jarrah to it (photW
15-16, next page). J used the same process shown
in photo 15-1S to flatten and then split the
jarrah nng I put aside the separated jarrah nng
with its smooth surface and proceeded to
flatten the lathe-mounted half.
160 THE ART OF SEGMENED WOODTURNING
15-16 Notice the SU8|slots where the parting
tool cut through at the glue joints.
15-18 This is just one of many clamping steps.
1 20 When spnng clamps are too small, bar
damp (an be used
15-17 Sanding DOCKS UuSbe very straight and
flat.
15-19 Smal l beads of hot melt glue can
temporarily hold nngs during flattening.
'S21 Notice the masking tape \h8
Qro
tN'
portholes from drlppmg glue
s(fo
re I
remo
ved the second half-ring of jarrah I
fla
ttene
d it usmg a large sanding block with 80-
gnt
paper
attached (photo 15-17). It is critical
tha
t the
mating glue surfaces be perfectly flat. A
tlOy inaccuracy on both surfaces can combine to
create a detectable flaw in the seam. Make sure
your sanding block is perfectly straight. As
sandi
ng-block
sizes increase. flatness and
straigh
tness become even more critical.
I
glued
one of the thin jarrah rings to the lathe
mounted ring of maple using many spring
damps. as shown in photo 15-18.
I u hot-melt glue to temporarily attach one of
the feature rings to the jarrah ring (photo 15-19). 1
|c it round, flattened a glue surface. removed
it tram the jarrah ring. flattened the jarrah ring,
and glued the feature ring onto the assembly.
I added another maple ring and jarrah ring (the
ones that had been set aside) to the feature ring.
The next layer is the porthole ring. I flattened
the surface, matched the diameters, removed it
fom its waste block. and glued it to the
assembly using bar clamps as in photo 15-20.
Using the waste block that had removed from
the porthole ring. I mounted the other full-sized
maple ring, assembled the upper third of this
section, and then added it to the upper surface
of the porthole ring as shown in photo 15-21.
Building Sunrise Segments
I bUilt two additional feature rings, one for the
Jer portion and one for the top portion of the
vessel. The construction of these small sunrise
depictions was a little diferent from disc
sander-built elements.
I first constructed half-round segmented rings
using the previously discussed half-ring
technique (Chapter 7). I disc-sanded the half
rings to form whole rings, but did not glue the
halves together.
I
dryclamped the glued halves together with
two unglued purpleheart splines separating
them. Photo 15-22 displays three of these rings.
With the pairs of half-rings hose-clamped
PORTHOLES ANO LARGE TURNINGS
1j22 This requires a drill press and a sharp
forstner bit.
'5-23
Check the fit asyou turn@
together I drilled 1/2-inch holes in the ring
centers and inserted a piece of 1/2-inch dowel,
as shown Dthe foreground ring.
The dowel provided a means of accurately
centering a Forstner bit on the dHpress, to
create a larger. smooth-sided hole in the other
rings. The two halfrings in the photo show the
result of removing the hose clamp.
I needed to transform these rings into
rectangular segments, but first I needed to fill
the center holes. I used the half-rings with their
halfholes as guide templates to turn plugs, as
shown in photo 15-23. The turning had to be
very accurate, any gap would be easily visible.
162 THE ART OF SEGMENTED WOODTURNING
1j2 These steps eventualJy result in a
segment
1j2 Because of the vessel profile, this ring
must be dimensioned quite thick.
To glue and clamp the cylinders, I used a hose
clamp Ushown on the left side of photo 15-24. I
aligned the grain of the plug so that it would be
horizontal in the vessel wall Notice that the dry
purple heart splines are no longer separating
the two halves. I separated the two half-rings
with the band saw and smoothed the surface on
the disc sander, as shown Dthe center of the
photo. Then I used the table saw to make a
parallel cut as shown on the right side of the
photo Using the miter saw I made miter cuts to
transform these assemblies into ring segments.
Next I cut narrow splines with horizontal grain
to separate the segments. Two different sized
rings were needed: one for the top portion and
one for the bottom portion of the vessel. The
larger of these two rings is shown in photo 15-25
B it is prepared for stacking to the other layers.
'5-26 Almost invisible glue lines are a result C
maintaining grain orientation.
:-2 Follow the gluing steps clockwise, starting
from the lower left.
Another Method
of Inserting Diamonds
Other than building conventional rings of
mesquite for the main body of the vessel, only
one additional chore remains. Three of the
mesquite rings have small Indian blanket
diamonds embedded within the segments.
These are much larger than the router-created
diamonds shown in Chapter ).did not want
any
noticeable glue Hnes so I laminated
these
designs into the ring segments as
follows:
First, I built the Indian blanket designs using
the technique described in Chapter 9. I used
the
miter saw and disc sander to create the
bordered
diamond shapes shown in photo
15-26
This
photo also shows a few blocks of
me
squtte
.
left and right sides of segments that
I cut
SnJ
PORTHOLS mDLRGE TURNINGS
15-28
A straight edge can be used reveal a flat spot on the profile.
reassembled with the diamond design
surrounded. I used the band saw to split the
blocks horizontally, the red line (photo 15-26)
represents this cut line. I numbered all the
blocks so I could reassemble them in their
original configuration, which would match the
grain and hide the glue lines.
The next photo (15-27) shows the steps that
-used to build the diamond-embedded segments.
Uthe bottom left, an unassembled segment has
four 450 miter cuts surrounding the diamond
center. After checking the fit against a bright
light and adjusting as necessary, I did the gluing
i steps: first one side as shown in the top
le of the photo, then the second step as shown
i the top right. After I achieved a perfect fit
surrounding the diamonds, 1 cut the ends
opposite the angled cuts flush with each other.
During gluing I knew that if the ends were
even, then the inside of the angle would be
correctly aligned. When applying the clamp
glue squeeze-out obscwes the view of the inside
miter corners, therefore aligning the trimmed
Cdi the best way to position them. In the
bottom right. I show a completed block ready to
be miter-cut into a ring segment.
More Large-Vessel Techniques
The remainder of the project consisted of
assembling and stacking rings. I built the vessel
fom both the top and bottom directions and
made the final glue line just below the center
feature ring. Creating a smooth profile with a
consistent wall thickness on such a large vessel
can be chaUenging. When turning smaller
vessels, we usually finish the outside shape and
then turn the inside to a consistent thickness.
Because of the depth and limited access, when
turning large and deep vessels I usually rough
turn both inside and outside surfaces as I stack
the rings, and try to finish-turn the inside
before 1 have to reach excessively. This can lead
to problems HI do not frequently check
diameters against my blueprint. If two different
profiles (inside and outside) exist, then the wall
thickness will not be consistent. For that reason
I try to maintain as much wall thickness as
possible for as long as possible, giving me
maximum options during the final shaping.
Large vessels tend to have very subtle curves,
which can make it easy to unintentionally
produce a flat spot on the profile. is not
always that easy to see and create the desired
subtle curve. One trick, and this might sound a
little funny. is to use a straight-edge on the
curve. By placing a ruler against the vessel
profle and rolling it from point to point. you
can easily detect flat spots (photo 15-28). A
straight line against a curved line should only
make contact at one tiny spot at a time. Uthe
ruler makes more than point contact anywhere
along the vessel profile, you immediately know
where a little more shaping is required.
With the bottom half completed. I prepared the
upper halffor its attachment (photo 15-29. next
page). completed as much of the fimsh work as
possible. all that remamed was the top rim of
ebony, left thick to provIde secure attachment to
its waste blok.
16
4
THE ART OF SEGMENTD WOODTURNING
'5-29 The top half Isalmost ready for the last
gluing step.
'5-30 Reverse mounting at this stage allows the
top to OeJID$hed
15 31 A hand herd cabtnet scraper and power
. andtng ftnlshes the inside seam
I knew from experience that once 1 assembled
the two h<lves, the final shaping of the top rim
would be difficult. Because of the height (30
inches) and relatively small base (less than 6
inches), without the aid of a tailstock or large
steady-rest. the top would vibrate excessively. I
avoided this by reverse-mounting the top half
as shown in photo 15-30. The mounting
technique used I-inch MDF for the headstock
side and 3/4-inch MOF for the donut-shaped
retaining ring. Four lengths of all-thread rod
held the turning in place, with thin rubber pads
protectng its surface. To ensure perfect
centering I turned a shallow groove into the
l-inch MDF, which held the vessel half in the
correct position. I used a fairly slow lathe speed
to finish the top shaping and sanding.
With matching inside and outside diameters, I
joined the two halves. The base faceplate
remained mounted on the vessel bottom. The
deanup of the last glue line required very litle
turning. I was able to clean up the inside seam
with a curved handheld cabinet scraper,
followed by power-sanding. The outside seam
required a little shear-scraping followed by
sanding. Photo 15-31 shows my arm reaching
into the vessel as I power-sand the inside seam
with a 3-inch disc. You can see the pwer cord
(under my arm) with a piece of masking tape on
it, this is a sort of depth gauge indicating the
approximate location of the sanding disc inside
the vessel.
Aer multiple applications of oil finish, it was
time to remove the base waste block and
faceplate. Reverse-mounting the entire
vessel
without the aid of a tailstock was not an option.
so I routed off most of the waste block. then
used a 3-inch, drill-mounted sanding disc to
finish the bottom. Iwould have preferred
lathe
turning the base. but without a tailstock or
large steady-rest. it was just too risky.
large turnings such as B#w$;;can be
scary for several reasons. The rotational
speed
of large diameters can be intimidating. but
perhaps the scariest aspect is the posslbihty Q!a
serious mishap wiping out weeks oftllne
and
effort. The greater the risk. the greatel thE
PORTHOLES AND LARGE TURNINGS 165
15-32 DECONONTHE COCBOLO (42 inches tall) -large pieces require creative solutions.
reward. I enjoy the challenges of large turings
but I a always relieved when a project such as
this is out of the studio.
Large turngs do not need to be vessel forms.
Architectural forms and furiture items are also
possibilities. One of my recent large projects,
DEC!0N ON THECOCOBOLO, is shown in photo
15-32. The title was chosen because of the nature
of the game of chess and for the cocobolo
squares that compose part of the playing
surface. The table top is 43 inches in diameter
and for structural strength. the wall thickness
of most components averages 3/4 inch. This
project required more than segmented
wOO<turing skills. The chairs swivel on
bearings
and the table top contains thidden
dovetail
ed drawers for storing the playing
pieces.
which are all segmented, lead-weighted,
and leather-based. ThiS project was treated U
several
different proJects, which I tackled one at
a
time.
First I created the playing pieces, which are
larger than official tournament size (5-1/2 inch
king), then I built the chess board in proporton
to the pieces. I researched the recommended
proporton and found the little-known
specification that the largest base diameter
should be 78% of the square dimensions. The
table top with its drawers was next. at which
point had to decide between bar-stool height
or normal table height. After building the three
pedestals, I finished the project by building the
chair seats. As I jokingly tell people. this was not
done in a weekend. It was, however. a very
satisfying project that had been bouncmg
around i my head for a long time. It Illustrates
the idea that if you can think it. you can bwld it
166 THE ART OF SEGMENTED WOODTURNrNG
16 \1 TURBULENCE (34 Inches tal1) segments |nmoIor
Segmented
Ri bbons
There seems to be no end to the number of
segmented design possibilities. At first glance,
most people would never guess that the form
Jm uDNc(photo 1601)waS created on a lathe.
It takes experienced woodturners a few
moments to visualize the method of
construction: each of the corners or changes in
direction represents one half of a turned,
bottomless bowl. TU8u1wcis nothing more
Ueight. stave-constructed bowls. The bowls
were tured with identical diameters, wall
thicknesses, and 45 slope angles. l then cut the
eight bowls into sixteen half-bowls using a
band saw. Aer using a disc sander to dean up
the saw cuts, I reconnected the half-bowls end
to end to create the twisting loop. This
technique is not for the timid, many potential
disasters lurk in the shadows. But for those who
are interested, the following is a step-by-step
description of a project using this technique.
Jwith most woodworking projects, it starts
with a plan. Questions have to be answered:
Ho many corners or half bowls are required?
What diameter should the bowls be?
What wood species to use and what (if any) is
the desired pattern of segments?
For this project I am going to construct a loop
consisting of six half-bowls.l decided to build
the three bowls with about a lo-inch diameter,
using 24 staves in each bowl. The exact
dimensions were not important as long as all
three bowls are identical. A quick calculation
[(10 x 3-1416) 24) told me that the wide end of
the staves should be about t.3 inches. There was
wneed to draw a blueprint for the bowl
construction, since I knew that the slope angle
bCNLNMDKIBBONS
l
5"
t 1
J
-
!
,J"
,
1
.1

~
.
1 -
.!
1
L
`
l
5"
l
!D
1 t
J
' 1

f
. 5"
I
`

16
-
02 The layout of lamination layers should
create a bricklay effect between staves.
had to be 458 and I knew the required width of
the staves. I did, however, need to figure out the
layout of the laminations from which to cut the
staves. Each stave was cut from a seven-layer
lamination, but these laminations were glued
end-grain to end-grain, unlike most
laminations. This was necessary so that when
the staves were cut from the lamination, they
could be glued together side-grain to side-grain.
Side-grain glue joints are much stronger and
mOTe dependable. With this design, for the short
end-grain glue joints within the staves to
overlap, two different laminations were needed.
Photo 16-02 shows these particular layouts
The boards that I selected were only about 5-1/2
inches wide. Therefore, the laminations from
which I cut the staves could not be longer than
about 5-1/2 inches. After a little arithmetic, I
determined that I could cut five staves from
each 5-1/2 inch long lamination. Three bowls
with 24 staves per bowl meant that I needed 72
staves. To be on the safe side. I glued together 16
laminations, resulting in a yield of 80 staves To
produce the individual lamination components,
I cross-cut the required pieces from three
different types of wood (Texas ebony. maple,
and eucalyptus). To msure extremely consistent
Widths and very smooth glumg surfaces. }
168 THE RTOFbIGMINTID VOODTURNJNC
16-03 The accuracy and smoothness of the
components is essential.
16-04 To accurately cut staves, the laminations
must be smooth and consistent.
16-05 Test flttmg and adjusting saw settings can
take SOOie lOC.
attached these strips to particleboard trays
using double-sided tape and passed them
through a drum sander. Photo 11-12 i Chapter 11
shows a similar procedure. This might seem like
over-kill, but I wanted the very best jOints and I
wanted the overall widths of the finished
laminations to be very consistent. Each small
lamination consisted of seven layers of end
grain to end-grain. End-grain jOints will soak-up
lots of glue, so I made sure to apply plenty of
glue to both surfaces of each joint. The cut-ofts
are shown in photo 16-03 before being glued
and the completed laminations are shown i
photo 16-04. To clean up the laminations, I
handheld one side against a disc sander and
then passed them through the drum sander
using 36-grit paper.
The next step was to cut the compound miters
from these laminations. The required angles are
listed in the compound miter chart in the
appendix. His a simple matter of looking up 24
staves with a 45" slope angle in the chart. The
specified blade angle is 5.3" and the miter is
84.68". Because the width of the laminations
was short, I decided to cut the staves on my
sliding-compound miter saw; I could have just
as easily used a sled on my table saw. There are
two goals: to produce tight-fitting glue joints
between the staves, and to produce a 45" slope
angle on the bowl sides. The 45" angle is critical
to success. There is a little extra thickness to
play with during the shaping of the bowl on the
lathe, but I recommend trying to achieve a slope
angle as close as possible to 45" during the
assembly of staves. There are 48 miter angles
(two for each stave), so the opportunity for
cumulative errors is great. To dial in my saw
settings, I made a series of trial cuts from
3'4-inch MDF After cutting 12 staves. I taped
them together and rolled them into a
half-bo
wl
shape. Then I used the go
"
angle between my
disc sander table and disc to check the fit.
Aer
three attempts with slight saw adjustm
ents.

was able to hold the 12 taped staves

against
the
disc and confirm that I was very close to my
goal. In photo 16-05. I am checklllg thiS
fit
Tht
seams have to be tight and the ends
of tht
assembly have to fit flush against the d:>t At
tl1':' lllnt, I could havt cut another t2 staves and
formed a complete cone shape to further
.onf,"l\ my saw settmgs, but I was confident
that Itwas close enough, so I proceeded to cut
the good stuff Photo 16-06 shows the miter saw
set-up that I used.
Notice
the MOF compound-mitered stop block
This tSdifferent from a typical pointed or flat
stop block. It was cut with the same angle B
the staves. The advantage of this stop block is
that as my lamination is cut shorter, it becomes
Impossible to position properly against the
back fence. As you can see in the photo, the
ltion has more contact with the stop
block than it does with the back fence. I can rely
upon this style of stop block for the needed
support and angle. Hdoes, however, require
extra care U keeping the saw-bed clean of
debris, which would othelWise become trapped
] n the lamination and the stop block.
It is temptng to glue the staves together
without sanding the gluing surfaces. This might
DOK with many of the joints, but it only takes
one imperfect seam to ruin the entire project, so
sand the stave Sides. The staves were at the
desired angJe, so only a smidgen of material
needed to be removed. If yOUT sanding disc has
seen a lot of use, then this would be a good time
Vinstall fresh paper. To adjust the sanding table
to the correct angle, I test-sanded several pieces
with pencil marks along the top and bottom
edges (the edges closest to the inside and outside
Othe staves). When could lightly touch a stave
Vthe spinning disc and remove equal amounts
Cthe pencil marks, ke that I was not
significantly changing the stave angles. Htakes
a very light touch to remove imperfections left
by the saw blade. I used a bright light to inspect
each surface before glUing. Photo 16-07 shows a
stave being lightly sanded
A mult-step gluing process is needed to
complete the cones. Because of the 45 slope
angle, a gluing jig as shown in Chapter )] will
not work. Photo 16-08 shows the first two steps.
To accomplish steps three and four (photos on
next page) I used a combination of duct tape
and rubber bands. Using duct tape B a
bCmND RB80NS
16-06 The compound mitered stop block
providesstabtt1yand 3ccu|acy.
1bg
16-07 Think of the disc sander as a big ta5|
that gently |moVs imperfecions.
16-08 L|u|ngthe staves together IS 3 mut| step
process
170
THE PCF SEGMENTED WCODTURNING
16-09 If applied correctly. duct tape can
be used as a clamp.
16-10 These are awkward shapes to
clamp.
16-11 The last gluing step uses rubber bands anchored
by little hot-glued blocks of MOE
clamping aid requires a little practice. The
attached tape in photo 16-09 is stretched across
the glue joint so that when the joint is closed,
the tape will become even more stretched
(photo 16-10). The tape's elasticity applies
pressure to the outside of the joint. Rubber
bands are then used to close the inside of the
joint. The key to success is to have near-equal
pressure applied to both sides of the seam. After
trying this a few times dry, you should be able
to use this damping method effectively. Of
course without a perect fit between the staves,
it matters little what type of damping method
you use.
Before gluing together the groups of 8 and 12
staves, assembled them dry and checked the
accuracy. The half-cone shapes had to be near
perrect, an oval shap would likely require too
much removal of material to create a round
shape. A little disc-sander touch-up earJy in the
assembly process can avoid the need for a major
adjustment towards the end. Before gluing the
completed halves together, one last touch-up on
the disc sander was required. The simplest
damping method for this shape and size is,
once again, rubber bands. To prevent the d
from slipping on the cone, attached a few
pieces of scrap MOF with hot melt glue. photo
16-11 shows a completed glue job.
It's finally time to do some turning. Aer
flattening the large ends of the cones on the
disc sander, I mounted them to a large waste
block with a few beads mhot-melt glue. This
allowed me to rough-tur the outside and to
turn a flat spot on the smaIL end, to which I
glued a small disc of 3/4-inch MF Once this
connection cured, I turned the MDF disc so it
could be used in a four-jaw chuck (photo 16-12).
There was only a small amount of gLuing
surface holding the MDF disc on the
cones. so
I
avoided any aggressive turning.
With the cones now mounted with
their
small
ends toward the chuck, my first goal was to tu
the outside of each to a g

slope. After a IJttle

rough turning. Idetermined which of the three
cones had the smallest outside diameter and
then turned the other two cones to match To
determine the slope angle. I used a q
=
square
conjunction with a straight ruler Photo 16-13
shows my square and stra1ght edge as chtd1
the angle Accuracy was cntlcal. if one bowl is
q1` and another IS 41, the total error will be 4-
and a contmuous loop will not be possible.
Along WIth checking the slope angle, the
outside diameters must be checked and
adjusted to be equal. The rim of the bowl should
be tured square to the slope angle. Keeping the
tim square makes it easier to measure the
diameter. After confirming the diameters and
slope angles, the sharp edges can then be
rounded. To achieve consistency, it is likely that
a bowl will have to be remounted, and perhaps
more than once. Notice in photo 16-13 the red 7
on the MOF disc, this is so I can re-chuck the
bowl with the same jaw orientation. Also notice
that the MOF disc does not bottom out inside
the chuck. It is supported on the face of the
jaws. This is just one more little trick to increase
stability. Te more accurate you are at this
stage, the easier it will be to reconnect the half
bowls later.
Turning the inside suraces required creating a
consistent wall thickness of about 1/4 inch. A
thinner wall was certainly possible and perhaps
more esthetically pleasing. but was
uncomfortable about sacrificing strength. The
same 45 square/ruer technique can be used to
check the inner surface. If the wall thickness at
the rim is 1/4 inch and both inside and outside
slope angles are 45-, then in theory, the wall
thickness should be consistent. However, I still
recommend checking it with a thickness gauge.
I power-sanded all the accessible surfaces and
then reverse mounted the cones on a piece of
MOF. To ensure centering, I turned a recessed
groove into a large MOF waste block to accept
the large end of the cones and used a donut
shaped retaining ring to secure them for the
finish turning of the small ends (photo 16-14).
After a little more turning and sanding, the
cones were ready to be cut apart.
As segmented turers, we rarely have an
opportunity to easily repair a defective glue
line. This is one of those rare opportunities: if a
seam between the staves is discovered to be less
than perfect, it can be corrected by choosing
that seam as the location for the band-saw cut.
SEGMENTED RIBBONS
16-12 This waste block is ready to be mounted \|
a four-jaw chuck.
16-13 A straight edge and 45 square are used to
check the slope angle.
1614 Attach the retalnmg nng. turn the donut
nng. and then secure the cone shape
172 1HE1R1OI 5EGMEH1EDVOODRNIHG
16-15 lstIc elbows make a great design aid.
When cutting the bowls apart, avoid the
mistake of messing up the pattern. To
understand what I mean, look at the center strip
of pink wood in photo 16-14. The segments
alternate from large to small. The bowls must
be cut so that the pattern continues with the
new alignment. otherwise when reassembled
two segments of the same size vbe
positioned next to each other. Compared to
TURULENCE (page 166), this is a simple shape to
assemble. When constructing more complicated
forms, it helps to have a model to visualize.
Many different shapes can be constTUcted using
goG plastic plumbing elbows. Photo 16-15 shows
a possible shape that could be built using half
bowl shapes. A model can certainly help you
avoid a major glue-joint direction error.
After careful cutting on the band saw and a
little touch-up on the disc sander, the half
bowls were ready to be joined. Clamping such
odd-shaped pieces together can be a challenge.
My solution uses a combination of small blocks
of scrap wood, hot-melt glue, and TUbber bands.
Ilyou study photo 16-16, you will quickly
undfrstsnd the technique. A key to success is to
16-16 This method requires equal pressure on
both sides of the glue joint.
apply equal pressure on both sides of the glue
joints with the bands. Alignment of the ends, of
course, is also important. The goal is to
minimize sanding after the glue has cured,
which will also help hide the transition
between sections. I divided the gluing into three
steps. First, I joined the two sets of halves which
comprise the horizontal curves. Afer sanding
these joints, I connected them with a vertical
section at each end. To make sure that
everything would come together on the fl
joint, 1 first dry damped both ends. Then I
undamped one end, applied the glue, and put it
back together. Once this seam had cured. I glued
the final seam together. While applying the
TUbber bands. you will wish you had another
pair of hands. so if possible enlist a friend.
The sanding of the seams was relatively easy
using 3-inch power-sanding discs, although
t.he
inside corners did require a little handwork.
All
in all, this is not a difficult project
.
but it
does
require a multitude of woodworking and
woodturning skills. Perhaps the most fun I>
hearing people say, How did he do that?
.
.
bLGNiN1iD K|88ONS
1-1] CPR 16 (10 in. 7 10 i n. 7 16 in.) -|Owyou knOwhOwto do that.
The last and certainly not the least important
task was to name the piece -titles are
important. They provide a verbal handle for
people when they refer to your work, and they
can help tell a story or convey a message. This
particular piece was a rush job in order to
include this technique in the book. So what
better name than ... Capter t(photo 16-17).
This technique opens the door to many other
possibilities. Instead of half-bowls, quarter
andlor three-quarter bowl shapes could be used.
Instead mbowls, donut-shaped components
could be used. There is no end to where our
imaginations can take us.
17
4
THE ART OF SEGMENTED WOODTURNING
"
\{ OT LAFTorLICHT(36 inchcsta||) 6.soo |nd|V|dua| p|cccsof vood qndp |o|q!.. |eu|}|J|\'|
17
Fi nal Thoughts
People
that meet me after viewing my work
often
remark that I must be very patient. I have
never thought of myself as a patient person,
quite the contrary. Progress on turning projects
always seems too slow. There is always a feeling
of urgency \ofinish a turning once I have
started, but I have learned to be methodical in
my approach. so perhaps I have become a more
patient person because of my experiences. I
have certainly leared that it does not pay to
rush a project. Hurrying results in mistakes and
possibly accidents and actually slows down
progress. One of the most asked questions is.
"How long did it take you to make that?" I
sometimes reply. "Does it matter?" We have
to eat, but there is a diference between turning
to live and living to turn. MO OI1oH(hoto
1701), with its more than 6000 individual pieces
of wood, certainly requires one to be passionate
about the art and perhaps less concerned about
the time required or the fnancial bottom line.
There is an old saying that the best woodworker
is the one Who best hides his mistakes. U
segmented turing constuction, there are very
opporunities to hide mistakes. All the glue
JOInts and every square millimeter of surface
are subject to close scrutiny. The intricacy of
segmented work almost begs people to closely
examine it. People sometimes even seem
challenged to discover a flaw. The most exacting
assembly techniques in the field of
woodworking are required and compromises
usually lead to disappOintment. The late Ray
Allen used to chalJenge people to find a loose
joint in his tungs, he was very proud of his
work for he knew that imperfections did not
exist. He was absolutely right about the need for
perect jOints, no matter how large or complex.
n matter how beautiful the wood, a turning
Wlll lose
much of its appeal if a viewer detects
less than perfection in its construction. To help
FINAL THOUGHT
17>
you achieve jct!ec!iOn,and to heIjenslITe \ha!
your turings. stay !oge1hetfor a very long time,
keep these things in mind:
Use only d1ywood (under 10% moisture
content).
Avoid intersections longer than 1 inch of
oppOSing wood grain orientation (shorter if
possible).
A1ways consider the shape as the most
important design element (more important
than wood type, color combinations, etc).
Never glue two surfaces together that do not
fit perfectly and remember, a sanded joint will
usually be stronger and have a better
appearance.
Do not nsor to filling a joint imperfection
Re-do or replace as necessary. The defect will
almost always show and you will inevitably
regret the lack of pertection later.
Avoid an abundance of oily wood to oily wood
glue lines.
Keep, and inform others to keep your turnings
away from excessive exposure to direct sunlight
(the colors will fade).
Everything evolves, and woodturning ais no
exception. It was not that long ago, when the
cover of Fine WoodWorking magaZine displayed
two segmented vessels by Addie Draper and
Bud Latven. I dusted off that old issue recently
and could not help but compare their
magnificent 1985 work to a recent AAW Instant
Gallery display of member's work -evolution
is alive and well. Learning and discovering are
big reasons why this form is so enjoyable.
Solving the puzzles of new design challenges
keeps your brain active and rewards you with
fantastic feelings of satisfaction. To quote a line
from an old Bob Dylan song, "He not busy being
born is busy dying." I feel my own jourey of
discovery has just begun, endless possibilities
keep me intrigued. Good luck with your own
jourey! I hope my efforts with this book have
provided you with a foundation and that your
own discoveries Wcontnbute to the
advancement of this wonderful art form.
176 THE ART OF SEGMENTED WOOOTORNNG
Appendix
Mathematical Formulas
and Concepts
A circle contains 360.
CzcumtezeDceequals diameter multiplied by pi
(J.1416)
D\ametetequals circumference divided by pi
(3.1416)
Segment length equals circumferences divided
by number of segments
Segment angle equals 360. divided by number
of segments divided by 2
Calculating
Compound Miters:
Miter Angle (M)
= inverse tan h = (cosS tan(360 ""2N)])
Blade Angle (BA)
= inverse tan (cosMA . tanS)
M is the miter angle
S is the slope of the vessel (measured from
horizontal to side)
BA is the saw blade angle (the bevel cut)
Wis the number of staves
Construction
of Platonic Solids:
Shape Miter Angle'
Tetrahedron 54735
Hexahedron (cube) 45000
Octahedron 35-264
Icosahedron 20.g05
Dodecahedron 3q17
This is the edge miter of the individual
components as listed in Magnus Wenninger's
book, 5]heruuAOdels.
Compound Miter Angles
for Stave-Constructed Forms
6 Staves 8 Staves
Miter Blade Miter Blade
Slope" Angle Angle Angle Angle
0 60.00 0.00 61-50 0.00
5 60.09 2.50 6758 1.g1
10 6038 4.g8 6781 381
'5
60.85 J.44 68.19 5.68
20 6152 985 68.73 752
2 62.38 1Z.20 69.42 9.31
30 63.43 14.48 70.27 11.03
35 6469 16.67 71.26 12.68
40 66.14 1875 7240 1424
45 67-79 20.70 73.68 1570
50 69.64 22.52 7509
n
05
55
71.68 24.18 76.64 18.27
60 7390 25.66 7830 1935
65 76.29 2695 80.07 20.29
70 78.83 28.02 8194 21.08
75 81.50 28.88 83.88 21.69
80 84.27 29.50 85.89 22.14
85 8712 2987 8793 22-41
90 90.00 30.00 90.00 2250
Slope is the angle from the horizontal surface
to the vertical side assuming that the form is
pOSitioned with the small end pointed down.
Compound Miter Angles
for Stave-Constructed Forms
10 Staves 12 Staves
Miter Blade Miter Blade
Slope" Angle Angle Angle Angle
0 72.00 0.00 7500 0.00
5 7
2.06 1.54 75.05 1.29
10
7
2.26 3-08 75-22 258
15
7
2.58 459 549 J.84
20 7302 6.0
7 7587 508
^b 7J.59 7.50 76.35 6.28
30 74.28 8.89
7694
!
35 75.10 10.21 7
1
.62 854
40 76.02 11.46 78.40 9.58
45
n
06 12.62 7927 1055
50 78.20 1369 80.23 11.44
55 79.44 14.66 81.26 12.24
60 8077 15.52 8237 1295
65 82.18 16.26 8354 13-57
70 8366 16.88 84.76 1408
75 85.19 1737 86.03 14.48
80 86n 17.72 8
B
4 1477
85 88.38 1793
88.66 1494
90 go.oo 18.00 90.00 1500
JC FORMULAS AND CONCEPTS 1
77
16 Staves 20 Staves 24 Staves
Miter Blade Miter Blade Miter Blade
Angle Angle Angle Angle Angle Angle
7875 0.00 81.00 0.00 82.50 0.00
78.79 097 81.03 0.
7
8
8253 0.65
78.92 1.94 81.13 156 82.61 130
7912 2.89 8130 232 8275 194
7941 3.83 81.53 307 8295 2.56
797
8 4
73 81.83 J.
79
8).20 3.16
80.23 5.60 82.19
H9
83.so 374
80.
75 6.42 82.61 5.15 8J.84 4.29
8134 po 8J.08
5n 8424 4.81
8199
1
.93 83-61 635 84.68 530
82.71 859 84.19 6.88 85.16 574
8).49 9.20 8481
1
.36 85.68 6.14
8432
n
3 8547 n9
86.23 649
85.1
9
10.18 86.17 8.15 86.82 679
86.11 10.56 8690 845 87.42 7
.05
8T05 10.86 87.65 8.69 88.05 724
88.02 11.08 88.42 8.86 88.69
1
.39
8g.01 11.21 8g.21
897 8934
1.47
90.00 11.25 go.oo g.oo go.oo T50
. Slope is the angle from the horizontal surface to the vertical side assuming that the form is
positioned with the small end pointed down.
178 THE AT OF SEGMENTED WODG
Descrip!onoITurnngs
p
. ,
fArHfU
36" tall
-European beech and red batu
P 7 PEARLS FROMTHEFOREST
34" taU
-many exotic woods
1-01 ALia's GARDEN
p.8 35" tall
-more than 4000 individual pieces of wood.
-12 trumpet shapes attached to a center-
positioned icosahedron.
-maple, ebony, purpleheart, bocote,
flamewood, bloodwood, canarywood, pau
ferro.
102 MOBIUS SONATA
p.lO 20" wide
-made from three and one half bottomless
bowls
-Bird'seye maple, Texas ebony, gabon
ebony, macassar ebony.
2-02 EMTSPIRAL
p.12 18" tall
ebony, maple, and tuJipwood.
3-02 CRIBBAGE BOARD
P
14 16" diameter
-Argentinean mesquite, Texas ebony,
maple, bubinga.
5-00 C VASE
p.18 12" diameter
caretto, purpleheart, gabon ebony.
5'04 TUE'SAp
P
25 4" tall box
-pink ivory, tulipwood, bloodwood, ebony.
7-00 BLC & WHmT
P 3
0
-
9" diameter
-gabon ebony, holly.
8-01 AUWOOD BOWL
p.62 10" diameter
-applewood. purpieheart, ebony.
8'02 FOR THE BOY KING
p.62 36" tall
-flamewood, assorted exotics. ebony, maple,
pau ferro.
8-03 MyRTE MOONS
p. 62 35 tall
-icosahedron center sphere, outer ring
constructed from router-mitered turned
rings.
-myrtlewood. pau ferro. flamewood. ebony.
maple, canarywood, purpleheart, walnut.
8-07 ZIGZAGBox
p.66 3 diameter
-ebony, pink ivory, holly
9-00 JOHN'S TuRNtNG
p72 12.5" diameter
-Bubinga, bloodwood, holly, ebony
10-00 CHESS JfCH
p.88 5.5" tall king
-stave construction, lead weighted
-curly maple. ebon-x veneer
-Part of DEL0NONTHE COCOBOIO (15-32)
10-01 THREE WEDDING GIf
p.88 10" tall
-stave construction
-curly maple, ebony, maple burl, spalted
maple. rosewood
11-00 MADHATTU'S TEACUP
p.IOO 6.5" tall
-laminated stave construction with
alternating orientation
-Iacewood, purpleheart
ll-lO CHIW'sT(close-up)
P104 7" diameter
-multi-generation lamination.
rly maple, yellowheart, purpleheart,
myrtlewood, jarrah. ebony.
11-19 SCRAMUD m x:
p.110 10 tall
-multi-generation lamination.
-over 900 pieces of wood.
-curly maple, pau ferro. ye11owheart,
purpleheart, myrtlewood. ebony.
11-25 WATERMONCACUS
P
113 21"ta11
-constructed hom vertically laminated
segments.
-purpleheart, hoiJy, ebony.
10 OF TURNINGS
'79
11-26 TEACP FOR AUCE 13-16 TALKING W WOOD
P1I3 4" tall P13S .taU
-laminated stave construction, offset -over 1500 pieces of wood.
mounted. -holly, ebony, carob, mesquite, bloodwood,
-purpleheart, holly, ebony, diamond wood yellowheart, purpleheart. jarrah, maple, pink
(handle). ivory.
11-27 CURVES 14-04 COSMOS with SATEWTE BOXES
P1l3 29" tall p.I)8 )6" tall
-<onstruction used laminated segments -icosahedron style sphere surrounded by
with angled orientation free spinning rings.
-curly maple, pau ferro. ebony. diamond -many exotics.
wood.
14-01 ANDY'S Toy
1l'4 FORGAUDI P

139 9" diameter

P117 1tall -an unturned icosahedron.
-router created diamonds. -pine and paint.
-individually turned staves.
-<anarywood, pau ferro, holly, ebony, 14'0) THEJUGGUR
rosewood. P139 34" tall
-icosahedron style sphere, multi-
11-49 BLACKBERRY SWIRL lamination container at top.
p. l 6." tall -<urly maple. ebony, and many exotics.
-made from angled rings cut from a single
nng. 14-18 PLATONIC MOLECULE
-purpleheart, holly, ebony. P 146 2tal
-icosahedron style sphere with attached
12-01 QULTED BOWLS smaller spheres.
p. 122 .tall -jarrah, ebony, Bird's-eye maple.
-bricklaid segments. 901 pieces each. purpleheart, canarywood, bloodwOQd,
-curly maple, ebony, rosewood. purpJeheart. tulipwood, cocobolo, black walnut.
yellowheart, jarrah. bloodwood, mesquite.
14-24 FOR JOHN GUNN
13-06 OVA BoWl #gl P149 14" diameter
p.128 5" tall -sphere transformed into vessel.
-<reated by cutting a bowl shape In half -ironwood, tulipwood, ebony, other exotics.
and rejoining.
-mesquite, ebony, yeUowheart. rosewood, 14-25 2Q HOUR PITCHER
pink ivory. p.lS0 20" !a
-sphere transformed into vessel form.
13-07 ROCK 'N ROUlR -curly maple, ebony, and many exotics.
p. 1)1 16" diameter donuts
-almost 2000 individual pieces. 14-26 MIDNIGHT SNOW
-bubinga. maple, and many exotics. p.lS0 17" tall
-sphere with plug decorations transformed
1)'14 INEGRATON into vessel form.
p. l]4 12" diameter donuts. bony, holly, curly maple.
-walnut and curly maple.
14-27 SPIS BOWLS
13-15 GroWTH p.ISI life size
P

135 27" tall -many exotics.

-mitered rings, screwed together.
-myrtlewood and ebony.
180 THE AT OF SEGMENTED WOODTORNG
'5-00
P152
front
cover
15'32
P165
16-01
p.166
16'17
p. '73
'7-0'
P174
BROKN SPIRITS
30" tall
-more than 3000 individual pieces of wood.
-unique porthole construction as described
in Chapter 1[.
-Argentinean mesquite, curly maple, ebony,
jarrah, purple heart. pink ivory, holly,
tulipwood, flamewood. bloodwood.
canarywood. pau ferro. black walnut.
DECEPTION ON Df COCBOW
43" diameter table top
-swivel seats. hidden storage drawers.
segmented playing pieces.
-myrtlewood. walnut. curly maple,
bloodwood. cocobolo, Bird's-eye maple.
wenge, ebony. yellowheart.
TURBULENCE
34 tall
-16 half-bowls reassembled into a twisted.
segmented ribbon.
-many exotics.
CHPTER 16
10" x 10" x 16"
-6 half-bowls reassembled into a twisted,
segmented ribbon.
-Texas ebony. maple, eucalyptus.
TfIOF LGH
36" taU
-router created heart shaped portholes.
-many exotics.
Bibliography
Brown. Emmett E. and Brown. Cyril,
Polychromatic Assembly for Woodturning,
Linden Publishing Co .. 1982
OrlginaUy published by the Society of
Ornamental Turners, England
Constantine. Jr., Albert. Know Your Woods,
MacmiUan Publishing Company, First Scribner
Paperback Edition. 1987
Cox, Jack. Beyond Basic Turning, Linden
Publishing, 1993
Daper, Addie and Latven. Bud, Segmented
Turning, Redefining an Old Technique, Taunton
Press. Inc, Fi n Woodworking Magazine,
September/October Ig85, pages 64-67
Hoadley, R. Bruce, LnderstandingWood.
Taunton Press, inc, 1980
Lincoln, William A.. World Woods in Color,
Linden Publishing, 1986
Nish, Dale I,Artistic Woodturning, Brigham
Young University Press. 1980
Nish, Dale L,Creative Woodturning, Brigham
Young University Press, 1975
Rannefeld. Clarence, Laminated Designs in
Wood, lark Books. 1998
Shuler, Michael, Segmented Turning. Swirling
patterns by cutting and reassembling a single
bOard. Taunton Press. Inc. Fine Woodworking
Magazine, MaylJune. lg89. pages 72'75
Smith, William, Segmented Turning. Schiffer
Publishing Ltd., 2002
Springett. David, Woodturning WiZardQ, Guild
of Master Craftsman Publications Ltd, 1993
Wenninger, Magnus, Polyhedron Models,
Cambridge University Press, 1971
Wenninger. Magnus, Spherical Models, Dover
Publications, Inc., 1999
Originally published: Cambridge [Eng. I
University Press. 1979
About the Author
Malcolm Tibbetts is a professional segmented
woodtumer living with his wife Tere in South
Lake Tahoe, Califoria_ After raising two
children and after more than thirty years as an
executive in the ski industry, he now creates
lathe-turned art objects fulltime for several
galleries. His appreciation and experiences
with wood assembly span fifty years going back
to the days spent in his grandfather's shop as a
youngster. His innovative work resides in
prestigious collections throughout the world
and his work has received many woodturing
awards. As a member of the AAW (American
Association of Wood turners) his contact
information is publicly listed in the AAW
Resource
Directory_ Feel free to contact him
with your comments or questions through his
website, .tahoetumer.com.
AROUT THE AlJTHOR
THE ART SEGMENTED WOOIRNNG
Index
Pc0cD\5,BV00D_,1
P:tcDc\,:DQ0t\BD0c01,1
P:t,cBDDc0,10tccBDD_,t
P0c,D\c0u5c01.22
P_DDcD\:
015c_DcD\0t:D_5,12
t0\B\:0DB, q
A!cD, KB,Q,1[
PDcHcBDP550cB\t0D01W000-
\utDct5,5B1c\y_U0cDc501,1b
PQQc,1BV0tc0U5c01, 2
Pt0Huc0BP 50:05,D00K0D,1q]
P5,D\c0 u5c01,22
P5QcD,:D1\c0U5c01, 22
P55cDD,01Bt_ctD_5,1Q
bBD05BW.
10tBD_c0t:D_5,11,12O
10tDBKD_tDD0D,1[I1[2
10t5cQBtB\D_tU_5,1b2
5Q!:\\D_t:D_5W\D,
bBtcBDQ5,10t Bt_c5c_DcD\5,11Q
bB5c:
0D5\tU\:0D 01, Q
_uD_01,bO
01cD0D,Qb
01_D\Wo0,Q
Qu_10t,O,Q[
52c01, bq
W\10B\D

QU_,QbO
W\\BQctcQu_,-Q
W00010, Q
bB5cD0K,cD0D_,b
bccc,5QB\c0,u5c01,22
btc,D:1c0U5c01, 22
btc,5QB1c0,u5c01,22
bBcXW00d,1BV0tc0U5c01, 2]
bB0cBD_c.
BcuB\c5c11D_0.Q2, 1q1
0cDD\0D 01, 1[b
10tc0DccD\DcDD_5,12O
bC00WC00:
1BV0Ic0 u5c01, 2
D1cB\utc DD_,[
5QDctc01,1qb
b0 0\c.1BV0tc0U5c01.}]
b0W_0u_c.
10QDOD0cID_,1
\UDlD_0QcDD_5WI\R,
bt0WD, Lt: Q
bt0WD, LDDc\\,Q
buDD_B:
c00t01,1
U5c01,2],[O
LBD0cD0D5,DQ0t\D0c5,b[
LBQct:
10tWB\DcXDc55,12Q
U5cUCD\ctD_,1
LBDBtW000.5QDctc01,1q[
LB0D, 1BV0c0u5c01.2]
LBtVD_:
0DB1Dc,11
Vc55c!100\,1]O
LBU5:
B\1BcDDcD\01,2Q
10t_U:D_ tD_5,1q
10t_U:D_5Q:Dc,Q]
LcP\ct:D_:
010Q0D:D_:D5QDctc.1q
01tlD_5,12
LDctty,u5c01,22,112
LDc55 D0B0,52c01, 1b
LDUX:0U-]BW:
10tc0Dc5,1[O
10t5QHctc,1q]. 1qq,1qb, 1q[
u5c01,b,Qb
LDUX:|BD:
10t\utDD_00Du\5,1q
10t \uDD_5QDctc,1qq, 1q,1q[,
1q. 1O
L:tcc:
ctcuD1ctcDcc01,b[,1[b
0c_tcc5D, b[
01BDc\ct01, b[
LBDQD_:
B_B1D5\5H0QccD_,2,[, 2,Q
BD0t:D_0tcD\B\0D,2-
1cB\utc tD_5c_DcD\5,[
Dcc010t,2[-2
01D:\ctc0t:D_5,1]
01VcDcct5c_DcD\5,[
012:_-5B_QB\\ctD5,1O2
\ccHDQUc501, 2
u5D_B\c 10t,,2,Q
W\D00uDc50c0 \BQc,]
W\DMDF tD_5,q-
W\DtUDDctDBD05,2,q
W:\D5Qt U_OBDQ5,
LBDQD_]:_,U5c501,q
LBDQ5.\
_
c501,2
L0 0D00, BV0tc0U5c01. 2]
L0DQ0uD0D:\ct5:
BD_c10t, 1b
_uD_01, QQ,QQ
5BW5c\\D_510t, [b
LuDc:
0cDD:\:0D01, 1Q
D\ctBD_c01,|[b
Lu\\D_15\:ctcB\0D01,[O,[Q
LBD0BctB\c_uc,U5c501,2[
OcD5:\, 01W000,2
Oc5_D.
OtcX-B0,bb
Dd:BDOBDXc\,bbb[
OBDc\ct:
cBcuB\0D01,b[
0c1D:10D01, 1[b
OBD0D051\BQc5:
ctcB\10D01,bb
BDDB\0D01,1b
W\D0:5c5BD0ct,11O
O5c5BD0ct.
B0]U5\DcD101.1bQ
BV00D_cB\W\D.11
1B\\cDD_ W:\D.qQ,2
10 1cB\utcH_ qb,2
10t
g
Uc5QUcc2c,1O2
10t B1tD_5,q1
10t|0:D1D_5c_DcD\5,1Oq
10tDBX:D_5D0W1BXc5,1qQ
10tD\ctc0tD_5,1]112,1]
0D5c_DcD\ cD05.
5\tB\c@01, Q
O00ccBHct0D,D:\ctBD_c01, 1[b
O0Du\ 5BQc,01D:\ctc0 tD_5,11
O0\50D,Vt_D:B,11b
O0Wc,\tcB0c0,10tB55cD0!:D_
5Qctc5, 1q
OtBQct.P00c, [
OtUD 5BD0ct.
CcBD5UQ5\tQ5,[b
10t0BD0D05BQc5.111
10t 0DcD50DD_,1,1O,
!>>
10t gUc5QUcc2c,1O[
10tBDDB\c01Bc5,1q
|_10,1]2
Ot\,
D
|
\BD c01, j]
\_\Dc5501.q1
Ot1D_5Dc0,10tW000,2O
Ouc\\BQ,10tBDQD_5c_DcD\5,
1bQ-1[O
Ou5\cX\tBC\:0D,1DQCt\BDc01,2
OcD_,01WBDU\,[
LD0D-X,DBX:D_1t0DWBDu\.[
LD0D
8V0tc0 U5c01,2]
10Vc55cDB5c,Qb
:D0BD0D05DBQc,111
:D1cB\UtcDD_.[
1D0VBD0W,12Q
5QDc01,Q],M
5Q Hc01,M
LD0D,DBcB55Bt,U5c01,b[
LD0D,cXB5,DBDDB\c0 00W.1b[
L1ccD, \P5c_DcD\c0\UDD_,b1
LD0w, QuDD:D_,B5 D00cD_
0cV:c,1[2
5W0t\D,OBV:0,0c5:_D\c0DD:QUc01,
P+
LQ0X,u5c501,2[
LUcBQ\u5,:DBDDB\c0D0W,1b[
cB\utc tD@:
B510 BQ0D1,[
c0D5\tuc\:0D01, b-bb,[-[
0:55BD01D_01,2
_U:D_01, [q-[
1DQ0t\BDcc01,b[
5c_DcD\510t,q-qb
W\DQ0t\0c5.b[
01!0HD5TUtD:D_,]
D5H.
Du11D_01,b
1DB c0B15,b
10tQt00UC\:0DD0W5,12b
D0D_055,V:t\uc501,b
QUtQ05c01,b
BDcW000,1BV0tc0 u5c01,2]
B\Dc55.
DcKD_W\D5\tB_H\c0_c,O
J
1DQ0OBDcc01,1
B\\cDD_.
01Bt_c t:D_5,1bO
W\D05c5BD0ct,qQ,2-]
0t5\DctD\,0cD\ctc0W\D00Wc1,1b1
UDI!utc,01!utDc010tD5,1b
LBD_cu\\D_.
015c_DcD\5,[,b. 12q
01\DD5\tQ5,[[q
LBU0:.PD\0D0.0c5:_D\cDD:Quc01,
PJ
LcDcB D5Dc5._tccD BD01. b
L5DD,Lc5,Q
LcDD,|0HD,1q
LUc:
c11c0\ 0D5BtQc0_c5,2}
5B1c\01,2Q
5QtcB0D_01,2
5Qucc2c0u\ 01,qO
\yQc501,2b-2[
LUcD0K5,\cDQ0tBt,QQ
UE]ODI5,ED@I0DIOED-@I0D,
LUE5QUEEZE,VOD@,1
LUE,OgUtEIDDE,1
LUD@:
IODEIHE,1O
CEDDE55WDED,}g
ED@tDIOED@tD. 1}1b
KEgOUI5 WDED,}b
OEIUtE7D@,}b
OD-VE55E5,bg
OCO5DEtOD,1g2
OOgWOO5,2,
33
OU@E,DOW, 5DQD@VE55EWIDb
LtD OtEDIIOD:
COHQOH5E,
HOtDCEO,22
DEIUtE tD@,}b
DHDIEDOW,1}
D 5E@HEDI5, 11
D 5IVE5,11,bO1
DCOHQIDIgO, 1,21
HD@EHEDIO,
OHOD5DE,11O,112
O5DE5,1
OVE55ED5E,
MDQDE.CED5 U5It\Q5,}
MtI,EOt@E,1g
MEI.EEC OD@UEDE5,11bU
MO

DD,bIEDED,11
MOWD:
OtHIEtD@]@,157
Ot5DD@]@,qg
OtIDE5W5E,
HOtIDCE O. 1
WDED5WD@5IVE5,2
MO-D,HQOTDCEO, 31
MOE5,IQEtE,1
MO

5E U@ O,bO
VOIE U5EO, 2
DHOD5DQE,111
DEIUtEtD@,}
tODEH5O.2g
5QDEO, 11b
MO5ECHQ5:
OtCgDEt5,12
Ot@UD@5QDEtE,1g2
U5EO,
MOI-HEIUE,U5E5O, 2
CO5DEtOD:
EDODO,1
@UD@O,1g2
D5DEtC5DE, 11gO
HIEtD

EO, 1}
ItUDCIE, g O
DDDDKtI:
5E5@D5OUICE,}
DEIUtEtD@,}
DI WOOIUtDD@, EDE, 12
tODWOO,U5EO, 22
1ItD,
1@:
VOIE U5EO, 2
DOV 5DQE.112
HDIODO, 1 g
Ot CHQD@,gb
OtCtO55CUIID@5E@HEDI5, 1O
,,'
OttUH5DEt,1O-1O, 1, 1b
I7
Ot@UD@,}g},g, 12g
Ot HIEtE DD@5,12
OtHEt1D@5E@HEDI5,1}
OtHEtD@ItD@E5,1g1
OttOUID@HODDOICDE5,111
Ot 5DD@,}b.1O
Ot5WD@5E@HEDI5,1O2-1O
Ot5WD@5IVE5,1
Ot5QIID@tD@5,12
Ot ItHHD@5IIQ5,1Ob1O
HQOTDCEO,31
1ODI,Ot VE55ED5E,
JODIUtE,CU5E5O, 21
LCtWOO,HEU5EO, 22
LHDI\OD5:
CUIID@DIO,112-11g
E5@DO, 105
OI5VE COD5\tMC\OD,1g
@UEOt.O
HU-@EDEtIOD, 1Og-1O)
5I\D@O, 1g
WIDO, 101
LIDE:
5 C!ND@IOO, 11
5CHQDEV\CE,b2
5ZEO. 13
LIVED,U,, 1}
L@DID@, HQOtDCEO, 1g
LHD,U5EO, 2?
NCDDEOO5,5IO, 1
NtODE, \H1IEU5EO, 22
NDO@Dg,1HIEU5EO,22
NDO@Dg,HOUDID,VOtEU5E O,

NE:
HDEDOWO,1}
HDOD O, 1g
5QDEO. g2,
NEVEDEEt,5QIDEO, Ub
NEDt5EgE
Ot5QDEtE QU@5, \gg
VOtE U5E O, 2
NQE:CUtg:
5D5EQU@,}
HOD 5DQEO, 11O
DOV 5D,112
5QDEO,
5IVE5O, 1
VOIE U5EO, 2
NQE:Dt:
VOIE U5EO,2
WIDEXOIC5,2g
NQE,5OI,HIEU5EO, 22
NQE,5QTE,VOtE U5EO,2
NIQUEItgE5@D5:
D OtDOE5,
OD U@5,1gb
N5KD@IQE,5 @UE tE55I,75
NLl:
OtHKD@IE5\ CUI5,1
5WD_O,
U5E5O , g
NE5QUIE:VOtEU5EO, 2
N5IE5:COQD@WD, 12,1,1}
NIEtD@E:
E\DIODO, 1
HE5UtEHEDI O,132
NIEtEDCE,Ot CtO55CUI 5E,g
NIEt5W:
UXtgEDCEOt,1
5E@HEDICUIID@WID,2
VEt1C CCUtCgO, ]2
AIEt5:COHOUD:
CCUIODO,b-O
U5EO, b
AO\5IUtECODIEDI: HE5UID@
C, 2O21
AOt15ED-IEDOD,Ot VE55ED5E,
Ag7IEWOO,VOtEU5EO,2
5D:E,,11b
LK:
Dt7E5O, b
H\IEU5EO,22
LCDEOD,HttD@EO,1}
LgWOO,@UD@O, 2,2}, 1
LtEDIIOD,O2@-5@QIIEtD5, 1O
L5_E-OtD@E,U5EO, 22
LV 5DQE5:
D HDIEDO7,112
IECDDgUEO, 12
UK: HIEU5EO,22,2g
H 5DEt,DHDIE 5IVE5,
11
7IDOO,Ot5QIIDtD5, b,b2
b,1Og,1
EDI@OD5,D5t lU@9,150-151
EI5HHOD,VOtEU5EO, 2
, EDIODO }
DKVOt}, VOtEU5EO, 2
IODC5O5,EDIODO,1
U@5:
OtD5E,
OtECOt0I\VE EECI, 1
Ot 5QDEtE.1gq
O!gCDtOHICWOOIUtDD@,E\DE,

OgUtEIDDE@UE,U5E5O, 2}
OQI,HIEU5EO, 22
OIDOE5'
D EIUtEID@,}
IUtDD@O,1
OWEIIOO5,15IO, \g
OZ5E5.LEDE,E5@DIECDDgUEO, g
tOUCIODIUtDD@,12
UtEDEII:
VOtE U5EO. 2
DDEtD@5,11b
DHOD5DQE,11O
1DCIUtE5,154
5QDEO. 1g2
VP@UE:
OEDIHEWID,g
U5E5O,22}
KEWOO, HIE U5EO,22
KE5\tIOt:
\tDEHEI,15
HQOIIDCEO, 17
KE5tIOtgDZt:
O@UE,2
OMF, 55
OWOOU5I.\}
KEItEV IOO:
OtCUI 5E@HEDI5,g
IE5,2.1O
WID5DD@]@,gg
KEVEt5EHOUDID@
Ot@EVE55E!,1g
O5E@HEDE tDDOD, 171
IECDDQUE O, b
XD@5,D\tD@0,11
X05cW000:
BtB2llBD,2
c0c0D00,2
LB5t D0BD,2,gb
BDcWw0, 2
U0D0UtB5,2
gBU ctI0,2
\Ugwo,2]
X0U\ct,0t0lBD0D0shaps, 111
XUDj0D\,
preparation 0,]b]g
\ccDDgUc50, qO
XUDDctDBD05:
cBDglD@WtD,2b,qO,qb,g.Og
for OBDgD@5c@DcD\5,1bO-1(O
0t@1UD@tD@5,12q
not with polyurethane @Uc,qg
oBDIc 5BW,0t5g\\D@tD@5,1g
5Bc\@Ul0cDc5, !b
oBD0D@:
@t\5W\D.b
ID50cVc55c,1bq
g0Wct,b
[email protected],12
5BD0D@@,q]
5BD0D@DO,11b
5BD0D@5cBcf.
BgglcB\0D0,gg
0t 5c

DcD\c0DW,1212b
U5c0,bb
5cDUct, N Uc,g
octBgct: 0BD0D0QDtc0,1,g
0BD0D0g0Dtc0, g
5ctBgct:
caDlDc\,0tD50cclcBDUg,1bq
caDDc!,5D00\DD@with, b
0aD0D0-g0D\, gb
t0UD0, 1
5DcBt,5DO0\D@W\D,b
5\cWBt\,bbb
5ctcW5,wo,for B55cDD,1]
ocUg\UIB 0ID,B\\tBt0D 0,b
oc@DcD\BD

c,0cIDlt0D 0, bb,1[b
bc@DcD\cD5:
05c5BD0D@0,b
teeDBD05BD0lD@0,bO,11-12
@UD@ W\D0Ut 5BD0D@,12q
5c@DcD\Dc@t,0cDt\0D0, bb
5c@DcD\ cD@\:
cBcUB\0D0,b[
0cD\0D 0,bb,1[b
Dg0I\BDcc0,[g-bO
Segment W0\D,0c1D\0D0,bb
oc@DcD\c0c0D5\tU\0D:
0cDc0,g
cccDc0, b1,12
D0D0\0Dj0, 1
of t1DD0D5,1b[
0
_
D5\c,12
o0U\DWc$\D0BD0c5@D5,bb
oc@DcD\c0 tD@5,
B5 half.rings. qO-q1
cU\\D@0,]b[
0cDc0,1
0tjl\\D@0,]g
0tcB\UtctD@,q
B\tcDD@0,qg
@UID@0, q[
lBt@c, 11g
DUDDet0, qO,p
5BWD@0t0ct0,b1
oc\2DBD.1Dc0D,g
5DUlct,NXc,g,technique 0, 11b12O
5@DB\Utc,0wo turner, b[
Sketch:
c0DgU\ct0t,[O
graph gBgct0t,bb
Dg0ttBDcc0,bb,b[
0!0DD5UDD@,[O-[1
[email protected],bg
o0ccctDBl,@e0Dc\tj0,1qg1O
o0U\DWc5\D0BD0@D5,bbb[
ogDctc:
Dcc0Dc5Vc55c,1qb1qg
[email protected]
5gl\\D@0,1qb
\UtDD@0, 1q-1qq
Wl\DBD@lc0th@5,1qb-1qg
5gDc:
0tBD@c0 DD@5,11b
in c05BDc0I0D,1q2
IBDDB\0D 0, g]
5gtB @Uc,U5c50,2[
5gtD@cU,OBVd,)g
5\BVc5:
0cDc0,11
@Uh@0,g]-ggg
DUDDct0,b,bg
Vc55clc0D5\tUct0Dwith. bg
o\0gD0cX:
c0DQUD0W\ctc0,1bg
0tct055CU\ 5lc0, j
0t5BD0D@]@,q
0t5BWh@ 5tBVc5, g2
0t\tDD u@]1@,1Og
0t2l@-2B@gBttctD5,1O2
DBtcBl0t, 1
glV0tD@,1Ob
settin
g
0, 2
Strai
g
ht edge:
0t

BU

@UtVc,1b]
\0cB\Dc55,O1
5UQct@Uc,U5c50, 2[
oDDc\t,Dg0t\BDcc0,b
BDc5BW:
ct055cU\5c00t,
5c@DcD\cU\lD@WtD,]]]
size 0,!
BcXc0\,ccBD1D@W1,bb
Bgc,B5@Uctc55\,gg
caX, D\c0U5c0,22
cDgB\c,0tg0t\D0c5,1]
Tetrahedron, DtctBD@c0, 1[b
cXDc55D@,WltD0IUD 5BD0ct,1]
DD5\tg5,CU\\D@0,[,[][q
\lc5,Dg0OBDCc0, 1[]
0015,IorWwd\u1D@,1]
[email protected]\ctc0,5BWD@0, 1qO-1q:
UgW000,BV0tc0 U5c0, 2]
UIcBtc5 H@Uc,U5c50,2[
VcDcct:
B55cDD0,b
clBDglD

0, [
@UlD@0,1q-1
0DBgc,g]
oWBlDU\,g
5c@DcD\50.b
Vc55cOtD:
BD0tH@BD@lc,11b.11g
c05c0,b]
00DU\,1]1
lDBlshaping 0, b]
1DQ0t\BDcc0,bbq
0Qeh-b
0V0l0,12g
QWcIcBIVD

0,1O
5l0gcBD@c0,bg
Vc55c5,Bt@c,\UDD@5gcc00, 1bq
V005,@lUc0tlD@,2[
WB\DcXDe55:
0c\ctDDB\0D0, gb
DB2D@g[
Dg0t\BDcc0, bq
D0VBD0W,12g
In porthole ring, 1b
DBDB@cDcD\0, b2,bq,1b]
05c@DcDtc0tDDoD,1[1
WBDU\:
0clD@DBcK, [
5gDc0,g
U5c0,22
WBDU\,VcDcct0,D 0VB5Bgc.112,
u,
WcD

c:
0r 5gDctcgU@5,1qq
\D 5gDc,1q2
U5c0,22
WcDDlD@ct,NB@DU5,1q]
W000:
choice of. 22
c0l0t0,2q,bq
0cD5t0,2],[]
0lDcD50DD

0,2
0tD@5Ded0t,20
lB\\cD u@0,2
@tB1D0tcD\B\0D0, [b
lD5gtBt0Dt0D,bq
U5B@c0,11b
D0l5\Utcc0DtcD\0,1g-2O
D0VcDcDt0.1g,21
00c\0.2q
0llDc550, 2
gtcgBtB\0D0,2q-2,bb
source tor,g1
5\0tB@c0. 2O
W0000U5\.Inhalation 0. 1[
Yc0WcBt\,BV0tc0 U5c0, 2]
ZcDtBW000,Dtc0U5c0,22
Z@-2B@ 5Bg85:
CIcB\0D0, bb
gIccl5c\\D@1D,1O1
5DBt

Dc550,1O2
tc01c, BV0tc0U5c0, 2]