Connection Methods For Timber Engineering

1
CONNECTION METHODS for TIMBER

ENGINEERING.
TRADITIONAL CONNECTION METHODS.
The traditional connection methods consist mainly of:

DOWELS, BOLTS, LAG SCREWS, SPLIT RINGS, SHEAR PLATES,
NAILS & SPIKES.
2
These connection types have all been used for a long time, but
they are all suffering from developing large deformations
during loading. In the past this has not been
considered a deficiency.
However, with more stringent performance requirements the

above jointing methods will no longer be satisfactory for
structures larger than houses.
Unfortunately, not much testing information is available on the

strength properties of the mentioned connections. In some cases
it is not even known if the data quoted in the design code are
based upon a strength criterion, or are governed
by some deflection limit (slip).
In most cases only single connectors have been used and the
effect of shear-lag has not been investigated sufficiently.
(connectors in a row)
3
With the introduction of “Limit States Design” and of reliability

considerations some years ago it was - due to lack of data - not
possible to do more than an approximate soft conversion
on the code section dealing with fasteners.
This was not even confirmed
by relevant testing.
The trend towards better performance requirements also raises

questions about the adequacy of the traditional
jointing methods for larger structures.
The deficiency in the connectors can in many cases be traced to

the use of oversized holes, too short end distances,
or shrinkage and swelling caused by
changing climatic conditions.
Few of the traditional connection methods can withstand

reversal of the forces which is an important
consideration for earthquake design
as well as in determination
of wind effects.
What I am trying to say politely is that our knowledge of

the strength properties of the traditional timber
connections are seriously absent.
4
CONTEMPORARY CONNECTION METHODS.
In the late 1950’s or early 1960’s three new connection

methods were introduced:
A) Truss Plates
B) Glulam Rivets
C) Glued-in Steel Rods Parallel to the Grain.
A) The truss plate invention changed the manner in which roof

trusses for houses were built.
Rather than construct rafter roofs on the job site, the new trusses
are factory assembled and shipped to the job site for installation.
The idea rapidly engulfed all of the
worlds developed countries.
5
The annual sale of trusses in North America is in

excess of $ 2.8 billion.
The trusses are ideal for houses, but should be limited for use in
buildings with clear spans of less than 10-12m.
6
B) The second innovation -the Glulam Rivet - is designed

specifically for use in large buildings.
It consists of a predrilled structural steel plate in which glulam

rivets having an oval shank are driven into the timber member.
7
The close spacing of the rivets and the conical shape of the
head creates a very stiff joint, which is able to
withstand reversal of the forces.
The glulam connection is capable of transferring bending

moments and can therefore be used for
statically indeterminate structures.
8
9
10
C) In the late 1970’s the development of a new connection.

method was reported in Scandinavia.
It consist of inserting steel rods into holes drilled parallel to the

grain of the glulam member.
(i.e. end grain).
Using Epoxy glue or Resorcinol glues it was possible to reliably

transmit a load from the rod into the glulam member.
This was the first reporting of a successful application of gluing

steel rods to timber.
It was used only for cases where the steel could be placed
parallel to the grain.
11
In Russia Mr. S. B. Turkowskij worked with the idea of

enhancing the strength properties of glulam
using reinforcing bars.
He started by placing the re-bars perpendicular to the grain in

places where the glulam was subjected to bearing forces.
Later he also used re-bars placed at a 300 angle to the grain to

reinforce the glulam members for high shear stresses.
12
13
14
In New Zealand some recent work with steel rods was

reported for knee joints.
Here the rods were again placed parallel to the grain in the
column and rafter members, but steel weldmends were
used to guide the forces past the knee.
15
However, there are some disadvantages when the rods are

placed parallel to the grain of the glulam.
One disadvantage is that the rod is often placed in a horizontal
plane so the gluing, therefore, has to be injected under some
pressure to avoid the forming of air pockets. This would
impair the bond between the steel and the wood
unless special techniques are used.
Another disadvantage is that the rods will be placed close to the

top and bottom of the glulam member thereby creating skin
type joints where only the top and the bottom of the
glulam carry the load from the rods.
16
GLUED-IN RODS IN CANADA.
At U.B.C. an investigation to expand the use of glued-in rods

was commenced in the late 1980’s.
The goal was to develop one or more reliable connection

methods that would meet today’s performance
requirements; they are:
HIGH STRENGTH
HIGH STIFFNESS
AVOID BRITTLE FAILURE
TOLERATE REVERSE LOADING
LOADS TRANSFERRED VIA SPECIFIED PATH
SIMPLICITY OF DESIGN
EASE OF MANUFACTURING
CONSTRUCTION FRIENDLY
ATTRACTIVE APPEARANCE
NO FIELD WELDING
NO FIELD GLUING
PROVIDE FOR CORROSIVE ENVIRONMENT (IF
NEEDED)
FIRE PROTECTION
COST
17
NEW INFORMATION REQUIRED.
Behaviour of the glulam and the glued-in rods (strength and

stiffness, for different sizes and lengths).
Behaviour of glued-in rods at an angle and perpendicular to

grain.
Behaviour when subjected to climatic changes.
INTENDED END-USE.
The primary applications were “beam to column connections”

and “knee joint connections”
18
19
It turned out that several other types of connections also could

be derived from the testing conducted.
The basic idea was to use steel anchor plates glued to the
corners of the glulam members and then use straps between the
adjoining anchor plates to transfer the forces
from one member to the next.
The anchor plates would have pre-welded rods on their

underside which would be inserted into drilled holes in the
glulam forming a 30 0angle. Epoxy glues were used.
20
21
The advantages of using glued-in rods at an angle are that the

rod will engage more of the glulam cross section to
the carry the load and at the same time will act as
shear reinforcing for the glulam member.
A disadvantage is that large compression perpendicular to grain

forces are generated where the rod is bent.
COMPRESSION PERPENDICULAR TO THE GRAIN.
According to ASTM - D 143 the perpendicular to grain strength

could be established by a simple test as shown. The stress
is equal to P/A; (force divided by area).
22
While that method is simple, it is - unfortunately - a deception.
A small experiment will illustrate the problem.
Closer examination of the compressing perp. behaviour reveals

that four different component are active in providing
the bearing capacity of glulam.
The active components are:
1) The area of the plate.

2) The length of the plate edges parallel to the grain.
3) The length of the plate edges perpendicular to grain
4) The magnitude of the indentation imposed.
23
A deeper insight can be had by testing bearing plates

of different sizes.
24
25
26
From these tests it was then possible to obtain the contribution

of the four types of behaviour observed.
27
A formula was developed from the data and the predictions were
compared with the test results.
The proposed method of prediction was then compared with the

ASTM method.
The above comparison showed that for small bearing plates the
ASTM method is grossly conservative, while for large
bearing plates it is un-conservative.
28
INCREASING THE BEARING CAPACITY.
A substantial gain in the bearing capacity of compression perp.

can be achieved by combining the usual bearing plate
with one or more rods glued into the glulam
member perpendicular to the grain.
29
Testing was conducted and showed that it was possible

almost to double the carrying capacity.
30
The minimum end distance was established and a design method

was prepared by which the effect of the rod could be
combined with the effect of the bearing plate.
EFFECTS OF CLIMATIC CHANGES FOR

GLUED-IN RODS.
Because steel and wood have very different expansion

coefficients the behaviour of the two materials
glued together was investigated.
31
The specimen chosen consists of a 178 x 250 x 450 mm glulam

block having a 16mm dia x 905 mm long rod glued
into the center of the block.
32
Three types of tests were conducted:
Control test (n=6),

Moisture Content (M.C.) test (n=17)
Temperature test (n=7)
33
The M. C. specimens were cycled 5 times from 10% to 30%

taking a total of 3 months to complete.
34
The temperature specimens were subjected to five cycles of

variation of the temperature from
-30 0 C to +50 0 C.
The results showed that neither of the two “torture” methods

had weakened the specimens.
35
TESTING FULL SIZED SPECIMENS.
Using the new information we could now design practical joints

equal to those of existing steel structures and test the
performance of the glued-in rods.
Racking tests were performed first and found to be satisfactory.
Then one-sided beam to column tests were done which also

passed with flying colours.
36
37
Tests with beams having columns top and bottom were also
carried out and showed that they amply passed our criteria,
except for interstory drift caused by wind loads.
38
39
This was followed by testing of knee joints with different slopes.

40
41
The final test was a long-term load experiment in which the

design load was present for seven months while the
climatic conditions changed four times.
42
43
44
45
SUMMARY.
It was shown that the traditional connections methods for timber

do not live up to today’s needs for stricter performance
requirements
Three types of contemporary connection methods were

described, showing that it is indeed possible to
implement better jointing methods.
Research work conducted at UBC dealing with the use of

glued-in rods was explained.
The advantage of having the rod forming a 300 angle is that a

greater portion of the cross section will contribute to the
transfer of the forces; thus providing grater reliability.
The drawback is that additional bearing stresses are created.
However, a review of the behaviour of the bearing stresses

show that our present concept of the compression perp.
stresses is underrating their actual capabilities.
Two simple calculation methods are presented that will enable

the design engineers to take advantage of the
real compression perp. strength.
THE USE OF STEEL RODS WILL ENABLE US TO DESIGN

TIMBER CONNECTIONS WITH THE SAME STRENGTH
AND RELIABILITY AS THAT OF STEEL.

Connection Methods For Timber Engineering

Uploaded by

Copyright:

Available Formats

Connection Methods For Timber Engineering

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Connection Methods For Timber Engineering

Uploaded by

Copyright:

Available Formats

1

CONNECTION METHODS for TIMBER

TRADITIONAL CONNECTION METHODS.

The traditional connection methods consist mainly of:

However, with more stringent performance requirements the

Unfortunately, not much testing information is available on the

With the introduction of “Limit States Design” and of reliability

The trend towards better performance requirements also raises

The deficiency in the connectors can in many cases be traced to

Few of the traditional connection methods can withstand

What I am trying to say politely is that our knowledge of

CONTEMPORARY CONNECTION METHODS.

In the late 1950’s or early 1960’s three new connection

A) The truss plate invention changed the manner in which roof

The annual sale of trusses in North America is in

B) The second innovation -the Glulam Rivet - is designed

It consists of a predrilled structural steel plate in which glulam

The glulam connection is capable of transferring bending

C) In the late 1970’s the development of a new connection.

It consist of inserting steel rods into holes drilled parallel to the

Using Epoxy glue or Resorcinol glues it was possible to reliably

This was the first reporting of a successful application of gluing

In Russia Mr. S. B. Turkowskij worked with the idea of

He started by placing the re-bars perpendicular to the grain in

Later he also used re-bars placed at a 300 angle to the grain to

In New Zealand some recent work with steel rods was

However, there are some disadvantages when the rods are

Another disadvantage is that the rods will be placed close to the

GLUED-IN RODS IN CANADA.

At U.B.C. an investigation to expand the use of glued-in rods

The goal was to develop one or more reliable connection

NEW INFORMATION REQUIRED.

Behaviour of the glulam and the glued-in rods (strength and

Behaviour of glued-in rods at an angle and perpendicular to

Behaviour when subjected to climatic changes.

The primary applications were “beam to column connections”

It turned out that several other types of connections also could

The anchor plates would have pre-welded rods on their

The advantages of using glued-in rods at an angle are that the

A disadvantage is that large compression perpendicular to grain

COMPRESSION PERPENDICULAR TO THE GRAIN.

According to ASTM - D 143 the perpendicular to grain strength

While that method is simple, it is - unfortunately - a deception.

A small experiment will illustrate the problem.

Closer examination of the compressing perp. behaviour reveals

The active components are:

1) The area of the plate.

A deeper insight can be had by testing bearing plates

From these tests it was then possible to obtain the contribution

The proposed method of prediction was then compared with the

INCREASING THE BEARING CAPACITY.

A substantial gain in the bearing capacity of compression perp.

Testing was conducted and showed that it was possible

The minimum end distance was established and a design method

EFFECTS OF CLIMATIC CHANGES FOR

Because steel and wood have very different expansion

The specimen chosen consists of a 178 x 250 x 450 mm glulam

Three types of tests were conducted:

Control test (n=6),