ANALYSIS OF CUTTING FORCES IN STRAIGHT-KNIFE

PERIPHERAL CUTTING OF WOOD

Piotr Iskra
Postdoctoral Fellow
E-mail: [email protected]

Roger E. Hernández*{
Professor
Centre de Recherche sur le bois
Département des Sciences du bois et de la Forêt
Pavillon Gene-H. Kruger
2425, rue de la Terrasse
Université Laval
Québec (Québec) Canada G1V 0A6
E-mail: [email protected]
(Received September 2011)

Abstract: Effects of rake angle, wavelength, and depth of cut on cutting forces during straight-knife
peripheral milling of sugar maple were determined. Progressions of normal (FN) and parallel (FP) force
components during a typical up-milling cycle were determined. Results showed that at the initial step of
the cut, FN was negative, ie the knife edge pushed the workpiece downward. The part of the knife path that
remained visible on the machined surface was always created in such a situation. As the knife edge
advanced in the cutting path, FN increased, reached a maximum negative value, then decreased to 0 N/mm,
turned positive, and continued upward. FP was always positive and increased as chip thickness increased.
Positive FN and FP reached a maximum just prior to emergence of the knife from the workpiece at about the
maximum thickness position. However, when chip splitting occurred, the maximum positive FN and FP
occurred before that maximum thickness position was reached. Maximum negative FN and FP increased as
rake angle decreased and as wavelength increased. Maximum positive FN increased as rake angle and
wavelength increased. Maximum positive FN and FP were also affected by depth of cut but to a lesser degree.
Impact of these maximum cutting forces on production of defects was analyzed, and ways to decrease them
were discussed.
Keywords: Cutting forces, peripheral cutting, up-milling, sugar maple.

INTRODUCTION peripheral cutting cycle, the cut occurs essen-

tially parallel to the wood grain, and chip
In peripheral cutting, material is removed from
thickness is minute. This condition favors the
the workpiece in the form of individual chips.
production of Franz type II chips, which fail in
These chips are formed by intermittent engage-
shear, tend to form continuous spirals, and are
ment of knives mounted on the periphery of a
associated with good surface quality (Koch
rotating cutterhead. The resultant surface con-
1985; Hoadley 2000). This initial portion of
sists of a series of individual waves produced
the cutting path will remain visible on the new
by the successive engagement of each knife
surface. As the knife movement progresses, chip
(Koch 1985). Peripheral up-milling is the most
thickness increases and attains its maximum
common process used for planing wood along
value just before the knife leaves the workpiece.
the grain. At the beginning of an up-milling
However, when the knife is close to leaving the
cutting zone, the cut is done at a certain angle to
* Corresponding author the grain. As a result, at the end of the cutting
{ SWST member cycle, a Franz type I chip, which is formed by

Wood and Fiber Science, 44(2), 2012, pp. 134-144

# 2012 by the Society of Wood Science and Technology
 Iskra and Hernández—CUTTING FORCES IN PERIPHERAL CUTTING OF WOOD 135

splitting ahead of the tool, is often produced wood density (Woodson and Koch 1970;
(Koch 1985; Hoadley 2000). Axelsson et al 1993), moisture content, and tem-
perature (Axelsson et al 1993) have been studied.
The resultant force applied by the tool on the
However, data on forces produced in peripheral
workpiece during the cut may be decomposed
cutting are still scarce, and additional methods
in a parallel (FP) and a normal (FN) component should be developed to facilitate interpretation
in relation to the surface produced (Fig 1). Mag- and comparison of results.
nitude and direction of these components depend
on a number of factors related to the cutting tool, Palmqvist and Johansson (1999) presented a
cutting conditions, and workpiece characteristics method for measuring forces during straight-
(Koch 1985). The normal cutting force compo- knife peripheral cutting. Changes in force com-
nent plays an important role on surface quality. ponents during up-milling a single chip were
In planing along the grain, normal forces can act described. These authors demonstrated that the
downward, compress the superficial tissues, and parallel force increases gradually during the cut,
might provoke cell crushing and raised grain whereas the normal force is downward at the
(Fig 1) (River and Miniutti 1975; Stewart 1980; beginning and becomes upward from the middle
Stewart and Crist 1982). Detriments of crushing to the end of the cutting cycle. The normal force
superficial tissues have been extensively explained attains its highest upward value close to the end
in previous works (River and Miniutti 1975; of the cutting cycle.
Hernández and Naderi 2001; de Moura and This study evaluated effect of rake angle, cutting
Hernández 2006). In contrast, normal forces can depth, and wavelength on behavior of normal
also act upward and increase the risk of torn and parallel force components in peripheral
grain as well as distort and weaken the structure straight-knife planing of sugar maple wood.
of superficial tissues (Fig 1) (Palmqvist and
Johansson 1999; de Moura and Hernández 2006).
Thus, the normal force component should be as MATERIALS AND METHODS
low as possible. It is therefore very important to
know how cutting conditions affect cutting forces. Sugar maple (Acer saccharum Marsh.) wood, a
diffuse-porous hardwood commonly used for
Several studies have been done on forces indoor applications, was selected for this study.
generated in orthogonal cutting. Thus, effects Four air-dried flat-sawn samples were stored in
of rake angle, cutting depth (Woodson and a conditioning room at 20 C and 40% RH until
Koch 1970; Axelsson et al 1993), knife wear they reached 8% equilibrium moisture content.
(Axelsson et al 1993; Hernández and de Moura After conditioning, samples were straight grain-
2002; Hernández and Rojas 2002) as well as oriented and cut to 150 mm long (L), 50 mm
wide (T), and 19 mm thick (R). Average and
standard deviation of basic density of the sam-
ples were 610 and 17 kg/m3, respectively. Each
sample underwent all cutting tests. Up-milling
was performed with a straight-knife cutterhead
that had 70 mm of cutting radius and was
mounted on the horizontal shaft of a milling
machine. Only one of the four knives on the
cutterhead was set for cutting. The remaining
knives served as counterbalance. High speed
steel knives were freshly sharpened with three
Figure 1. Parallel FP and normal force FN components different geometries. Rake and knife angles of
during up-milling. Diagram shows that FN can act upward the knives were 10-50 , 20-40 , and 30-30 ,
(positive value) or downward (negative value). respectively. Clearance angle was always 30 .
136 WOOD AND FIBER SCIENCE, APRIL 2012, V. 44(2)

Roughness of the rake face of the knives, near was considered negligible and not measured given
the cutting edge, was measured with a Micro- that tests were performed with a straight-knife
measure confocal microscope. A surface of cutterhead. Ten successive knife cuts (or cutting
40 mm2 was analyzed per knife. Data were col- cycles) for each pass and sample were examined.
lected with Surface Map 2.4.13 software using This corresponded to the analysis of 480 knife
an acquisition frequency of 300 Hz and a scan- cuts for the entire experiment. All negative paral-
ning rate of 3 mm/s. Average roughness (Sa) was lel forces as well as their corresponding normal
determined using Mountain Software (Besancon, forces were set to zero as proposed by Palmqvist
France) based on ISO (1997). A cutoff length of and Johansson (1999). No additional filtering was
0.25 mm combined with a Gaussian filter (ISO performed on the original data. According to
1996) were used to calculate mean roughness Palmqvist (2003), the pushing action of normal
(Sa) of the rake face for each knife used. force was represented by negative values and the
pulling action by positive values (Fig 1).
Samples were fed parallel to the grain at 0.625 m/
min. Rotation speed of the cutterhead was set to
obtain 15 and 27 knife marks per 25.4 mm of
RESULTS AND DISCUSSION
length. This corresponded to 1.75- and 0.93-mm
wavelengths, respectively. Two cutting depths were Typical progressions of normal, parallel, and
also tested: 0.25 and 0.50 mm. Variables tested in resultant cutting forces during one cycle are
the experiments are summarized in Table 1. shown in Fig 2. The theoretical undeformed
chip shape and the instantaneous progression
During machining, wood samples were fastened
of the chip thickness are also represented. Chip
to a Kistler (Winterthour, Switzerland) 9257B
shape, derived from the parametric equation of
quartz three-component dynamometer, which was
movement of the tool and workpiece as well as
fixed to the feeding table of the milling machine.
instantaneous chip thickness were calculated based
Given this assembly, forces during peripheral
on formulas described by Martellotti (1941).
milling were measured in relation to the feed
direction and not in relation to the tool motion in At the initial step of the cut, FN was negative, ie
the workpiece (as usually referred to when orthog- the knife edge pushed the workpiece downward
onal cutting). A charge amplifier Kistler type (Fig 2). For all cutting conditions studied, anal-
5010B equipped with a 180-kHz analog low-pass yses of the knife cuts showed that the part of the
filter was used to amplify and condition input knife path that remained visible on the machined
data. A short time constant (1 s) was chosen to surface was always created under such conditions.
allow measurement of the fast time-variant cutting As the knife edge advanced in the cutting path,
forces. Normal (FN) and parallel (FP) components normal force increased, reached a maximum
of the cutting force signals were digitized at a rate negative value, then decreased to 0 N/mm,
of 16,000 samples per second and stored into a PC turned positive, and continued upward. FP was
disk memory for further analysis. Lateral force always positive and increased as chip thickness

Table 1. Variables tested in the experiments and related cutting parameters.

Variables Values

Rake angle (deg) 10, 20, and 30

Wavelength (mm) 0.93 1.75
Depth of cut (mm) 0.25 0.50 0.25 0.50
Feed speed (m/min) 0.625 0.625 0.625 0.625
Cutting speed (m/s) 4.95 4.95 2.62 2.62
Length of a single cutting path (mm) 5.93 8.39 5.94 8.40
Average chip thickness (mm) 0.039 0.055 0.074 0.104
Maximum chip thickness (mm) 0.066 0.098 0.104 0.165
 Iskra and Hernández—CUTTING FORCES IN PERIPHERAL CUTTING OF WOOD 137

Figure 2. Progression of parallel and normal cutting forces during an individual chip length along with the theoretical chip
shape and chip thickness. Resultant force is also represented by successive vectors.

increased. FN and FP should both reach a maxi- theoretical maximum chip thickness position. As
mum just prior to emergence of the knife from a result, a splitting phenomenon occurred, causing
the workpiece (at the maximum thickness posi- earlier release of the cutting force components.
tion). However, Fig 2 shows that maximum
A detailed analysis on the maximum force com-
positive FN and maximum FP components do
ponents reached during a complete cycle of cut
not correspond exactly with maximum chip
was conducted. Because the normal force com-
thickness. This slight offset could occur because
ponent changed from negative to positive during
the chip shape shown is theoretical (undeformed),
the cutting cycle, both maximum negative and
and it is understood that the acting forces tend
positive forces were considered separately.
to stress and deform the workpiece body and
the chip. In other cases, maximum FN and FP
occurred before that maximum chip thickness
Parallel Force Component
was reached. This occurred when the FN compo-
nent exceeded tensile strength perpendicular to The contribution of the parallel component of
the wood grain before the knife reached the the cutting forces to machined surface quality
138 WOOD AND FIBER SCIENCE, APRIL 2012, V. 44(2)

may appear less important compared with that

<0.0001
<0.0001
Table 2. Analysis of variance (rake angle, wavelength, depth of cut) for the maximum force components obtained during a peripheral up-milling knife cut of sugar

0.3967
0.1017
0.0003
0.2121
0.0001
p value
associated with the normal component. How-

Maximum positive FN
ever, the parallel force may be related to produc-
tion of defects such as torn edges when the tool
leaves the workpiece at the end of the milling
operation (Palmqvist 2003). The magnitude of

0.72
2.30
8.39
1.56
93.45

10.01
267.71
F-ratio
the parallel force, in particular its maximum,
will affect the occurrence of this defect type.
To determine the influence of independent vari-
ables on maximum parallel force, an n-way
analysis of variance (ANOVA) was performed.

<0.0001
0.0026
0.8511
0.0004
0.0201
0.6800
0.1168
p value
The three independent variables (rake angle,

Maximum negative FN
wavelength, and depth of cut) and their two-
and three-way interactions had a statistically
significant effect on maximum parallel force
(Table 2).

9.19
0.04
8.01
3.94
0.17
2.16
549.37
F-ratio
Within the range of cutting conditions studied,
effect of rake angle on maximum FP was the
highest, followed by wavelength, and finally by
cutting depth. Maximum FP decreased almost
<0.0001
<0.0001
<0.0001
<0.0001
<0.0001

<0.0001
0.0012
linearly as rake angle increased (Fig 3). The
p value

highest value of maximum FP was obtained at

the lowest rake angle (10 ), highest wavelength
Maximum FP

(1.75 mm), and highest cutting depth (0.50 mm).

However, the triple interaction analysis showed
398.8
136.6
3363.6
1672.1
46.0

10.6
that the effect of rake angle on this force 44.2
F-ratio

depended on wavelength and cutting depth vari-

ation. Thus, Fig 3 shows that effect of rake angle
on maximum FP was more pronounced for
the 1.75-mm wavelength than for the 0.93-mm
Degree of freedom

wavelength. However, the effect of rake angle

was greater for the 0.50-mm depth of cut than
468
479
2
1
1
2
2
1
2

for the 0.25-mm depth of cut (although this

effect is seen only for 10 of rake angle). Finally,
effects of cutting depth and wavelength on par-
allel force decreased as rake angle increased.
Thus, at 30 rake angle, the effect of depth of cut
Rake angle*wavelength*depth of cut

on FP was nil, whereas that of wavelength was

still present.
From a practical point of view, it is obvious
Source of variation

Wavelength*depth of cut
Rake angle*depth of cut
Rake angle*wavelength

that a torn edge during machining should be

avoided. This can be accomplished by using
high rake angles and low wavelengths. Varia-
maple wood.

tion in cutting depth at higher rake angles

Depth of cut
Wavelength
Rake angle

would not have any effect on torn edge inci-

dence. If rake angle is low, for instance 10 ,
Total
Error

torn edge occurrence can be controlled by

 Iskra and Hernández—CUTTING FORCES IN PERIPHERAL CUTTING OF WOOD 139

Figure 3. Maximum parallel force component as a function of rake angle, depth of cut, and wavelength.

decreasing either wavelength or cutting depth. wood surface). The minute thickness at this point
These observations confirm what is normally produces a chip that is easy to deform. The knife
carried out during machining to obtain edges then reaches the lowest point of the wavelength,
free of defects. at which the angle between the instantaneous
velocity vector and wood grain equals zero. At
A set of figures illustrating one cutting cycle at this knife position, the peripheral cutting is
different rake angles when all other cutting equivalent to 90-0. From this point, cutting prog-
parameters are kept constant is shown in Fig 4. resses against the grain until the cutting edge
At the beginning of chip formation, the cutting leaves the workpiece. As the cutting cycle
edge enters into the wood at a minute chip thick- develops, the chip becomes thicker causing
ness and at a low angle formed between the higher resistance to its formation. When rake
instantaneous velocity vector and the wood angle of the cutting tool is low, its rake face
grain. Cutting is hence occurring at an angle that could also take part in pushing the chip forward,
is considered as “following the grain” (assuming causing its further deformation. As rake angle of
that grain is exactly straight and parallel to the the tool increases, its rake face will push and
140 WOOD AND FIBER SCIENCE, APRIL 2012, V. 44(2)

Figure 4. Effect of rake angle on cutting forces.

 Iskra and Hernández—CUTTING FORCES IN PERIPHERAL CUTTING OF WOOD 141

deform the chip to a lesser extent. Instead, the ANOVA showed that maximum negative FN
chip slides over the rake face. As thickness of was significantly affected by rake angle, wave-
the chip increases, the chip becomes more and length, and the interaction between these two
more stiff, and consequently the stress generated variables (Table 2). Effect of rake angle was by
perpendicular to the grain may exceed the trans- far more pronounced than that related to wave-
verse tensile strength of the wood. As a result, length. Effect of rake angle on maximum nega-
splitting ahead of the tool edge may occur. As tive FN depended slightly on the wavelength
rake angle increases, it becomes more likely that considered. As expected, the highest (absolute)
splitting may occur causing stress releases and, negative FN was obtained at 10 rake angle
as a result, release of the force components. This (Fig 5, see negative values). As rake angle
explains the sudden releases of parallel and nor- decreased, the cutting edge tended to push more
mal force components shown in Fig 4 when perpendicular to the grain, which resulted in
rake angles were 20 and more. However, split- higher FN. Furthermore, the contributions of
ting and releases of the parallel force component friction between the rake face of tool and wood
were well pronounced only when wavelength chip as well as the pushing effect of the rake
was 1.75 mm. No splitting was observed when face on the chip increased. These effects should
wavelength was 0.93 mm, although rake angles have decreased as rake angle increased. How-
were 20 or 30 . Also, the release was observed ever, Fig 5 shows that the lowest (absolute)
at the end of the cutting cycle. This supports maximum negative FN was recorded when rake
the theory of splitting ahead of the cutting edge angle was 20 and not 30 . This could be
when the chip is thick enough to generate explained by differences in sharpening quality of
resistance that exceeds tensile strength perpen- knives used in the experiment. In fact, the sole
dicular to the grain. Figure 4 also clearly shows cutterhead used had a knife slot ground at 20 .
the progression from type II to type I chips as The knife used for 20 rake angle was thus only
rake angle increased from 10-30 . This progres-
sharpened on its clearance face. Conversely,
sion is visible in the second portion of the cut-
knives used for 10 and 30 rake angles had to
ting path in which sudden releases of parallel
be sharpened on their rake and clearance faces.
and normal forces occurred (type I chips).
This resulted in different rake tool face qualities
between the two types of knives. Mean roughness
Normal Negative Force Component (Sa) of the rake faces was 6.2 mm for knives
working at 10 and 30 rake angles and 5.3 mm
Normal negative force is considered critical for the knife working at 20 rake angle. The
for defects left on the newly created surface. smoother rake face of the latter knife probably
This force component is responsible for the contributed to decreasing maximum negative FN
amount of transverse compression that occurs by decreasing friction between the rake face tool
when the cutting edge attacks the workpiece. and the wood chip. It is therefore hypothesized
Excessive compression may produce cell crush- that, within the range of values studied, the rela-
ing at the surface and subsurface as indicated tionship between rake angle and maximum
previously (River and Miniutti 1975; Stewart negative FN should be linear.
1980; Stewart and Crist 1982). Compression
might also cause raised grain, which could neces- Effect of wavelength on maximum negative FN
sitate fine sanding operations (Palmqvist and was slight and variable depending on the rake
Johansson 1999; Palmqvist 2003). The normal angle considered (Fig 5). Wavelength could
negative force component is therefore impor- rather be used to assure that the maximum nega-
tant given that it primarily occurs in the first tive FN is reached far away from the visible part
stage of the cutting cycle. That is when the of the cutting path. Thus, for 6 of the 12 cutting
visible part of the wood surface is produced conditions studied, negative FN attained its
(Fig 2). maximum value within the visible part of the cut
142 WOOD AND FIBER SCIENCE, APRIL 2012, V. 44(2)

Figure 5. Maximum normal forces for both negative (pushing) and positive (pulling) components as a function of rake
angle, depth of cut, and wavelength.

(Table 3). This corresponded to the six conditions Normal Positive Force Component
tested at 1.75-mm wavelength. When wavelength
Normal positive cutting force is probably the
decreased to 0.93 mm, maximum negative FN
principal contributor to torn grain production
was observed out of this area (or at the limit).
given that this component acts upward, ie
As expected, depth of cut did not show any pulling the chip from the workpiece. As the cut-
effect on maximum negative FN given that this ting cycle progressed, the knife edge left the
parameter was not implicated at the first stage of visible region and traveled along the cutting
the cutting cycle. The analysis shows that it is path against the grain (Fig 2). After a certain
therefore of interest to use high rake angles to period, normal force turned from negative to
diminish the crushing action on the visible part positive. From this, the positive FN component
of the cutting path. Small wavelengths can also acted perpendicularly to the grain in which
decrease this action by transferring maximum tensile strength of the wood is not particularly
negative FN far away from the visible part of high. As soon as this force exceeded tensile
the cut. Changes in depth of cut did not have strength perpendicular to grain, the chip was
an important impact on this phenomenon. split following the grain. As reported by Stewart
 Iskra and Hernández—CUTTING FORCES IN PERIPHERAL CUTTING OF WOOD 143

(1969), cutting against the grain also causes deflec-

negative to positive
Table 3. Proportions of the cutting path when the normal force component is in the visible portion of the path, reaches its maximum negative value, and changes from

Proportion for each event was calculated from the number of points acquired during a cutting cycle. Precision is hence given by the distance between two points, ie it depended on the frequency of data acquisition
Changes from

tion of fibers. Based on this, deflected and loosened

82
63
93
83
50
40
35
25
55
40
37
29
fibers are more likely to be torn out before they are
cut. Because of the torn grain phenomenon, the
Proportion (%) of the path when FNb:

wood surface can be damaged severely and can

require reworking or scrapping (Palmqvist 2003).
Reaches its maximum

Therefore, to limit torn grain incidence, it is essen-

negative value

tial to decrease the pulling normal force.

41
30
23
17
14
13
19
14
18
27
21
17
Factors that significantly affected the maximum
positive FN component were rake angle, wave-
length, and the triple interaction among rake
Is in the visible
part of cutting

angle, wavelength, and depth of cut (Table 2).

14
13
26
20
14
13
26
20
14
13
26
20

The most significant factor affecting FN was

rake angle. However, at this position of the cut-
ting path, the effect of wavelength became more
Total number

important than that observed when FN reached

of points

its maximum negative value. Thus, Fig 5 shows

22
30
43
59
22
30
43
59
22
30
43
59

that maximum positive FN increased as rake

angle and wavelength increased. Variation in
negative to positive

depth of cut did not show a significant effect

Changes from

on maximum positive FN. This factor showed

18
19
40
49
11
12
15
15
12
12
16
17

erratic behavior, which was probably caused

by the quite low range of values studied (0.25-
Number of points (time) when FN:

0.50 mm). Splitting should be in fact more

Reaches its maximum

governed by the chip thickness variation. Within

negative value

the range of values studied, chip thickness was

9
9

3
4
8
8
4
8
9
10
10

10

more affected by wavelength than by cut-

ting depth variations. Average chip thickness
increased 41% when depth of cut varied 0.25-
Results obtained during a peripheral up-milling knife cut of sugar maple wood.

0.50 mm and 89% when wavelength increased

Is in the visible
part of cutting

from 0.93-1.75 mm (Table 1). Depth of cut

could rather be used to assure that the change in
3
4

3
4

3
4
11
12

11
12

11
12

FN from negative to positive is reached far away

from the visible part of the cutting path. Thus,
Length of a cutting

the probability of production of torn grain

path (mm)

should be decreased by limiting the pulling

5.93
8.39
5.94

5.93
8.39
5.94

5.93
8.39
5.94
8.4

8.4

8.4

action of the tool. Table 3 shows that changes

in FN from negative to positive happen later for
the 0.25-mm depth of cut than for the 0.50-mm
(in this case, 16,000 samples per second).
cut (mm)
Depth of

depth of cut. Similarly, these changes are also later

0.25

0.25

0.25

0.25

0.25

0.25
0.5

0.5

0.5

0.5

0.5

0.5

at lower rake angles than at higher rake angles.

negative to positive.a

Analysis shows that it is therefore of interest to

Wavelength

0.93
0.93
1.75
1.75
0.93
0.93
1.75
1.75
0.93
0.93
1.75
1.75
(mm)

use low rake angles and short wavelengths to

diminish the pulling action of the cutting tool.
Small depths of cut can also decrease this action
by transferring the change in FN from negative to
angle
(deg)
Rake

10

20

30

b
a

positive far away from the visible part of the cut.

144 WOOD AND FIBER SCIENCE, APRIL 2012, V. 44(2)

CONCLUSIONS REFERENCES

Measurements of cutting forces when up-milling Axelsson BOM, Lundberg ÅS, Grönlund JA (1993) Studies
sugar maple wood showed that at the initial step of the main cutting force at and near a cutting edge. Holz
of the cut, FN was negative, ie the knife edge Roh Werkst 51(1):43-48.
de Moura LF, Hernández RE (2006) Evaluation of varnish
pushed the workpiece downward. The part of the coating performance for three surfacing methods on sugar
knife path that remains visible on the machined maple wood. Forest Prod J 56(11/12):130-136.
surface was always created under such condi- Hernández RE, Naderi N (2001) Effect of knife jointing
tions. As the knife edge advanced in the cutting on the gluing properties of wood. Wood Fiber Sci 33(2):
path, FN increased, reached a maximum negative 292-301.
Hernández RE, de Moura LF (2002) Effects of knife
value, then decreased to 0 N/mm, turned posi- jointing and wear on the planed surface quality of north-
tive, and continued upward. FP was always pos- ern red oak wood. Wood Fiber Sci 34(4):540-552.
itive and increased as chip thickness increased. Hernández RE, Rojas G (2002) Effects of knife jointing and
Positive FN and FP reached a maximum just wear on the planed surface quality of sugar maple wood.
prior to emergence of the knife from the work- Wood Fiber Sci 34(2):293-305.
Hoadley RB (2000) Understanding wood. A craftsman’s
piece, ie at the maximum thickness position.
guide to wood technology. The Taunton Press, Newtown,
However, when type I chips were produced as a CT. 280 pp.
result of chip splitting, maximum FP and positive ISO (1996) 11562. Geometrical product specifications
FN were recorded before the theoretical maxi- (GPS)—Surface texture: Profile method—Metrological
mum thickness position was reached. Maximum characteristics of phase correct filters. International
negative FN and FP increased as rake angle Standards Organization. British Standards Institute,
London, UK.
decreased and as wavelength increased. Maxi- ISO (1997) 4287. Geometrical product specifications (GPS).
mum positive FN increased as rake angle and Surface texture. Profile method. Terms. Definitions and
wavelength increased. Impact of depth of cut on surface texture parameters. International Standards Orga-
maximum FP and maximum positive FN values nization. British Standards Institute, London, UK.
was less important. Koch P (1985) Chapter 18 Machining. Pages 1688-2281 in
Utilization of hardwoods growing on southern pine sites.
Volume II. Processing. Agric Handb 605. USDA For
From a practical point of view, occurrence of torn Serv Forest Prod Lab, Madison, WI.
edges should be avoided by using high rake Martellotti ME (1941) An analysis of the milling process.
angles and short wavelengths. These cutting con- Transactions Am Soc Mech Engineers 63:677-700.
ditions also should diminish the crushing action Palmqvist J (2003) Parallel and normal cutting forces in periph-
eral milling of wood. Holz Roh Werkst 61(6):409-415.
on the visible part of the cutting path. Short wave-
Palmqvist J, Johansson G (1999) Cutting forces in periph-
lengths will transfer the maximum negative FN eral milling of wood. Pages 751-760 in Proc 14th
far away from the visible part of the cut. Changes International Wood Machining Seminar, Paris, Epinal,
in depth of cut will not have any impact on such Cluny, France.
defects. Conversely, use of low rake angles and River BH, Miniutti VP (1975) Surface damage before glu-
ing weak joints. Wood Wood Prod 80(7):35-36, 38.
short wavelengths should decrease the pulling
Stewart HA (1969) Effect of cutting direction with respect
action of the knives when FN turns positive. Small to grain angle on the quality of machined surface, tool
depths of cut can also decrease this action by force components, and cutting friction coefficient. Forest
transferring the change in FN from negative to Prod J 19(3):43-46.
positive far away from the visible part of the cut. Stewart HA (1980) Some surfacing defects and problems
related to wood moisture content. Wood Fiber Sci 12(3):
175-182.
Stewart HA, Crist JB (1982) SEM examination of subsur-
ACKNOWLEDGMENTS face damage of wood after abrasive and knife planing.
Wood Sci 14(3):106-109.
This research was supported by the Natural
Woodson GE, Koch P (1970) Tool forces and chip forma-
Sciences and Engineering Research Council of tion in orthogonal cutting of loblolly pine. Res Pap
Canada and by the Fonds Québécois de la SO-52 USDA For Serv Southern Forest Experiment
Recherche sur la Nature et les Technologies. Station. 29 pp.