Engineering, 2010, 2, 727-732
doi:10.4236/eng.2010.29094 Published Online September 2010 (
Copyright © 2010 SciRes. ENG
Rolling Deformations and Residual Stresses of Large
Circular Saw Body
Bolesław Porankiewicz1, Jari Parantainen2, Karolina Ostrowska3
1University of Zielona Góra, Zielona Góra, Poland
2Stresstech OY, Vaajakoski, Finland
3Poznań University of Technology, Poznań, Poland
Received May 21, 2010; revised July 19, 2010; accepted September 5, 2010
Rolling path squeezes and rolling residual stresses of large diameter circular saw body for wood, generated
by rolling pressure from 10 up to 120 bar were examined. X-ray diffraction, Barkhausen noise (BN) and Full
Width of the peak at a Half Maximum (FWHM) (o) methods for evaluation of residual stresses were used.
Dependencies of a tangential rolling residual stresses inside rolling paths upon rolling pressure p (bar) and
rolling area A (mm2) were evaluated. The rolling pressure, as large as 60 bar, resulting in the rolling squeeze
as high as 0.04 mm2, and, tangential residual compression stresses inside a rolling path, as large as
TI =
822 MPa, was considered to be the largest for the practical application.
Keywords: Circular Saw, Rolling Squeeze Area, Rolling Squeeze Width, Rolling Squeeze Depth, Rolling
Pressure, Tangential Rolling Residual Stresses, Radial Rolling Residual Stresses, X-Ray
Diffraction, Barkhausen Noise, FWHM.
1. Introduction
Circular saws rolling use to be widely applied method of
initial tensioning, aiming at increase the dynamic stiff-
ness of saws for wood and secondary wood products
machining. This method is not devoid of negative aspects.
It has to be mentioned that the rolling may cause neces-
sity to correct flatness when stresses distribution inside a
circular saw body is not correct. There are several ways
of evaluation of rolling effects, like: - a depth d (mm), -
an area A (mm2) of a rolling path, - a light gap between
deformed blade and a straight edge rule [1], - a static
stiffness [2], - a compression stresses inside a rolling
path. However, as a final measure of effect of a saw
blade rolling was recognized as a shift of natural frequ-
encies and critical rotational speeds of several initial vi-
bration modes [3,4]. From available literature important
facts concerning with amount of tensioning necessary to
insert in a saw blades of different dimension and for dif-
ferent applications in order to get stable work are known
[5-7]. However, from practical point of view there are
lack of informations in published works, about rolling
pressures used, plastic squeeze of circular saw body ma-
terial and rolling path shape [2]. The goals of actual work
were: to exam the amount of squeeze in a circular saw
body using different rolling pressures p (bar) and meas-
ure presence of residual stresses using three different
2. Experimental
The circular saw body, before cemented carbide tips
soldering, by thickness of tS = 3 mm and by diameter of
D = 620 mm, made of 75Cr1 low alloy steel, was rolled
in industrial conditions, with use of rolling machine Arga
T08 on 12 different paths. The pressure p (bar) in hy-
draulic cylinder, was from 10 bar (145.04 psi) up to 120
bar (1740.456 psi) with of 10 bar (145.04 psi) increment.
The depth d (mm) and width w (mm) of a rolling paths
were measured with use of profilografometer type ME10.
X-ray residual stress measurements contained totally 25
points from the blade using modified d(sin2
) [8] me-
thod. X-ray measurements were performed using XS-
TRESS3000 diffractometer manufactured by Stresstech
OY, by following conditions:
Device: G2R (#7147)
Radiation source: CrKa
Diffraction line angular position,
Copyright © 2010 SciRes. ENG
according to Bragg’s law 2q: 156.4° (211)
Spot size: 1 mm and 2 mm
Exposure time: 20 s and 8 s
Tilt angles: 4/4 tilts,
oscillation ±5°
Young’s modulus: 211 000 MPa
Poisson ratio: 0.3
Calculation: Cross correlation, constant background [8].
Measurement method: Modified d(sin2
) [8]
The measurement directions can be seen in Figure 1.
The angle of
= 0° corresponds to tangential direction
and of
= 90° to radial direction.
The FWHM (°) values were calculated from the x-ray
diffraction peaks in order to get indirect information for
the presence of the residual stresses through micro hard-
ness and plastic deformation (dislocation density) layer.
The FWHM (°) values are average ones from
(°) an-
gles examined.
The BN measurements were performed in the same ra-
dial path, like during the X-ray diffraction ones, using
following conditions:
Device: Rollscan 300
Sensor: S1-138
Magnetizing voltage: 4.0 Vpp
Magnetizing frequency: 100 Hz
Analyzing frequency: 70-200 kHz
Higher hardness and/or compressive stresses decrease
the BN and vice versa [9]. It has to be mentioned that
using the BN itself was not possible to evaluate absolute
values of residual stresses.
Outside rolling paths Rockwell hardness was meas-
ured (according to PN-EN ISO 6508 standard) with pre-
liminary load of 98 N and total load of 1471 N, in places
shown in Figure 2, by 5 repetitions. For every place out-
side rolling paths an average value and standard devia-
tion were calculated.
3. Results and Discussion
The rolling squeeze cross section shape for the largest
rolling pressure, shown in Figure 3, was approximated
with use of a second order polynomial function d = f(a1 +
a2 · wi + a3 · wi
2). It was evaluated by a Formula (1) with
correlation coefficient R and standard deviation SD(mm),
as large as 0.91 and 0.0067 mm respectively.
d = 0.010209 0.037531 · wi + 0.00709 · wi
2 (mm) (1)
In the rolling squeeze cross section shape, several up-
casts can be seen, what indicated possible wear of rolls
and/or bearings in the rolling machine. The depth and
width of the largest up-cast were as large as 11.4 m and
590 m respectively. Results of measurements of rolling
squeeze depth d (mm) and width w (mm) were collected
in Table 1 and illustrated in Figure 4 and Figure 5,
From Figure 4 it can be seen that rolling squeeze
depth d (mm) increased with growth of the rolling pres-
sure p (bar) with rather large dispersion. The width w
(mm) of the rolling squeezes, shown in Figure 5 in-
creased with a growth of the rolling pressure p (bar) until
50 bar. For larger pressure opposite tendency can be no-
ticed with large dispersion. The area of the rolling
squeeze A (mm2), shown in Figure 6 was evaluated by
integrating the surface limited from the bottom by the
squeeze shape and from the top by d = 0. Large disper-
sion in the dependency A = f(p) can also be seen espe-
cially for rolling pressure larger than p = 60 bar, namely
p = 70 bar, 100 bar and 110 bar. The reason of large dis-
persion of rolling depth d (mm), rolling width w (mm)
and rolling squeeze A (mm2) was probably the wear of
rolls or bearings in the rolling machine used. According
to the work [6], the rolling squeeze area of 0.04 mm2,
applied for a circular saw diameter D = 400 mm, saw
blade thickness tS = 2 mm and collar diameter of dC =
100 mm, resulted in 2.4, 29.4 and 14.5 times increased
Figure 1. Direction of X-ray diffraction measurements.
Figure 2. Places for Rockwell hardness measurements: -
outside rolling paths nos. 1-25, t - tooth area, rp - rolling
Figure 3. The observed (red) and predicted (blue) shape of
cross section of the rolling path for maximum rolling pres-
sure p = 120 bar.
Copyright © 2010 SciRes. ENG
Table 1. Rolling pressure p (bar), rolling depth d (mm),
rolling width w(mm), rolling area A (mm2).
p d w A
(MPa) (mm) (mm) (mm2)
10 2.02 3.24 0.0027
20 1 3.95 0.0015
30 3.4 3.9 0.0075
40 5.6 5.48 0.0176
50 9.3 6.33 0.029
60 15.3 6.11 0.0443
70 9.1 5.03 0.0318
80 23.2 4.21 0.0463
90 21.6 5.59 0.0642
100 18.8 4.99 0.0312
110 17.7 4.69 0.0312
120 49 4.95 0.123
Figure 4. The average rolling path depth d (mm) in depen-
dence from rolling pressure p (bar).
Figure 5. The average rolling path width w (mm) in depen-
dence from rolling pressure p (bar).
Figure 6. The average rolling path area A (mm2) in depen-
dence from rolling pressure p (bar).
natural frequencies of (0,4), (0,3) and (0,2) vibration
modes respectively, while the vibration modes (0,1) and
(0,0) natural frequencies were dropped down as much as
18% and 23% respectively by rotating speed of 3600
min-1 (the first digit in the brackets ‘0’ is the number of
nodal circles, the second digit ‘4’, ‘3’ and ‘2’ is the
number of nodal diameters). However, in the work
quoted above, no information on used rolling pressure
and rolling squeeze cross section profile were given.
The residual stresses evaluated with use of the X-ray
XSTRESS3000 diffractometer, inside and outside rolling
paths, in both tangential
TI (MPa) and radial directions
RI (MPa) were shown in Figure 7. From Figure 7 it can
be seen that tangential residual stresses inside rolling
TI (MPa) did not change their values smoothly ac-
cording to an enlargement of the rolling pressure p (bar),
what was clearly seen for points nos.: 4, 8, 16 and 22.
The dispersion can also be seen for tangential residual
stresses outside rolling path
TI (MPa) (points nos. 5, 11,
12, 17 and 21). The reason of that might be a rather large
dispersion of the residual stresses in the saw blade before
rolling and/or large dispersion of rolling squeeze profile.
A saw in which body are present such a large, and highly
differentiated residual stresses generated during manu-
facturing process, has small chance for a smooth and
effective work in future even if using hammering will be
exactly flattened.
From Figure 7 it can be seen decreasing tendency of
the tangential residual stresses outside rolling paths
(MPa), with an increase of the rolling pressure p (bar).
Starting from point no. 11 up to point no. 25, the tangen-
tial residual stresses outside rolling paths
TO (MPa) did
change from compression to tensile. This was due to the
total influence of rolling effect on tangential residual
(MPa) along saw body radius. It was also
assumed that in the saw blade examined before rolling,
there were average tangential residual stresses distribu-
(MPa) with randomly distributed dispersion.
The average total rolling effect, as a function of distance
T = f(L) was approximated with statistical Formula (2)
by correlation coefficient R and standard deviation SD
(MPa) as large as 0.85 and 43.6 MPa, respectively.
T = 79.861196 + 3.278002·L 0.0141832 · L2 (MPa) (2)
The tangential compression stresses inside rolling
paths after correction
TIK (MPa) were calculated as dif-
ference between the tangential compression stresses in-
side rolling paths
TI (MPa) and the average rolling effect
on the tangential residual stresses outside rolling paths
T (MPa), described by Formula (2), and approximated
by statistical Formula (3) in dependence upon rolling
pressure p (bar), by R = 0.96, SD = 40.8 MPa and upon
rolling squeeze area A (mm2) by statistical Formula (4),
by R = 0.89 and SD = 68.7 MPa.
=161639.174 -161769.431 · p0.001152 (MPa) (3)
= 54277.706 -55542.535 · A0.00184 (MPa) (4)
It has to be mentioned that the dispersion of the resid-
ual tangential residual stresses in the examined saw blade
Copyright © 2010 SciRes. ENG
before rolling (outside rolling paths
TO (MPa)) oversha-
dowed examined relation.
Looking at Figure 8 it can be concluded that an in-
crease of rolling pressure above p > 70 bar stops the in-
crease in tangential compression stresses inside rolling
path. One can conclude that in analyzed case, the maxi-
mum rolling squeeze, giving increase tangential com-
pression stresses inside rolling path was A = 0.04 mm2.
In the published papers [1,2,5-7], there was no informa-
tion about dispersion of residual stresses inside saw
blades before rolling.
In examined case, after correction, the tensile stress
inside saw blade before rolling, at point no. 5 was evi-
denced, as large as +1.9 MPa (Figure 8). The compres-
sion stress in point no. 21 was also very low, as small as
5.7 MPa.
The radial residual stresses inside rolling paths
(MPa) were smaller (Figure 7). They oscillated on level
of about <332, 484> MPa. From Figure 7 it can also
be seen the presence of a residual radial tensile stresses
RO (MPa) for all points outside rolling paths. No sig-
nificant correlation between residual radial tensile
stresses outside rolling paths
RO (MPa) and distance L
(mm) along saw body radius, according to increasing
rolling pressure p(bar) was recognized. The residual ra-
dial stresses outside rolling paths
RO (MPa) did oscillate
in the range of <+4, +266> MPa with an error of < 35, 46
> MPa.
It can be seen from Figure 9 that the BN measure-
ments followed general shape of stresses distribution
showed in Figure 7, excluding total rolling effect
(MPa). The BN measurements allow recognizing places
with large and small values of compression stresses.
Places outside rolling paths shown in Figure 9 on posi-
tions L = 108 mm, L = 114 mm, L = 134 mm, and L =
144 mm were having the largest of all the BN values,
what effect was not seen in Figure 7 and Figure 8. The
reason of that was lower Rockwell hardness of the saw
body material closer to the rim, what show Figure 10,
for points nos. 17, 19, 21 and 25. According to Figure 9,
the hardness in point no. 23 was too large, however, not
the same measuring path for the BN and the Rockwell
hardness as well as large dispersion of the saw body
hardness resulted in such a difference. For places outside
rolling paths on positions L = 0-4 mm and L = 12-18 mm,
shown in Figure 9, low BN values can be associated with
large saw body hardness. For places inside rolling paths,
on positions L = 6 mm and L = 22 mm, shown in Figure
9, slightly higher the BN values, can be associated with
small rolling residual compression stresses. The BN
measurements results for places outside rolling paths
shown in Figure 9 were characterized with large disper-
sion. The BN measurements technique would be useful in
manufacturing conditions, to control residual stresses
distribution in saw blades after different operations. This
conclusion can also be supported by the fact of many
times lower price of the Rollscan 300 device in com-
parison to the X-ray XSTRESS3000 diffractometer. Still
the priority benefit is in the time consumed in performing
the measurements. The whole X-ray measurement pro-
cedure can easily last hour or two, but the BN measure-
ment is usually done in few seconds.
Results of measurements of the FWHM (o) outside and
inside rolling, were collected in Table 2 and illustrated
in Figure 11. From the plot shown in Figure 11 it can be
Figure 7. The plot of residual stresses (MPa) along saw
blade radius, in tangential and radial directions, * - inside
and ¤ - outside rolling paths.
Figure 8. The plot of residual stresses in tangential direc-
tion after correction, along the saw blade radius, * - inside
TI (MPa), and ¤ - outside
TO (MPa) rolling paths.
Figure 9. The plot of the Barkhausen noise along saw blade
radius, * inside and ¤ outside rolling paths.
Figure 10. The plot of the Rockwell hardness of the saw-
body, along saw blade radius, in points number No., outside
rolling paths, t - tooth area.
Copyright © 2010 SciRes. ENG
Figure 11. The plot of FWHM (o) along saw blade radius in
tangential and radial directions, * - inside and ¤ - outside
rolling paths.
seen that the FWHM (o) values were larger for places
inside rolling paths than for places outside rolling paths,
what was opposite in comparison to the BN measure-
ments results shown in Figure 9. The difference between
small and large residual compression stresses inside roll-
ing paths can not be recognized using Figure 11, what
indicates that the shape of the plots from Figure 7 and
Figure 11 was not followed each other. The FWHM (o)
measurements allow recognizing places with large and
small residual stresses, but with much lower precision if
compare to the BN measurements. The increase of the
FWMH (o) values inside rolling paths might also be oc-
curred by work hardening effect of the rolling process.
4. Conclusions
The experiment and measurements of rolling effects and
analysis of results obtained allow concluding that:
1) It is recommended to apply the rolling pressure up
to 60MPa by use rolling machine Arga T08, giving roll-
ing squeeze as large as A = 0.04 mm2 and tangential re-
sidual compression stresses inside rolling paths as large
TI = 822 MPa.
2) For rolling pressure p (bar) from 70 to 120 bar, no
increase of the tangential compression residual stresses
increase inside the rolling paths was observed.
3) For rolling pressure p (bar) from 70 to 120 bar large
variation of rolling squeeze depth d (mm) and width
w(mm) was seen. By the largest rolling pressure (d =
0.049 mm, w = 4.95 mm, A = 0.123 mm2) significant up-
cast of 11.4 µm in depth and 590 µm in the width was
evidenced, indicating possible wear of rolls and/or bear-
ings in the rolling machine.
4) Large dispersion of tangential
TO (MPa) and radial
RO (MPa) residual stresses outside rolling paths, added
to the saw blade before rolling, with maximum value of
155 MPa and 274 MPa respectively was evidenced.
5) The BN measurements allow recognizing the pres-
ence of small and large compression stresses inside and
outside rolling paths, but this information is mixed with
effect of hardness change distribution.
6) In one measuring point, outside rolling path after
correction positive, residual tensile tangential stress of +
1.9 MPa was recognized.
Table 2. Stress and FWHM (o) values, evaluated with use of
X-ray diffractometer.
Stress Tangential
= 0o
Stress Radial
= 90o
= 0o
= 90o
(MPa)(MPa)(MPa) (MP)  (
o) (o)
1 64 10 225 46 3.33 3.28
2 5338 332 13 3.56 3.46
3 60 11 190 41 3.35 3.30
4 6257 424 8 3.56 3.49
5 51 13 274 39 3.31 3.27
6 6268 458 18 3.53 3.45
7 26 11 215 42 3.29 3.25
8 6457 484 15 3.55 3.50
9 19 12 189 46 3.33 3.26
10 68110 483 11 3.52 3.43
11 77 17 266 40 3.33 3.29
12 68310 477 18 3.53 3.48
13 113 12 217 40 3.28 3.25
14 6786 464 15 3.54 3.45
15 81 18 175 41 3.28 3.25
16 66017 462 19 3.52 3.43
17 144 20 181 40 3.29 3.26
18 6749 453 20 3.50 3.43
19 69 15 86 36 3.29 3.27
20 6719 435 21 3.52 3.44
21 155 21 115 35 3.23 3.22
22 63111 392 25 3.43 3.37
23 67 22 30 36 3.26 3.22
24 68711 377 34 3.48 3.42
25 59 26 4 28 3.22 3.24
7) No significant rolling effect on radial residual
stresses inside
RI (MPa) and outside
RI (MPa) rolling
paths was seen.
5. Acknowledgements
The authors were grateful for the support of the Stress-
tech OY Company in performing X-ray diffraction and
BN measurements. The authors were also grateful for the
support of the Poznań Networking & Supercomputing
Center (PCSS) calculation grant.
6. References
[1] P. F. Lister and G. S. Schajer, “The Effectiveness of the
Light-Gap and Frequency Measurement Methods for
Evaluating Saw Tensioning,” Proceedings of 10th Wood
Machining Seminar, University of California, Forest
Products Laboratory, Richmond, 21-23 October 1991, pp.
[2] R. Szymani and C. D. Mote, “Circular Saw Stiffness as a
Copyright © 2010 SciRes. ENG
Measure of Tension,” Forest Products Journal, Vol. 27,
No. 3, 1977, pp. 28-32.
[3] J. Rhemrev and L. Trinchera, “Improving the Stability of
Thin Circular Saws though Applied Research,” 32nd An-
nual Meeting the Forest Products Research Society, At-
lanta, 1978.
[4] G. S. Schajer and C. D. Mote, “Analysis of Roll Ten-
sioning and its Influence on Circular Saw Stability,”
Wood Science & Technology, Vol. 17, 1983, pp. 287-302.
[5] U. V. Münz, “Means of Testing and Designing Circular
Saw Blades,” Proceedings of the 13th International
Wood Machining Seminar, Vancouver, 17-20 June 1997,
pp. 105-116.
[6] G. S. Schajer and C. D. Mote, “Analysis of Optimal Roll
Tensioning for Circular Saw Stability,” Proceedings of
7th Wood Machining Seminar, University of California
Forest Products Laboratory, Richmond, 18-20 October
1982, pp. 264-283.
[7] G. S. Schajer and C. D. Mote, “Analysis of Optimal Roll
Tensioning for Circular Saw Stability,” Wood & Fiber
Science, Vol. 16, No. 3, 1984, pp. 323-338.
[8] P. S. Prevey, “X-Ray Diffraction Residual Stress Tech-
niques, Metals Handbook,” 10 Metals Park: American
Society for Metals, 1986, pp. 380-392.
[9] R. L. Pasley, “Barkhausen Effect - An Indication of
Stress,” Southwest Research Institute, San Antonio,
Texas, 1969.