Professional Documents
Culture Documents
Abstract
The coordinate measuring machines (CMMs) have proven to be reliable, flexible and very much suitable for determining the acceptability
of manufactured parts. In this paper, techniques for evaluating circularity and sphericity error from CMM data are presented. The form
error can be evaluated directly from CMM data by employing circle/sphere as assessment features and using normal deviations. The CMM
data can also be transformed by applying appropriate methods that not only suppress the size but also introduce distortion. The form error
is evaluated from the transformed data by employing limacon/limacoid as assessment features and using linear deviations. The methods
for handling CMM and transformed data are given in this paper. The proposed methods are validated using the data available in literature.
© 2003 Elsevier Science B.V. All rights reserved.
0924-0136/03/$ – see front matter © 2003 Elsevier Science B.V. All rights reserved.
doi:10.1016/S0924-0136(03)00187-0
G.L. Samuel, M.S. Shunmugam / Journal of Materials Processing Technology 139 (2003) 90–95 91
2. Circularity and sphericity errors While evaluating the circularity error using coordinate
data, the assessment circle is established with O0 (x0 , y0 ) as
The International Standard ISO 1101 [2] defines circu- center and radius r0 as shown in Fig. 2(a). The CMM data
larity error as the radial distance between two concentric is shown in Table 1. The normal deviation of a point (xi , yi )
circles separated by minimum possible distance and con- on the profile from the assessment circle is determined as
taining all the measurement points on the given profile as given below:
shown in Fig. 1(a). However, sphericity evaluation is to be ei = [(xi − x0 )2 + (yi − y0 )2 ]1/2 − r0 (2)
done with reference to two concentric spheres containing all
points of the data set and having the minimum separation, The circularity error is determined using Eq. (1).
as shown in Fig. 1(b).
For the purpose of assessment, it is convenient to think 3.2. Sphericity error
of an assessment feature superimposed in such a way that
While evaluating the sphericity error, the normal devi-
the maximum distance between it and the actual feature
ation of a point from the assessment sphere with center
concerned is the least possible value. Once such assessment
O0 (x0 , y0 , z0 ) and radius r0 as shown in Fig. 2(b) is consid-
feature is established, the circularity or sphericity error is
ered. CMM sphericity data is shown in Table 2. The normal
evaluated with reference to these assessment features as:
deviation is given by
Error ∆ = |emax | + |emin | (1) ei = [(xi − x0 )2 + (yi − y0 )2 + (zi − z0 )2 ]1/2 − r0 (3)
92 G.L. Samuel, M.S. Shunmugam / Journal of Materials Processing Technology 139 (2003) 90–95
Table 1
CMM data of a circular profile [16]
S. no. xi (mm) yi (mm)
1 70.0150 50.0000
2 68.7900 58.4734
3 65.4060 65.9372
4 59.5675 72.7493
5 51.3791 77.7452
6 44.7944 79.6013
7 40.8903 79.9958
8 32.0312 78.9306
9 27.2296 77.1385
10 20.3993 72.7076
11 16.1556 68.2304
12 12.7184 62.4905
13 10.6380 56.0806
14 10.0183 49.2149
15 11.4275 40.8264
16 14.1050 34.8682
17 18.8168 28.7427
18 24.6321 24.2200
19 31.6833 21.1862
20 39.1626 20.0207
21 45.5204 20.5021
22 55.3996 24.2692
23 62.3561 30.0114
24 67.3540 37.6492
25 69.6190 45.2028
Fig. 2. Assessment using coordinate data: (a) circularity error; (b) spheric-
ity error.
ri = di − dmin + d0 (5)
−1 yi − y R
θi = tan (6)
xi − x R
Table 2 Table 3
CMM data of a sphere [9] Transformed circularity data for the CMM data given in Table 1 (xR =
40.0067 mm, yR = 49.9909 mm, rR = 30.0083 mm, d0 = 0.01 mm;
S. no xi (mm) yi (mm) zi (mm)
reference circle: 1, 10, 17)
1 0.81814 −0.43615 −0.38525 S. no. ri (m) θ i (◦ )
2 −0.80885 −0.53045 0.27905
3 0.03749 0.90643 −0.42737 1 42.5173 0.0174
4 0.86935 −0.27403 −0.42518 2 41.3988 16.4203
5 −0.50786 −0.07441 0.86060 3 24.3585 32.1216
6 0.75348 −0.67012 −0.00610 4 43.7024 49.3210
7 0.05867 −1.00773 −0.00198 5 28.0847 67.7185
8 −0.17486 0.60442 0.78297 6 29.1711 80.8153
9 0.71121 −0.50123 −0.50042 7 52.1145 88.3132
10 −0.47493 −0.51466 0.71992 8 52.782 105.4077
11 0.06016 −0.74129 −0.67432 9 38.3151 115.2042
12 0.05337 0.02006 1.00515 10 42.5173 130.7984
13 0.29869 0.87920 0.38988 11 60.1041 142.5940
14 −0.15488 −0.86697 0.48585 12 49.0629 155.3895
15 0.38994 −0.00060 −0.92592 13 27.6287 168.2855
16 0.00776 0.73922 0.68741 14 32.6533 181.4823
17 0.62090 0.52006 −0.59341 15 46.8601 197.7795
18 0.00952 −0.03493 −1.00376 16 27.4535 210.2785
19 −0.44592 0.76584 −0.47827 17 42.5173 225.0787
20 0.01069 0.99511 0.11265 18 42.8477 239.1802
21 −0.93371 −0.17351 −0.33160 19 17.3805 253.8828
22 −0.48121 −0.69836 0.53717 20 16.3081 268.3867
23 0.50394 0.86933 0.03322 21 34.0605 280.5907
24 0.80924 0.07332 −0.59398 22 10.0000 300.898
25 0.84569 0.06876 −0.53570 23 12.1482 318.2045
26 0.36179 −0.07902 0.93410 24 37.4254 335.7106
27 −0.78624 0.20326 −0.59063 25 31.1207 350.8152
28 −0.46529 0.69616 −0.55167
29 −0.93402 −0.02267 −0.35813
30 0.43016 0.49319 −0.76122 The transformed data is given in the spherical coordinates
31 −0.46929 0.11501 0.88119 as:
32 0.60788 −0.41130 0.68580
33 0.70712 0.46598 0.53998 ri = di − dmin + d0 (8)
34 0.67398 −0.38068 −0.63956
−0.06245 −0.50544 −1 yi − yR
35 0.86696 θi = tan (9)
36 −0.09577 0.96216 0.26433 xi − x R
37 0.29318 0.79379 0.54237
38 −0.39873 −0.68009 0.61867 −1 zi − zR
−0.30833 −0.07805
βi = tan (10)
39 0.94937 xi − x R
40 −0.70424 −0.64359 0.31476
41 0.67010 −0.17187 −0.72841 where (ri , θ i , βi ) represents the point in the polar form, dmin
42 −0.12359 −0.70688 0.70559 the minimum of di values and d0 chosen conveniently. The
43 −0.09694 0.64820 −0.75729 transformed data is shown in Table 4.
44 −0.13014 0.04876 0.99815
45 −0.28485 0.86262 −0.43728
46 0.16745 −0.11142 −0.98603
47 0.00946 −0.00746 1.00873 5. Evaluation using transformed data
48 0.15374 −0.33634 −0.93141
49 −0.44828 0.85926 −0.25853 5.1. Circularity error
50 −0.61673 −0.28843 0.73859
Table 4
Transformed sphericity data for the data given in Table 2 (xR = 0.001386
mm, yR = −0.000511 mm, zR = 0.000654 mm, rR = 1.00475 mm,
d0 = 0.001 mm; reference sphere: 15, 35, 37, 50)
S. no. ri (m) θ i (◦ ) β i (◦ )
1 2.892516 331.92551 −22.630814
2 7.385204 213.18689 16.042796
3 3.520519 87.720345 −25.247120
4 4.744676 342.50897 −25.076246
5 2.148883 188.25681 59.104265
6 7.010473 318.32050 −0.3842970
7 8.851511 273.25510 −0.1496052
8 4.502541 106.24343 51.151781
9 2.819969 324.80035 −29.977900
10 4.277047 227.18747 45.741929
11 3.892918 274.53638 −42.249303
12 6.051866 21.589701 86.814434
13 6.866947 71.326988 22.741355
14 5.280115 259.77658 28.858496
15 4.747238 0.0000000 −67.249518
16 9.395520 89.506317 42.872129
17 3.846185 40.039986 −36.284063
18 5.038199 283.29636 −90.000000
19 8.339309 120.27138 −28.357010
20 1.945131 89.464592 6.4178506
21 7.337647 190.48158 −19.258731
22 3.864665 235.33429 32.306696
23 5.110749 59.982666 1.8567226
24 5.819051 5.2218342 −36.241849
25 2.659742 4.6903501 −32.339316
26 3.682601 347.71091 68.438292 Fig. 4. Assessment using transformed data: (a) circularity error; (b)
27 5.732323 165.49475 −36.009188 sphericity error.
28 4.092739 123.81676 −33.372072
29 2.100106 181.35712 −20.979435
30 4.013573 49.026047 −49.361139
31 5.100687 166.21010 61.171694 6. Results and discussion
32 3.000626 325.88953 43.086139
33 3.269524 33.464756 32.518304
34 3.388419 330.52365 −39.646803
Data sets in Tables 1 and 2 show the CMM data for
35 4.747238 262.79456 59.566083 circularity and sphericity features, respectively. Data sets
36 2.847405 95.762978 15.243843 in Tables 3 and 4 are the data obtained by applying ap-
37 4.747238 69.828720 32.626278 propriate transformations on the CMM data sets given in
38 1.931578 239.51169 38.084669
39 1.000000 194.05539 71.400189
Tables 1 and 2, respectively. These data sets are taken
40 5.049290 222.34476 18.211659 to bring out the differences in approaches to be followed
41 4.031829 345.62714 −46.563569 for processing CMM data and transformed data. Table 5
42 5.740304 259.96664 44.500332 shows the results of minimum zone evaluation of circular-
43 2.485199 98.618797 −49.118635
44 7.336913 159.46350 81.985185
ity and sphericity. Various parameters of the assessment
45 9.314523 108.34682 −25.714897 feature such as the coordinates of the center and radius
46 6.691022 326.26227 −78.558469 are also given in Table 5. The simplex search techniques
47 8.133594 319.28290 90.000000 suggested by Shunmugam have been used in the present
48 2.366958 294.40216 −68.413730
49 4.283524 117.60987 −14.956114
work for evaluating the circularity and sphericity errors
50 4.747238 204.97614 47.260787 [14].
The values of form error evaluated from the CMM data
and the transformed data are nearly the same as shown
5.2. Sphericity error in Table 5. Also these values are well in agreement with
the error values published in the literature. The form error
The surface distorted due to the transformation is well value for the CMM data is based on the normal deviations,
approximated by a limacoid (ref. Fig. 4(b)), an extension of whereas for the transformed data it is based on linear devia-
limacon [17] and the linear deviation of a point (ri , θ i , βi ) tions. However there is a small difference in the form error
is expressed as: values obtained, due to the approximation error which is
negligibly small. The use of normal deviation for the trans-
ei = ri − (r0 + x0 cos βi cos θi + y0 cos βi sin θi + z0 sin βi )
formed data will lead to incorrect results and hence should be
(12) avoided.
G.L. Samuel, M.S. Shunmugam / Journal of Materials Processing Technology 139 (2003) 90–95 95
Table 5
MZ evaluation of circularity and sphericity
S. no. Reported method Present approach
Form error Parameters (mm) Form error (m) Without transformation With transformation
Parameters (mm) Form error (m) Parameters (m) Form error (m)
1 Circularity x0 = 40.0008 [16] 29.2860 [16] x0 = 40.0007 29.2816 x0 = −5.9602 29.2809
y0 = 50.0016 [16] y0 = 50.0015 y0 = 10.6228
r0 = 30.0001 [16] r0 = 30.0000 r0 = 34.2762
2 Sphericity – 7.660 [9] x0 = 0.0025 7.66023 x0 = 1.1181 7.66011
– y0 = −0.0001 y0 = 0.4149
– z0 = 0.0005 z0 = −0.1726
– r0 = 1.00543 r0 = 5.3717