10 31。43 28。80 - - - - -

9 31。96 28。90 27。27 - - - -

8 33。58 29。10 27。67 25。78 - - -

7 33。62 30。05 27。79 25。88 24。24 - -

6 35。69 30。69 28。43 26。13 24。54 23。40 -

5 41。04 32。96 30。05 27。00 24。72 23。69 -

4 - - - 28。08 25。08 23。61 21。79

3 - - - - 26。81 24。35 21。90

2 - - - - - 25。10 22。44

1 - - - - - - 22。80

0。5 - - - - - - 23。18

Fig。 15。 Error ratios of analysis values to design values。

Therefore only the compressed deformation of the urethane- covered roll needs to be determined。 To that end, analyses according to the contact angle between the bending roll and the sheet were performed。 This contact angle is illustrated in Fig。 14。

The analysis results for the contact angle between the bend- ing roll and the sheet are listed in Table 5, and the errors of the analysis values compared with the design values are shown in Fig。 15。

3。1。5Analysis results and discussion

The curves for the error ratio of analysis values to design values show different patterns, as indicated in Fig。 15。 There- fore, a new process parameter incorporating the main process parameters (contact angle, material thickness, and forming radius) needs to be defined so that the curves for the error ratio of analysis values to design values regarding the various mate- rial thicknesses could have similar patterns without regard to those material thicknesses。 Thus, in the present study, the new process parameter,『contact angle/(material thickness×for- ming radius)』, was defined。 The values of the new process parameter are listed in Table 6。