Impact of thickness measurement on predicting residual stresses due to winding

Abstract

Axisymmetric winding models were initially developed to deal with web thickness variations that were manifested in the machine and cross machine direction. The web thickness variation is a prime input to these models. Small web thickness variations can integrate during winding to produce large residual stress variation in the wound roll. Web thickness measurement in process lines employ sensors that scan over the web width while the web is moving in the machine direction. Spatially this provides a measure of web thickness in a zig-zag pattern. This thickness variation is used as a control feedback parameter that control a forming or coating die lip to reduce the web or coated web thickness variation. The scan rate of the sensor will typically be set to capture thickness data at a frequency higher than the frequency response of the control/actuator system. Is thickness data captured spatially at this rate for the process sufficient for input to the axisymmetric winding model? This publication addresses this problem. The web thickness variation of an existing web was fully mapped in the machine and cross machine directions. This data was interrogated to determine what thickness data would have been recorded at varied scan rates. The full map of thickness data was input to the winding model as well as data that would have resulted from various rates of zig-zag scanning. Results show that as the rate of zig-zag scanning increase, that the residual winding stresses converge toward the stresses produced by the input of the full thickness map. Model results were compared to test results for validation.