PROCESS-CONTROL FOR A COMPONENT-DOMINANT OPERATION - A CASE-STUDY

A plant manufacturing 14-inch TV electron guns was experiencing high r ejection (0.92%) of a sub-assembly of components grid 1 (G1), grid 2 ( G2), pre-focusing (PF), 4 (G4) and grid 5 (G5)), called the 'electron optic assembly (EOA)' owing to the defect known as a 'G3-G5 gap outsid e specification'. The estimated loss is as high as 0.69 million rupees (US$42 000 per annum at the usual production level of 2.5 million. Co llection of jigwise data revealed that: the process was found to be se t at levels lower than the nominal in all cases except jig number 8; d ata did not exhibit the state of statistical control except that of ji g numbers 1, 3 and 8. Analysis of process behaviour and a thorough stu dy on the assembly drawing of components G3 and G5 resulted in the est ablishment of a relationship formula between the G3-G5 gap and the com ponent dimensions. In analysis of further data collected and using the relationship formula, it was established that the process is 'compone nt dominant'. Two types of solutions were arrived at: namely, long ter m and short term. In long-term solutions, the tolerances for dimension s A, B, C and D of components G3 and G5 were arrived at scientifically to +/- 15 mum instead of the existing +/- 50 mum, so as to control th e G3-G5 gap. Hence for further procurement of components G3 and G5 it was decided to give the suppliers the revised requirement of +/- 15 mu m. In short-term solutions, the required dimensions of the G3-G5 block in the jig were calculated for all the four different combinations of components G3 and G5 and suppliers 'X' and 'Y' such that rejection ow ing to the G3-G5 gap being outside specification is minimal for the ex idsting stock of G3 and G5 in the store. Implementation of the recomme nded block dimension for a particular combination (which was in produc tion during the period of the study) resulted in a reduction of reject ion from 0.92 to 0.22%, which will save 0.525 million rupees (US$32000 ) per annum.