Online statistical process control with NDE and computers

Abstract

It is shown that, for statistical process control (SPC) using ultrasonic instruments for nondestructive evaluation (NDE) directly interfaced with a computer, a variable in every part can be measured at production line speeds and used in groups, typically of five, to generate points for control charts. The process control computer can then analyze the control chart (held in its memory and updated in real-time) to indicate out-of-control conditions. An indication can be produced in the time to manufacture from 5 to 40 parts, not in hours as with manual SPC. The computerized speed cannot be used directly because the statistics of control charts yields Type I errors (calling good production bad) of 1% probability per control chart point. However, an algorithm has been developed yielding an error probability of three parts in 100000. The algorithm is based on a Monte Carlo study of the statistics of control chart run rules. The Monte Carlo simulation is presented.<>