STATISTICAL PROCESS-CONTROL - ASSIGNABLE CAUSES AND DATA FORECASTING

A mathematical model is used to evaluate chemical process data and isolate assignable causes. Any remaining non-assignable cause (process randomness) is then treated by the classical Shewhart method. This methodology enables design of process control strategy, improvement of process consistency, and control ofthe process at the "State of Statistical Process Control" by demonstrating that the process is stable, predicatable and capable.