Results May Not Vary

摘要

Controlling manufacturing processes so that products meet customer specifications is difficult. Understanding and quantifying the impact of process inputs on the outputs is critical to process control. In such situations, factorial experimentation can be the key to understanding the impact of each process input on the process outputs. Metalor Technologies, a Swiss-based metallurgy enterprise with subsidiaries in 15 countries, used a well-designed factorial experiment to better understand its silver powder production process. Implementation of the results led to a 200% improvement in product quality and control in its production process. Using design of experiments, Metalor set out to determine how density and surface area are affected by three key inputs: reaction temperature, ammonium and stir rate. They illustrate a 50% decrease in process variation, leading to higher quality powder that meets the customer's specifications. In addition, rejected lots were reduced by 75%. Their approach is generally applicable and can be used to improve product quality and process control in many manufacturing operations.