STATISTICAL ANALYSIS OF RADOME THICKNESS

In the 1980's Satellite Communications were being deployed in all the services. Satellite dishes need protection from the elements, natural and battlefield. One program required antennae that could survive not only extreme weather exposure but also nuclear blast waves at a specified distance. The engineers and scientists calculated that a protective radome of a specified wall thickness and variability be used to protect the dishes. The problem was that the process for making the radomes had more variability in wall thickness than allowed by specification. Rejected radomes were piling up at the supplier.

Steve was helping to manage the subcontract for the radomes and on the side was teaching the staff there Statistical Process Control (SPC). He showed them how to determine the Cpk, the process capability and reached the conclusion that they would never have a yield much better than 60%.  

Armed with data and analysis Steve decided to apply to the DoD for a waiver. When the scientists compared the data to their current analysis they decided that an even better solution was to open the tolerances via specification change order. Within a few weeks the new specification was approved and nearly all the rejected radomes were accepted saving the supplier considerable money and significantly reducing schedule risk to the prime contractor.