Modeling of screening effect to number of thermal cycles by investigation of space hardware thermal vacuum test

Abstract

Thermal vacuum and thermal cycle testing are widely complemented in space flight hardware and components as an effective way to screen out potential defects in manufacture and integration. The number of cycles is a test condition which affects the screening effect. The more cycles makes higher screening effect, however, it needs more testing cost. The quantitative analysis of screening effect versus number of cycles is an important issue which will help to decide the appropriate number of cycles for the space programs. There were researches to model the screening effect versus number of cycles by investigation of nonconformance detected during thermal cycle testing of aviation electronics in a statistical way, but there was few report related on space experience. Nevertheless, the confidence of model for space flight hardware was much less reported. This paper proposes a method of modeling the screening effect with number of cycles by applying the nonconformance sampling of space hardware. Validation of proposed model is discussed and its statistical confidence for accounting limited samples is calculated. From the proposed model, it was indicated and confirmed that over 99 % of space flight hardware and components procured on the basis of JAXA design standards are nonconformance-free after 8 thermal cycles in 85 °C test temperature range with 90 % confidence.

Keywords

• thermal vacuum test
• thermal cycle test
• number of cycles
• test effectiveness
• space hardware
• nonconformance data
• lognormal
• weibull
• exponential
• extreme value