A critical evaluation of RGA testing

Silanized glass ampoules containing 12 accurately known gas mixtures were carefully prepared by a laboratory. Ampoules were sent as blind samples to four laboratories for residual gas analysis (RGA) using mass spectrometry. Results showed gross inaccuracies and inconsistencies, particularly for water vapor. Even relative measurements of water vapor level (i.e. ordering of samples from dry to wet) were inaccurate. There was great inconsistency in results between the analyses of the same atmosphere by different laboratories. The deletion of this test from MIL-STD-883 until an accurate, repeatable test method is established is recommended