Frequently pressure vessels are subjected to a proof test for the purpose of establishing a safe allowable working pressure. Various techniques are commonly employed. They include the use of electric resistance strain gages, brittle lacquer, deflection dial indicators or a combination thereof. In addition, a test to destruction is often used for the purpose of establishing a safe working pressure.

The purpose of this paper is not to investigate the adequacy of the numerous tests for evaluating a safe working pressure but to analyze the significance of the data thus obtained. For simplicity the discussion is limited to uniaxial and biaxial states of stress only.

A given test may produce ample data when the vessel is to be used under a given set of conditions, while the same data may have little significance when the operating conditions of the same vessel are altered.

In so far as the author is aware, the ASME Unfired Pressure Vessel Code contains the only published rules for conducting a proof test. For this reason reference is made to these rules throughout the paper; not in the way of criticism but as a basis for discussion.

Since many designers use the rules of the Code for vessels which fall outside the scope of the Code, either with respect to pressure or material, a critical review can serve a useful purpose.