Testing a pressure vessel is dangerous

Stored energy, and why external pressure tests are done differently.

Testing a pressure vessel is dangerous because a vessel loaded from outside has a large amount of energy stored in the surrounding fluid, energy that was there before the test began and that the collapsing shape suddenly stops resisting the moment it buckles, releasing it all at once in a way a slow, quiet material failure elsewhere in engineering rarely does.

Water that keeps pushing after the shell gives way

Testing most structures to failure is comparatively safe because the failure draws mostly on energy the test supplies moment to moment. A beam that yields bends further and slower as it goes, and a bolt that snaps releases only the modest energy stored in stretching it. A tensile test is limited by how fast the machine can pull, and once the sample parts the machine stops adding energy.

An external pressure test loads the vessel with water pressing inward on every surface at once. Right up to the instant of collapse, the vessel's shape is holding back a volume of pressurised fluid ready to rush into whatever space the shell stops occupying, and that surrounding tank or ocean keeps pushing after the vessel has started to give way. The vessel itself may weigh very little, yet the implosion can damage nearby equipment and injure anyone close by, and even a small housing tested at a modest pressure can carry enough stored energy in the water around it to be dangerous.

Easing the cap off a well-shaken bottle of fizzy drink slowly, letting a thin hiss escape first, releases built-up pressure harmlessly. Let the cap slip from a thumb and the same gas rushes out in a fraction of a second, launching the cap and spraying the contents far beyond where the slow release reached. The amount of gas stored inside was identical; only the speed of its escape changed. A buckling vessel behaves like the slipped cap, with no chance of a slow, controlled release along the way.

Chambers, remote control and staged tests

External pressure tests are therefore run in a dedicated chamber built to contain an implosion, with the vessel isolated from anyone nearby, pressure controlled remotely, and instruments capturing the moment of failure without a person standing beside it. Testing a bracket to destruction on an open bench is a different posture entirely, and bringing that casual posture to a pressure vessel because the part looks small is the mistake that turns an informative test into an accident. The instruments need the same thought, since a sensor or camera close enough to capture useful data has to survive, or be sacrificed to, the event it is recording.

A vessel held well below its expected buckling pressure carries much less of this risk, which is why qualification usually proceeds in stages: proof pressure first, held and monitored, with a destructive test to collapse reserved for a separate, specifically prepared trial. Every earlier article in this set described a way of holding a shape correctly, and this last one is a reminder that finding out how much margin a shape really has deserves the same seriousness as the collapse it is meant to prevent.