Hydrostatic testing of a high pressure reservoir (non-DOT rated)

[Geof180]

Good evening all,

I am looking at making a determination as to whether or not further NDI is useful once a hydrostatic test (completed every three years) is performed.

This NDI would consist of either a mag-particle or x-ray inspection.

Anyway, the regular operating pressure of the bottle is 3000 psi. It is proof tested to 5000 psi...burst pressure is 6667 psi.

I am inclined to say that further NDI is not required because any flaw in the bottle (either in the walls or in its welded seams) will be found during the hydrostatic test.

What about proof testing at -40ºF? That way, the concerned party would be absolutely sure their bottle is in tip top shape...

Geof

 

[zeusfaber]

Probably depends on the material the bottle is made of.

There have been some cracking problems in the necks of aluminium SCUBA bottles in recent years, which have led to a requirement for any made of the affected alloy to have an eddy current crack detection done every 30 months (concident with the inspection/hydrostatic cycle).

Steel bottles don't seem to be affected.

Not sure what sort of cracking was involved, so can't really help on whether the problem might read across to bottles which live in a kinder environment (less frequent deep charge/discharge cycles and immersion in salt water - all good stuff for a lightweight bottle made without a corrosion allowance).

I imagine the worry is one of fatigue growth - a hydrostatic test only proves that no cracks are catastrophically large now. It says nothing much about how big they might be in a couple of years (after perhaps 2000 charge/discharge cycles).

A.

 

[Rjeffery]

Additionally, Mag particle (MT) of the welded seams could show potential stress cracks. Straight beam ultrasonic (UT) testing of the field between welds (especially along the bottom or low point) could show internial corrosion that may not be evident.

Most Steam and "mud" drums on marine boilers are routinely surveyed using either or both methods after a period certain.

This is a very "belt and suspender [braces]" methodology.

 

[Geof180]

The bottle is made of 4130 steel.

Because I do not know the actual thickness of the wall, I constructed a crude graph showing the internal tensile stresses developed within the wall of the vessel. (The whole wall thickness issue is another story in of itself)

During the hydrostatic test, the material is not allowed to plastically deform, therefore the stress must be kept below the yield stress of the steel.

Even examining the situation using LEFM and various crack lengths, I was not able to get a theoretical fatigue failure that would occur in my lifetime...or the lifetime of the airplane... with the charge/discharge or hydrostatic testing cycles..

I understand that performing NDI would definitely tell me if any sort of inclusion or flaw is present, but there are other facilities (DOT certified) performing hydrostatic tests on the same bottles without the NDI.

As the engineer, I am just trying to give my customer the best response. At this time, I am inclined to say the additional testing is not required.

Geof