Pressure Testing the First Tank

How much pressure can this tank hold?  It's necessary to know which will fail first, the tube or the crimp and at what pressure.  Based on the yield strength of 40,000 psi (from the supplier's catalog) the tube can withstand up to about 1600psi.  However, this tube had several gouges in it so it probably wouldn't make it to the maximum yield pressure.

This is really primarily a test of the crimping technique.  I felt reasonably certain the crimp would be weaker than the tube, but I had no idea how to predict at what pressure it might fail.  The crimp appears similar to using a retaining ring, but how similar?  To find out what pressure it would fail at, and in what way it would fail, I set up this test apparatus:

600 psi test
The pipes on the right are on top of a positive-displacement plunger pump.  Pressure is set and maintained by an adjustable relief valve.  On the left end of the rigid piping is a ball valve that can isolate the test article while the pressure is adjusted to the intended test pressure.  The test of the tank starts when this valve is opened and the tank is pressurized through the flexible tubing.  In the photo above the ball valve is closed (hard to tell because of the viewing angle) and the gauge is at 600psi.  The tank is hanging freely from a steel wire in a plywood stand and a piece of 1/8" clear plastic hung over the front as a splashguard (hydrotest failures are usually not very energetic, but the jet of water at the failure point can travel at high speed for few feet).

The test followed a systematic procedure.  Beginning at a pressure 400psi, the tank was first visually inspected for leaks and obvious changes, the diameter measured, and then the tank pressurized to 50psi more than the previous test.  The tank was held at the test pressure for 5 minutes, depressurized and inspected.  This was repeated until failure at 1100psi.
Failure end

The failure was due to the crimp deforming until the end piece could slip past it.  This end first started showing deformation at about 850psi.   At 1100 psi 4,000 pounds of force acted on the end piece.  It didn't slip out completely because the pressure dropped immediately and the crimp caught in the o-ring groove.  Further insight into how it failed required cutting open the end of the tube to examine the crimp and end piece.

The top photo shows the same pieces as above.  The bulge at the top of the picture above is facing the camera.  There are deep score marks between the crimp groove and the opposite side of the o-ring groove.  Everything else is mostly intact.  I thought the ridge between the o-ring groove and the crimp groove might be collapsed or bent but it appeared to be unchanged.  I didn't record "before" measurements to make precise comparisons, but any changes were too small to detect visually.

Failure parts
Interior of tube at failure

Here's the inside at the same point.  The crimp was deeply scored and only protruding into the tube .005"-.008" (hard to measure due to bulging of the tube wall.)

Meanwhile, on the other end of the tank...

Tank other end

hardly anything happened !!!  The bulge had a profile similar to the end that failed, but the diameter increased only about .003".   An examination of the components shows no damage.

Other end parts

Note the absence of score lines compared with the failed end.