# Speedy Pressure Vessel Fatigue Cycling

Gents,

I am looking for a design modification to accomplish 250,000 cycles of pressure loading in a short time period. We are going from 0 (1atm) to 10psi but currently only getting 10 cycles in 120 seconds. Enclosed is a diagram of the system.

Details:
• Rate goal is 3hz
• Power Source is air at 120 psi
• Pressure Vessel is 4.75 cubic ft
• Fluid inside the pressure vessel is 75% antifreeze and 25% water
• Actuator1 is 1quart size and filled with air
• Actuator2 is 2 quarts in size and filled with water
• Actuator1 is linked to actuator2 with a steel ram
• Actuator2 is linked to the pressure vessel with a 0.25in (6.3mm) hose
• The control system box has a 120psi inlet and a 1/2" air hose
• Actuator1 has two 1/4" inlet/outlet lines
• Actuator1 can pressurize either side of its diaphram through the 2 lines face
• Actuator2 has an open side of the diaphram and a fluid side

The theory is that the rate of pressurization is much too slow due to the small diameter input lines. What should the correct configuration be? Some ideas if we keep the same power source is to increase the line size (to what?) and to employ a pressure tank very close to the system to enable large (fire hose size?) input lines. If that is the case adequate control will also be and issue. 3hz will require just 0.16 seconds for each stroke of compression or tension on the actuator 2 rod with transition (turn around) time going to zero. Other ideas include using and electric motor with some sort of cam system to create the correct pressure over that number of cycles.

So in a nutshell. . .
What is the physics solution behind the problem?
What is the optimal and most economical setup to get the testing done?

V/R
Tomax

#### Attachments

• PressureVesselSystem2.png
38.5 KB · Views: 389

chemisttree
Homework Helper
Gold Member
You need to find out how much mass change of the glycol/water will cause the tank to reach 10 psi. A piston pump with the head removed and attached directly to the inlet of the tank and driven by any power source operated at 200 rpm will suffice. In this configuration the new "head" for the piston pump is the tank itself. The piston displacement should equal the mass change required to produce 10 psi. You will have to degass the whole thing carefully and be sure to equip the tank with a bleed at its highest point.

It will be loud and low!

I like it

c-

The idea is simple and straight forward. I will keep the data logger as is but will not need the control. I think perhaps we could also fine tune the overshoot by having some adjustment in the stroke length -- perhaps a threaded connection.

V/R
tomax

chemisttree
Homework Helper
Gold Member
Slightly oversize the pump. Add material to the top of the cyl and Dremel it away until you hit the sweet spot. Epoxy blob leaps to mind. Place the pump on the bottom of the tank to ensure that any cavitation-induced degassing does not hide on/in the cylinder/piston.

C-?

It was brilliant!

As far as tweeking the amount of mass moved, I am thinking about a cam on a rod (camshaft system). The downside of this is that we have many tests with varying psi requirements around 10 psi. To change the stroke of the actuator we would need a variable cam setup.

The second requirement would be a high return speed. The question is, how much force is needed to quickly return the actuator to zero position given fricton and resistance. Perhaps we would need an opposing spring on the actuator rod to impose a significant preload.

tomax