Torsion Spring Life Cycle Test Machine - Need some advice

[Shockdesigner331]

Hello all. I need to design a quick and Simple torsion spring tester.
About the spring to be tested:
-Deflection angle required = 105°
-Anticipated Torque required to rotate to 105° = 70 in-lbs
-Number of cycles needed = 100k

My idea (by no means am I thinking this is the best way, but probably the easiest):
I have access to a pneumatic ROTARY ACTUATOR that I can control the angle of deflection to meet what I need and it has more than enough torque (actually overkill) but I have no experience in automating the rotation cycles. There are limit switches that could be purchased for the actuator.
Anyone have any ideas? If I could just supply air to the one side for 1 second, release, then supply air to the other port for 1 second and release - then repeat this cycle...

 

[FJTOM]

Why not use an electric motor driving a pulley or wheel with an eccentric pin. Pin connected by Heim jointed rod to spring tester.(ie crankshaft/rod assembly). Motor rpm, or motor and pulley system if speed reduction is required, should be easy to time to get to 100K without specific counting. For instance 100K/1800rpm = 55.5 minutes run time.

 

[Shockdesigner331]

Thanks for the response. I thought of an arrangement like that, but then I'd need to buy the motor, a couple of pulleys to reduce the RPM's, shafting and a more complicated stand. Trying to see if I can use what I have.

 

[handleman]

How about a pneumatic relay or mechanically actuated valve? You'd have a fully pneumatic control circuit with no need for any electronics. The act of the actuator reaching the end of its stroke trips the valve.

-handleman, CSWP (The new, easy test)

 

[DesignerGuy16]

You also might think of setting a motor to a cam to compress it once a revolution. Get a vfd on the motor and use a potentionmeter to adjust the hz to a known RPM that's reasonable and let it run. Calculate what size cam to deflect the spring (or have the cam move something else) to cycle the spring.

I did a similar setup to life cycle test some compression springs that were compressed with a linkage setup. The end results were very accurate vs. claims from the field. Actually ended up going from 200,000 cycle life (someone really screwed the pooch on the spring design on this one), to 7,000,000+ and still going with a simple logical fix.

Just an FYI, my test I ran at 450 RPM (was way over actual cycle time, but was within reason and didn't create any wierd effects on the spring during testing. At 450 RPM it still took ~8hrs to get 200,000 cycles. So you'll be in for a long test I'm sure. Even with that I checked it every 45-60 minutes so my % error was relatively low in the general scheme of things.

Have fun!

James Spisich
Design Engineer, CSWP

 

[SteveWag]

"If I could just supply air to the one side for 1 second, release, then supply air to the other port for 1 second and release - then repeat this cycle... You can do just that with a simple P&B relay, CNT-35-96 and one or two solenoid valves. Might be a test of the actuator though.

Steve

 

[KENAT]

thread404-221229 budt gave some good links on pneumatic logic.

KENAT, probably the least qualified checker you'll ever meet...

 

[MintJulep]

Simple control using the two limit switches and two three-way solenoid valves would do the trick.

 

[IceStationZebra]

One nice thing about using a switching type arrangement (as opposed to a motor/cam set-up) is that it will stop when something breaks, which you need to plan for unless you intend to watch it all the time.

Also, don't forget to include a cycle counter somewhere.

ISZ

 

[Shockdesigner331]

Thanks guys.

 

[zekeman]

All you need is a latching relay together with your limit switches to make the oscillator. When the rotor goes to one end LS it actuates the latching relay to power the pneumatic in the opposite direction allowing a reversal to the oposite LS; you now have the simplest form of oscillator using your idea.

 

[Shockdesigner331]

Thanks zekeman, I was just thinking about how the limit switches would open the solenoid valves. The latching relay answers all my questions in that regard. I just did a few quick internet searches for all the stuff that I need, looks like this will be pretty simple (and cheap). Anyone have any recomendations with regards to reliable makes of solenoid valves?