5 replies to this topic

[Andrew Murphy]

Hello All,

This is the first time I've posted on the Tech Forum, I normally subject myself to the terrors of the RC forum .

I do some engineering/mechanic work for a friend who participates in the S2 class in SCCA. We feel as though we've pretty much tuned everything there is to tune on his car except that we've NEVER done anything on his shocks. We got a recommended setting from a reliable competitor [yeah we're foolish enough to trust him!] about three years ago and we haven't touched the settings since. The shocks are triple adjustable Ohlins.

I was thinking about making a shock dyno to assist our evaluation of the various settings so we'll have at least some clue what we're doing at a test session.

Here's the basic layout I'm considering:

One end of the shock will be fixed. The other end will be connected to a 2.00" stroke pneumatic cylinder of known bore diameter. The air feed will be regulated by a high quality laboratory spec valve. To perform the experiment the operator will flip a switch or valve to allow the regulated air to activate the cylinder and compress the shock 2.0". So, you have a known force from the cylinder and a distance over which the shock is stroked. Now, what I want to do and haven't figured a good way yet is how to time the process. I'm not sure what kind of 'off the shelf' unit I can get or rig up to accurately time the motion from 0" to 2.00". Any suggestions?

DREW

[MclarenF1]

the problem with that is you will only get two points on your damping curve (line). So you won't be able to determine what the shape of the deflection vs. time curve is i.e. the nonlinearity. I think it would be worth your time to measure the displacement with either a string pot (fairly cheap) or a linear position transducer (more expensive but more accurate). You should be able to make a fairly simple data acquisition unit or purchase a cheap single channel box (micro measurements, national instruments, HBM, ect). I think this will produce much more usable results.

[Andrew Murphy]

Originally posted by MclarenF1
the problem with that is you will only get two points on your damping curve (line). So you won't be able to determine what the shape of the deflection vs. time curve is i.e. the nonlinearity. I think it would be worth your time to measure the displacement with either a string pot (fairly cheap) or a linear position transducer (more expensive but more accurate). You should be able to make a fairly simple data acquisition unit or purchase a cheap single channel box (micro measurements, national instruments, HBM, ect). I think this will produce much more usable results.

With regard to the 'single point' issue:

My plan was to take multiple measurements at differnet air pressures [i.e. different force] but keep the damping factor on the shock constant, thus building the force vs velocity curve. To calculate velocity I need time hence my question.

To summarize the opration of the machine:
1. set air regulator to 10 psi
2. flip switch to activate machine
3. record time to compress shock 2.0"
4. calculate force and velocity, add to graph
5. increase air regulator to 20 psi
6. repeat steps 2-4
7. increase air regulator to 30 psi
8. repeat steps 2-4
9. repeat until you have measured up to 100psi

DREW

[Greg Locock]

If you are happy with fixed stroke testing why not use a large flywheel and do a cyclic test?

Real shock absorber machines use hydraulics, I can't imagine that a pneumatic system is going to have the power and speed simultaneously to calibrate a shock.

Figure on speeds up to 2 m/s for a circuit car, and resistances of the order of 1000 N/m/s, so crudely you might need 2 kW of pneumatics, not impossible, but not easy.

Any idea how you were going to measure the force in the shock?

[Andrew Murphy]

Originally posted by Greg Locock
If you are happy with fixed stroke testing why not use a large flywheel and do a cyclic test?

Real shock absorber machines use hydraulics, I can't imagine that a pneumatic system is going to have the power and speed simultaneously to calibrate a shock.

Figure on speeds up to 2 m/s for a circuit car, and resistances of the order of 1000 N/m/s, so crudely you might need 2 kW of pneumatics, not impossible, but not easy.

Any idea how you were going to measure the force in the shock?

I was going to calculate force by taking the sum of pneumatic pressure and piston area.

I can't say the idea is original. I've seen shock dynos that are pneumatic that use a 2.00" bore cylinder. That said your point has merit.

DREW

[Greg Locock]

If you've seen a system that works, fine, go with it.

You may need to compensate for pneumatic friction, seal drag in the pneumatic cylinder, and the inertia of the piston and linkage. So at a rough guess you'd get F=pressure*Area-mass*acceleration-Friction-1/2*Cd*rho*pneumaticvelocity^2

Where Cd is some sort of discharge coefficient of your pipework.

You might be able to knock up a strain gauge based load cell fairly cheaply, parts are available from RS.