Measuring thrust (and NOT torque) of rotating shaft with strain gauges

[1gibson]

I'm trying to determine a strain gauge configuration that will allow me to measure axial strain (thrust, in tension) of a rotating element. Shaft, or coupling spacer.

I want to cancel out the torque loading which will be significant.

It seems that a Full bridge type III with 4 strain gauges would work. Two of them 90* on opposite sides of the shaft to measure axial strain, and the other two at 45* to cancel torsion.

I've seen the application as described, except 0* orientation to cancel bending... Which is negligible in this loading scenario. Please see link, last item on the page.

http://zone.ni.com/devzone/cda/tut/p/id/4172

Will this arrangement reject strain due to torque if using a 45* orientation?

Additional information:

1) As you can see at the bottom of that page, someone (not me) asked a few months ago if the 0* arrangement would cancel torsion, I assume that it will not.

2) It is not an option to measure anything related to a bearing, or mounting flange, this would be too intrusive. The measurement must be derived from strain gauges on the rotating shaft. (This is why I asked in the instrumentation forum and not in a mechanical forum!) I have done some research, closest I came to similar application was measuring thrust AND torque on propellers for boats, but I want to remove the torque so this was not 100% applicable. Same with automotive components, they are concerned with material stress of the combined loading, I am not.

3) Plan to use a very small data logger that will rotate with the shaft. I mention this, because obviously getting the readings from strain gauges on a rotating element would be a little tricky and I want to avoid the response of "doesn't matter because you won't be able to wire it up."

4) Thanks in advance for any guidance.

 

[BrianE22]

Put two gages in the axial direction. Put them in opposite sides of the bridge. For about 30% extra output use 2 poisson gages (each near the axial gages but at 90 degrees to the axial gages). Mount those in the other two legs of the bridge.

If all gages are positioned perfectly then the bridge will only be sensitive to axial tension/compression. Misalignment of the gages may make them sensitive to torque.

If the axial strains are small relative to the torsional strains then it may be tough to make the transducer insensitive to torsional strains.

 

[1gibson]

Appreciate the response, so you are saying that 0*/90* will not be sensitive to torque? And trying to use 45* is not necessary?

Most cases, axial thrust value(lbs) will be around 2-4x torque value(ft-lbs)

 

[BrianE22]

Yes, you do not need torsion (45 degrees to axial direction) gages.

It's not so much the lbs and ft-lbs that is important. It's the strain due to each that is important since that is what the gages respond to.

Have you looked at Vishay - MicroMeasurements web page? It's been awhile since I have but they used to have quite a bit of info on gaging.

 

[1gibson]

Thanks again, yes I've looked at the Vishay site they have some nice dual element 0/90* strain gauges. I've also emailed them re: this application.

 

[WilkinsM]

HI 1Gibson,

I'm new to the forum so sorry if this has already been said or is teaching you to suck eggs, but you may have issue with the logger on the shaft depending on the RPM i.e. G Load you you be subjecting it to?

It may be an option to look at a telemetry system or a slipring to bring the signals off the shaft and record / process them locally? It might be worth googling "Datatel Telemetry" or looking at one of the miriad of slipring suppliers to make the acquisition simpler and enable you to monitor it real time?

Hope this is useful

Matt.

 

[Skogsgurra]

Yes Wilkins, that is old hat to most of us. And, if you read the OP you will find this: "Plan to use a very small data logger that will rotate with the shaft. I mention this, because obviously getting the readings from strain gauges on a rotating element would be a little tricky and I want to avoid the response of "doesn't matter because you won't be able to wire it up."

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[WilkinsM]

Hi Skogsgurra.

Good point, I didnt read the OP with enough detail, however I think my point may still stand, I have never managed to find a data logger that is small enough not to affect the rotor dynamics on a small shaft & has a high enough sample rate to be really useful.

Sorry to hijack the thread but if you know of one I would be interested!

Matt.

 

[1gibson]

This is the smallest I've found:

http://www.omega.com/pptst/OM-CP-BRIDGE120.html

Dimensions: 42 H x 68 W x 20 mm D (1.7 x 2.7 x 0.8")
Weight: 60 g (2 oz)
Reading Rate: 20 Hz to 12 hours

All solid state, so I can't imagine any issues with loads from rotation.

It will work for some applications I have in mind, installed inside a rigid coupling spacer (which transmits torque and thrust.) More difficult to set up on a small shaft OD, I agree.

Appreciate your responses.

 

[ScottyUK]

1gibson,

Don't under-estimate the long term effects of high speed rotation. We used a rotating earth fault detetctor on the shaft of our generators. The unit was solid state and rotating at 3000 rpm. They had an indifferent survival rate, considering the cost of them they had a fairly poor rate to be honest. I know they didn't contain any electrolytic capacitors, which are the most vulnerable to rotational damage, but we always suspected that the bond wires of the semiconductors were the next most vulnerable component. The whole things wa spotted in resin which we always failed to remove without destroying the circuit in the process.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[1gibson]

Good to know, was this recent? I'd assume most of the components in the one I linked are surface mount, which should be less likely to suffer than older "pin mount" PCB components.

This would be for short duration testing, intermittent use, non critical application. Max 3600 rpm, more often 1800/1200. So if I go this route I'm pretty optimistic.

Unless yours used surface mount components as well, then all I can say is that I plan to place it in the middle of the rotating coupling, so velocities will be lower than mounting to the OD of a shaft for example.

Can hold it in place with foam so there is minimal shock from vibration.

 

[ScottyUK]

This is a current product, although not a particularly new design. The whole thing was embedded in (what I assume to be) epoxy resin within a machined aluminium housing, so pretty rigid. Must have weighed half a kilo or so. The big factor working against it was being bolted to the circumference of a shaft about 18" in diameter, so the acceleration was much higher than on a small shaft or on one mounted on the axis.


----------------------------------

If we learn from our mistakes I'm getting a great education!