Direct Torque Measurement versus Speed derived torque

[Porsche356supersport]

I have a question which I hope people can give some answers to.

Today the torque of an engine or transmission is construed from speed sensors fitted to either the crankshaft, input and/or output shafts.

I've learned that many companies now want to measure torque directly using either the strain or the shear stress in a rotating shaft.

Why is a direct torque measurement better than a speed sensor derived torque measurement when speed is a component of torque?

Is it today with transmission losses that make speed sensing less accurate than direct torque measurement.

You can see an example of what I mean in the link below.

 

[patprimmer]

Speed is not a component of torque. Torque times speed equals power.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[IceStationZebra]

The link you supplied shows a system that measures shaft strain to derive torque, which combined with speed gives you power.

I don't work for an automotive company, but I have dealt with sensors like this throughout my career and I have never heard of deriving torque/power from speed measurements alone.

ISZ

 

[GregLocock]

You could /estimate/ the power, and hence torque, instantaneously from the MAF and the engine rpm, since if you measure a few engines you can work out what the average relationship is. You can also try and work out bmep by looking at torsional vibrations, form the crank toothed disc.





Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[Porsche356supersport]

Hi there, thanks for your replies so far.

If it is not the case that the speed sensors are used to derive engine or transmission torque then can you identify how this is accomplished today.

Then having identified the above can you indicate why companies are now seeking to measure torque directly on a rotating shaft.

What I'm trying to identify here is the reason to move from the current system to direct torque measurement - is it measurement error, lack of accuracy due to the interference of other torque effects. etc..

I understand how the MDI technology works (see also ABB Torductor, Magcanica,NCTE Gmbh, Transence Technology, etc..) but I am trying to figure out (which they don't identify) why this would necessairly be better?

 

[ivymike]

A common way to measure torque is to measure the force required to react the torque at the mounting structure for the engine or dynomometer.

A reason to want to measure shaft torque directly is that you cannot always capture all the force reactions in a real-world (outside of the lab) setting, and sometimes you'd like to know what torque really is. Take a gas compression skid as an example - there are many ways for the torque of the engine to be reacted at the engine and compressor mounts, and if there is a rigid structure connecting the engine to the compressor (as you have with a transmission) then there is very little (or no) external reaction required. Without a direct shaft torque measurement, you can't precisely know engine output power, which is something you might like to know for controlling your equipment.

 

[patprimmer]

Can twist in a drive shaft be measured with two sensors or chopper wheels that measure a certain spots on each end of the shaft.

The degrees change of alignment or the time delay from one to the other as a fraction of the time for one complete revolution should provide enough data provided you know the torsional rigidity of the shaft.



Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[Porsche356supersport]

I had a look at this both an optical and magnetic method but both suffered from issues of contamination (specific to optical) and temperature coefficient, packaging space and alignment issues. The areas I have an interest in are the crankshaft, input/output shafts and half shafts. You really need a small package that can be glued/clamped on with signal transfer via RF, inductive coupling or electromagnetic impedance. That still doesn't answer my question - why is it preferred to measure torque directly on the shaft versus a computation of torque. Also torque sensors are generally not fitted in high volume applications today (I'm not talking about testing applications) so how is torque being derived today! Again thanks for all the contributions.

 

[GregLocock]

My reply above indicates how OEMs can give the 'keen' driver a power estimate (which after all, they cannot argue with). It is not a direct measuement for sure. I've used strain gauges to measure driveshaft torques for 14 years, they are instantaeous and accurate. Like all real time instrumentation, the signal can require a lot of post processing.

What makes you think OEMs are measuring torque directly on prodction cars? My guess would be to improve the responsiveness of systems like torque vectoring.



Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[IceStationZebra]

Torque transducers are relatively expensive and somewhat delicate. The only reason I can think of to use one in a production application is because a computation isn't dynamic enough.

ISZ