Monitoring torque on a shaft.

Used the ABS in an instrumented car,it came from the centre of a rearwheel axis ,the signal was a coil pickup,16 sinewaves/rotation going up to ~100 volts at high speed. For safety I picked it off that line with a 1 MEG resistor,which i fed into a small cap. That nicely converted it to a constant voltage over a large range, and on a CMOS smitt trigger it worked from .5 KM/H on up.

I see in another post you say that this is a FWD car. That might make the shaft strain vs torque a bit trickier to measure (smaller diameter, shorter shafts to work with).

Here are two ideas:

1) Take a look at the CV joint adjacent to the transaxle. See if there might be a way to place a thin sensor between the CV joint and the flange to which it is attached. Most half axles have enough axial play (usually splines) to accommodate a fraction of an inch.

2) Measure the axle twist by attaching a sleeve around the axle, fixed to the axle at one end. The sleeve will not experience any torque and will not twist while the axle does. At the free end of the sleeve, measure the displacement between it and the axle.

--
Paul Hovnanian     mailto:Paul@Hovnanian.com
------------------------------------------------------------------
If everything is coming your way then you\'re in the wrong lane.

Influenced by the 40foot caravan you're towing, the dingi you have on the roof and the frontwind blowing with 20 knots.

Rene

--
Ing.Buero R.Tschaggelar - http://www.ibrtses.com
& commercial newsgroups - http://www.talkto.net

A modern car has fuel injection, which is precisely metered to match the amount of air taken in to create a clean combustion. All this under control of the engine management. It seems that there should be a tight correlation between injection puls shape (or length) and the torque generated. Action of e.g. a turbocharger is "automatically" factored in as the management adjusts fuel injection to match intake air pressure. Same management also does "air mass flow" measurement somewhere. In short - there is a cpu in your car that alread knows it all. Just tap the right signals :-)

--
 - René

Not to forget the air resistance due to the car actually moving...

Nevermind. It was Friday and I wasn't thinking.

Iwo

Why not measure how much the engine moves?. The torque being delivered to the gearbox output shaft, will directly relate to how much the engine mounts deflect.

Best Wishes

Modulo cam shape, internal drag, ... I already have GPS data, to graph injector width vs speed/delta-z. It diddn't really work.

On Tue, 14 Mar 2006 16:11:09 GMT in sci.electronics.design, "Roger Hamlett" wrote,

The engine also moves for other reasons, like going over a bump.

Would think these factors are reproducable - hence a lookup table might help?

--
 - René

I would think so too - but not without a fair bit of time on a dyno, and calibration every so often for engine wear, ...

Primary measurement would be so much nicer.