Hello,

I would like to try and build another dyno. This time the dyno would be part of the car. The basic idea is to install a load cell on the car where it could read the torque required to move the car forward. They make load cell U joints but the u joints are expensive. I also thought about making some motor and tranny mounts with cells, however the best location I can think of is the torque arm.
A mount would be made so that the load cell would measure the force in the load cell as the car accelerated. Im trying to get an idea how sensitive the dyno would be . Has anybody ever broke a torque arm mount and countinued to drive? Will the car still drive with a broken mount or will the torque arm hit the car floor pan.? What I am thinking is, since the springs and absorbers push nearly at the center of the axle they cannot create a moment of torque and they will not stop the axle from rotating. Also the control arms are pinned so they shouldnt stop the movement either. In order for this dyno to work the torque arm needs stop nearly all of the axle roll.
Will it work?
Thanks for your help

 

Yes.
Yes, you can drive it, but it sounds like the end of the universe is upon you. I had the front mount for one break open, and release the mount. Any accel, or deccel, and it'd hit either side.

 

Thanks, Thats just what I wanted to hear.
That means that even small changes in the torque going to the drive wheels could be detected.
Where would be the best place to get the rpms for finding instant horsepower? The sensor would have to be on the driveline side of the transmission how about the VSS?
Also, I have noticed the torque arm seems like it is preloaded and exerts a force into the side of the tranny mount. Is this just to keep it snug in the mount? I dont think it will be a difficult project. It will be neat to see how much horsepower the car makes under acceleartion and how much power is needed just to push the car down a flat road.

Thanks again Grumpy!

 

How the hell would you calibrate this thing? Surely you're not getting anywhere near 100% engine power at the torque arm, maybe you get what, a 10% reaction? You really need to measure the driveline, not reaction forces restraining it. The relationship between engine power and the reaction at the torque arm is also certainly nonlinear, and probably impossible to predict accurately with a home-built setup.

 

The torque at the wheels is equal to the lever arm mulitliplied by the force the bracket exerts on the lever arm.
If the distance is known from the axle center to the loadcell on the torque arm ,and the force in pounds at the bracket is also known then the torque will be equal to the distance X the lbs of force.
Since the torque arm is what keeps the axle from rolling. Either the torque arm moves or the car moves forward.

 

I'm a mechanical engineering student, I get your relationship of axle housing torque to the torque arm. But, how are you relating how much the torque arm resists axle housing torque to the power at the axles? You said it yourself, either the axle torques or the car goes forward. The car DOES go forward. Your means of measuring power output would only be valid if the car did NOT move, AND only if the car did not "squat" on the suspension when you accelerated. In reality, the problem you are trying to solve is all dynamics. A simple torque = force x distance won't even get you close. The reaction forces from the driveline power go to too many things other than the torque arm (chassis flex, suspension compression, etc) Also, once the driveline power has overcome axle torquing (and other static reaction forces) and is now mostly being used to power the axle shafts, the relationship between axle torque and power output at the axle shafts will be much, much more complex. Now you have multiple dynamic reaction forces. It would be pretty hard to pick only one of these reactions, ie at the torque arm, and then predict total driveline power output from this. You need to measure power output at the driveshaft or axle shafts.

 

you might want to check out Land and Sea dyno's they have an on vehicle dyno that places the load cell on the drivshaft. It looks pretty safe.

http://www.land-and-sea.com/rotary-s...shaft-dyno.htm

 

computer fart (double Post)

 

That shaft dyno would definitely be the ticket for accurate onboard dyno'ing, if you want more than a novelty setup.

 

Could you place load cell(s) instead of engine mounts? Seems like if you got it set up right, you should be able to measure it pretty precisely.

 

Even with load cells as motor mounts, you are still measuring only a few of many reaction forces, not output power. You will be measuring how much power it takes to restrain the engine from spinning on its mounts, nothing more. There are so many reactions to the driveline power in the car, that using only one or a few measured reactions would not be enough to recreate the whole picture of driveline output (even considering all of the reactions won't do it, since the majority of the power moving the car is not going into reactions at supports). I have seen guys start V8 engines sitting flat on the oil pan in junkyards. They can even rev the engine a little and it won't fall over. Obviously the amount of force it would take to restrain the engine is not even close to its power output. This is not a problem you can solve with simplistic statics calculations, it's all dynamics.

 

I really like the looks of the torque transducer from land and sea. They are just a little over my price range for this project.

I had to think on this one, you bring up a good point with the chassis flex. After giving it some thought I still think it will work. The car squats down because of the weight transfer from the front of the vehicle to the rear. If I’m thinking about this right, as the wheels turn forward they create a torque that lifts the front end of the car. The torque comes from overcoming the inertia of the car. The load cell inline would still account for the torque in the chassis squat. Look at it in the most extreme case, the axle spins free. If the axle was able to spin then Nothing would happen. The car would not go forward. And the weight will not transfer.

I really like the idea of making some brackets with load cells. It would not be hard to make some for the motor mounts. It is the tranny mount that is holding me up. The mount for the tranny might have to have a load cell as well??? Thanks for your input Craig



The torque that is required to restrain the engine from spinning is the same as the torque in the driveline. The only thing I can see that would slightly change the results are the hoses and lines attached to the motor. I don’t think they would alter the results significantly. The engines you saw sitting on their oil pans had no load. Apply a load, and they would have the same torque trying to rotate them as the engine applied on the load.

 

Hmmmm. Yeah you're entirely correct about engines with no load having no force counteracting the load. I'm retarded, of course you can't measure the work done by an engine without loading it.

 

What if you tap into the front wheel anti-lock sensors, or install some from another application. Or you could use some on the rear, but then tire spin might be an issue.

Couldn't you calculate the force needed to accelerate a known weight (the car) over a certain time, which you would also know, up to a speed, which you know from the wheel sensors?

Which is basically what the accelerometers do, using the g-force vs time. They are pretty accurate over shorter distances. The integration gets them slightly out over longer distances.

This involves more calculations, than actual torque measurement. But it would still be good for a base value.

Or

Seems to me that if you measure the acceleration of the drive shaft, you could somehow measure the torque you are putting out that way.

Good luck btw, was impressed with the other dyno.

 

Check this out, I remember I was interested in doing something like this a couple years back for realtime optimization and turned up this effort. Looks real cool, and might just fall in the price range by now.
http://home.att.net/~jroal/dynorep.htm

 

Thanks Craig, great article. Looks like it would be easy to do, At first I thought 3 cells would be needed 2 for the motor mounts and one for the tranny, Looks like they made it work with one. I also didnt think of the effects of dynamics as the car turned or hit a bump. If the torque arm cell does not work for me I will try and go with the motor mount setup.
I have been thinking more on the cell in the torque arm and the only reason I can see it not working is if any of the suspension would stop the rotation of the differential under acceleration. If the suspension does alter the affect under hard acceleration I might have to build some brackets for the control arms and springs so the differential will always be free in its mounts. Another potential problem I can see is where the pivot point for the differential is located. If I remember right the control arm bolts a few inches forward from the center of the axle. I think there might be a slight problem from hitting bumps and creating a torque in the torque arm from the off center pivot.

If I get some time this weekend Im going to put this together\
Thanks again everbody for your help

 

12hp at 50mph. 3400lbs. cd of .31 and a frontal areara of 13.4 sqft. ahh al you need is a way to monitor the vss and then calculate with an acceleromter when you started actuall moving then you can simply run a dyno with a few peice of software. wiat a minute. this could all be done on a pic. im going back to the lab.

 

ECM852 baby, it's all there. But then again, that's PC software.

All it comes down to is F=ma. But if you actually measure some stuff, then it gets pretty cool...

 

Has anyone thought about trying to measure deflection in the driveshaft?

They do this in aircraft.

 

I've been thinking about measureing torsional deflection in the driveshaft and it should be pretty easy with a traditional crank trigger style pickup.. I would just find the modulus by trial and error. Maybe one of you EE's can work up some circut to measure the difference between the two interupts.


Anthor thought... Some vipers came with a torque gauge. Anyone know how they calculated it? Measuring torsional deflection is ok in a aircraft where your making 3000ft*lbs but in a car it might be hard to measure such a small angle.

 

It can all be done in excel with deltaRPM/deltaT, or deltaMPH/deltaT, and either a constant load (psuedo dyno), or on the road (known mass). The RPM and MPH are in the data stream.

Also a G-tech would work if you want to do this on the road, and save the hassle.

If you're set on putting something on the car, then the two load cells at the motor mounts would work. Just find the difference between the two to automatically cancel out road bumps, and do it in a straight line. You'd have to rigidly mount the load cells to either the chassis or the engine, to prevent the engine from parallelogramming in the engine bay, though. Pin joint it to the other. As long as there's minimal engine deflection, other torque resistances can be neglected (hoses, tranny mounts, etc...).

 

I've already got the calculations built into my ECM852 program. Feel free to take a look at the source code to figure it out if you want.

Basically it takes the MPH, car's weight, coefficient of friction (in the form of a 'max mph', a bit crude but nonetheless), and then looks at d2U/dt2 to determine F=mA.

 

How about a load cell, in an engine mount?.
Measuring line pressure in the tranny?.

 

ah, but theres a error too hard to get away from...

mg*cos the angle of the hill your traveling on.

Close enough for goverment work though..

 

Hills won't matter, not when it's torque that you're measuring. The two load cells at the motor mounts works in most situations if you just take the difference between the two cells to negate bumps. It won't work when going around a corner, unless you somehow mount them across the center of mass of the engine.

 

I was referring to the moates software.

 

The SSR's also have torque gauges, may wanna check that out as well