Power Loss in the Drive-Train

[Mike DiFranco]

I am trying to test the power loss that occurs through the differential to the wheels on our last years car. I am looking into doing some sort of experiment in which I can hook up either an electric motor or a shaft to the rear sprocket and add weights to simulate an input torque. I am looking to measure the output torque at the hubs/wheels.

In our case the sprocket is attached to a torsen style differential basically the university special as I am sure most of you are familiar with.

I am looking for a low budget test that is fairly simple. I do not have access to a chassis dyno at this time. I do have the engine dyno data from last year's car. If anyone has done a test like this / has an idea of how to do one I would appreciate the help. Thanks!

[Mike DiFranco]

I am trying to test the power loss that occurs through the differential to the wheels on our last years car. I am looking into doing some sort of experiment in which I can hook up either an electric motor or a shaft to the rear sprocket and add weights to simulate an input torque. I am looking to measure the output torque at the hubs/wheels.

In our case the sprocket is attached to a torsen style differential basically the university special as I am sure most of you are familiar with.

I am looking for a low budget test that is fairly simple. I do not have access to a chassis dyno at this time. I do have the engine dyno data from last year's car. If anyone has done a test like this / has an idea of how to do one I would appreciate the help. Thanks!

[billywight]

The only loses you should see would be due to friction in the bearings and any inertial effects. The diff really is not a function of the losses unless the wheels are rotating at different speeds (such as in a corner) in which case you wouldn't really be worried about losses anyways.

Friction should be fairly negligable and inertial effects could be calculated rather easily if you have CAD models of everything.

[The AFX Master]

Certainly you will need to measure pinion/chain/sprocket power loses. as billy said, bearings are negligible (although i can`t say that regarding CV joints that are on bizarre angles).

Focus on measure chain efficiency. also if you do a steady state study, you can neglet inertias. But inertias are rather easy to calculate

[Patrick]

If you have a chassis dyno you can do a "negative horsepower run." This will show power losses as your drivetrain angular velocity decreases. If you could measure axle angular velocity with an encoder or something and spool up the drivetrain from the transmission input shaft you could also determine the power lost in the drivetrain.

-Patrick DeGrosse Jr.

[The AFX Master]

Good idea, If you do it also take into account tire drag at 0% slip ratio with your tire data. But Mike says thath he haven't access to a chassis dyno yet.

a simple setup of pulleys and weights acting over a sprocket and chain assembly is fair easy to do and will give you an idea regarding chain power loses.

[billywight]

You shouldn't neglect the inertial effects in your calculations as they can be quite large. Our 06 car made 85hp on the dyno at steady state, but only 65hp on an acceleration pull...

CV's (or tripods) at large angles might have an effect as AFX says, though I would expect them to be rather small in comparison to the power loss through the chain drive and the inertial effects.

A good way to realize the effect of inertial forces is to write a spreadsheet that calculates the energy necessary to accellerate the car from one speed to another (from corner exit to the end of a strait for example). Calculate energy necessary for rotational and linear accelleration seperately, and assume a time for which this takes place (this will give you power). Play around with changing mass and moments of inertia to see the effects of each on power loss - this may lead you to some different (or more specific) design goals.