Something I've never understood.

I wonder about this from time to time, but I always forget to ask.

We talk about driveline losses as a percentage. For instance, many people say to subtract 20% from your flywheel HP to get rearwheel HP, or conversely; to do the reverse math to get FW HP.

Here's the thing though...

If you have a 200hp engine, then 20% is 40hp.

If you have 400hp, then 20% is 80hp.

I cannot see any logical reason why the hp requirements of a transmission should double just because the engine HP did. I mean, it's a transmission, not the IRS.

So how do you determine ACTUAL drivetrain power loss? Or to be more accurate, how do you estimate the drivetrain loss with a reasonable degree of precision without using both an engine and a chassis dyno and subtracting?

You can do a "coast down" test with the software in most chassis dynos. This will allow you to get a better estimate of drivetrain loss. But it doesn't really account for the different behavior of the drivetrain while under load, so it is not in any way definitive.

Rich

Might want to try a search.... I know I've posted on this many times. There are two components of driveline loss - inertia and friction. The inertia load is a function of how much each rotating part weighs and how fast the engine can accelerate them. That's going to be a fairly "constant" value.

Friction on the other hand is drectly proportional to load. Increase the torque being transmitted by a factor of two and you increase the friction load on the bearings and gear faces by a factor of two. Hence, friction loads are proportional to the power being transmitted.

Looking just at the transmission, the inertia loads will not change with power, but the friction loads will. Since the friction loads are the greater percentage of the loss, the power losses through the tranny will increase almost (but not quite) in proportion to the power being transmitted.

Throw in a fluid coupling, like a torque converter, and you have yet another issue of hydraulic losses, which increase exponentially as load increases.

There is a coastdown test on ws6.com an he got losses in the range of 18-20HP. That would roughly represent the inertia loads (with a little bit of friction thrown in, but not much). I have engine dyno vs. chassis dyno pulls on my original M6 setup, and by looking at the HP losses for power loads ranging from roughly 500 to 800HP, and using simultaneous equations to seperate friction from inertia, I got similar results.... maybe 20HP "constant" loss, with the rest being proportional to load.

Great explanation Fred, Thanks.

In reference to the frictional loads you mentioned and the way it increases; is that in reference to a mauan transmission as well?

I ask that because I can visualize the action in an auto, but not in a manual.

There's gear friction and bearing friction in a manual tranny. Gears involve sliding surfaces.... lubricated, yes.... but still a friction load. Locking it for 1:1 eliminates the gear issues.

Just to keep things in perspective.... my M6 setup lost 12.1-12.6%. Its generally believed that the 12-bolt accounts for 7% of that loss, so there's not a whole lot left for the other components. Bending the power 90-degrees really eats it up.