drivetrain losses

joey.lisano

New member
This conversation spilled over from BAD STI's journal.

I believe drivetrain losses are not constant and dependent upon power input. A 300whp and a 400whp STI cruising down the street should have roughly the same drivetrain losses since they are using the same amount of power. At WOT, however, the 400whp car will have more drivetrain losses due to the increased power input.

Efficiency in gears is measured just like other efficiencies.
power output = efficiency * power input
output torque * angular speed = efficiency * input torque * angular speed
power lost = (1-efficiency) * input torque * angular speed

The engine makes a certain amount of a torque at a designated engine speed. Torque times angular speed is power. Only a true nerd would have his tachometer read in rad/s, so there will be some unit conversions to get between ft-lb, rpm, and hp. Regardless of the units, if you input more power, you're going to have more losses in the gears. Spur (straight cut) gears and helical gears (found in the transmission) are typically 98-99.5% efficient. On the other hand, automatic transmissions use a planetary gearset as well as a torque converter instead of a clutch. Bevel gears (found in the differential) have a wide range of efficiencies, but for intersecting axes, we can assume 98-99%. Correct me if I'm wrong, but we have three differentials, whereas FWD and RWD cars only have one. That means you're going to have more losses from the gears. From the picture below, you can see there are quite a few gears engaged within an open differential, and it only gets more complicated with more advanced differentials.

In addition to the gear losses, you have a bunch of other things. Viscous dissipation in the oil, friction everywhere, wheel bearings, rolling resistance on those 295 tires :shock:, etc.

As defined, gear losses are linear; ie, if you double the power input, you get double the losses. However, there are nonlinear losses like the viscous dissipation and rolling resistance. As a result, you cannot using scaling. If a stock STI on your tuner's dyno measured 2XX whp, you cannot use the 2XX whp and 300 bhp to reserve calculate what your bhp based on your whp. You could do that if drivetrain losses were a linear system, but they aren't. It may be a good prediction, and it may be great to brag about, but it is conceptually wrong nonetheless. Of course, you're average person doesn't understand the difference between whp and bhp, so in that case you may want to quote them on the bhp number and just say XXX hp, or maybe XXXX hp. :tard:

types of gears
gear_types_640px.jpg


manual transmission
transmission-5speed-gears.gif


a classic, open differential.
torsen3.gif


limited slip differential
torsen2.gif


Torsen differential
RealTorsen.jpg
 

HolyCrapItsFast

Drinks beer!
Yeah... I spent all day at work trying to exonerate myself and as the day whent on I was realizing I have lost touch with my engineering roots. I was begining to remember bits when I was a student. You lose track of things when your career has you focused on only one or two aspects of engineering for 20 some odd years.

I now sit in your camp.

Now if I read you correctly this attacks both the misconception that drive train loss is a constant and the miconception that some have about drivetrain loss being a fixed percentage (not the same as a constant)

For example lets assume we have a gear set with a 1:1 ratio for simplicity and you apply a power of 300hp at a speed of 5000rpm. In the Ideal 1:1 wolrd you would get the same out but since this is not an ideal world you have windage, churning, sheer and rolling losses which are all forms of friction. For argument sake lets say it is 15%. That would mean that my theoretical output would be 255hp.

Now lets apply 500hp at the same RPM into the same gear set. By many peoples reasoning the power out would be 425hp for a drive train loss of 15%. This seems absurd to me. The percentage would increase as power increases and I believe this is what you suggest because the power losses due to windage, churning, sheer and rolling losses are not constant.

Am I close?
 
Last edited:

Paul.c

New member
This is no debate!! Everyone is agreeing. However, it is very informative. Thanks for the info guys.
 

IGOTASTi

System Operator
Staff member
Wow we have such a great group of guys!
 

HolyCrapItsFast

Drinks beer!
I would say any information is great. I would certainly be interested to here more on how power can be beter judged via consitant 1/4 mile times. After all power is a measurement of work done over time and the work can be easily calculated form mass and distance traveled.
 

joey.lisano

New member
Sure, post up some links. I think I've heard of the 1/4 mile calculation before, but I figured the launch might be too much of a factor.

If you need to edit my first post to add more info, feel free. You can even delete everything I wrote, and type, "I love lamp." :D
 

berdugo

New member
great info Joey :):thumbsup:

I lol when I read the 295 tires, a WBK and 295 is in my future modifications, it's just waiting for installation :lol:
 
Top