drive train losses

[Alan Pritchard]

Hmm, I do believe a 15% figure is a sensible loss to work with...
Frictional losses in bearings, end loads on helical gears, friction between
gears, pumping losses of gearbox oil,  cv joints are particularly
inefficient, losses thorough tyre distortion and friction, air drag on
rotating components, there are a lot of losses to be had, especially through
awd...
Agreed though, a flat percentage loss figure needs to be taken with a pinch
of salt. Some of these figures will increase as engine output does, some
will remain constant. So perhaps there should be a basis of 2 figures, a
fixed and a percentage...
Though I could happily believe that a 1000hp skyline could lose 150hp
through its drivetrain...


Best Regards,
Alan Pritchard

Network Administrator
Mechanical Design Engineer
Seaeye Marine Ltd.
+44 (0)1329 289000

 -----Original Message-----
From: 	Mike Arman [mailto:armanmik at earthlink.net] 
Sent:	26 August 2004 13:27
To:	quattro at audifans.com
Subject:	drive train losses

 << File: ATT53719.txt >> 

I have to take issue with the idea that you can "lose" 100 hp in the drive 
train.

Where does this energy go? 100 hp is (at 33,000 ft. lbs per hp) over 3 
million foot pounds, at 746 watts per hp, it is 74.6 KW, enough to run your 
whole house, central A/C, big screen TV and all. This energy simply does 
NOT "disappear" in the drive train.

What's going on with these numbers is this -

In the old days, the car companies would run their engines on a test stand, 
and engine X would make 200 hp. That was without an alternator, a power 
steering pump, an A/C compressor, any smog stuff, no air cleaner, no fan, 
no mufflers, essentially anything that takes power to run, so of course, 
they got some pretty impressive figures.

Once the engine was installed in a car, with all accessories, with 
mufflers, and add AIR DRAG at higher speeds plus rolling resistance of the 
tires, plus the windows are down (more drag), plus 300 lbs of "junk" in the 
trunk, plus 400 lbs of passengers, plus 15 gallons of gas at 6 pound per 
gallon, well guess what - all that takes horsepower to move, and the result 
is there's less "extra" horsepower after we use up what is needed to move 
the car and contents.

So as a very rough guesstimate, we might say it takes 40 hp or so to hustle 
this barge down the turnpike . . .

Which means the 200 hp engine has only 160 "left", and the manufacturer can 
say "Well, we figure 20% "overhead" for drive train losses, air drag, low 
tires, a trunk full of bricks, etc." (And that's why your 200 hp car acts 
like a 160 hp slug in the real world . . . )

And *we* seize on this magic "20% in drivetrain losses" figure completely 
out of context . . . because it gives us some really big numbers for free, 
just by working the "formula".

And we try to work this "formula" backwards . . . we put a car on the DYNO 
- no air drag, A/C turned off, weight is not an issue, etc., and we 
discover that we are getting 160 hp at the rear (or front) wheels. Hot 
diggedey! If there's a 20% loss in the drive train, then we *must* have 200 
hp at the flywheel!"

No, we don't. On a dyno, or on the street, we are not going to get 
appreciably more HP at the flywheel than at the drive wheels - it will be 
within a very few percent . . . because energy cannot be created or 
destroyed, just changed from one form to another.

And since the 20% figure is a guesstimate and unverifiable, lets use 25% or 
30% as out "drivetrain losses" figure instead - which "gives" us an even 
bigger number "at the flywheel". (Gee, if we had 100% drive train losses, 
logically, the horsepower going in would be infinite . . . )

What goes in, comes back out again - 3 million foot pounds or 74.6 KW or 
100 hp is NOT going to disappear between the engine and the ground - it has 
to go SOMEWHERE. Some of it is lost as friction in the transmission - and 
that friction translates to heat, which is why we see transmission oil 
coolers. Some of it is noise - gear whine. Some of it is tire flex - which 
is why tires heat up - but NOTHING like 100 hp - not even close. MAYBE one 
or two percent.

Besides, if car designers could pick up 20% in efficiency by attention to 
the drive line, and given government pressure for fuel economy (not to 
mention consumer pressure), don't you think they would be paying very close 
attention indeed to drive train losses? Hey, guys, here's 100 hp going to 
waste! Every automobile company on the face of the earth would gleefully 
sell their grandmother (or yours) for a 20% improvement in fuel efficiency.

What you are measuring on the dyno is the output of the engine without 
needing to take into account the work needed to move the car - that's where 
the horsepower goes, into moving the car - NOT into drivetrain losses.

Look at the specific HP output of engines - 1 hp per cubic inch is pretty 
common, and is even attained by pushrod engines these days. 2 hp per cubic 
inch is turbo territory (remember a 2.3 liter engine is only 138 cubic 
inches, so 280 hp is pretty respectable), 3 hp per cubic inch is race car 
and high revving motorcycle engine territory, so 500 hp out of 138 cubic 
inches is 3.6 hp per cubic inch - which is NOT likely. Engine life also 
goes down drastically at high hp per cubic inch figures - think dragsters.

While it is absolutely possible to increase the horsepower of almost any 
production car engine, doubling it (especially if you are starting with a 
fairly well designed, modern engine anyway) is improbable. Sure, if you 
eliminate smog controls, add boost, port and polish, add more boost, 
advance the timing map, and so forth, yes, you'll get more horsepower - but 
drivability and reliability are the FIRST things to go in this - it is a 
tradeoff. Carried further, the engine gets very peaky (nothing below 5,000 
rpm, 800 hp between 5,000 and 5,500 rpm, big explosion at 5,501 rpm), 
requires 300 octane gasoline, is hard to start, won't idle, and blows up 
every mile or so.

If these gains were readily achievable, every car manufacturer would copy 
them - and with our Audi 5 cylinder engines, they've had between 15 and 20 
years to look at it - they're not stupid, they have a LOT of money to work 
with, and they do know a good thing when they see it.

Best Regards,
Mike Arman