Eric Bryant

The only reason I brought it up is because it becomes rather significant when trying to determine drivetrain losses on a car equipped with an automatic transmission - I wouldn't consider it to be important at all on a vehicle with a manual.

I'm definitely not trying to get into a "esoteric pissing contest", especially not with you I just didn't want someone to get the wrong impression (especially in an advanced tech forum) that the standard "x %" drivetrain loss factors can be applied to a vehicle with an auto with the expectation of good correlation.

brain

The main reason for differences between a dynapack and a dynojet are due more to the way they come up with their hp number. A dynapack is an actual loak-brake dyno, while the dynojet is an inertia dyno. The dynapack will measure torque, and calculate hp. The dynojet will calculate tq, based on its calculated hp, from acceleration of a known mass (the drum) to a measured speed over a certain amount of time. The dynojet is more popular, but the dynapack is more closely related to an engine dyno. Another interesting note; dynapack claims that the straps tightness on an inertia dyno has been loosened slightly and seen gains of 10 hp. Makes sense, as the tighter the straps, the more the tire is loaded.

OldSStroker

Good points all.

Anyone ever tried back-to-back chassis runs with 10-15 psi difference in rear tire pressure? It can make a measurable difference. Tire slip on the rolls also makes a difference.

Eric, Injuneer and MadMax are all making valid points.
(I really don't see an EPC going on here, BTW.)

Superflow once ran a test during an AETC where they engine dyno'd a BBC one evening, then bolted it back into the drag car it lived in and ran it on their AutoDyn chassis dyno (in inertial mode, at the owner's request). They then plugged in numbers they had developed for a drag Powerglide, 9 inch rear and drag slicks to estimate the flywheel hp/torque. While the difference was something like 15-20%, with their known measured efficiencies of the drivetrain components, they were within a percent or 2 of the flywheel hp. They stressed that you have to know the driveline efficiencies.

On a (rented!) Viper, running in 5th gear (that's over 200 mph @ power peak), they corrected pretty close to the advertised hp. That was easier with a manual and good radial tires and only a few rpm/second near power peak. Remember, OEM flywheel hp/torque numbers are from step tests.

My $.02

brain

I guess I am still a bit lost here. I agree with Lt4 fd. How can frictional losses eat up so much power? Lets say I have an 1200 RWHP drag car. Does it really east up 200 HP? One type of loss is inertia, which would be equal regardless of power, so lets say its about 45 hp. By adding so much hp, friction is eating up an extra 155 HP? That just sounds a bit excessive. Fred, even your car lost less power on the gas with the manual, I know that nitrous isn't exact everytime, but shouldn't it have at least matched the losses, and not been less?

Mindgame

I once read somewhere that the reciprocating assembly accounts for about 35-45% of the total frictional loss in an engine. Of that, about 60% of the loss can be attributed to the rings and piston skirt.
So, it's not too much of a stretch to imagine a 12-15% loss just to the rings a piston skirt. Not to mention all the other losses we've included when we look at the entire drivetrain.

-Mindgame

Highlander

Engine loses will be seen on a engine dyno because they are part of the engine...

Mindgame

uh huh
They'll also be seen in the output at the rear wheels because the losses are accumulative... engine, drivetrain, pinion, etc..

-Mindgame

Highlander

They cant be seen or measured via dynos if both are present on both dynos...

When doing the difference it will cancel each other.... Engine are always present in both... therefore cannot be seen when changing from chasis to engine dyno.

Mindgame

Wasn't talking about going from one dyno (engine) to the chassis dyno. I was talking about just the chassis dyno... that's what I thought the topic was about?

-Mindgame

Injuneer

You really have me lost with that statement. You appear to be saying I lost less HP (manual trans) because the percentage went down. That's not the way it works.

N/A: 486 - 425 = 61HP (61 / 486 = 12.5%)
1-st N2O: 633-555 = 78HP (78 / 633 = 12.3%)
2-st N2O: 762-670 = 92HP (92 / 762 = 12.1%)

As flywheel HP increased, the actual HP losses increased, but they decreased as a percentage of flywheel HP. If you attempt to simultaneously solve the equations for loss:

Fixed loss + (HP x variable loss) = total loss

you come up with the result that the fixed (inertia) loss is in the range of 8 - 9 HP, and the variable (friction) loss is in the range of 10.8 - 10.9%.

When you consider that the published efficiency of a 12-bolt rear axle is 93%, you have accounted for 7% (53HP in the max HP example) of that 10.9% loss. The friction losses would have to come from the rest of the drivetrain. And I agree, that nearly 4% loss (30HP in the max HP example) seems high for the trans being in direct drive. So, all the trans bearings and the idler gears, and the u-joints, I guess some tiny amount for the T/O bearing add up to this.

In the 762fwHP case:

Inertia - 9HP
Rear axle - 53HP
Trans, etc. - 30HP

Total losses - 92HP

(Keep in mind, this is just one set of data from one specific vehicle, on one set of test equipment...... these are just examples, and rough numbers. It would be interesting to see similar data from another car/test equipment)

The original question was about drivetrain losses. The only way I know to measure this with any degree of accuracy is to measure the output at the flywheel (engine dyno), then at the rear wheels (chassis dyno), attempting to keep the test methods relatively consistant.

You have to remember that the when the engine is installed in the vehicle, other things are changed. My engine dyno tests were run with most of the intake track installed, the headers (but from the collectors back, it was attached to the shop dual exhaust muffler/stack setup - in the car it flows through a 3" y-pipe with Borla XR1 mufflers in place of the cats, and a 4" Mufflex), no serpentine belt (alt, ps, idlers still on car - no A/C in the car), the coolant came from the engine dyno system, not the radiator (different pumping losses), etc..... so it is reasonable to assume that a percent or two of the flywheel vs. rear wheel results are from other than drivetrain losses.

Any discussion of "internal" engine losses - pistons, rings, etc - would appear to not be relevant to the original question. We are only looking at the losses from the flywheel to the rear wheels.

Fred

OldSStroker

"Indicated horsepower" is what the engine is producing from burning the air/fuel mixture. This isn't directly measurable, nor is it useful in propelling the vehicle.

Engine dynamometers measure BRAKE torque (and calculated horsepower) at the engine flywheel. Internal friction losses, pumping losses and losses to heat are already factored out from the indicated hp; they don't get to the flywheel. We are assuming that the engine is dyno tested with exhaust, inlet plumbing, accessories, etc. just as it is installed in the vehicle.

FWIW, friction and pumping losses can be measured (and are by OEM's) by motoring the engine, or spinning it with a big electric motor and measuring the torque it takes to spin it. I doubt that many engine builders have this capability, but I wouldn't put it past the highest end race teams. This is the easiest way to see friction reduction from coatings, clearances, oil grade and type, even bore/stroke relationships on friction (longer stroke = more piston speed).
The chassis dyno, or "rolling road" as the Brits say, measures what gets to the dyno rolls, or the road. The difference (power at the flywheel - power at the road) equals the drivetrain losses. Drivetrain includes transmission, u-joint, bearing, ring/pinion, tire deformaton, etc. losses.

Inertia plays a significant part in measuring power. OEM hp and torque numbers are from step tests, where the engine is not accelerating. Most aftermarket engine builders measure with the engine accelerating from 200 rpm/second to maybe 1000 rpm/second. The higher the rpm/second, the lower the output because power is lost spinning up the components. Here is where lightweight (low inertia) components pay dividends.

To accurately assess drivetrain losses, the engine acceleration on the chassis dyno should duplicate what was run on the engine dyno. About the only chassis dyno I'm familiar with that can do this easily is the Superflow Autodyn. It uses an inertia wheel and eddy current brakes and can be programmed to any acceleration (assuming the engine is strong enough to accelerate the dyno roll that quickly). At least run your car in a gear which closely duplicates the acceleration of the engine dyno run. If comparing to factory numbers, run in the highest gear the dyno will handle.

More of my $.02