Estimating RWHP from FWPH?

[Mike]

Doc, I think you nailed it. There can't be any argument that whatever the driveline loss is, at a fixed speed that loss is fixed; whether you have a 500 HP motor or a hampster. We try to talk HP on chassis dynos, but as I understand it, they measure torque (on the drum or whatever the drag device may be) and CALCULATE HP, based on time/i.e accel. of the drag device. Obviously, the greater the acceleration, the greater the power required (power loss) to accelerate the device and rotational items like wheels (hence the 'bigger tire/wheel combo = lower dyno HP reading' theory), but I maintain you're not measuring HP, you're measuring the acceleration rate. The reduction from your engine dyno HP is a combination of actual drive train loss (fixed) PLUS the HP required to accelerate that drive train (which will vary with the 'rate' of acceleration). You wouldn't get a real HP reading until the RPM stabilized/0 acceleration, which 'never hoppens'. You quit the pull.

Al Wohlstrom's dissertation above seems to say the same thing.
I would suggest that the "HP" reading off a chassis dyno isn't really a measure of HP but of the 'accelerative' force available in one's engine/drivetrain combo. It provides a good index of comparison in how fast your car will accelerate. It is directly related to HP, but it is NOT steady state HP. This would support the logic that the bigger the motor, the bigger the 'apparent loss' due to it's greater acceleration.
To get an actual HP reading, I believe you would have to run it up to near the Rev. limit, increase the drag until the RPM starts to drop, then read the highest steady state torque/power figure you get.
RE. increased heat losses due to increased friction from higher powered engines, I believe those would be relatively insignificant.
In a nutshell, for bragging rights, I'd lean towards engine dyno figures; for racing discussions, chassis dyno numbers would be more relevant.
You can get an accurate estimate of quarter mile performance from RWHP (it's all acceleration), but I bet you couldn't calculate the top speed of a car (steady state)from it's chassis dyno HP figure. Any Bonneville types out there have any chassis dyno numbers vs. salt flat speed?

That's my 2 cents worth.
Merry Christmas to all.
Mike

 

[Garrett Clark]

Hey, guys,

This is a very interesting discussion. Let me add a new opinion that I belive is actually a combination of the statements made so far. This was supposed to be short, but I see that that didn't happen. I'll try better next time.

Lets start with this statement: Power loss in a transaxle is not affected by the amount of power that the engine produces.

This is true, and it isn't...The power lost in a transaxle is a function (not necessarily linear) of the amount of power that is being put through the trans, among other things. No matter how much power your engine makes, it takes the same amount of power to move your car down the road at a constant speed, and the transaxle will consume the same amount of power at that speed. BUT if you floor the throttle and begin to accelerate, you apply more power to the transaxle, and it will consume a larger amount of power--even before your speed has increased a single bit.

How is that? Is partially because of the way gears work. When torque is transferred from one gear tooth to another, the force is passed between the teeth normal (perpendicular) to the tooth face. By necessity, this is not perfectly tangential to the direction of rotation of the tooth. This means that some of the force is wasted pushing the tooth toward the gear centerline. This is one element of loss in the transaxle. This force is applied to the bearings in the transaxle and increases the friction losses in the bearings. A similar thing produces the thrust loads in helical gears, and this torque loss is why racers use noisy straight-cut gears.

So this means that the more power you put into a transaxle, the more power the transaxle will consume. (Again, I'm NOT saying it is linear, which would produce a constant percentage loss.) Thinking more about a dyno pull, the engine will be operating at the maximum power it can produce, and a higher power car will lose more power into the same transaxle at the same speed than a lower power car.

While I'm typing, I'd like to review the relationship between torque and horsepower. Horsepower is simply torque multiplied by RPM (with a conversion factor). Torque is the rotational analoge to force, and we all remember F=ma. If we are trying to get the most a, we need the most F. Simple. Why would anyone care about power? Well, there is a reason...

Power is basically the rate of production of torque. A human with a cheater bar can easily produce, say 100 lb-ft or torque. This is embarrasingly similar to the flywheel torque output of my Integra Type R. So why can't I just twist my car by hand from 0 to 60 by applying a torque wrench to a lug nut? Becaucuse I can't produce that 100lb-ft of torque at a fast enough rate to accelerate the car after it gets to, say 0.1 mph.

Cars take advantage of the high rate of torque production from an IC engine (aka horsepower) by using a torque multiplying gearbox to increase the torque available at the wheels (basically a longer cheater bar). This requires the engine to spin at a fast rate. My type R can spin at these rates. My arm cannot. So to really accelerate a car, we need a torque delivered at a high rate of speed, a combination of torque and rpm...aka horsepower.

Garrett

 

[Gary Gibbs]

Good luck getting buy in from others /ubbthreads/images/graemlins/grin.gif

 

[Pete McCluskey.]

This might seem an overly simplistic question. By why not just use a dyno that measures RWHP?

 

[Garrett Clark]

Pete,

The problem is that it is not possible to measure horsepower directly becasue "power" is not a fundamental unit--time, distance, temperature, and mass. Power is simply a useful conglomeration of these fundamental units, and is defined as 550 ft-lbf/sec. (Force (in lbf, pound-force) is not really a fundamental unit either, but we won't get in to that now.)

One horsepower is the amount of power required to lift a weight of 550 pounds 1 foot in one second. So we can't measure horsepower, but we can hook an engine up to a cable and lift a known weight while we measure time and distance. We then calculate power.

There are many other ways of calculating power. One is to remember that Energy = power * time. This is the relationship that is in play when we are calculating power from the car's ability to accelerate a rolling drum, or from a car's quarter mile performance.

Garrett

 

[Howard Jones]

Garret, Isn't that what I said?

 

[Garrett Clark]

Howard,

Here is what you said:

"I guess I have always looked at dyno sheets that say somthing like 400Hp@6000RPM and believed that the motor makes 400hp when the engine is running at 6000RPM. BUT really thr sheet is saying that it took 400HP to get the engine to 6000RPM in a given time @ a contant load."

Assuming we are talking about a Dynojet or similar dyno, I think you pretty much have it. The only thing I would add is that the dyno actually looks a the change in energy of the spinning drum over a series of short periods of time, rather than across the entire pull, and the load need not be constant.

Now, if we are talking about a different type of dyno, say an engine dyno that operates at steady state, all of this is moot. Heavy wheels (or a heavy flywheel) wouel have no influence on "power" for this test.

Garrett

 

[Stew Stau]

Just my two bits worth. On big bikes (GSXR 1000 etc), the losses through the gearbox are almost always from 12-15hp. No percentage anything. If the bike produces 165hp at the crank, it's probably 150-153 at the wheel. If it's 180 at the crank, it's 165-168 at the wheel.

So with a transaxle behind a V8, I'd expect a similar thing. Quotes of 40hp or so sound pretty logical. Mind you, I've seen two supposedly identical chassis dynos produce up to 20% different figures on the same car the same day. The old drag strip comparison, despite my misgivings of lowering a finely tuned piece of endurance racer to such an agricultural test, is probably the bottom line. If your XXXXlb car headbuts the horizon in XX.XX seconds, then that's how fast it is. If it is faster than the guy with 600hp, then either he's not got 600 or you've defeated physics.

Stew

 

[Garrett Clark]

Stew,

The more power you put through a transaxle, the more power it will consume. I'll try only once more to explain it...

Let's take a hypothetical transaxle that consumes 10 hp when connected to a 100 hp engine. The peak power reading on the chassis dyno is then 90 hp. Lets take that SAME EXACT transaxle and hook it up to a 90 hp engine. Dyno reading is 80, right? Lets follow that line of logic all the way down to a 11 hp engine. Chassis dyno reading is 1 hp. Hmmm. Still with me?

Next, remove the transaxle and turn the input shaft by hand. It would be impossible to turn unless you can generate at least 10 hp, right? Do you still like the "constant power" theory? How about this: the "constant power" theory says that the same transaxle will continue to consume 10 hp while it sits silent and motionless on my garage floor. If you are thinking that this is getting absurd, then you and I are in full agreement.

I'll bet everyone would agree that the transaxle will consume very LITTLE power when turned by hand--when there is very LITTLE power going through it. Then why is it so difficult for the armchair engineer to believe that the same transaxle will consume MORE power when MORE power is being put through it? I honestly don't know.

The more power you put into a transaxle, the more you lose. It probably isn't a constant percentage, but a percentage is a much better approximation than a constant amount of power.

Now then, back to reality. You are absolutely correct that a faster car is faster, regardless of what the dyno sheet says. A faster car is also faster, no matter what the engieering theory says!

Garrett