On a manual car, that's pretty easy. ON the dyno, you throw it into nuetral after the pull anyway, which basically isolates the pistons and any pumping losses from the engine itself. Since the engine is only getting net power measured anyway, it probably would be a good way to do it... for a manual transmission. The dyno operator will then brake the dyno. Not the vehicle driver.

Good point on the gear teeth. it might be negligible, it might not, but you're right - the gears are behaving differently under decel than accel. Nice catch!

 

I'll use an imaginary dyno jet to try to explain this. Anything motor side is irrelevant on the deceleration side.

Angular (rotational) Acceleration is basically measured by a 3000 lb roller and how well the car (engine plus drivetrain) can accelerate that roller. So the net output AFTER the drivetrain losses is what is left to accelerate the dyno roller.

Then, by throwing the car into neutral, the dyno rollers start to slow down slowly. Assuming a perfect, frictionless roller (which I know it isn't but humor me), the only thing slowing it down is the drivetrain friction. In this sense, you are able to get a pretty accurate representation of the amount of total friction in the drivetrain.


I'm really glad you posted that. it takes more torque to accelerate the drivetrain faster too - not just accelerating the dyno roller faster. I had overlooked that myself. Though I wonder - how much is that impact vs overcoming the MOI of the dyno roller? Probably a very small fraction?

As to doubling the power. If you double the torque, at best you'll see a 41.4% improvement in performance, since F=ma AND d=1/2at^2. You can't cut the time in half by doubling the acceleration rate. You can only cut it by 1/sqrt(2). That would be in a perfect system, too. And since power doesn't translate linearly to acceleration, you'll get even less from doubling power.

 

if you put the car in neutral, you will not know how much power you need to acceletare the engine, right? on the other hand the argument concerning engine braking seems valid. i can only repeat, that if you measure it the way im saying, you come up with the manufacturers advertised numbers. i also remember many years ago, they would just brake the engine at a certain rpm. what came out was the right numbers as well.

anyway, i think the friction loss of the drivetrain and the loss due to acceleration of drivetrain and engine (incl. clutch, wheels and brakes) have to be treated separately.

 

Joe,

Pardon me but what I saw in original discussion, and still see is a "Total Lack of use of Common Sense" of trying to validate how on a Chassis Dyno you can use its data (RWHP HP/TQ) and MFG's (Engine Dyno's HP/TQ) and say the lost of xx.x HP & TQ between the two sets of numbers is due to "Parasitic Drivetrain Lost.

All use Peak HP to calculate Drivetrain Loss number. So if peak HP is increased by either Air in or Air out, that moves RWHP number closer to Engine Dyno Rated HP number, does the amount of HP lost due to drivetrain parasitic lost go down, because parasitic lost decreases? If so is drivetrain lost a constant fixed lost or a moving target lost based on increasing or decreasing RWHP?

Yes I kept going back to "A GM rated 480 HP Crate Engine and also for a GM rated 505 HP Crate engine to a much lesser extent. What I was trying to point out with my intake or exhaust examples changes to engines that Hurt HP, to your as stated above on how engine (is rated by MFG) is goin to play a part in how trying to use Chassis Dyno numbers, Vs Engine at Crank numbers.. Math is real, but can it come up with a "fixed HP Lost number" In "Real World, is it Realistic or not ".

My bad on using my experience in world as examples. What I should have used is another vehicle, one that came with a 505 HP Engine Dyno rated engine in it, a C6 Z06 Corvette instead of two of my cars, one with GM rated 480 HP, the other with a 505 HP one, and the differences that changes to just Intake Air in volume & speed to engine an Exhaust air out volume and speed could and does change Chassis Dyno HP & TQ number(s) you say are use to come up with using Laws of Physics. "Drivetrain parasitic Lost(s), stands on it's on, "Right".

MFG Tolerances of Drivetrain's components could be different car from car, causing slight differences in HP/TQ numbers of same optioned models rolling of assembly line . Either way "Drivetrain Lost" is for the most part fixed a fixed number car to car.

So can you use the Numbers between Engine's Rating HP/TQ on a Engine Dyno and Chassis Dynos to say the difference between the two sets of numbers is a xx.x "Fixed" number car to car due to Parasitic of Drivetrain's components o each? YES or NO.

"In the real world, for most of us, HP/TQ lost to Drivetrain is a XX.X fixed number. Changing number would require a component an/or components to be modified or replaced with one(s) that create less Parasitic Drag." Can we somehow change Parasitic Lost of Drivetrain components to HP/TQ without making any changes to them? Yes or No.

I think you can, trying to use Chassis Dyno numbers compared to MFG Engine Dyno rated number.

Now using your knowledge of Math, Physic, Science, Could you show me the "Math" you use to calculate "Driveline Lost" using the below HP/TQ numbers in each of the 3 examples below? I'm interested in will 'Drivetrain Lost" change or stay a Fixed Number.

Note: By the way, I'm in total agreement about what you say in your statement you made about "EXACT HP", they want match up engine to engine in same make an model car, optioned out the same. I just used "Intake Air in" an "Exhaust Air out' as just 2 reasons why trying to say you can calculate Drive Train loss based on Chassis too Engine Dyno Numbers cannot be accurately done.

I know personally that a C6 ZO6 Corvette that comes standard with a LS7 505 HP on same dyno, same day, slightly different time though, WANT make the same HP/TQ numbers. So if you use as Rated numbers Vs as installed numbers, using Parasitic lost as reason even if both cars are Tuned to get max HP/TQ as installed want have the same HP lost to some extent. Maybe one's drive train has more parasitic lost than the other one, so XX.X of each's numbers seems to be attributed to Drivetrain Lost.

With no change to internal engine components I'm going to change Engine's RWHP number(s) with two basic engine upgrades to the afore mention engine. One that changes amount of Intake Air in, an One that changes Exhaust air out.

Question is will either or both change "Drivetrain Lost" numbers using your math as each change is made? I Don't think Drive Train Lost number(s) will remain fixed. Your math, plugging in RWHP numbers that have changed, Vs Engine Dyno numbers, should show me and others if I'm right or wrong in my assumption.

I'm not versed in book Physics. More of what you call "Real World Experience". Maybe you can educate me with your skills. I have two real beliefs, "Stupidity is Terminal", "Ignorance can be cured". I want to be in the "Cured Column".


1st: You have the 505HP rated LS7 engine tuned to make max power as installed so you know as delivered RWHP/TQ Numbers. That way you know what your power gains are for two most common bolt of upgrades.

Engine Dyno's 430 RWHP.

Question to you: What amount of lost between LS7 Engine Dyno numbers of 505 and Chassis Dyno numbers of 430 RWHP can be attributed to Driveline Lost. Show me the math to come up with the number of HP lost contributed to Driveline.

Question is will either or both examples below change "Drivetrain Lost" numbers using your math as each change is made?

I Don't think Drive Train Lost number(s) will remain fixed. Your math, plugging in RWHP numbers that have changed, Vs Engine Dyno number. should show me and others if I'm right or wrong in my assumption.

I'm not versed in book Physics. More of what you call "Real World Experience". Maybe you can educate me with your skills. I have two real beliefs, "Stupidity is Terminal", "Ignorance can be cured". I want to be in the "Cured Column".

Bolt on option Number 1: What do you do if you want a few extra ponies for the least amount of money. The shop doing the Tuning will be glad to help. They will recommend the Air Intake/Breather/Filter they think works best.
So they install it, an re-tuned depending on how much if any the Tune has changed due to more air coming in with intake change.

Engine now makes 442 RWHP. RWHP/TQ numbers have now moved closer to Engine Dyno numbers. I've only change the amount of intake air that the engine can pull in through the new Air filter/intake component swap out. Drivetrain lost could in no way have changed.

Plug in the new numbers, and show me the math, that "Drivetrain Lost" will remains same as 430 RWHP number.


Bolt on option N0. 2 Exhaust manifold swap out for Headers. Now you have the fever, an you want more HP. Shop again says we have just what you need, Headers or a complete Exhaust System swap out. That change should allow the engine to breath out better, so more air can flow in. You know more air in & more air out is what making HP is all about, especially if intake air an/or exhaust flow, is not increased to point you decrease intake or exhaust velocity to much. Again, have I changed the number of Horses lost that was previously claimed was due to Drivetrain? Yes or No.

Exhaust changed an engine retuned.

Engine now make 460 RWHP. Show me the math of what Driveline lost will be now.


I've now changed RWHP upwards twice. I've not changed Engine's Crank HP. So the question become is Drivetrain lost a fixed number that can be calculated on a Chassis Dyno. Your math should show me an any other it's truth.

Note: HP is a calculation derived from Torque.

 

engine output is affected by too many variables, for example oil viscosity, exact fuel octane, and humidity. That is one reason why "going back to the dyno for 10 horsepower" is not a good idea, as you may gain or lose some ~3.5% just based on random situation.

Therefore, you cannot go directly from chassis -> engine -> Manufacturer expectations, at least not directly. There needs to be some variance, some standard deviation, some normal distribution involved. If you are close enough, well that is close enough.

example A
Day 1, car put down 300

Day 2, I change the air filter and car put down 310.

I come back saying "look the air filter is worth 10!"

But the car would have made 310 or 308 on Day2 even with the old air filter (or insert mod) because atmosperic pressure, air temp, and humidity are more favorable that day (example). Even if you ran back to back the increase in oil temp or air temp will affect the next run, even the trans fluid can have an affect as it goes up in temp. No way to tell when you split the small % differences. If you made repeated attempts across multiple days using both mods (on and off, then off and on) you could establish a graph/plot and through that determine within a confidence interval how much "HP" the mod was worth. But who does that?>!