Hi, I am trying to make a tweak to replicate my car..

I would like to convert rear wheel power to engine power.

Does anyone know how to calculate to get it within a 0-10hp accuracy, can provide more information if needed.

Take wheel horsepower. Muiltiply it by a random number between 1.05 and 1.3 depending on the config of your car (FF, FR, MR, RR, 4WD, Turbo, NA etc etc etc). Maybe add between 1 and 5 just for luck.

In other words, there is no direct conversion, you just guess the percentage losses and multiply the wheel figures by that amount.

Quick google search suggests it's +20% with manual trans +30% with auto

Although I'm not sure how accurate that would be.

A racing car transmission is more efficient than a shopping car's. A front wheel drive is more efficient than a front engined rear wheel drive car. A four wheel drive car has more losses than almost anything, and a 4WD auto is worst of all.

So, as I said, pick a number from the ether and use that...

Tristan is right. If you want to get complicated though, you'd be able to figure the wheel-torque out by accelerating in the lowest possible gear (least drag) without wheelspin. Using the vehicles mass and acceleration forces (F=MA), you find the torque at the wheels. Torque at the wheels corresponds to engine torque proportional to the gear ratio, minus losses. Now you've figured the amount of loss that exists in the LFS transmission. Put your car on a dyno, now you have your real cars wheel-torque. Adjust accordingly.

Power is transfered from engine to wheels. There are looses... that depends from car to car, gears, diffferentials, etc. But power looses must be low... maybe 10-20%.

So, the short answer is: same power in engine and wheels.
Long answer is: Know as much mechanics as you can and your car and evaluate power looses...

I have dyno test results and drivetrain information I just don't know the formulas to calculate the mechanical loss.

I used the +20% method and am getting acceleration accuracy within 0.1-0.3 seconds.

And? Sounds like you have succeeded lol.

Just fiddle with it until it is +/- a tenth.

Quote from spanks :

And? Sounds like you have succeeded lol.

Just fiddle with it until it is +/- a tenth.

Yeah but then it's a matter of grip, I would feel more satisfied if I had the power spot on, then I could adjust grip accordingly.

All the dyno tests I did with any of my cars included drivetrain losses on the results sheet. So whoever did your test should have acces to specific figures for your vehicle.

Green line on the graph below:



Uploaded with ImageShack.us


Configuring LFS cars to my road cars' spec always returned 1/4 mile results surprisingly close to those achieved in real life.

The drivetrain losses determined by a chassis dyno are just guesses as well. No more or less accurate than best guessing a percentage between 5% and 30%.

But as it's printed out, people believe them as gospel. Don't.

Ah yes, I've heard of those dynos. I have yet to see one anywhere around here though
Do you know how they work? I could imagine you do the measuring run, then put the car in neutral and let the wheels spin down, in which case you can measure the negative acceleration of the wheels. This would measure drag in a lot of drivetrain components, but isn't completely reliable.

Also, holy shit. Over 33% loss on your Evo?

Exactly that - they measure the drag of the wheels on coastdown, and use that figure as the power loss. But of course it's hopelessly inaccurate and isn't a fair test at all. So ignore it. Ask to have it removed from the charts, and just have power at the wheels shown, so you can do your own analysis. Besides, what you have at the wheels is more important than what you have at the flywheel.

A little off topic but nevertheless, an Evo X with 380bhp but only 255 of those reach the wheels? Unless I've mis-read, that's appaling..

Point proven. They've measured power at the wheels, so one can assume it's about right +-20hp.

Given a 4WD car with far too many diffs and pointless electronics that spoil it, you could expect a 25-30% loss in the drivetrain. If 25% then your flywheel figure is going to be around 340hp.

But if you've ACTUALLY got 380hp, then with 33% loss I'd start looking for a fault somewhere.

The Corvette Z06 loses about 13%, from a bit over 505 hp to a bit over 440 rwhp. Not sure if that is unusually good for a car.

Quote from RasmusL :

Also, holy shit. Over 33% loss on your Evo?

Not my Evo, only giving an example.

Quote from S14 DRIFT :

A little off topic but nevertheless, an Evo X with 380bhp but only 255 of those reach the wheels? Unless I've mis-read, that's appaling..

That's 4WD for you. Those kind of losses are pretty standard. Fortunately one does not drive with a dyno graph

Quote from JeffR :

The Corvette Z06 loses about 13%, from a bit over 505 hp to a bit over 440 rwhp. Not sure if that is unusually good for a car.

I'd say it's a tad optimistic. Unless someone is getting extra freebie horses from Chevy.

Over here US horses have a reputation for being a little weaker than EU ones (same cars have better dyno results in US than in EU)



@Tristancliffe: should I start producing dyno graphs which "guesstimate" stock (as in not tuned) cars' drivetrain losses so precisely that they match manufacturer declared power outputs +-5hp?



Dyno testing is not perfect, but when done properly is much better for estimating drivetrain losses than having a guess in the modest 5-30% range.

I don't believe it is. A dyno is only accurate to within around 5%, and the method used for determining the guessed drivetrain loss is tenuous to say the least. I'll stick with educated guessing, and most likely end up with better (more accurate) figures than the dyno's output.

Not sure what you mean by producing the graphs for standard cars bit. I guess it's you trying to be smart.

Too old to try that. Plus I wouldn't dare being smart in a language that is not my own. I'll guide you step by step through my thought process:

a) take stock car
b) take car manufacurer's declared power from a brochure (it is a flywheel horsepower figure)
c) dyno test it using the same dyno as one used in the example graph above
d) compare the manufacturer figure with the figure obtained during a dyno run
e) based on the difference in the two figures draw conclusions as to the accuracy of using a dyno to indicate car's drivetrain losses

In one sentence:

If a dyno produces flywheel hp figures that almost exactly match manufaturer's figures, then it must be pretty accurate at measuring the drivetrain losses.


Same place, same dyno:


Uploaded with ImageShack.us

Manufacturer data: 305 hp, 440 Nm

Keep typing...

They (the dyno coastdown test) guessed right. Doesn't make my claim that it's still a guess, and the test isn't really valid, any less true.

If you told me to guess the wheel power of a 305hp Focus (Front engined, front wheel drive) I'd have said about 15% losses, which is 259hp, which is within 5% of what your dyno figures stated - i.e. just as accurate (bearing in mind how inaccurate chassis dynos are, and how variable engine outputs are in mass produced cars).