ALFAGTV6.COM

Agreed on the issue of how much loss to use - 18% is what the tech suggested. The only data I'll consider 'real' is when I install the new engine - then I'll simply be able to compare RWHP figures to arrive at a % improvement.

Cheers,

Daniel wrote:Ok, stupid question time - Are the losses really a constant percentage.
If you had 150hp atw and calculated (@18%) 177hp at the engine, that's 27hp in driveline and system losses.

One part is, but other isn't. Same car with only 70 hp engine may show something like 25% overall losses..

Jim, for purposes of the book I would just use rear wheel horsepower to avoid arguments.

Most 80's cars seem to have about an 18% loss when comparing rear wheel horsepower to advertized horsepower using a Dynojet. This includes most BMWs, Alfas and the Porsche 928. Now I don't know if they really loose that much, or if the factorys exagerated the flywheel ratings.

The Dynojet brand of dynos are very very consistant because they just measure the spin up time of a heavily weighted roller. The types that use hydraulics to provide resistance are notoriously inconsistant.

Newer cars may show only an 8% loss because they have more efficient drivelines, or more realistic factory numbers, I don't know which.

Daniel, the losses are very close to a constant percentage, not a straight number. This is largely because of the increased friction in the transmission and ring and pinion gear. With more power on the gears they have more friction and develop more heat thus drawing more power.

Greg Gordon wrote: The Dynojet brand of dynos are very very consistant because they just measure the spin up time of a heavily weighted roller. The types that use hydraulics to provide resistance are notoriously inconsistant.

Not all, Rototest is very accurate and have excellent repeatability.

Greg Gordon wrote:Daniel, the losses are very close to a constant percentage, not a straight number. This is largely because of the increased friction in the transmission and ring and pinion gear. With more power on the gears they have more friction and develop more heat thus drawing more power.

Sorry for arguing, but in that case there wouldnt be much free rolling resistances. Some dynos use to determine free rolling curve after the power shot and add that on wheel power to estimate the engine power in some level. Free rolling losses aren't prosentage of engine power.. Biggest losses comes from tyres in both, the prosentage and not prosentage part, and so on... Sometimes the prosentage just happens to be the only available value to use, right or less right..

Just try to take power curves out with all different gears. Do you get same numbers?

What Kimmo just issued in his last post is something whaich wasn't covered at all so far.
This is relevant for dyno's accelerating masses. when the acceleration is fast, the effect of engine and driveline inertia start playing a role.

the best dyno's (like rototest) use slow acceleration or steady state in which case the inertia is irrelevant.

if you measure the power in different gears on a dynojet type you can actually calculate the inertia pretty good.

Mats, I was not saying that ALL brands of dynos that use hydralics are inconsistant. I have never even heard of the Rototest type.

Kterkkila, I am not sure which of my points you are arguing. Are you saying that dyno losses are not a percentage or that the transaxle doesn't cost horsepower.

Well, the way you wrote it it looked like you meant that.

www.rototest.com

I agree with you Greg, there will be more losses through the transaxle through friction but the inertial losses in the driveline are virtually the same (unless you've beefed it up for the turbo conversion ).
So if, in the example I gave, you are going to lose another 27hp (20kW) through the transaxle in frictional losses, you'd better get a good trans cooler going or it'll last 2 minutes.
I know of a few cars with 500+ Hp atw and they still claim 20% losses for 600+ Hp at engine. Surely these can't be blowing out that much heat from the gbox and diff.
I think the percentage has to reduce but to answer Jim's question - use RWHP, at least it will be more consistent.

What a great thread.

I am still digesting Michael's dyno graphs though. I can't get over the quantity of torque the 3.0 puts out at low revs. That's just very respectable. Puts my daily-driver WRX to shame.

Daniel, you are correct about the losses from the driveshaft being the same. The losses from the transaxle and to a lesser extent the axles (unless they are perfectly straight) will increase with an increase in power.

The losses in the transaxle are somewhat misunderstood. It's not just spinning the thing that costs power it's the friction in the gears. This is especially true with helical cut gears that most cars use. Not only does the friction between the gears increase with an increase in power, they also cause a fore/aft bearing load that increases with power. The same is true in the differential.

This does create a lot of heat. A lot of cars that run sustained high levels of power do use gearbox and differential oil coolers. Alpina used a belt driven pump and oil cooler for the old E28 BMWs for this reason. A 600 hp car will generate huge amounts of heat in the gearbox, however it's probably not possible for it to put out 600 horsepower for 2 mins straight so I doubt it's a factor. Think about it, you would have to be at max power rpm at full throttle for 2 mins.

Just how much is the power loss it? I don't know. As I said, it may be lower than 18%, but that's the number that consistantly works with rear drive European cars of that era. In the book "How to tune and modify engine management systems" they dyno test a stock BMW 325ic. It has a claimed 189 flywheel horsepower and dynos at 151.4 at the wheels for a 21% loss. This is consistant with other results with similar cars. Now if the loss IS lower then we have to conclude the factory horsepower claims are higher than actual power.

I prefer to stick with rear wheel horsepower to eliminate this unknown variable.

I think the administration should move these posts to a new topic titles driveline losses. We are sort of hijacking Michael's post.

Side note: I had a temp sender in my gear box a couple of years back and with my standard 2.0 carb engine the gearbox temp quickly rose to 130*C within a couple of laps. I'm sure the GB is capable of moving HUGE amounts of heat to the ambient air but I'm convinced that there is a "constant" and a percentage part of the loss, not just a percentage.

I think on the old forum Ben Bishop said that it is not just a percentage, but for practical purposes, it is close enough if we just treat it as a percentage.

Alright, I get the point! I knew this would happen when I asked the question. Now we got a problem! Just talked to a very knowledgeable dyno guy here and he brought in a whole bunch of variables. Some time ago, they did a test with a Honda CRX. They dynoed the pi$$ out of the poor thing, changing tire sizes, tire pressures, gear oil types and what they found was that while engine power shown (derived by coastdown method) remained consistent, wheelpower varied all over the place, +-5% or more!
So where does all this leave us? I thought wheelpower would be a simple bare and true measurement but I was disappointed because it can be so easily doctored! I'm all ears guys!
Jim K.

That's exactly what I said. Wheel HP is not very scientific