S4/RS4 Drivetrain Losses

[Nige_RS4]

Andiroos post on the AP22 thread set me off this morning on the search for an accurate drivetrain loss figure for our cars. I believe his estimation of 28% drivetrain losses might be too high for the quattro system, as this is considered one of the most (if not the most) efficient AWD systems available. Does anyone have any reliable information for our cars ?

BTW, rolling roads measure wheel HP, then it is left up to the owner of the rolling road to 'convert' to flywheel HP using their own figures. Basically, they can fudge the figures if they so wished, hence why it is important to take your car to a reputable RR. AMD are know to be good, but I'm not sure if they adjust the figures for the different types of AWD systems out there ?? [img]images/graemlins/rant.gif[/img]

[DavidT]

Some rolling roads can calculate transmission losses.

I went to Well Lane Turbo Centre, Batley (when's the next one AL B?) and they freewheeled the car down from max revs in the test gear to calculate the losses, then did the power run to give the flywheel bhp figure.

Anyone else had this and is it accurate ?

[wazza]

G-Force uses 25% for the 996TT 4wd system. They have bench tested the flywheel output of an engine, then transplanted that straight into a car and onto their dyno.
They did not test our cars on the rollers for setup (I think), so they just used the 25%. [img]images/graemlins/confused.gif[/img]

[tanoga]

Audi quote the RS4 as having 380bhp which is "at the flywheel". We can only assume that this is a bench tested figure and is reliable. If you then put a box standard RS4 on a rolling road that is reliable you can then work out the transmission losses from the "at the wheel figures" having regard for Audis quoted figure "at the flywheel".

[Andiroo]

G-Force are the most knowledgable guys in the business and they work on 28% for the Quattro as of yesterday, which is comparable to the Skylines. Like Karl says, the best way is to see if a bog standard RS4 has been ATW measured on G-Force and this will be the definitive answer [img]images/graemlins/wink.gif[/img]

Andiroo

[wazza]

So, you were down there yesterday and didn't say? [img]images/graemlins/suicide.gif[/img] [img]images/graemlins/grin.gif[/img]

[Andiroo]

Sorry mate, they were busy yesterday so asked me to limit number of people attending. Will be there again so no worries [img]images/graemlins/wink.gif[/img]

Andiroo

[Andiroo]

Hey Nige, do you still feel that 28% loss is too high? Have you managed to find any data to contradict this yet?

Andiroo

[Dippy]

I went to Well Lane Turbo Centre, Batley (when's the next one AL B?) and they freewheeled the car down from max revs in the test gear to calculate the losses, then did the power run to give the flywheel bhp figure.

Anyone else had this and is it accurate ?

As far as I know that method can be a very accurate method, however I would expect that most tuners use a 'book' value instead.

From my thoughts and limited research, 28% for the Quattro system is about right.

[DavidT]

At G-force my car was 207 bhp at the wheels, which they calculated at 280 flywheel bhp. I dunno how they did it though [img]images/graemlins/confused.gif[/img]

At Well Lane the graph showed 278 flywheel bhp.

Pretty consistent.

[jeffw]

AmD had mine down as 275 at the wheels with flywheel as 395BHP....something like 30% loss.

[Nige_RS4]

Hi Andy, I found quite a few references to the audi quattro drivetrain, but NO hard FACTS on definitive drivetrain losses. One thing I did find a number of times was references to the validity of the aforementioned freewheel method of calculating drivetrain losses - here is one of the extracts ;

Variations in horsepower loss vary due to different tires, different pressures, different vehicle weights, and even different methods of strapping down the car - that's why I have heard everything from 10% to 35% for estimated power losses "in the drivetrain" on a chassis dyno. I put the drivetrain in parenthesis, because this loss isn't all coming from the drivetrain, some of it (again, it's variable) is being contributed by the dyno and the tire/roller interface. Do you think an auto manufacturer would really accept a 35% power loss from their drivetrain engineers? Some dynos say they can measure your drivetrain loss by doing a coastdown test. They tell you to get the vehicle up to a certain speed. pop the car in neutral, and let it coast down - they measure the time to decelerate, and this is supposed to be the drivetrain loss. Think about this: Under acceleration, the gears are smashed hard against each other, the bearings are being side loaded, the pinion is trying to unscrew itself out of the rear-end, etc. During the "coast down test" everything is unloaded and freewheeling as it decelerates and he gears are being driven on the opposite sides of the teeth - the car is now being driven by the wheels and there is no load at the input shaft of the transmission. The frictional loads and losses are not the same as during hard acceleration. If the loads aren't the same, then what good is the test. Sure, you've got data, but it's worthless.

Another difference in power numbers, are due to different rates of acceleration. If you use different methods of testing that give you different acceleration rates, you WILL get different numbers. The Dynojet has no real provision for changing the rate of acceleration. It is a fixed weight, you move that weight faster, it equals more power - simple. If you have a load type (non-inertia) dyno, you can vary the rate of change of the engine. Different loads and rates of acceleration will produce different numbers, and different vehicles will respond differently as to just how much difference you will see when changing the rate of acceleration. The bottom line is: If you want consistency in your data, choose one dyno and testing method and stick with it. Forget about comparing numbers between different dynos and vehicles with any rule-of-thumb percentages. Don't believe that there is a hard-fast percentage you can use to determine crank horsepower - it'll ALWAYS be a guess unless you actually measure it there.

So it would appear that unless you dyno an engine and then RR it, it is never going to be that accurate (hence why some cars seem to produce substantially more HP/torque than others with the same mods - anyway I've put together a list of RR do's & don'ts - please feel free to add to these (or shoot me down in flames if you think I'm talking b*llocks);

1. Try and use a reputable tuner - AMD, G-Force, Well Lane (+ others)
2. Always use the same tuner, as HP figures are never going to be 100% accurate, but increases in HP following upgrades are more relevant.
3. Check with the tuner what figures they use for drivetrain losses and where they get the figures from (4WD systems from different mfrs are not the same) - make sure they use this figure on future runs on your car.
4. Remember, peak HP figures mean nothing, the way the power & torque is delivered across the rev range is much more important (except when bragging in the pub [img]images/graemlins/nodder.gif[/img])

Nige [img]images/graemlins/thumbs.gif[/img]

[Andiroo]

Couldn't agree more Nige. My peak bhp is being investigated as we speak coz it's down on what everyone thinks it should be, but the torque band from 3000rpm to 6000rpm is awesome, showing 700nM peak at 4750rpm on G-Force's RR. Linked to higher boost I find it hard to complain about the power delivery, even though the bhp seems to be down [img]images/graemlins/laugh.gif[/img]

Andiroo

[fade2grey]

RR plots are only barely comparable, there are loads of harsh reading\high reading roads in the country. Even the way they tie down the car & the temps on the day have a big impact on the figures. Your tire pressures are one way you can alter the figure yourself! Getting a base line from RR#1 & then comparing to the graph from RR#2 doesn't nesc follow.

That said, they are great if you go to the same place pre\post mods to compare or a gang of you go to the same place on the same day. [img]images/graemlins/smile.gif[/img]

A

[Dippy]

Remember, peak HP figures mean nothing, the way the power & torque is delivered across the rev range is much more important (except when bragging in the pub [img]images/graemlins/nodder.gif[/img])

Nige [img]images/graemlins/thumbs.gif[/img]

Great post Nige. [img]images/graemlins/thumbs.gif[/img]

But I suppose that 'bragging in the pub' is what MOST people want the bhp figure for isn't it?