[[email protected]]

Re: ([email protected])

Quote, originally posted by [email protected] »

The MAHA creates its own correction factor based upon the individual car. IMO that would be the most accurate. I don't know enough about it though to really understand how it works. Apparently though - it has been accurate to 1-2 hp on determine HP on stock cars.

Wheel horsepower is the amount of power being transmitted from the tires to the rollers, or to the road. This power is significantly lower than crank horsepower (at the output shaft of the engine) due to several factors including:
clutch slippage
gears intermeshing at the input shaft to the transmission
gears intermeshing between gears in the transmission
drag induced by transmission fluid
gears intermeshing at the output shaft to the transmission
gears intermeshing inside the differential
gears intermeshing between the differential and half-shaft
gears intermeshing at the CV joints
Wheel bearing drag
tire deformation
tire slippage
Many of these factors behave very differently during a coast-down and some aren't applicable at all. There is no way to accurately measure crank horsepower with a chassis dyno. Period.

 

[rastta]

Re: ([email protected])

Quote, originally posted by [email protected] »

Wheel horsepower is the amount of power being transmitted from the tires to the rollers, or to the road. This power is significantly lower than crank horsepower (at the output shaft of the engine) due to several factors including: clutch slippage gears intermeshing at the input shaft to the transmission gears intermeshing between gears in the transmission drag induced by transmission fluid gears intermeshing at the output shaft to the transmission gears intermeshing inside the differential gears intermeshing between the differential and half-shaft gears intermeshing at the CV joints Wheel bearing drag tire deformation tire slippage Many of these factors behave very differently during a coast-down and some aren't applicable at all. There is no way to accurately measure crank horsepower with a chassis dyno. Period.

I'm well aware of that and in fact, and in fact have questioned the MAHA's results as well. However, it is an interesting concept and one that would provide yet another dyno comparison to those already done. I'm sure we could debate dyno's, calibrations, conditions etc. until the cows come home (or are sold for meat) but the fact remains that it would be good to see what different types of dyno's are showing.

 

[[email protected]]

Kim Collins Quattro Sports in the UK is going to baseline an R32 next week on their MAHA LPS 3000. -Pete, Stratmosphere

 

[[email protected]]

Re: (bmxvr6)

Quote, originally posted by bmxvr6 »

Except nobody does a "hard launch" in 4th gear at 40mph where a dyno pull is started.

No they don't, but they also don't have constant slip conditions keeping the car in AWD mode either.
Either way, it seems that the Haldex system is going to give variation across the board on different dynos. So long as the car is consistently tested on the same dyno we should be able to at least measure differences in certain mods.
Below is the graph showing the R32 haldex torque split front and rear (in Red and Orange) and for comparison sake an S4 torsen torque split front and rear (light blue and light green).

You can see that the Haldex system actually went rear biased at first and where the lines intersect when to a 50/50 split and then gradually back to a front wheel bias.
The S4 torsen torque split stayed equal for the most part throughout the run with a mostly rear drive bias.
Just for the heck of it below is the dyno curve comparison on the same dyno of a WRX vs our R32.

- jamie

 

[Daemon42]

Re: ([email protected])

Quote, originally posted by [email protected] »

No they don't, but they also don't have constant slip conditions keeping the car in AWD mode either.

Right. The point we were trying to make though, being that in 4th gear at virtually any speed,
it'll be 90+% front biased. Won't see the kind of torque split this dyno run represents
in out on the street except in like 2nd gear.

Quote »

Either way, it seems that the Haldex system is going to give variation across the board on different dynos. So long as the car is consistently tested on the same dyno we should be able to at least measure differences in certain mods.

Agreed.

Quote »

Below is the graph showing the R32 haldex torque split front and rear (in Red and Orange) and for comparison sake an S4 torsen torque split front and rear (light blue and light green).

That S4 split is freaky. Torsen isn't supposed to work like that. A torque biasing diff is supposed
to put equal torque where there's equal load (which the dyno should be supplying). It should only show
an imbalance when there's unequal load, up to about 3-4 times multiplier.. (one side will never
show more than 4 times the torque of the other). Very odd. I gotta think if there's any way
to achieve a static bias with a Torsen or other torque biasing diff.

Quote »

Just for the heck of it below is the dyno curve comparison on the same dyno of a WRX vs our R32.

Truly priceless, although it'd be a little less dramatic if the torque curves were on the same scale
as the horsepower and the Y scale didn't *start* at 88 lb-ft. Still, I like to use the point at
which the torque curve rises above 150lb-ft as a good measure of "low end torque", (looking
at other VR6 and 1.8t dynos and such), and you gotta love the look of the R32's low end.
ian



Modified by Daemon42 at 3:59 PM 3-19-2004

 

[Chapel]

wow
thats impressive
the R32 makes off-idle torque greater than the WRX makes peak!
Impressive throughout.

 

[BluHeaven]

Re: (Chapel)

so... since rotational mass/weight affects hp put to the ground... and the dynapack measures off the hubs, how much power would be lost with the wheels/tires back on the car?
just curious

 

[l5gcw0b]

Mmmmmmmmmm, torquey.

 

[crazy88]

Re: (l5gcw0b)

does anyone make a haldex "boost" gauge? it would be cool to see the f/r power distribution as you're driving

 

[Forge US]

Re: (crazy88)

Quote, originally posted by crazy88 »

does anyone make a haldex "boost" gauge? it would be cool to see the f/r power distribution as you're driving

^ That is a really cool idea! http://****************.com/smile/emthup.gif
Also, mentioned earlier was the 1-2 rpm difference between the front and rear wheels on a Haladex system to maintain full lock. Pardon my ignorance, but is it possible for a change in the size of either the front or rear brake rotors enough to change the rotational inertia of the rotational mass?

 

[[email protected]]

We stock a computer from Germany that replaces the Haldex unit with a modified Haldex computer. 50/50 full time split. Works here and with the Audi TT. -Pete, Stratmosphere

 

[Chapel]

oh, THATS cool
I love when you can force AWD systems to do your bidding.
I know plenty of people who have Skylines that lock in 10/90 for most of their track work

 

[[email protected]]

Re: ([email protected])

Pete,
Does that computer allow for changing of bias? So say I wanted 40/60 haldex split instead of 50/50, would this be possible w/that "computer"?
We've been hearing about folks doing work on Volvo's in Europe involving Haldex.

 

[[email protected]]

No "bias pot" but I think it has a defeat switch to allow stock function to operate. -Pete

 

[TTigg]

Re: ([email protected])

Details? can't see anything on your site about it
TIA

 

[[email protected]]

It's not on the site. Only got a couple as a one-off. Can't get a steady supply so it's only been something that customers/dealers know about. -Pete

Modified by [email protected] at 12:59 PM 5-7-2004

 

[Daemon42]

Re: ([email protected])

This is the same controller that Forge was messing with a while back.
Much was made of it on Audiworld more than a year ago, and in the end, after
a few people tried, it turned out to not be all that, and Forge decided not
to sell it.
It basically attempts to lock the Haldex clutch all the time, and it was
shown that mileage suffered horribly and power delivery on the track left
something to be desired as well, because drivetrain loss went way up. Keep
in mind, there is no center differential with Haldex, but when going around a corner
the rear tires need to turn slower than the fronts, so with a locked center coupling
something has to slide on the ground instead. This is fine if you're driving
at 10/10ths and you want that slippage to be the rear end hanging out, but
the rest of the time it's just power robbing drag, and will produce
severe stress on the drivetrain. Basically this piggyback
unit would be good for some occasional entertainment value but
should be left off most of the time.
In other news, there's a rumor that another tuner is not that far
from having actual more aggressively tuned Haldex programming available
for the R32.
ian

 

[[email protected]]

Re: ([email protected])

Thanks Pete for the info.
I'm guessing it's a matter of time before someone develops a way to tune Haldex so the end user can play with it.
A 30/70 or 40/60 split with a bigger rear sway bar would definitely help in dialing out the lack of turn-in/understeer.

Ian..I remember that. Essentially the car became a bigtime dog with it split/locked up. The way to go though if/when someone does this would be dialing it in under certain conditions of slip and maybe having a more aggressive initial bias setting.

 

[Daemon42]

Re: ([email protected])

Quote, originally posted by [email protected] »

I'm guessing it's a matter of time before someone develops a way to tune Haldex so the end user can play with it.

I doubt it'll be user adjustable, but I do know that Haldex tuning is being worked on.

Quote »

A 30/70 or 40/60 split with a bigger rear sway bar would definitely help in dialing out the lack of turn-in/understeer.

Again, I have to point out that this car has no center diff, and all power going to the rear
comes through the front diff first. So there will *never* be a static 30/70 or 50/60 f/r split.
Nor is locking the center coupling, truly a 50/50 torque split, unless there's exactly the same
load and traction front and rear (which is unlikely given that 60% of the weight of the car
is on the front tires) In fact, this is something that bugs me about every magazine review
of the R32 where they say "Haldex can send up to 50% of the torque to the rear". False.
When it's locked, it can sent nearly 100% of the torque to the rear but it'll
only do so when the front has no traction and 100% of the available traction is at the rear.
You might think.. "yes, but wouldn't that be true of any AWD system?" It's not.
The reason this is a significant distinction is because while cars like the WRX (open center with
viscous coupling) and S4 (Torsen center) both start with a *true* 50/50 torque split, neither
one of them can send 100% of the torque to the rear should the need arise. The viscous
coupling will only allow 70-80% at most, and Torsen actually goes belly up completely if one
end of the car is totally without traction (like an open diff), although under normal circumstances
it'll bias it 33/67 or 67/33 either way. A Haldex equipped car ranges from 100/0 to nearly 0/100
torq split depending on traction conditions, but when fully locked, its static bias is more likely
to match the weight distribution of the car (about 60/40 f/r).

Quote »

Ian..I remember that. Essentially the car became a bigtime dog with it split/locked up. The way to go though if/when someone does this would be dialing it in under certain conditions of slip and maybe having a more aggressive initial bias setting.

Ya.. basically what Haldex itself does. It's active a lot more than people think it is, as I've driven the R32
as a FWD and it's a very different experience. Can feel its influence at all speeds and while cornering
etc.
ian

 

[[email protected]]

Guess the true "tuning" then is going to be on how it reacts to slip.
Information I have is long since dated, but at least what they were trying with Volvo was to get somehow more of a rearward bias to Haldex. How or if they have even accomplished this I don't know b/c information on this is REALLY hard to find.
George had some information on it awhile back, but even that really didn't talk about specifics. I'm not overly holding my breath on this one.