Supercars, Car and Driver, Results - When one of you guys taking a GT40?

[Ron Earp]

Well,

I was keeping Sydney all weekend alone thus could not work in the garage. But, I can work on the computer and have the September issue of Car and Driver with the 15 Supercar shootout.

Essentially, it is a competition started 2 years ago to find the "baddest car in the land", their words, where the land is defined as the US and the baddest is defined pretty much by 1/4 mile times and some track work. They have a track portion but the horsepower and motor really decides the winner here.

As I've mentioned here before, as have others, I don't buy a lot of the high hp claims by lots of these folks. So, I took the data: weight, trap speed - for each car and calculated hp used through the quarter. I also estimated hp available assuming a 18% drivetrain loss and using the manufacturers flywheel claims.

The data are shown here Car and Driver 2002 HP Shootoutand I've highlighted the cars in red which don't "add up". That is, their claimed hp doesn't match well with their trap speed hp calculation.

The generally accepted formula for hp calculation from rear-wheel drive cars is:

hp = weight * (speed / 234)^3

You can perform it based on ET but the results are more erroneous due to traction etc.

Needless to say a lot of the cars are questionable, as you'd expect. But, check out the loser of the competition, the GMC truck. That car has the goods for sure, serious power.

NOW the real reason I started this is to find out what manufacturer of GT40s is going to step up for next year and enter a car. You supply the car and the driver, it only takes two days for the competition. And the exposure would be great. Superformance and Beck Lister has one of their cars here, so replicas/kits are accepted for sure. I imagine it will have to be one of the larger companies with current GT40 offerings out and about for purchase: ERA, CAV, Roaring Forties, etc.

It is of interest the 5th place car, the Photon, is a mid-engine design and it was indicated it could have easily won with a more powerful engine as the chassis was the best of the group. A GT40 next year maybe?

Ron

 

[Guest]

Hi Ron

Got to disagree with you on your analysis! You have assumed a straightforward 18% loss of power in the transmission. On the GMC truck this is in excess of 200 bhp! Eeek! Sounds like a transmission I would build.

Take the Renault 21 car with say 150 bhp at the flywheel and 115 at the rear wheels. This car has the infamous and much loved/hated UN1 box. A 23% transmission loss of power is clearly seen here.

GTD builder man comes along and steals this magnificent box and slaps it into his GTD (right hand drive chassis), throws out his chevy 350 he first thought of and mates it to a FORD 347 stroker kit engine with say 400 bhp at the flywheel. He would get 308 bhp at the rear wheels by your reckoning. So where on earth does this extra 57 bhp go?

I would say that it doesn't go anywhere as the transmission will only suck up a finite amount of power irrespective of what drives it.

Therefore you would need to re-run your spreadsheet with a fixed deduction of bhp for each type of transmission, be it auto or manual or whatever.

My car was dyno'ed on a rolling road at 306 bhp at the rear wheels. I have always assumed that I have had about 350 bhp at the flywheel. This has not been proven either by a pure engine dyno or computer simulation. Some only credit me with 340 bhp at the flywheel. Whatever, this gives a range of bhp loss for the Renault UN1 box at between 44 and 34 bhp.

Put this into Roy's car at a claimed 560 bhp at the flywheel (I await to be corrected) and he has in excess of say 510 bhp at the rear wheels. No wonder he breaks traction every time he dabs the throttle. Under your assumptions he would "only" have 431 bhp at the wheels.

The information bulletin that goes with this bit of chit chat is, be very careful of rolling road testing a good strong engine with a Renault box as Robin stripped his fifth gear under test conditions. Clearly my engine is not powerful enough to do this, but hey I still beat him on a regular basis!

Malcolm

[ August 12, 2002: Message edited by: Malcolm M ]

 

[Mark Worthington]

The subject of drivetrain losses got a lot of attention when the 99 Mustang Cobras came out and dynoed ~250-260 rwhp when they were rated at 320 crankshaft hp. Ford eventually fixed the problem, and stated that the 320 hp Cobra should make 271 rwhp, translating to a 15% loss through the powertrain to the dyno rollers. While it is customary to express this loss aa a percentage of crankshaft hp, the powertrain loss really doesn't change much with increased horsepower. I recall reading about some very detailed testing that was done to isolate sources of powertrain loss between the crankshaft and the dyno rollers, and the leading source of lost power was actually the tires on the rollers, followed by the differential, followed by the transmission. I believe the T5 tranny robs about 35 hp compared to the T56 which robs more like 45 hp (more and larger gears, etc.) Since the GT40 doesn't have a driveshaft with U-joints, and since the halfshafts and CV joints are pretty damn efficient, most of the losses will be attributable to tires and gearbox.

[ August 12, 2002: Message edited by: Mark Worthington ]

 

[Ron Earp]

This is something that I thought of as well but couldn't do anything about. As far as the truck goes, well, it is probably more.

See the truck is running a auto tranny from GM (well known for big losses) and is 4 wheel drive with 1 transfer case and 2 differentials (big losses).

At any rate I think the situation is still true - there are some cars there that do not put out what they advertise. Especially the blown NSX and the turbocharged Lexus.

Now, when is someone taking a GT40 to this thing?

R

 

[ChrisL]

Ron,

Not doubting your formula totally, but I think that it may probably only be applicable in some limited controlled situations. It doesn't take into account for example, the all important gearing.

Try doing the same quarter-mile in the same car, except skipping second gear. You'll see what a difference the gearing can make. Same engine, same horsepower, same losses but an enormous difference in results.

Most European supercars (and for that matter the original GT40's I would suspect) are geared for high speeds and not a quarter-mile so would run taller final drives, ie using a high gear all the time.

[ August 13, 2002: Message edited by: Chris L ]

 

Guys no offense but I think Ron is more worried about somebody putting a GT40 in the comp next time around rather than hp formulas.
Adrian

 

[Guest]

Ron

You need to persuade Roy and Paul to ship Roy's car over. This is really the sort of thing that turns them on! If only it was in the UK and you would get a queue of owners....

Malcolm

PS Mind you if they did go they would learn too much and then no-one would ever catch them at Brighton Speed Trials! Not that we could before.

 

[Ron Earp]

Adrian has it right, I figured some of the manufacturers would like the exposure and chomp at the bit to sign up. I think more than a few Cobra replica makers wanted in on the action (from Club Cobra's board) but were denied. A GT40 should do well and whosever's car is in the article will be getting lots of calls for sure.

Private folks cannot enter their own cars, has to be a manufacturer. Else I'd plan to go when done and bugs worked out.

R

 

The Shoot out I prefer is ran By Grassroots Motorsports. It is something I can relate to.
The cars total cost could be no more than $2002.00 and I belive all the entrants had a lot of cheap and fast fun.
VIC

 

[Guest]

Ron:

Your formula and drivetrain losses have been accepted by the automotive comunity for many years. They may be off a little either way, but your numbers are very good. The Lexus advertised fwhp is off by about 200. They either lie very well, or the car was very sick.

Jeff

 

[Jim Rosenthal]

I nominate Lee Holman (are you listening, Lee?) for the shootout for next year. Here's why:
1) he is a manufacturer
2) the "Holman GT" has been around long enough to be thoroughly debugged
3) the car will kick some serious butt regardless of what else enters
4) it is the shape and make we all know and love
Maybe if we all suggest this to Lee he would do it. Any seconds?

 

[Guest]

Jim, I just sent Lee an email for his amusement if nothing else. We'll see if he rises to the bait

 

[Gary Gibbs]

Warning very long and boring post!!!!!!

My 0.02 on the drive train losses. I am not a drive train engineer nor have I done this sort of testing so all of my statements have no test data for a backup.

We could get an idea of what is going on by looking at the physics. Kinetic friction forces (inside a gear box or differential for example) are proportional to the normal force applied. Just try sliding a block on a given surface the friction force resisting the sliding is proportional to the vertical (normal force) applied. This is the reason that aerodynamics provides a significant increase in a cars cornering speed. Aero provides downforce (increased normal force) on the tires and thus they provide more friction force for cornering.

Back to the gear box. The forces measured in the gears and bearings of a gear box are proportional to the torque applied to the shafts. Double the torque and the forces everywhere double.

If the friction forces arise from normal forces and are proportional to normal forces then the friction forces are proportional to the applied torque. Again you double the torque you double the friction forces.

Now getting to power. The power input to the box is equal to the torque applied to the shaft times the speed of rotation (times some constants to get the units to come out right). The friction power is equal to the torque due to friction times the rotational speed. This would lead to the conclusion that the ratio of power input to power dissipated due to friction would be a constant (torque in * speed)/ (friction torque * speed). Thus a percentage could be applied (ratios equal a constant percentage). Now this percentage I would expect would be different for different gears, transmissions, rear ends,etc. Maybe certain drive trains are similar, not sure. All of this friction related power goes into heat inside the gear boxes and must be removed.

There are likely other types of losses that are viscous in nature and would be proportional to speed or speed ^ 2.

Now lets see if this sort of jives with reality. Take a rear end gear for example (such as a Winston Cup 9" ford). It is difficult to argue that it consumes a constant amount of power under all conditions (including HP applied to the input shaft). Lets say that it consumed 20 HP. When it is sitting still it consumes zero HP. When the car is idling down the pit lane it would not consume 20 HP (cars cruising down the interstate consume something like 15-25 HP including wind resistance). Now when it is at speed on the track being feed 750 HP does it consume 20 HP? If one believes the above friction force discussion, then one would believe that the consumption of HP is proportional to the power feed in. Thus the average 9" rear installed in a Ford pickup would consume far less then the same 9" rear installed in a WC car. It also makes sense when you see the requirement to get rid of huge amounts of heat in a WC rear through use of pumps and coolers.

In summary I can believe the HP losses are proportional to HP for a given vehicle (thus use a percentage, e.g. 20%) and would expect the percentage to vary somewhat from vehicle type to vehicle type.

[ August 14, 2002: Message edited by: CCX33911 ]

 

[Guest]

When Ron or myself say "drivetrain losses", what we shold be saying is the difference between a car on a chassis dyno, and an engine on an engine dyno. The generally excepted conversion numbers are between 15% and 20%. Thus and engine on the dyno tat makes 500 hp would make about (17.5% conversion) 412 rear wheel hp.

Jeff

 

[Gary Gibbs]

Jeff,
I agree that these "drivetrain losses" to first order should be a percentage of the HP sent through them not just a constant value (e.g. 20 HP for all HP engines in that car).

 

[Mark Worthington]

G. Gibbs, you neglected inertial drivetrain losses in your analysis. A T56 robs more power than a T5 tranny because it uses more and larger gears, and crankshaft horsepower is expended in accelerating anything which has an angular moment of inertia to a given angular velocity. The injertial losses are essentially constant and will not change as a percentage of applied horsepower.

There are also non-inertial losses that are independant of applied hp/torque. Take, for example, airflow that is created in the vicinity of spinning wheels, CV joints, and tires. For a given speed, the airflow and attendant power loss is independant of applied torque and horsepower.

I also think the frictional losses may not be directly proportional to applied torque but I need a little time to think about it.

Bottom line is that if you measure powertrain losses as a percentage of horsepower output then you will find that, for a given powertrain, the percentage loss decreases with increasing horsepower.

 

[Ron Earp]

I'm not sure what the industry accepted value is, but it is certainly higher than the 5-10% quoted in one of the posts above.

I know for a fact that a chain drive motorcycle engine losses are around 8-11%. Of course, I'm sure and do agree this varys with transmission, power level, etc. But, a motorcycle is very simple system.

Of the two cars I've seen both the engine and wheel dynoed, the losses have been 15-18% at the wheel vs. the flywheel. This seems consistent with what a lot of the more hard core engine/car mags indicate. But, the thing is there are certainly differences in dynos and the ones I've seen of course can't use the same dyno as one is for chassis and the other for measuring flywheel hp. It isn't quite comparing apples to apples.

I don't know, seems 15-18% pretty normal. At any rate, some of those cars ain't making what they claim but some of ours are. One of you UK folks need to enter a car in this thing. The only problem is that I think the car has to be available, that is available for purchase. That would make GTDs difficult and a one-off twin turbo GTD not accepted.

R

 

[Neal]

A local tuner dynoed my Cobra and did a parasitic loss test before any pulls. The car was brought to a fixed rpm and then allowed to coast until nearly stopped. Loss was calculated at 17%. I'm surprized more tuners / testers don't do this.

 

[Guest]

Neal, I used to be very active in the Stealth/3000GT community and the procedure you described was SOP for all the dyno testing that was discussed on those forums. This was used to approximate the flywheel vs wheel horsepower which virtally everyone wanted to know.

 

[ChrisL]

G. Gibbs

Just another factor to consider. Kinetic friction in your sliding block example is highly dependent upon an existing force being applied vertically, in this case maybe gravity. No gravity, no friction, regardless of the amount of horizontal force applied.

For the introduced force to be a factor in the level of friction, the force would need to have a verticle component. ie. pushing diagonaly down on the block, otherwise the change in friction would simply be a fuction of speed.

You would therefor need to take into account how the two components of the resultant force are applied to the individual items in the gear-box to determine whether the friction is a function of torque and speed, or simply speed alone.

The viscosity factor, incidentally, would also be speed related with no added friction due to torque levels.

Been a while since I pulled out my physics books, so I might be way out on this.

Ahhh, Errrr, have we strayed somewhat from the original subject?