Remove torque from weight calc for NA cars

[cwbaader]

Joon, look up the equation for horsepower and you'll see how nonsensical your response is.

[gisqc]

The only thing that would be wrong with this calculation is that it doesn't take count of where the powerband is, I explain :

 

A torque monster with a 3500-6000 rpm powerband will get its hp average with torque (and since hp = torque with rpm), it would be penalized against a car that produces same hp with less torque but in the 6500-8000 rpm range.

 

Yes torque is a factor of a "flat" hp curve which is perfect scenario in a hp/weight series. But a lower rpm flat hp curve will have higher torque then a high rpm flat hp curve. But only one of these car have to average with the torque to "reduce" their allowed hp. And this is why the S54 is such a good GTS motor, flat hp curve but it's high enough to not get penalized with the tq averaging.

 

It prevents good turbo curves, but anyone willing to go extra mile for flat hp curve will shift powerband up in rpm and get around this rule anyway.

 

 

good points, but here is the bottomline.

torque really is only an indicator of a "flat HP" curve. a "torque monster" will be the flat hp curve competitor. all it means, when you have LOT of torque, is that your HP curve might be flatter than someone else with the same peak HP.

 

If you understand the actual situation here, averaging HP and torque, does only one thing. It uses a car with a greater torque than HP , when the units are measured in Ft/lbs for torque, and tries to equalize its advantage over a car with a peakier HP curve.

 

Hum, what I meant is take the SAME flat hp curve on a 3000-6000rpm powerband and on a 5000-8000 rpm powerband, one will have to average torque with hp (because big hp at 3000rpm takes alot of torque) while the other will only use hp numbers (because you need less torque to make the SAME hp at 5000rpm). Which means engine turning more rpms have an advantage in this way of calculating.

[cwbaader]

" Which means engine turning more rpms have an advantage in this way of calculating." Now if you factor in the applied torque/hp through gear ratio things get more one sided toward a high rpm motor. Where a high torque motor will run a 3.9 or 3.7 gear, a high RPM motor will use a 4.1 to 4.5 gear. Where the torque motor runs out of hp/tq half way down the straight, the high RPM motor is pulling its hardest. Do the math. CB

[cstreit911]

one of the things that really should be addressed is absolute weight, which is far beyond the advanges of a "flat" Hp curve, which is now being equalized by the crude HP vs torque averaging ( for only cars with greater numerical torque values than HP) (if you remembered we fixed part of the problem from day one, from NOT averaging the cars with lower torque than HP which was a HUGELY smart move by the way)

 

anyway, the correct way to incorporate equalization in the HP /weight ratios would be to simply use average HP from 3 data points. Hp at peak torque, HP at max HP and HP at redline. this would be the correct way to do it, at slightly more work.

BUT, a 3200lb car vs a 2700lb car racing would be an unfair race for the heavier car. You can make up for some of it by tire size, but generally, there should be a range of weights too that correspond to the HP /weight ration. this is MUCH more important than the HP/torque averaging. maybe there are classes of cars 2500 to 2700, 2701lbs to 2900lbs, 2901lbs to 3100lbs. those weight ranges have points associated to it, similar to using Rcompound, vs slicks.

 

Mark,

 

Interesting points but you say " 3200lb car vs a 2700lb car racing would be an unfair race for the heavier car."

 

Let me pose this to you... At Road America would you rather have a 355HP 3200# car, or a 2700# 300HP car?

 

I'd choose the former. At Autobahn, I'd choose the latter. Point being that at high speed tracks HP is king... Short tracks, notsomuch. I think some of this averages out.

[rlipner]

Perhaps if the current rule is unfavorably skewed for one class (GTS 1) make an adjustment for that group but leave the other classes as is.

[JSG1901]

At this point, this is just discussion for discussion's sake as we're not going to do anything with power-to-weight ratios or calculations for 2013.

 

Having said that, it's a discussion we've had a lot of over the past howevermany years. The problem is, we have a lot of folks who like it the way it is, and a few who are sure what we're doing is not quite right, but we don't have numbers to actually prove it either way.

 

What I would love to see...and what I can't seem to find anywhere...is a calculator into which we can feed real data (including horsepower and/or torque curves--not just max numbers, but the whole curve--redline, weight, gearing, final drive, and a Cd number), which can compute acceleration all the way up through all the gears.

 

I can't believe there isn't one of these already available on the web but if there is, I sure can't find it. SURELY, one of you guys has the engineering chops to write a spreadsheet capable of doing these kinds of calculations. Given that, we could plug in some real cars with real horsepower/torque curves and see what difference these things actually make.

 

Without something like that, I don't see how any change we might even consider would be any less arbitrary than what we have today.

 

So...who's up for it?

[sstecker]

that calculator exists. i did those test with the bosch motorsports lapsim - you can download the demo version and use it to test any hp/tq curve you want with any gear ratio, rpm switch points/etc, tire sizes, cd, etc..

 

i tested with a perfect flat motor - 200whp from 1000-9000rpm - otherwise identical cars. i then had it compute straight line acceleration times from 60km/h to 140km/h. car A switching at 5250rpm (TQ monster) and car b (high rev) switching at 9000rpms.

 

edit i also tested those against a flat 200tq motor switch gears at 5250rpm

[cwbaader]

And the results were????

[bmwjoon]

Scott, I have something that might work on my old laptop from college. No joke, it's from the 90's. It's a drag racing program. You put in Cd, gear ratios, hp, and torque numbers, redline. You can even put the time to shift and it generates 1/4 mile times.

 

I'll see if I can dig it up.

[bmwjoon]

Wow, I actually remembered the name of it, googled it and found it !!!

 

It's called Cartest. It was an old DOS program that was on the PC inside our student autoshop. I have a copy of this now.

 

There is a new version called Cartest 2000. It has a "lap test" feature that creates a virtual lap time.

 

It's only $39 bucks so if there's interest should I buy it and try it?

 

Worth a shot?

[JSG1901]

This is a pretty interesting question, as we've seen. In the interest of keeping everybody's identities private, I'm going to try not to identify any of the players in the following.

 

I have a friend who is an engineer with a major company and who has spent a fair chunk of time working on GTS cars over the years. Along with a lot of data from those other cars and setups, this person developed a spreadsheet which has been used quite successfully in analyzing acceleration based on actual horsepower and torque curves and actual gearing.

 

Separately, I have a dyno sheet from a car which is of the sort we're wondering about here. To only give key numbers, this car makes a maximum of 150 hp at 5500 RPM and 189 ft-lb of torque at 3600 RPM. Pretty grossly over-torqued compared to a "normal" dyno curve.

 

I asked my engineer friend to use the values from this dyno sheet along with the car's stock gearing to see what we could figure out. We imagined the back straight at Mid-Ohio, so the calculations were based on a 60mph entry speed to a 3000-foot straight. The question we were trying to answer was not how fast the car was going but how long it took to reach the other end of the straight since, ultimately, that's what matters.

 

First he played around with different rear end ratios to see what that might do. The best of all he tried (based on actual rear ends easily available for this car) only improved the result by 0.1 seconds.

 

Then, the test: 60 mph rolling start. 3,000 ft long straightaway

 

Under the current rules, this car could run in GTS1 at 3132 lbs, or in GTS2 at 2455 lbs. We tried it both ways.

 

1) GTS1 @ 3,132 lbs: 21.62 secs

2) GTS2 @ 2,455 lbs: 20.57 secs

 

Alternatively, using the rule that was proposed here which would ignore torque, for GTS1 this car would only need to weigh 2775 lbs and would reach the end of the straight in 21.09 secs, a bit more than half a second faster.

 

For reference, I asked for numbers from a couple of known competitive GTS1 and GTS2 cars my engineer friend had the data from. These are all BMWs and use data from actual dyno runs along with GTS minimum weights. They came out like this:

 

GTS 1 Car (141 hp): 21.49 secs

GTS 1 Car (131 hp): 21.67 secs

GTS 2 Car (190 hp): 19.79 secs

 

What I get out of all this is the car we're talking about would be quite competitive in GTS1, at least in terms of acceleration, if it carried that much weight. And, although we didn't run the numbers on this, keep in mind that what we're talking about here is time down the longest straight on the Mid-Ohio track, a length which is more or less representative of most long straights. With the high torque low down, I'd expect this car to be quite fast out of slower turns, which might be an advantage over some competitors.

 

But, on the car in question, this is a lot of weight.

 

I asked my engineer friend, having worked his way through all the data, what he thought and, for that matter, what he would do if it were his problem to solve. His analysis:

 

"The problem is that this car falls right between GTS1 and GTS2. It would be hard to ballast that car up enough to run in GTS1 with the current rules and it's probably a little slow for GTS2, so it probably wouldn't be terribly competitive there either. However, the proposed no-torque rule would make this a killer car in GTS1. I'm familiar with this kind of car and engine, and the numbers on that spreadsheet are pretty stout for that particular motor. I'm guessing it's had some work done to the motor along with some tuning. ECU tuning is worth nearly 20hp on that engine. If it were mine, I'd put back in the factory chip to drop the power/torque a little bit to get it to fit nicely into GTS1. You can get an ECU out of a junkyard for $100 and you'd be done with it. Use the stock ECU for NASA and the tuned ECU for other racing bodies. Immediately, you'd be within a tenth of a second or so of the straightaway times of the fastest cars from last year at an achievable weight."

 

For the record, I know the sources of all this information and have complete trust in the person doing the analysis. What this all looks like to me is not enough difference in advantage either way to be worth opening the Pandora's Box of potential new issues that would come with a wholesale change in the minimum weight calculation methodology. That's pretty much what we decided several months ago but it's nice to see some form of more tangible confirmation.

 

As far as I'm concerned, for now at least, case closed.

[cwbaader]

So, Scott, what you are saying is that the double torque issue is a non-issue as far as the computer simulation is concerned.

[JSG1901]

So, Scott, what you are saying is that the double torque issue is a non-issue as far as the computer simulation is concerned.

Sorry...I'm not sure what you're asking.

[cwbaader]

Looking at the projected terminal speed of the chosen examples, the program is saying that double torque is a non issue, while it would make high torque cars more competitive. Yes?

[JSG1901]

Looking at the projected terminal speed of the chosen examples, the program is saying that double torque is a non issue, while it would make high torque cars more competitive. Yes?

I must be dense today...sorry. What "projected terminal speed of the chosen examples"? I assume you're talking about my post from yesterday. There aren't any projected terminal speeds in there. I don't even know what they are, since I don't have the spreadsheet.

 

And, I'm not sure what you mean by "double torque." There was a proposal to ignore torque, but double?

 

My understanding is that, yes, as many have said, if we were to ignore torque in calculating minimum weights, cars with extraordinarily high torque in relation to their horsepower would have a definite advantage. I think that's your last statement.

 

Again, sorry if I'm misunderstanding.

[cwbaader]

I surrender

[JS154]

At this point, this is just discussion for discussion's sake as we're not going to do anything with power-to-weight ratios or calculations for 2013.

 

Having said that, it's a discussion we've had a lot of over the past howevermany years. The problem is, we have a lot of folks who like it the way it is, and a few who are sure what we're doing is not quite right, but we don't have numbers to actually prove it either way.

 

What I would love to see...and what I can't seem to find anywhere...is a calculator into which we can feed real data (including horsepower and/or torque curves--not just max numbers, but the whole curve--redline, weight, gearing, final drive, and a Cd number), which can compute acceleration all the way up through all the gears.

 

I can't believe there isn't one of these already available on the web but if there is, I sure can't find it. SURELY, one of you guys has the engineering chops to write a spreadsheet capable of doing these kinds of calculations. Given that, we could plug in some real cars with real horsepower/torque curves and see what difference these things actually make.

 

Without something like that, I don't see how any change we might even consider would be any less arbitrary than what we have today.

 

So...who's up for it?

 

Fandozzi may be able to help you do the modelling you are interested in.

[rabbit_diesel]

This is a pretty interesting question, as we've seen. In the interest of keeping everybody's identities private, I'm going to try not to identify any of the players in the following.

 

I have a friend who is an engineer with a major company and who has spent a fair chunk of time working on GTS cars over the years. Along with a lot of data from those other cars and setups, this person developed a spreadsheet which has been used quite successfully in analyzing acceleration based on actual horsepower and torque curves and actual gearing.

 

[snip]

 

I agree with you JSG: "This is a pretty interesting question, as we've seen." Would your friend be willing to generate time estimates for a different car? I think I have some example engines that demonstrate the point which you are discussing.

 

Car details:

body: VW Rabbit

weight with driver: empty = 1970 lb, full = 2020

tire size: ~1006 rev/mile

tire type: slicks

this car in GTS-1, using 20 lb/hp, HP limit is: 98.5

 

Engine 1 (using HP & TQ average rule):

final drive: 3.94

gears: 1st = 3.45, 2nd = 1.94, 3rd = 1.29, 4th = 0.91, 5th = 0.71

peak HP = 83 @ 4350

peak TQ = 107 @ 3700

assigned HP = (83+107)/2 = 95 [does not exceed limit (was closer the previous year)]

torque rise (using 4th @ 4500, to 5th @ 3500) = 11%

       Engine 1
       HP    TQ
3000    48    84
3250    61    98
3500    70   105
3750    76   106
4000    79   104
4250    82   101
4500    81    95
4750    65    72  

Notes: The rev-limiter on this engine rolls off smoothly. The body of the car has enough frontal area and drag that with Engine #1 the car will reach terminal velocity on a long flat straight-away before hitting the rev-limiter; the car has neither speedo nor tach, however I estimate the RPM to be 4200).

 

Engine 2 (straight HP, but not allowed by GTS rule):

final drive: 3.94

gears: 1st = 3.45, 2nd = 1.94, 3rd = 1.29, 4th = 0.91, 5th = 0.71

peak HP = 98 @ 4350

peak TQ = 126 @ 3700

torque rise (using 4th @ 4500, to 5th @ 3500) = 11%

       Engine 2
       HP    TQ
3000    57   100
3250    72   116
3500    83   125
3750    90   125
4000    93   122
4250    97   120
4500    96   112
4750    77    85  

Engine 3:

final drive: 3.94

gears: 1st = 3.45, 2nd = 2.12, 3rd = 1.44, 4th = 1.13, 5th = 0.89

[before this engine wore out, I did not use fifth, because it was not useful.]

peak HP = 98 @ 6750

peak TQ = 76 @ 6400

rev-limiter = 9000

torque rise (using 3rd @ 8000, to 4th @ 6300) = 28%

       Engine 3
       HP    TQ
5500    71    68
6000    87    76
6500    95    77
7000    95    71
7500    93    65
8000    89    58
8500    83    51   

 

My predictions:

Engine #3 is fastest, and Engine #1 is slowest.

Engine #2 is only a little behind Engine #3.

 

Will

[rabbit_diesel]

Lest anyone forget, or not know:

 

Way back when the GTS class was begun, it was strictly weight to power ratio --- there was no averaging of HP and TQ as there is today. The inclusion of TQ came later, and there was a HUGE blowout on the website/forum when that change was rammed through over the objections of the people that did not want it.

 

The GTS website/forum that existed back during that period is long gone, however a few years ago when I was looking into it I found copies of the information on the 'way back machine' [*1]. Also, some people posted on this forum that they remembered the incident of the TQ rule change.

 

Will

 

[*1] I have not looked on the 'way back machine' recently, so I do not know what still remains of the GTS website/forum.

[rabbit_diesel]

Wow, I actually remembered the name of it, googled it and found it !!!

 

It's called Cartest. It was an old DOS program that was on the PC inside our student autoshop. I have a copy of this now.

 

There is a new version called Cartest 2000. It has a "lap test" feature that creates a virtual lap time.

 

[snip]

Thank you for writing to tell us of this program. I found out the company is giving away the DOS version, so I downloaded and installed CarTest 4.5 ("http://www.cartestsoftware.com/cartest4.5/"). I see they also have an online demo of the CarTest 2000 version ("http://www.cartestsoftware.com/cartest2000_demo_applet.html").

 

Putting in the parameters needed for JSG's '150 hp & 189 ftlb' engine example, and then doing '60 mph rolling start 3000 foot straight' simulations, I find that the CarTest software gives results within a few hundredths of a second of the 21.09 and 21.62 and 20.57 figures that JSG's engineer friend generated.

 

Will

[rabbit_diesel]

At this point, this is just discussion for discussion's sake as we're not going to do anything with power-to-weight ratios or calculations for 2013.

 

Having said that, it's a discussion we've had a lot of over the past howevermany years. The problem is, we have a lot of folks who like it the way it is, and a few who are sure what we're doing is not quite right, but we don't have numbers to actually prove it either way.

 

What I would love to see...and what I can't seem to find anywhere...is a calculator into which we can feed real data (including horsepower and/or torque curves--not just max numbers, but the whole curve--redline, weight, gearing, final drive, and a Cd number), which can compute acceleration all the way up through all the gears.

 

I can't believe there isn't one of these already available on the web but if there is, I sure can't find it. SURELY, one of you guys has the engineering chops to write a spreadsheet capable of doing these kinds of calculations. Given that, we could plug in some real cars with real horsepower/torque curves and see what difference these things actually make.

 

Without something like that, I don't see how any change we might even consider would be any less arbitrary than what we have today.

 

So...who's up for it?

 

Well, it has been a while and I notice that no one has posted the results of any more simulations. So, I will.

 

It seems to me that the best way to make a comparison for demonstrating the reasonableness of the HP+TQ handicap for excessive torque, is to make tests with sets of examples where ONLY the torque is varied; and to hold everything else the same. At a minimum: a control example, an example without the handicap, and an example with the handicap.

 

BMW 325i

power = 150 bhp @ 6000 (at wheels)

final drive ratio = 3.73

tires = [whatever was in the CarTest database for 325i]

gears: 1st = 3.83, 2nd = 2.20, 3rd = 1.40, 4th = 1.00, 5th = .81

[i am not familiar with BMWs. Can someone tell me whether these gear ratios are at least close to correct?]

 

I selected the 3.73 final drive ratio because it was shown in the NASA Spec E-30 rules. I did not apply much thought to what ratio, because JSG reported that his friend's simulations showed that ratio made little difference. The CarTest software selected the shift points.

 

For a 3000' straight-away and 60 MPH starting speed:

#10) 189 ftlb @ 3500, 2775 lb --> 20.99 sec

#11) 148 ftlb @ 4500, 2775 lb --> 21.11 sec

#10a) 189 ftlb @ 3500, 3136 lb --> 21.50 sec (HP+TQ handicap)

(Note: each case above only used 3rd and 4th gear, because to shift into 5th made the elapsed time greater.)

 

Notice that the difference in elapsed time between car #10 and car #11 is only .12 seconds. How close do these need to be before most people would agree they are competitive? In any case, what is the justification for applying a handicap so large that it puts what had been the winner by .12 seconds --- behind by .39 seconds? Seems awfully heavy handed.

 

JSG's friend simulated cars that are quite similar to my #10 and #10a:

150 hp @5500 RPM and 189 ft-lb @ 3600

#1) GTS1 @ 2,775 lbs: 21.09 secs

#1a) GTS1 @ 3,132 lbs: 21.62 secs

(I am guessing these runs used 3rd, 4th, and 5th.)

 

-----

 

So, how might cars #10, #11, and #10a do on a race track? The CarTest program can also do performance simulations for various race tracks.

 

#10 @ Watkins Glen ---> 2:50.5 sec

#11 @ Watkins Glen ---> 2:51.9 sec

#10a @ Watkins Glen --> 2:53.7 sec

 

#10 @ Laguna Seca ---> 2:19.1 sec

#11 @ Laguna Seca ---> 2:20.0 sec

#10a @ Laguna Seca --> 2:21.7 sec

 

#10 @ Road America ---> 3:7.6 sec

#11 @ Road America ---> 3:9.2 sec

#10a @ Road America --> 3:12.0 sec

 

The difference between cars #10 and #11 varies from .9 to 1.6 seconds per lap. Notice that in every case the HP+TQ handicap puts the car that had been the winner, behind by 1.7 to 2.8 seconds. What is the justification for applying a handicap so large? Seems awfully heavy handed.

 

-----

 

Does anyone wish to suggest that the relative time differences shown by the CarTest software are not reasonable? Okay --- please post the supporting data, and tell us why it is better than the results from the CarTest software.

 

Will

[cwbaader]

And just look at all the GTS 1 and 2 entries at the nationals.....0 GTS 1 and 7 GTS 2, Pitiful!

[Zach H.]

feel free to correct me, but from what i have heard the GTS3 winning car at nationals had 275ftlbs/265hp......

[cwbaader]

Zach...265/275 does not skew the numbers. Look at 150/185....suddenly your hp number jumps way up.

[tcdesign]

feel free to correct me, but from what i have heard the GTS3 winning car at nationals had 275ftlbs/265hp......

 

As the driver of that car you are way wrong... we were not even close to having the same torque as hp let alone more.

 

-tony