In addition to driving an IT car, I’m also a member of the IT Advisory Committee, the committee that advises the CRB on all things related to the Improved Touring rules.

One of the things that the ITAC has been wrestling with over the past couple of years is how to account for engine torque.  We all know that big V8s (some of which were just classed into ITR last month) are torque monsters, and that other cars are “gutless,” like the high-revving 4-cylinders in the Honda S2000 and Acura Integra Type R, and fast-spinning rotaries like the Mazda RX-8.  Here is a quick look at some ITR cars and their published specs (including my own car):

When these cars were added to ITR, the ITAC added/subtracted some weight to some of them to try to make up for the torque effect.  But the fact is that the amount of weight added/subtracted was, as one ITAC member likes to put it, a POOMA (“pulled out of my ass.”)  And we hate that.  We want the weight-assignment process to be repeatable and consistent, and POOMA doesn’t really help with that.

One of the struggles that the ITAC has is that it doesn’t have a whole lot of information to go on when assigning a race weight to a new car listing.  We don’t have dyno charts, but we do have the HP and torque peak numbers … the manufacturers all publish those.  So, I sat down to try to figure out whether or not there was enough information in those numbers to determine something about the effect of torque on the racing potential of a car.

So, to start, I had to get beyond the conventional wisdom and try to put some math into the thinking.  The conventional wisdom says things like (all things I have seen written in the past week):

• “Horsepower sells cars, torque wins races.”
• “High torque gives you better exit speed off the corner.”
• “A flat torque curve means you have to shift less often.”
• “You can’t feel horsepower, you can only feel torque.”

I’m sure there are more.  The problem is that none of these apply any math to anything.  They certainly do make the case that a car with lots of torque should get a weight penalty though, don’t they?

Before I begin, I want to give a big caveat.  This topic tends to be one that has a lot of old-wives tales, gut feelings, and strongly-held beliefs associated with it.  There are “camps” related to this argument (those who believe that torque rules and HP doesn’t matter, those that believe the opposite, and … the rest of us.)  So after a couple of weeks of intensive study, I believe everything I say below, but I’m totally willing to admit that I’m wrong.  I’m just a somewhat geeky race car driver, not a physicist or a mathematician.  Someone smarter than me might very well get through two paragraphs of this and figure out what I’ve got wrong here.  End caveat.

The more I read about this topic, the more I realized that some of that conventional wisdom, while true in some cases, is really misleading when it comes to classing IT cars (or really building any road race car).  First of all, in IT, you can run any differential gear ratio you want to.  Meaning that no car is forced to lug around at 1,500 rpm like we do on the street.  We can gear our cars to spend time in whatever part of the rev range we want to.  So the concept that the car with the low-end power is better isn’t applicable here.  Likewise, the idea that a wide, flat curve means that we have to shift less is also less interesting.  Who cares how many shifts you have to make 0-60?  We don’t go 0-60.  We don’t go 0-anything.  And race car drivers are good shifters.  Sure, every upshift is an opportunity for a mistake and does cost a small amount of acceleration for the duration of the shift, but it’s not a big effect (at least, I postulate that it’s not a big effect.)

Since each car in a given racing class can complete a lap in roughly the same amount of time, all cars in a class go the same average speed around the track.  There might be minor differences in straightaway speed vs. cornering speed, but such is the nature of a mixed-marque racing category like IT.  And further, given the freedom to pick any differential gear we want to, all cars have roughly the same overall gearing.  A quick scan through the IT listings for ITR shows that almost all cars with a 5-speed transmission have roughly the same percentage drop between 3rd gear and 5th gear (about 60%).  Therefore, I conclude that we all shift roughly the same number of times in each lap, regardless of the nature of our engines.  What probably makes high-revving cars feel “peakier” is that in order to keep the revs high, those cars use shorter gears.  Shorter gears mean that the revs rise more quickly as speed increases, which means that they have to shift more.  The good news for those cars is that they already have gear ratios that work well (“close-ratio gearboxes”), and at least in ITR, they are often 6-speed transmissions.  They might have to use 4 speeds of their 6 instead of 3 of the gears in a 5-speed, but it all works out.  I’ve even been told that a lot of S2000 guys don’t gear down, they just use 2nd, 3rd, and 4th, which actually have a closer spacing than 3rd/4th/5th in a lot of 5-speeds.

The other big revelation is that it’s the HP peak that’s the king of the road here, not the torque peak.  If we had a CVT, we would set it up so that we’re always at the HP peak, 100% of the time.  Put another way, if we had a CVT, we wouldn’t care about torque AT ALL.  Given a conventional manual transmission, we have to deviate from the HP peak on one side or the other, but our goal is to maximize our HP at all times, and that means, staying near the HP peak at all times.  I’m going to gloss over why I believe this to be true.  I won’t regurgitate what I’ve read online.  I found several compelling explanations, but this one by Paul Yaw is the best one I’ve seen.  Another fairly fun-to-read one is titled “Plato and Socrates Discuss Torque, Power and Acceleration.” See the bottom of the page.

Now, given that we all gear ourselves similarly, we can all use roughly the same range of revs in any given lap.  If I use a 3K RPM range (HP peak +/- 1500rpm), then so does the next guy.  It might be higher (S2000) or lower (Mustang GT) absolute revs, but the operational part of the curve is roughly the same width.

Since we’re forced, due to the use of a conventional manual transmission, away from the HP peak, that means that the more HP you lose as you move away from the peak, the lower your engine’s potential.  Which means that a flatter HP curve near the HP peak is what we want.  It defines a better engine for these purposes.  It’s what we need to know.  So if we want to understand the impact of torque, we need to know how it affects the shape of the horsepower curve.

Now, remember back up there when I said that all we had to go on was the torque peak and the HP peak?  We don’t have a dyno plot for every car, so therefore, we don’t know how flat the HP curve is at its peak … we only know the peak value and the revs it occurs at.  Or do we?  We also know the torque peak and its revs, and using that, we can calculate the HP at those revs.  So, we actually have two points on the HP curve, not just one.  And with two points, we have a line, and with a line, we have a slope.  A smaller slope means a flatter line, and a steeper slope means a more vertical line.  So using the HP peak and the torque peak, we can calculate the slope of the HP curve between those two rev points.

So, I did just that for several cars that race in IT.  And guess what?  Surprisingly, the monster low-revving V8s have a significantly peakier HP curve near the peak than the high-revving gutless 4-cylinders.  And that goes against all of the conventional wisdom!  A 1990 Mustang GT 5.0 loses 27hp for every 500rpm on either side of the HP peak, but a Toyota Celica GT-S loses only 7!  Comparing two cars with the same stock peak HP (190), an 3.8L Ford Mustang V6 loses 15hp for every 500rpm, but a ‘96 Honda Prelude VTEC loses only 10.

Now, those numbers assume that the curve is the same shape on both sides of the peak, which isn’t usually true.   What is usually true is that the torque drops off more dramatically after the HP peak revs than it does before the HP peak revs, which means that the slope of the HP curve to the right of the HP peak is steeper than it is to the left.  But that doesn’t really matter.  It just means that rather than have our HP peak in the center of our operating range, it should be somewhere right of center.  No big deal.

The bottom line, my big conclusion, the “so what” of all of this, is that what we think we know about torque’s effect on lap times of a race car might not really be true.  And that leads me to believe that the ITAC’s process of accounting for torque (adding big weight to a 5L V8, and subtracting weight from a high-revving 4-cylinder/rotary) might be all wrong.

Summarizing my argument:

1. We try to maximize our time near the HP peak
2. We all have roughly the same gearing potential
3. The faster the HP curve drops on either side of the HP peak, the worse off we are
4. So, what we need to know to evaluate two different HP curves is the SLOPE of the HP curve near its peak
5. We know two points on the HP curve — its peak, and the HP at the torque peak
6. We can calculate the slope of the line between the two points
7. For two cars with the same power-to-weight ratio, a flatter slope ultimately has more potential

Now, given that applying this logic leads to a conclusion that disagrees with conventional wisdom, I figure that either the gross assumptions here don’t hold true, or that there’s something fundamental that I don’t understand.  It’s also possible that conventional wisdom is about drag racing and street driving, not about road courses.  Can someone set me straight and tell me where I’ve gone wrong (or validate this conclusion?)  I’m thinking that IF I’ve got this wrong, the problem is that there’s a bigger penalty than just more shifting for the short gear ratios that the high-revving cars have to use.

I’m looking forward to your comments!

This entry was posted on Thursday, May 7th, 2009 at 8:30 pm and is filed under IT.You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.