Revisiting Torque

[TomL]

Josh -

This torque misunderstanding has bugged me for a long time. I had been thinking that, over the winter dead time, I would start a flame war by publishing a rant about this whole topic (You need to keep everybody excited when you can't race!) If it would have helped you, I'm sorry I didn't get around to it.

Your analysis using the only two numbers available - HP at peak and torque (and indirectly HP) at peak sort of confirms my point (wish I'd thought of that!) I would quibble with a couple of things. You say that the transmission ratio gaps are about the same for all cars, so that the number of shifts is about the same for everyone. But then you turn around and say that the high reving (shorter-geared) car increases its revs faster and thus has to shift more. Obviously, it can't be both. The first statement is the correct one. The reason the second statement is erroneous points to the main shortcoming in you analysis. The short-geared/high reving car increases engine speed faster, but not car speed. If one car shifts at 6000 with a 3.40 final drive and another shifts at 9000rpm car using a 5.10 final drive, they'll be shifting at exactly the same car speed (trans gears and tires diameter being equal).

Similarly, when you look at the power band, you need to consider the effective rev range not as a specific number of rpm (3000 in your example), but as a percentage of your maximum rpm. In other words, because of the differences in revs at which cars operate, the high reving car needs to have a broader hp range, at least as expressed in rpm. As in the example above, if your 9000 rpm car needs a 3000 rpm rev band, the 6000 rpm car needs only a 2000 rev band because of the longer gearing.

Since your analysis is based on absolute rpm instead of a percentage of the peak, it makes the low reving cars look a little worst than they actually are. A further and related problem is that the low reving cars tend to have a greater rpm variation between hp peak and torque peak, so your slope (hp/ 500rpm) is based on a broader range of absolute rpm than for the high reving cars. And if you again account for rpm as a percentage of peak this problem is exacerbated. Bottom line - the way your analysis is done makes the low rev cars look considerably worse than they actually are. If you could account for both these factors, I suspect you'd find that the power bands are fairly similar for all the cars.

That gets me back again to my original point - horsepower is what counts. I've heard all those torque mottos before. Just remember mine - "You can multiply torque. You can't multiply horsepower." (Not that it's original with me - I think it's Paul van Valkenburgh).

[JoshS]

A further and related problem is that the low reving cars tend to have a greater rpm variation between hp peak and torque peak, so your slope (hp/ 500rpm) is based on a broader range of absolute rpm than for the high reving cars.

Thanks Tom! Glad we agree on the conclusion, even if I missed a couple of points. I need to really digest the rest of what you said, but let me just quickly address this point: it's not true, at least not for the sample of cars I chose for my analysis. Here are the rev difference between the peaks for those cars:

Integra Type R: 700
SN95 Mustang: 800
Celica GT-S: 800
S2000: 800
Fox Mustang: 1000
Camaro 305: 1200
Prelude VTEC: 1500
300ZX: 1600
Z3 2.8: 2000
968: 2100
Mustang V6: 2500
RX-8: 3000

So we have both high-and-low revving cars at the top of the list (ITR & Mustang), and high- and low-revving cars at the bottom (Mustang V6, RX-8).

No clear correlation.

[JoshS]

Similarly, when you look at the power band, you need to consider the effective rev range not as a specific number of rpm (3000 in your example), but as a percentage of your maximum rpm. In other words, because of the differences in revs at which cars operate, the high reving car needs to have a broader hp range, at least as expressed in rpm. As in the example above, if your 9000 rpm car needs a 3000 rpm rev band, the 6000 rpm car needs only a 2000 rev band because of the longer gearing.

Good point. I replaced "500" with the revs at the HP peak, and things do get a lot more equal. I'm actually glad about that, since my previous analysis had actually inverted my expected results. Now it's really just showing that it makes no difference at all.

When I get a moment I'll update my dissertation with that point.

[GKR_17]

Where were you guys a couple of months ago? :-) .

Uh... I did... several times. Here just one: July 30, 2008

Maybe I should stop posting this because no one seems to listen...

A single torque value is not a good indication of the potential of an engine. The power transmitted to the ground at that instantaneous engine speed is what makes the car go. Horsepower is basically torque multiplied by RPM. Many of the cars with low torque are also turning higher revs, so they're still putting good power to the pavement.

What is really important is the full dyno plot combined with the transmission ratios,

[pfcs]

pfcs Online!
Senior Member
Join Date: Oct 2003
Location: long valley, NJ
Posts: 147



My physics courses were not yesterday, I don't have all the formula's memorized but the basics of what I am telling you are true. Fact not opinion... please do some research before you try to tell me I am wrong again.
__________________

I took physics a long time ago, but force=mass x accelleration (F=MA) still applies.
Or A= F/M. The greater the force, the greater the acceleration. In any gear, the force is maximized at the torque peak of the engine. If you shift up, the gearing multiplies the torque out of the trans, but the engine torque is lower, and the derived torque out of the transmission is less than the torque at the same output rpm (vehicle velocity). You can make it clear by thinking what happens to rear wheel torque when you shift to a lower gear-look at the torque curve and do the math: you're running a fixed ratio with a variable torque as you move along the torque curve, where will the torque at the wheels be max? At the engine torque peak. I know it's somewhat counter-intuitive, but it's the hard truth and simple physics. This does not mean we want to lug the engine, just that when you're at peak torque in any gear, the vehicle will have it's max accelleration at that velocity-if you're revving a gear past the point where the next gear would bring you to the torque peak, you're leaving accelleration on the table.
__________________
phil hunt

Above from 2007

TomL Offline
Junior Member
Join Date: Jul 2004
Location: Decatur , GA, USA
Posts: 20



Kirk, thank you for posting this question. I continue to be baffled by the importance a lot of people seem to give to the torque number. The rate of acceleration of a vehicle at any given speed is strictly a function of the available horsepower at that speed. So, in the absence of any other information, any choice other car other than number 3 (the 190 hp version) is going to lose.

Tom-its strictly a function of available (crankshaft) torque at that rpm. Period. However, horsepower is a better indicator of performance. This discussion is attempting to discern the possibility that looking at the shape of (area under the curve) the torque curve is germane to predicting the relative competetiveness of a given vehicle.

[TomL]

You're right, Josh. As soon as I posted the comment about HP/torque peak RPM variation being higher in lower rev engines, I looked back and realized that it wasn't much of a correlation at all. I think I got focused in on the S2000 and V6 Mustang and didn't do a proper analysis. Sorry.

However, I'm glad to see my point still bears out when RPM weighted. BTW, when I looked at the RX8s 3000 rpm variation, I back-calculated the the torque at 8500 and found that even at 8500, it still had 149 lbft, only 10 less than at 5500. What a broad torque band!! Ought to dominate ITR! (Steve Eckrich, forgive me.)

One other thought. In my discussion about high reving cars needing a wider power band, I was assuming everyone had about the same trans ratios. After looking further, I realized that in reality that's not always the case. If a particular car has noticeably closer ratios, then the needed rev band is also much smaller. As GKR noted earlier, that keeps you closer to the peak hp. In some cases (e.g., the S2000), the ratios are so much closer that the high rev car probably doesn't need to have a wider power band in absolute rpm than another low rev car with more typical trans ratios.

Phil Hunt -

I agree with most of what you said. But I'm not sure I followed all of your discussion. And I think you're still missing my point. I'll try again tomorrow when I feel like doing a couple charts to explain what I'm saying.

Knestis' original question related to using the torque number as a factor in setting IT weights, not anything about torque curve (and rightly so, since the torque curve info is generally not available.) One of my main points was that the torque number doesn't really tell you anything about the car's performance that you don't already know from the HP.

So I can better understand where you're coming from, are you saying that 160hp, 200 lbft car is going to accelerate faster than a 160hp, 140 lbft car? And if so, could you explain a little more why you think that's the case?

[JoshS]

Uh... I did... several times. Here just one: July 30, 2008

Sorry man, maybe I just wasn't ready to listen yet!

[Bill Miller]

Kirk,

I applaud what you're doing here, but honestly, you lost me when you said a 150/150 car would weigh the same as a 170/170 car.

I'm pretty much in agreement w/ Phil, and have advocated for a long time, that the area under the power curve is much more indicative of performance than some peak value.

[rx7chris]

Torque makes all the difference in the world. HP is just a figure derived from a torque reading. For example, gonna use motorcycles cause i can speak from experience. Say Bike 1 makes 52lbs of torque and 120HP Bike 2 makes 47lbs of torque and 130hp. Bike 1 and 2 are otherwise equal. Both bikes come out of a turn at the exact same time, Bike 1 will win the race to the next turn given that it's not at such a distance that the added hp of bike 2 allows it to catch up. Say if this turn is the bus-stop at daytona. By the time both bikes reach start finish they should be about equal again. If the turn was the carousel at sebring bike 1 would have a definitive lead by the time they reach the short-course start finish. Each side has it's pro's and con's and you can have a different advantage depending on track lay-out.

Torque can also be massaged with dyno time as well.

Don't try to measure it, you'll go insane with 300+ cars at an event.

[Andy Bettencourt]

Torque makes all the difference in the world. HP is just a figure derived from a torque reading. For example, gonna use motorcycles cause i can speak from experience. Say Bike 1 makes 52lbs of torque and 120HP Bike 2 makes 47lbs of torque and 130hp. Bike 1 and 2 are otherwise equal. Both bikes come out of a turn at the exact same time, Bike 1 will win the race to the next turn given that it's not at such a distance that the added hp of bike 2 allows it to catch up. Say if this turn is the bus-stop at daytona. By the time both bikes reach start finish they should be about equal again. If the turn was the carousel at sebring bike 1 would have a definitive lead by the time they reach the short-course start finish. Each side has it's pro's and con's and you can have a different advantage depending on track lay-out.

Torque can also be massaged with dyno time as well.

Don't try to measure it, you'll go insane with 300+ cars at an event.

But in IT things aren't otherwise equal. You can run a R&P that helps your problems and exploits your strengths.

[pfcs]

"You're right, Josh. As soon as I posted the comment about HP/torque peak RPM variation being higher in lower rev engines, I looked back and realized that it wasn't much of a correlation at all. I think I got focused in on the S2000 and V6 Mustang and didn't do a proper analysis. Sorry."
It's more illustrative to look at this by % of rpm difference than rpms difference. And the relative numbers (ft/lbs and hp). They reveal a whole lot about the area under the curve and the ballistics of said engine. Engines with more torque than hp will have the hp peak nearer the torque peak and are stones, running out of torque quickly, they generate little hp because hp is a product (in the math sense) of torque and rpm. There's little area under this curve.
Imagine other scenarios and you can imagine the neccessary tq curves and the implications.
AB: "But in IT things aren't otherwise equal. You can run a R&P that helps your problems and exploits your strengths."
And if the %difference between peak tq and hp is less than the %difference between 2 transmission gears you use, you're screwing the pooch-when you're out of torque, you're better drinking beer.

[JoshS]

Engines with more torque than hp will have the hp peak nearer the torque peak and are stones, running out of torque quickly ...

There's lots of conventional wisdom like this, but IT JUST ISN'T TRUE, at least for the fairly random sampling of cars I ran. There seems to be very little correlation between these factors.

[JoshS]

Torque makes all the difference in the world. HP is just a figure derived from a torque reading. For example, gonna use motorcycles cause i can speak from experience. Say Bike 1 makes 52lbs of torque and 120HP Bike 2 makes 47lbs of torque and 130hp. Bike 1 and 2 are otherwise equal. Both bikes come out of a turn at the exact same time, Bike 1 will win the race to the next turn given that it's not at such a distance that the added hp of bike 2 allows it to catch up.

Please prove it with math.

[Knestis]

Kirk,

I applaud what you're doing here, but honestly, you lost me when you said a 150/150 car would weigh the same as a 170/170 car.

I'm pretty much in agreement w/ Phil, and have advocated for a long time, that the area under the power curve is much more indicative of performance than some peak value.

I'm pretty sure that I didn't say that. What I suggested was that both should have the same "torque adjustment." The weight will ALWAYS be based on the power.

K

[lateapex911]

Found it.
Read this:
http://www.yawpower.com/tqvshp.html

[pfcs]

The point I keep trying to make is so easy to miss, and I'll bet people like Bill O'Reilly wouldn't admit they got it if they did!
Horsepower is a FUNCTION of TORQUE and VELOCITY (rpm in a crankshaft). To make a lot of horsepower, you need to have good torque at higher rpm. (Xtorque@4000rpm =Yhp; Xtorque@8000rpm=2Yhp). Understanding their interelatedness, and how their relative amounts and %rpm-spreads describe the area under the curve in question tell a lot about the dynamics of the engine and its likely response to tuning. (an airflow restricted engine [high tq/lowhp/small rpm-spread] is unlikely to respond much to IT tuning unless the restriction is in the exhaust system). Yes, you need a lot of horsepower to go fast but hp is still determined by torque AND-torque is what accellerates the car, even at 10,000rpm, hp simply describes how well it makes that torque at high rpm. Get it? You need to know BOTH to describe the dynamics of an engine. If you only know one parameter, hp is more informative; if you know both (tq@rpm/hp@rpm), you have much clearer info.
On another front-I actually heard Rush Limbaugh suggest that what killed Detroit was trial lawyers. Turn you r radios off please.

[quadzjr]

Kirk,

I'm pretty much in agreement w/ Phil, and have advocated for a long time, that the area under the power curve is much more indicative of performance than some peak value.

I agree as well, that makes 100% sense. however implementing would be a nightmare. Anybody have a factory issued dyno plot for an opel GT?

[JoshS]

I agree as well, that makes 100% sense. however implementing would be a nightmare. Anybody have a factory issued dyno plot for an opel GT?

Yes, area under the operational part of the curve is the right answer. In other words, a flat power curve near the peak is better than a peaky curve, because we base the vehicle weight on the peak. The closer you can stay to the max HP, the better. And I claim that you can't derive anything about the shape of the HP curve from the any published vehicle specs. That's the ITAC's problem.

[Bill Miller]

I'm pretty sure that I didn't say that. What I suggested was that both should have the same "torque adjustment." The weight will ALWAYS be based on the power.

K

Assume that the minimum weights are the same for the purposes of this discussion. Since we're talking about figuring out how to consider torque (or if we should), don't presume that given whatever the final process might be, the result would be that they are different.

Maybe I misunderstood what you meant by the above comment.

[Knestis]

Ah, gotcha. That wasn't an argument - it was an exercise to try to suss out how we think about the value of these things we think of as 'power' and 'torque.' Sorry for the confusion.

K

PS - Good to have you back in the fold, Bill...!