Torque versus Horsepower

[Anonymous]

The torque versus horsepower argument has been around for as long as there've been engines and continues to motivate and entertain gearheads to this day. Here's the question in another form: "Is torque or horsepower more important to acceleration and top speed?"

Answer: Yes! :roll:

As SM wrote in another thread, torque is a force, which can be directly measured. In fact, it is worth noting that nearly all dynos measure torque and then calculate horsepower using the following equation: Power=[torquexRPM]/5252. There may be a few dynos around that directly measure horsepower, but you and I will never strap our engines into them.

That's it then. Torque is the King. Long live the King. Now let's all go have a cold one.

But wait, you say! If torque is the end all and be all, then why do I keep hearing about the importance of horsepower? Simple, horsepower is what enables us to take advantage of gearing to multiply the effect of torque.

Let me illustrate with an example. My friend Dave has an '85 MR2 with the Toyota 4age engine making 112 hp at 6600 rpm and 97 lbs/ft at 4800 rpm, and has a redline of 7500 rpm. I have a Golf TDI making 90 hp at 3750 rpm and 155 lbs/ft at 1900 rpm with a redline of 4500 rpm. Each car has about 12:1 multiplication in 1st gear, but the TDI has only 2.5:1 in 5th while the Toyota has about 3.7:1.

From a dead stop it looks like the Golf is a sure winner, since it has 155x12=1860 lbs/ft of tire smoking torque. Meanwhile, Dave's MR2 is flailing like crazy trying to keep up with its mere 97x12=1164 lbs/ft at an ear-splitting 6600 rpm. Even with its 400 lbs lower weight (2450 vs 2850), Mr Two is well behind, since each foot pound of torque is trying to accelerate 2.1 lbs of car, while the Golf's torque is only moving 1.5 lbs.

So, I've jumped off to a quick lead in my race to top speed with Dave and am already starting to taste the beer he's going to buy as we row up through the gears. But then we get into 5th gear and something strange happens. My Golf is still making a peak torque of 155 lbs/ft, but that's only being multiplied by a miserable 2.5:1. Gulp...155x2.5=386 lbs/ft, versus Dave's 97x3.7=358 peak torque. Now the shoe is on the other foot, since Dave's lighter weight translates to a torque loading of 6.84 pounds, while the Golf is trying to accelerate 7.38 pounds per lb/ft available. "DANG!" I say as Dave goes zipping past with a big grin on his face. 8)

What went wrong? Simple, by revving to a higher rpm, Dave is able to multiply the torque he has available through steeper gears to achieve greater net acceleration and top speed. Even though Dave's torque available is lots lower than mine, he is ultimately able to out-run me.

Now let's examine another case -- my SCCA Formula Atlantic race car. When we started racing the car, the engine made 235 hp at 9100 rpm and 136 lbs/ft at 7600 rpm, redlining at 9800. [In the past I have incorrectly listed our torque as ~116, but I dug out the dyno sheets tonight.] We have continued to develop the engine over the years, and it now boasts 252 hp at 9600 rpm and ~140 lbs/ft at the same 7600 rpm, and can reliably rev to 10,500. From the equation above, we can see that it is delivering just 123 lbs/ft at redline (assuming 245 hp at 10,500).

So what does that mean? Well, at 7600 rpm the engine has about 1450 lbs/ft of torque-multiplied thrust available to accelerate a 1300 pound package through a pair of 15" wide rear meats. That's less than a pound of car per pound of available torque!

Furthermore, at 150 mph in 5th we only need to pull 14 mph per 1000 revs. That's only a little more than your road car pulls in 2nd gear. Even at redline in 5th gear with its 1.2:1 ratio and a 3.89 diff, we are multiplying the engine's 123 lbs/ft to 575 lbs/ft of torque. That's still less than 3 pounds of car for each pound of available torque.

Now you know why I like high revving engines.

PS - For those still reading, here are some numbers from a modern F1 car: 800 hp at 18,000 rpm. To solve for torque, use T=(HP*5252)/RPM. Assume 7th gear is about 6-7:1 total. Solve for available torque. Tire spin occurs at anything above about 1000 drive-wheel lbs/ft...

 

[jwhoss76]

Okay, I can't argue with you there. BUT in some situations Torque is king. Take towing for instance, a high revving engine will sream like crazy and will barely move the weight being towed. Now in a perfect world it would eventually be able to get the load up to speed but this world is far from perfect and hills and inclines will slow even the highest revving engine. But I don't expect a torque monster engine to run with the top fuel racers either. Once again it comes down to aplication and need. JMHO

 

[xctasy]

That's it! Take shaft torque, multiply it by the gear ratio, and check how much shaft torque. Then you can calculate potential acceleration. This works at 3 mph or 300 mph. That's how racers find ideal gear change rpm, and ideal gearing for drag racing, and ideal gearing for circuit racing. Hp is just a means of getting the calculations to work. Before Hp exists, torque was there.

I've sat down, and discovered that hp is just instaneous torque at a given rpm. In a quarter mile calculation you use the peak hp, but the rest of the calculation transforms this to a vehicles weight divided by shaft torque.

The rest is simple math.

 

[Anonymous]

That added low-end torque is only valuable below the high-hp-low-torque motors peak torque in first gear. The fact that it needs to spin is obvious and is a personal thing more than any actual hinderance.

I've always found this to be a good link for the hp/torque debate:
http://vettenet.org/torquehp.html


-=Whittey=-

 

[xctasy]

Amen. It's lovely. The guy simply explains things.

 

[Anonymous]

Im definitley no teckno expert but it seems to me that the recipricating motion and the weight of the internal parts like the crankshaft, flywheel and maybe the driveline play an important part.
If resistance were applied to a heavy rotating component it would not slow as quickly as a lighter one spinning at the same rpm with the same amount of power driving it.
If you were able to dyno the two at decreasing braked rpm the heavier component would show that it was producing more force and as a result more power. If this was done it the reverse and measured while accelerating I am not sure what the dyno would read up through the rpms (less for the heavier I think) but the lighter component would accelerate much quicker because it has less resistance to overcome.
No one measures power when a car is de-accelerating and it would only be practical you were trying to hold say for example high speed or a heavy load on a hill.
It seems to me that torque as I have described it (I have not included stroke and rod ratio ect) would be detramental to excelleration.
Remmember the comparison is with exactly the same power feeding both the heavier and the lighter. My conclusion is that in this example the "actaul power" is lost somewhere in the torque figures.

 

[Anonymous]

The torque advantage of the heavy parts, like a flywheel, only have an advantage when leaving a dead stop. As soon as the clutch is fully engaged there is no further advantage.

John

 

[Anonymous]

It been an old secret for some time that if you want to get a vehicle to accelerate quicker lighten the flywheel. I know someone who did this to a Datsun 1200 with dramatic results.
My machinist recommended that I do the same same to a Corrolla I once had him re-build if I wanted quicker acceleration.
Only reason I didnt do it to my 250 2V I had was that I had two 250 2V blocks and one of the flywheels was already 50% lighter than the other one (neither had material removed).
It does have a down side in that it does reduce the ability to hold rpm. Not that I noticed with all the torque that a 250 has anyway..

 

[Anonymous]

Correction about one third lighter

 

[Anonymous]

I should have made it more plain that I was talking about racing engines, and I apologise for any confusion my post may have caused in that regard.

Dmgdgoods76 raises an interesting point about low speed torque engines, though. Engines for tractors and tow vehicles emphasize low speed torque at the expense of horsepower. The 460 in my '76 E-250 van is a good example. IIRC, it is rated at 370 lbs/ft at 2000 rpm, but just 220 hp at 3700 rpm. In top gear (1:1 ratio) and a 3.73:1 rear end, it's putting down 1380 lbs/ft to the rear tires at peak torque (ignoring losses). To get 1380 pounds out of my Toyota engine at its peak torque of 140 lbs/ft at 7600 rpm, I'd need a rear end ratio of 9.85:1...!

OTOH, at that rpm the 460 is only making (370*2000)/5252=141 hp, and burns approximately .45lb*hp=10.6gal/hr. Meanwhile, the Toyota is making (140*7600)/5252=203 hp at the same torque output, and is burning 15.2 gal/hr -- 43% greater fuel consumption than the 460...!

Sooooo, now you know the real reason tractors and tow vehicles use large displacement, slow turning engines -- they are more fuel efficient for towing than their smaller counterparts.

 

[xctasy]

I did some calculations last weekend. A car off the line to 30 mph will often yield an actual accelration figure of 5 times the amount it can based on just the shaft torque released from 0- 30 mph. Take a Falcon 4.1 Cross flow EFI like Jacks, and workout the potential acceleration from rest to the 30 mph. The axle shafts may wheels 'see' more than 2365 lb-ft with an 84 Falcons 9.61:1 overall first gear, from an engine which produces 246 lb-ft at 2800 rpm. Torque at 4500 rpm is 170 lb-ft.

First gear change is at 34 mph at 4500 rpm. The potential acceleration figure is something like 0.07 g's calculated, but the actual g forces are 0.36g's.

Off the line, a heavy flywheel in a manual helps 60 foot times, but is a disadvantage for circuit racing. NASCAR racers rabbit on about heavy flywheels too.

 

[Anonymous]

Something doesn't add up in your example, Ex. If the torque multiplication sends 2365 lbs/ft to the rears, the car should see nearly 1-g initial acceleration from 0 (assuming it's launching at max torque). After all, doesn't the car weigh about 2500 lbs? If your actual G's are only .36, then it sounds to me more like you have a traction problem.

 

[Anonymous]

Yah, but while you're lugging around in 4th gear, the Yoda is rolling around in second which will destroy that 460 in ground-measured torque.

Speaking of fuel economy, ever notice that the little engines are always called economy engines and the big ones are supposedly powerful? Do the math on economy/displacement and power/displacement. The big engines are good for economy and the little ones are good at power. I think thats more designed in rather than due to disaplcement though. After all, not too many would be buying a 6mpg car for a daily driver, though some might buy 30hp hundred mpg vehicles...


-=Whittey=-

 

[Anonymous]

Execute you say a heavier flywheel helps 60ft times off the mark.
I probably agree, although it goes against what I said earlier. The reason for this is too light a fly wheel will also cause the engine to bog down (depending on vehicle) when the clutch is engaged and can make it stall. Once it is engaged it will accelerate quicker.
It is my understanding that the experts when testing manual cars hold the rmp at the torque peak (say 3500 rpm) and engage the clutch while keeping it at around thos revs and attempt not to lose traction (they dont dump it). Apparently this takes a lot of practicse to master and can seriously effect 0-60 mph times in a manual car. In this case a lighter fly wheel would not help at all as it would make it difficult to hold the rpm while engaging the clutch.
For all practical purposes I would not do this as it would be hard on a clutch unless I was very good at it. When I take off quick I prefer to engage the clutch at much lower rpm and then accelerate hard. In which case the lighter flywheel would have the advantage (so long as not to light)
Its all relative stuff with lots of variables. When I looked into having my flywheel machined down the machisist told me that 250s already have a light flywheel similar in weight to a Datsun 1200 flywheel (actually if I recall he said the 1200 flywheel was heavier)
Back to my old point the weight of the recipricating parts does (must) have a big effect on torque. Although it may not have all that much practical value in determining how much effect it has with other factors involved as well. I bet its must play an important part when designing trucks as they develope massive torque with comparitevly small HP.

 

[xctasy]

The axle shafts may (DELETE :arrow: wheels) 'see' more than 2365 lb-ft with an 84 Falcons 9.61:1 overall first gear, from an engine which produces 246 lb-ft at 2800 rpm. Torque at 4500 rpm is 170 lb-ft.

The movement of a manual EFI six cylinder Falcon from 0 to 31 mph (50 km/h...bl**dy metric car tests) takes a 3000 rpm clutch dump at the start, on up to 4100 rpm at 31 mph. The Lietz Correvet triggers at 3 seconds flat. The car input shaft to the trans 'sees' between 246 and 187 lb-ft at these rpm.


Wheel trust is (torque * O/all gearing*0.85) divided by rolling radius in feet. The axle shafts in a 4.1 Falcon 'see' 246*2.77*3.47*0.85. Hmmm, that's more like 2009 lbs.(Not 2365 exactly, but close. I forgot the drive train loss) Then this gets depleted by the feet of roll out (which is 6.43728 feet for a ER70H14 tire). Net result is actually 312.2.

G-force is wheel thrust divided by car weight, so this is 312.2/3763 lbs, which is 0.0083 g's

The potential accleration is g force *32.174 ft/sec/sec, which is 2.67 ft/sec.


The actual measured acceleration is 15.19 ft/sec, with 0-31 mph in just 3 seconds. Or 0.472 g's, man! Not 0.0083! There is 5.69 times more wheel trust producing vehicle accleration than the formula predicts. Thats because of the "flywheel" affect that produces 'sling shot factors' of between 4 and, in some cases, 7 times the predicted acceleration. Cool huh?

Do the calcs. Cars accelerating cheat the laws of physics because of the momentum of the spining engine. Where else do you get five times the g's?

Whittey, you are right about the fuel consumption. An ohv GM 3800 v-6, based on a hacked up 215 Buick V8, is more economical than a 3.0 liter Nissan ohc, and has better Brake Specific Fuel Consumption (BSCF), lower engine weight, and more power and torque eveywhere. There's something that Americans know that the Japanese still haven't cottoned on to.

There is, incidently, a cool way of determining on the road fuel consumption figures from BSFC figures, given the tire type, gear ratios, drag factors and vehicle weight. David Vizard did a cut on it way back in 1979 in a Car Craft Magazine. It's very easy to run play off two separate cars with no need to put them in a chassis. All you need is the BSCF figures. Neat!

 

[Anonymous]

Execute
If they dump the clutch at 3000 rpm wouldnt it just bag em up at the rear. I would have thought it would be faster to use the clutch a bit more progressively until it is engaged while still doing 3000 and keep the wheels just on the edge of adheasion. I believe this is supposed to be the best way to do it?

 

[Anonymous]

Execute, all those figures you quote are starting to scare me. I think I better get out of this section and go somewhere where I belong
Cheers Tim

 

[xctasy]

Like my work mates says, do the theory, take it to the extreme, then through the manual over your shoulders and use you own common sense. Too much study can wear you out!

The rest of my posts will skip the figures. I have to have some other devious algeraic toys to play with at home...my Stage 4 Math paper

Tim: ooops. I assumed they did a partial clutch dump. I duuno. I don't have a manual

The math is just to back up my case. Stole it from a refereance manual, and then just used real world Falcon EFI's, sucker 4AGE Corollas, an E49 Charger R/T, and a couple of ancient 164E and 144 carby ovloV's. That should take car of the bullsh*t factor.

 

[Anonymous]

Have you heard of that new Corolla Sportivo thing with a huge 141 kw and about 200 nm of torque out of a 1.8 litre engine. It doesnt accellerated fast for its big power and with it all right at the end of its big rev range. My 250 2V Cortina would gobble it up and I wouldnt have to rev it more than 4000 rpm. Ha ha.
Cheers Tim

 

[Anonymous]

A flywheel is like a battery in that it stores energy. In the case of a flywheel, that energy is in the form of inertia. A big 'ol heavy flywheel, once moving, has more energy (for the same RPM) than a tiny light flywheel. When you let out the clutch, you slow the flywheel down, which transfers its energy into heat (from the slipping clutch) and into movement. An engine with a heavy flywheel will feel more torquey when you let out the clutch, but it will climb the revs slower (you have to put that energy into the flywheel). Soooo, assuming you can keep traction, a larger flywheel will help in the short times but hurt at the big end (though powershifting may help a touch more with a heavy flywheel).


-=Whittey=-