Peak power vs Average Power (Engineers are welcome to this post)

"traderQAMobileTestAutomationMobileBoostOn" (el-peasant) 06/30/2016 at 10:09 • Filed to: None

1

36

While browsing your local CL, you see 2 Hupmobiles for sale - one with 500hp V36 and one with a 400hp V36. I can read your mind to see that you would take the 500hp car.

What you didn’t know, is that when both engines were placed on a dyno, their power curves looked like this:

Yeah, the orange Hup makes a maximum of 500 hp at the top of the tach, but it’s a complete dog until then. Meanwhile, the 400hp Hup has 400hp almost everywhere.

If you pick the orange Hup to have a better chance of smoking a Subaru Justy at every stoplight, prepare to be disappointed.

But every engine must be paired with a suitable transmission. To make the most of its short-lived >450hp, the proper transmission for the orange hup would need very close ratios to keep the engine in that 5500-6000 range. Even then, it would still be sluggish off the line. Meanwhile, the orange Hup would be well off with wider ratios.

Then there’s the factors of torque, and gearing...

To the engineers: Is there anything that I missed? Specifically, what’s the role of torque here? Besides that it combines with RPM to make HP.

SidewaysOnDirt still misses Bowie > traderQAMobileTestAutomationMobileBoostOn 06/30/2016 at 10:16

3

Oh come on.... You’ll never smoke a Justy. Sounds like the orange one needs a CVT, the greatest of all transmissions.

MasterMario - Keeper of the V8s > traderQAMobileTestAutomationMobileBoostOn 06/30/2016 at 10:16

2

Actually torque is not another factor. You could calculate out the torque given HP and RPM. But, what you have here is a good understanding of basic engine theory. Also your power curves are completely ridiculous, but they are that way to make a point.

RamblinRover Luxury-Yacht > traderQAMobileTestAutomationMobileBoostOn 06/30/2016 at 10:17

7

Those curves would make more sense as torque curves rather than horsepower curves. An engine maintaining a flat *horsepower* curve would be losing horsepower at a linear rate with increased RPM, offsetting the increase in horsepower (linear at fixed torque) coming with increased RPM. In other words, that engine would be the worst breathing engine known to man, but with better heads could probably quadruple the horsepower. But yeah, I think you meant to write “torque” on your graph, calling out the peak torque as the point of peak horsepower (or more correctly almost immediately *after* peak torque).

Ash78, voting early and often > traderQAMobileTestAutomationMobileBoostOn 06/30/2016 at 10:21

1

Those graphs are messing with my head. Can you please rename the orange one "Every Honda Ever" so it's more applicable to real life?

Berang > traderQAMobileTestAutomationMobileBoostOn 06/30/2016 at 10:23

2

I like that the 400hp engine has almost infinite torque at idle.

MasterMario - Keeper of the V8s > RamblinRover Luxury-Yacht 06/30/2016 at 10:24

0

That’s not quite right. HP is flat throughout so it’s not losing HP, though torque would be decreasing throughout. Peak torque does not equal peak power. Engines are developed for completely different reasons with completely different curves. An engine built for heavy machinery will have peak torque as close to idle as possible with peak HP towards the rev limit, this gives a flat power curve where you have a similar amount of power available through out the RPM range. Where-as sports cars generally have peak power and peak torque closer to each other.

traderQAMobileTestAutomationMobileBoostOn > RamblinRover Luxury-Yacht 06/30/2016 at 10:25

0

This is hypothetical, I just used extremities to better display my point.

traderQAMobileTestAutomationMobileBoostOn > MasterMario - Keeper of the V8s 06/30/2016 at 10:27

0

Is torque good for anything then? In performance applications.

Berang > traderQAMobileTestAutomationMobileBoostOn 06/30/2016 at 10:33

2

Well, to get 400hp at 1000rpm, you’d need over 2000ft. lbs. of torque at same rpm. So it is good.

MasterMario - Keeper of the V8s > traderQAMobileTestAutomationMobileBoostOn 06/30/2016 at 10:34

1

Torque and RPM are what you use to calculate your HP. Generally, when you flatten a power curve your peak HP will go down, that’s why you see things like farm and construction equipment with less than 200HP. For things like sports cars where you need as much power as you can get you push the peak HP up but sacrifice useable power over the RPM range, that’s why you see transmissions with so many gears now, engines can be more efficient when they only operate in a narrow RPM range. The ideal engine/transmission combination would be an engine that only has to operate at a set RPM and an continuously variable transmission handles all the speed changes. I think the problem right now is making a true continuously variable transmission that can handle the abuse of racing.

RamblinRover Luxury-Yacht > MasterMario - Keeper of the V8s 06/30/2016 at 10:35

1

Try reading what I wrote again. I was saying that an engine with constant power is wildly implausible because power is torque times revs, and there would have to be a linear power loss to keep overall power steady. That linear power loss would take place in the form of a linear loss in final torque, which is so obvious I cannot believe you would think I was saying something else. Linear power loss, linear torque loss, all part and parcel of the same thing. You do not have to explain any of this to me.

Second, I mentioned that the “peaky” power curve would have peak horsepower near that peak. That is only because it is so very peaky. Obviously. In fact, I clarified my statement with a mention that the peak power is right after the torque peak and not actually on it.

RamblinRover Luxury-Yacht > traderQAMobileTestAutomationMobileBoostOn 06/30/2016 at 10:36

0

Well, it’s a perfectly okay hypothetical, but those curves make more sense as torque curves rather than horsepower ones. That’s all.

MasterMario - Keeper of the V8s > RamblinRover Luxury-Yacht 06/30/2016 at 10:41

0

I thought you were talking generally and not specifically about the curves shown. They are wildly implausible. Though I’m still not quite understanding what you mean by “ there would have to be a linear power loss to keep overall power steady” ...power loss=steady power? Power is power, torque does not equal power. If you meant torque loss=steady power then yes, that’s right.

Aaron M - MasoFiST > traderQAMobileTestAutomationMobileBoostOn 06/30/2016 at 10:44

0

This is basically a hyperbolic explanation of the notion of “area under the curve”, generating more power more often makes the car faster than having peak numbers. It’s also why most cheap aftermarket exhausts can honestly advertise “+3 hp” while still making the car slower.

RamblinRover Luxury-Yacht > MasterMario - Keeper of the V8s 06/30/2016 at 10:57

0

I thought it was obvious I was talking about his example curves - however, I failed to say “torque curve 2" when I meant to, which would have made it plain. I thought I had. As to what I’m saying, think about what causes a torque loss with increase of RPM. Any sort of linear velocity related loss (say, fluid viscosity related) takes place in the form of energy per time. This is power. I am talking about the engine as a system which has a sum total power output after all factors are considered - it is entirely reasonable to talk about a linear power loss within the system.

Within that system , there must be a linearly increasing power loss to keep the sum power output from that system to a steady amount. It can be described as a linear decrease in torque per increase in angular velocity in terms of what you get from the system, but this is a simplification. Engines do not typically behave this way unless they are experiencing *losses* within their system to one or more factors, and the power of those losses can often be in the form of heat. Or, in the case of flow issues, the effective displacement of the engine starts going down, which is more complicated. It is useful to talk in terms of power losses within the system because that’s how you account for things - discussing things in terms of torque loss within the system is nearly useless.

I never said torque is equal to power. I said “part and parcel of the same thing” in regard to power loss vs. torque loss. A question of how they are accounted for - whether in terms of a final “torque loss” or in terms of a base amount of power the engine *would* possibly produce minus an increasing power loss to one or more factors. I thought I had made that very plain.

-this space for rent- > traderQAMobileTestAutomationMobileBoostOn 06/30/2016 at 10:57

0

Area under the curve is more important than peaks.

Sir_Stig: and toxic masculinity ruins the party again. > SidewaysOnDirt still misses Bowie 06/30/2016 at 10:58

0

Exactly, with a cvt all that is holding you back is how fast that engine can rev to peak power!

RamblinRover Luxury-Yacht > traderQAMobileTestAutomationMobileBoostOn 06/30/2016 at 11:03

0

Having a wide torque band is very useful because it means you don’t have to stay exactly within the range at which you develop useful torque when upshifting. If you upshift and that requires the engine to drop into low RPM, and it’s VTEC, yo, and has very little torque at that RPM, it will hurt your acceleration, regardless what the VTEC, yo, can wring out of the engine higher up. The wider the torque curve, the more forgiving the driving, because at the end of the day the torque put to the driveshaft (a multiple of some kind of engine torque) is an expression of the magnitude of force pushing you forward.

just-a-scratch > RamblinRover Luxury-Yacht 06/30/2016 at 11:03

2

An idealized DC motor (constant efficiency, linearly decreasing torque, constant voltage input, current proportional to power) would be close to the blue plot. I've never seen an IC engine that will do anything like that as a max power vs rpm curve.

SidewaysOnDirt still misses Bowie > Sir_Stig: and toxic masculinity ruins the party again. 06/30/2016 at 11:06

0

Throw a manual clutch between and you can keep it peaked and then it’s just about how well you can release the clutch.

MasterMario - Keeper of the V8s > RamblinRover Luxury-Yacht 06/30/2016 at 11:12

0

So basically you have been talking about overall thermodynamic power where you are taking losses due to friction, fluid viscosity, etc. and here I am just talking theory where Power=Torque x RPM

RamblinRover Luxury-Yacht > MasterMario - Keeper of the V8s 06/30/2016 at 11:17

1

Yep. Talking past each other. You were talking total outside the system, and I was talking about inside the system - where Power = torque X RPM also, but the native explanation to the engine designer is not “torque is decreasing” so much as “something is sapping my ideal power!” I apologize if I was a little salty, but your responses seemed to indicate you thought I was somehow talking about more advanced things than I actually understood - a sort of engineering homunculus. “You don’t know what you’re talking about” is about the only rage button I have left, though I try to keep it better in check.

ChooChooMotherFudger > traderQAMobileTestAutomationMobileBoostOn 06/30/2016 at 11:25

0

I think you are pretty on the money.

As for torque. Torque is just horsepower dived by the speed of the engine.
So if you used torque instead of power for this whole article with those graphs it would be the same point.

just-a-scratch > traderQAMobileTestAutomationMobileBoostOn 06/30/2016 at 11:26

1

Speaking strictly hypothetically, (see RamblinRover’s post) the superior option is almost always the blue plot. It’s very difficult to control a vehicle to stay in that narrow ~5000-6000 rpm range when called upon. A really great (and as yet non-existant) CVT might serve the purpose.

By conventional means, the orange one is nearly non-driveable.

There is a bit of debate out there as to whether area under the curve or max power is a better measure of performance. I say usually the peak power figure is more useful because the shape of the curves is comparable. After that, you have to start getting into what type of use the thing is going to see and a zillion little factors. That’s all a bit more than we can reasonably cover here.

Trust me, I’m an engineer.*



*Do not trust anyone just because they claim a credential. I’ve seen too much garbage engineering to be so trusting. I’ve even been the one making stupid mistakes on occasion to produce the afore mentioned garbage.

MasterMario - Keeper of the V8s > RamblinRover Luxury-Yacht 06/30/2016 at 11:51

0

No worries. I just wasn’t wrapping my head around what you were talking about thinking only about the equation Power=Torque x RPM

HammerheadFistpunch > traderQAMobileTestAutomationMobileBoostOn 06/30/2016 at 11:53

0

hey I wrote a thing about this once!

http://oppositelock.kinja.com/power-under-th…

it was even fp’d though it was reframed so the FP author got all the views and comments...

Sir_Stig: and toxic masculinity ruins the party again. > traderQAMobileTestAutomationMobileBoostOn 06/30/2016 at 12:07

0

Step 1: put 500HP engine in car

Step 2: install CVT in car

Step 3: ???

Step 4: 6000RPM EVERYWHERE!

I understand what you are getting at, but those engine values are wayyy outside what you would find in an ICE car.

If only EssExTee could be so grossly incandescent > traderQAMobileTestAutomationMobileBoostOn 06/30/2016 at 12:17

0

Torque and HP are very closely related to Bore and Stroke. It’s a balanced equation. Increasing stroke gives you a longer power stroke and thus more torque but you lose RPM and thus HP. Thats why truck and large equipment engines redline at like 3000rpm and have 300hp but make like 1000 torques. On the flip side, a huge bore and short stroke will let you rev to the moon but have no torque. That’s why exotics like Ferraris will have a torque number that’s like half of their hp number.

DynamicWeight > traderQAMobileTestAutomationMobileBoostOn 06/30/2016 at 12:19

0

Torque is power. Horsepower is a silly made up marketing concept. I really don’t care how fast my engine is spinning. All I care about is how much power is getting to the wheels, and that’s torque. Always remember, horsepower isn’t just Torque X RPMS, it’s Torque X RPMs / 5252. It is only used because it makes it easier to compare real world engines. It isn’t a “real” measurement in my mind.

And yeah, RamblinRover is right. The blue horsepower curve you drew is losing power as the RPMs increase.

Sir_Stig: and toxic masculinity ruins the party again. > just-a-scratch 06/30/2016 at 12:19

0

Engineering is like all science, it’s not a good design until someone else does the math.

just-a-scratch > Berang 06/30/2016 at 13:45

0

and it idles at, or near, 0 rpm.

just-a-scratch > traderQAMobileTestAutomationMobileBoostOn 06/30/2016 at 13:48

0

I’m starting to feel bad that the commenters are being critical of the question here. It’s an interesting subject to bring up. Most likely, there are a lot of people on the forum here that have similar curiosity. So, I hope the criticism is taken in the best possible way, not personally.

I suppose this is what happens when you ask for technical minded people to respond. Such responses require critical thinking.

LoudmouthOne > traderQAMobileTestAutomationMobileBoostOn 06/30/2016 at 15:13

0

100% they should start publishing HP in area under the curve as well as peak. Area under the curve would be a single number and eye opening for a lot of people.

samssun > traderQAMobileTestAutomationMobileBoostOn 07/01/2016 at 00:07

0

Torque will be power divided by rpm (torque is how “strong” your engine is, and power is how much strength it can apply in a given unit of time, or torque * rpm).

So an “ideal” (in terms of broad usable power) engine would have a flat torque curve. As you double rpm you apply the same push twice as often, so your power would increase linearly, making a 45 degree line that crosses torque at 5250 like always.

In reality that’s hard to do (Corvette aside). If you optimize for low end torque but sacrifice it up high, or can’t rev, you’ll have a flatter power curve (less push x more revs = same or only slightly more power).

If you optimize for high revs but sacrifice low end, you’ll have a steeper power curve, increasing torque/push at the same time you apply it more often.

If you have a tiny displacement Honda engine, which can’t breathe down low because its short runner are optimized for the high end, and you strap a massive turbo to it that won’t spool til 5000, you will have a very steep torque curve, and a nearly vertical power curve.

BiTurbo228 - Dr Frankenstein of Spitfires > traderQAMobileTestAutomationMobileBoostOn 07/01/2016 at 06:55

0

So pair the 500bhp Hup with a CVT and you can smoke anything you like :)

TahoeSTi > traderQAMobileTestAutomationMobileBoostOn 07/13/2016 at 13:10

0

torque is a measure of power, horse power is a measure of work done with that power. So horse power is work*time / 5252. Horse power should be represented as ft-lbf/min