Tune for torque or tune for power?

[Brodyj]

Tune for torque or tune for power?

Maybe I have not done enough research, but there are many engine tuning questions that I still am not sure about. I have not found many books nailing down basic concepts like the goal of tuning for torque and horsepower. It seems a lot of people are similar to me and go by what they heard and they believe it as the truth. Because our engine development is done, all this information is just to clear up some unanswered questions in my head.

1. Do you know any books or journals that talk about the goal of tuning power and torque, not how to tune for horsepower and torque.

2. From those books what did you learn? The consensus seems to be that in autocross you want your driver to have a lot of control by tuning for a flat torque curve.

3. How flat? And how wide? Torque variation within a certain percent? Over what RPM band? Over the years our team has always worked down low in a range from 5000 to 10000 rpm with a rev limiter at 11000. We would make about 50-50 hp on the dyno. Is this wrong? Other teams go much higher rpms, but we have never made much torque up there and power would go down. It seems that you would want your torque band to be as wide as the biggest difference between 1st and 2nd plus a little safety factor. What do you think?

4. Some people say, torque is all you care about and that peak power is insignificant. This does not make any since to me because you can gear a car with gobs of torque but it should not go faster than the same car with higher horsepower. Will it?

5. For max acceleration, do you want to drive around the rpm band that has the most power or torque.

Brody
Oregon State University

[Brodyj]

Tune for torque or tune for power?

Maybe I have not done enough research, but there are many engine tuning questions that I still am not sure about. I have not found many books nailing down basic concepts like the goal of tuning for torque and horsepower. It seems a lot of people are similar to me and go by what they heard and they believe it as the truth. Because our engine development is done, all this information is just to clear up some unanswered questions in my head.

1. Do you know any books or journals that talk about the goal of tuning power and torque, not how to tune for horsepower and torque.

2. From those books what did you learn? The consensus seems to be that in autocross you want your driver to have a lot of control by tuning for a flat torque curve.

3. How flat? And how wide? Torque variation within a certain percent? Over what RPM band? Over the years our team has always worked down low in a range from 5000 to 10000 rpm with a rev limiter at 11000. We would make about 50-50 hp on the dyno. Is this wrong? Other teams go much higher rpms, but we have never made much torque up there and power would go down. It seems that you would want your torque band to be as wide as the biggest difference between 1st and 2nd plus a little safety factor. What do you think?

4. Some people say, torque is all you care about and that peak power is insignificant. This does not make any since to me because you can gear a car with gobs of torque but it should not go faster than the same car with higher horsepower. Will it?

5. For max acceleration, do you want to drive around the rpm band that has the most power or torque.

Brody
Oregon State University

[Ryan Schoffer]

what is best (and what the judges want to see) is lots of area under your torque curve between the shifts, as well as shifts far enough apart so the driver doesnt need to shift all the time


so it becomes a balancing act between tuning the engine (and designing the manifolds etc.) for high peak power, or a wide powerband - the wider the powerband, the fewer shifts are needed, as well as the higher the top speed. if you tune for peak power, then your shifts will be very close

[Charlie]

You questions aren't easy, except for the last one.

Horsepower is what accelerates the car, the car with the most horsepower will always out-accelerate a car with less (no matter the torque output). This is of course assuming constant numbers (or average numbers) over the RPM range used during the run.

[Dan Deussen @ Weber Motor]

Charlie-

I have to disagree with your statement. Acceleration is caused by torque and not by horsepower. Of course, an engine with higher horsepower at a given RPM automtically has more torque at that given RPM because horsepower = (RPM * Torque) / 5252.

But acceleration time is calculated using the following formula:

t = ((WK^2)*(delta N)) / (308 * T)

where
(t) = acceleration time in seconds
(WK^2) = inertia
(delta N) = change in RPM
(T) = acceleration torque in ft-lbs

As a result, more torque at any given RPM will result in a higher acceleration rate.

[imajerk]

Torque. Useful torque. Power is a calculation... of torqueJ

Use Excel to create this:


List - car weight, driver weight, tire radius, weights bias, tyre coefficient of friction, all relevant gearing and figure out the maximum force that can be applied to the rear wheels before they slip.

So, you have now found at what point torque becomes useless. Go through every gear and work out how much force is going to the rear wheels verses speed (m/s km/h whatever). you can figure out your instantaneous acceleration, and also play with Cd (drag force) to see how that effects the cars instantaneous acceleration.

At the end of that you will see that having lots of torque low down and in the midrange is a waste as most gearing set-ups mean there is simply to much potential force where it cannot be used, and when the car gets to speed torque naturally drops off and is than overcome by drag.

Estimate what is your maximum top speed you will reasonably use (eg acceleration event) and try to play with the excel file to see how much torque you really NEED at various speeds.


Email me for a sample excel file on james1@DELETEstudent.monash.edu

[Charlie]

Sure horsepower can be derived from torque. That does nothing to prove or disprove its worth.

An inertia dynamometer measures horsepower by how quickly the mass accelerates. It cannot measure torque. So you see in another case you have to derive engine torque from horsepower with an engine RPM signal (Dynojet).

I do not think that a Briggs V-twin that makes 40 ft-lbs of torque from 2500-5000 RPM can out-accelerate a F4i that makes 35 ft-lbs from 6500-10000 RPM.

In Tune to Win Carrol Smith recommends gearing to acheive the most area under the horsepower curve to optimize acceleration. This can't be the same as the area under the torque curve as torque generally falls off well before this point, especially in a racing engine.

imajerk, I would be willing to bet that your excel file actually graphs power and not torque, without you realizing it. Force=power/speed with the correct conversion factors of course.

[Andreas]

On every motor forum I have visited, there have been this discussion about torque vs power. My favorite is: "Torque gives acceleration and power gives top speed"...

I agree that torgue gives acceleration, BUT, that is when we talk about wheel torque. The wheel torque and the wheel radius result in a propelling force. For a given wheel speed (which is proportional to the vehicle speed), the wheel torque is proportional to the wheel power. If we skip the losses, this means that for a given speed, the wheel torque is proportional to the engine power, since the power doesnt change int the gearbox. The torque is multiplied by the gear ratio in gearbox and I think that this is where people go wrong.

For a given gear, the car will accelerate quickest at the peak torque, but if you (at the same speed) gear down to where your peak power is, your car will accelerate even quicker because your (engine torque*gear ratio) increases (just as much as your power does).

If I hade a CVT, I would tune it to keep the engine to peak power RPM at the acceleration event.

Andreas Chalmers -02, -03

[Big Daddy]

That is exactly what a CVT does. Your gear ratio changes as you move faster. That is the reason all cvts I have dealt with were torque sensing, if you need more torque it closes the driven pulley lowering the ratio. As for tuning a cvt you want it to start to shift at peak torque and you want it to run at peak power. That way you have the best torque off the line and the best power after the car is moving.

[imajerk]

Charlie, of course my spreadsheet deals with power.

I find it simpler to explain to look at the torque, and try to get the right torque for the right speed, or if you prefer the right power for the right speed. This is a static analysis (i don't care about inertia for this example) - but still more useful than not knowing anything.

I have seen a particular fsae car with a very generous torque curve for midrange - but because of gearing etc it is absolutely excessive, instead rasing the torque higher up in the RPM is more useful.

There is a much better may to deal with transmission – and that involves getting rid of those silly clutches and gearboxes and simply tune for maximum hp and use... nah, that's too much informationhttp://fsae.com/groupee_common/emoticons/icon_smile.gif

[Brodyj]

Hey thanks for all the input. These are basic concepts that our team was always guessing on before, this stuff will help us next year.

Any good books on this stuff? I would like to hear more opinions if people have them.

What rpm range to you play in? 5000 to 10000? Or all the way up to redline? Why?

Brody

[Frank]

ahh, it is confusing,

think about this...

power is a unit of energy produced per unit time.. if my engine is making more energy per unit time than yours, then my engine has the potential to accelerate any mass faster

power is geared down to make torque

high rpm power usually makes a car more difficult to drive though, and usually requires a larger number of gear ratios available, and shifts required

i think i remember Carroll Smith writing/saying this somewhere

[MikeWaggoner at UW]

I thought rubber belt CVT's were just RPM sensing (snowmobiles, some ATV's etc). The other ones I've seen (Audi, Toyota, Honda etc) all seem to use complex controllers that don't just torque bias. Good explanation of how these CVT's work:

http://www.gates.com/brochure.cfm?brochure=1033&location_id=542

I think it would be difficult to tune this for ideal acceleration and driving, although I've always wanted to get rid of the spring/weight system and install a screw drive motor w/ controller and a throwout bearing...

-Mike Waggoner

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by Big Daddy:
That is exactly what a CVT does. Your gear ratio changes as you move faster. That is the reason all cvts I have dealt with were torque sensing, if you need more torque it closes the driven pulley lowering the ratio. As for tuning a cvt you want it to start to shift at peak torque and you want it to run at peak power. That way you have the best torque off the line and the best power after the car is moving. <HR></BLOCKQUOTE>

[Bam Bam]

If you don't understand torque vs power,

I would recommend some first year physics textbooks would get you off on the right foot.

[Big Daddy]

Mike all cvts on snowmobiles and atvs are torque sensing. The secondary pully has a coil spring and there are two sets of ramps and when you need more torque IE you start up a hill the motor will try to slow thus allowing the driver pulley to start to open and the secondary to close now granted this really isnt torque sensing its just what everyone calls it. As for the electric kind we had a guy build on for his senior project. The prob was that it is hard to find an electric motor that can shift the cvt that fast. Now Saturn's Vue is shifted by hydraulics and is very fast, but they use metal pulleys and a metal belt (not real sure on the wear aspect of that it still puzzles me) http://fsae.com/groupee_common/emoticons/icon_biggrin.gif

[schumafer]

for me it is simple,

you have to tune for torque at all rpms, then you will have more power

P=T*W

the more torque you have.. the more power you will have at any rpm.

of course, if your engine doesnt have too much torque, it needs more speed to produce the same power as if it has high torque (obvious).

so, what you care about is torque, that is what moves you. and if you want smooth drivability, then tune for a high flat torque curve.

hope i explained myself...

Fernando Gòmez
Universidad Simòn Bolìvar
F-SAE USB
www.formulasae.grupos.usb.ve (http://www.formulasae.grupos.usb.ve)

[flow_with_it]

my faith would lie with andreas's answer. torque is proportional to power. good torque and power curves will have plenty of both in the rev range that you are using most often, interms of formula student the mid range. The gearing is also crucial. before i confuse people anymore try this link, may have to read it a few times. http://www.allpar.com/eek/hp-vs-torque.html

[Pete M]

Why are you replying to a post from over 3.5 years ago? I think he's figured it out by now anyway.

In any case, to say something useful, ignore engine power and torque and look at tractive effort. Make a simple spread sheet that will give you a curve of power at the wheels (or torque at the wheels or thrust at the wheels, they're essentially the same) vs ground speed and optimise that. Engine torque is irrelevant. If i had an engine that only made 10 N.m of torque but reved to 100,000 rpm... that'll sure as hell accelerate the car quicker than what we're currently using, even though it's got a tenth the torque, so long as you do the gearing correctly.

[Chuckster]

Seen this same question time after time on performance sites.

Answser is almost too simple:

Power is rate of using energy-more HP is required to move (accelerate) something faster from Point A to B.

End of story-really! Honest,there is no more.

Back up antidote-been racing in Solo with a CVT car for over 15 years and have pumped the daylights out of all succesful road racers and snowmobile racers-always the same answer.

Nobody fast with a CVT ever ever tunes it to stay at the motor's torque peak-always the HP peak.

[ben]

I think it depends on the calculation you chose to make. If you do an energy method you'll use power and if you do an equilibrium of forces you'll use torque.

My preference was to use torque to calculate tractive effort curves and to use an energy method using power and drag to calculate top speed.

Power and torque are a fixed relationship to each other. All these debates are is a bunch of engineers arguing about how their own brain conceptualises the two parameters.

I hear what you're saying Charlie, but if you calculate a tractive effort curve based on the torque, you'll still be calculating the velocity by multiplying up the rpm using the gear ratios so you'll still demonstrate the point without needing to directly calculate the actual power from the torque curve.

Ben

[samphlett]

Wow, a thread back from the dead...

Given a perfect gearbox (like what you get on a diesel electric locomotive in the UK) power is all, so long as the shafts can handle the torque. For a disrete ratio'd gearbox there are benefits in shaping you torque curve to suit your ratios. Think of power as the rate at which you can drag a big mass up a hill. Or accelerate a big mass down a track. With a crap torque curve or a crap gearbox you'll be continually shifting to maintain whatever your peak power is.

[Charlie]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by ben:
I hear what you're saying Charlie, but if you calculate a tractive effort curve based on the torque, you'll still be calculating the velocity by multiplying up the rpm using the gear ratios so you'll still demonstrate the point without needing to directly calculate the actual power from the torque curve.

Ben </div></BLOCKQUOTE>

When you plot a tractive effort curve what you are really calculating, and looking at, is power. The units are just goofy and the scaling is a bit odd.

To bring Torque into the discussion (even though it is used in the calculation of the curves) is confusing, and technically, inaccurate.

The bottom line is you always want more horsepower, damn the torque. The tractive effort curve is proof of that, if you realize what you are looking at.

I am sitting here surprised that I agree with an answer I gave over 2 years ago!

[murpia]

This (http://craig.backfire.ca/pages/autos/horsepower) is quite good fun for this topic...

Regards, Ian

[The AFX Master]

Thanks for thee link Ian, although that`s a very basic guide, is a very complete one and fair easy to read. including the redneck vs ricer comparison http://fsae.com/groupee_common/emoticons/icon_biggrin.gif