Interpreting power curve

[Adam230K]

Hi Guys,

The way I interpret the curve of my polo below is that I really should be keeping it between 3-4k rpm, I even think that taking it to its 4500 rpm redline might be a good idea too. But when i've had various 'fun' run ins with other cars as I've been driving it, it just seems that my car slows down after 4k rpm.

As an example I was messing around with a 1.8 clio the other day, obviously my car and his are very closely matched so we were neck-and-neck alot of the time, but my point is that I noticed as I let the revs go beyond 4k he would start to edge away and as soon as I changed gear, I'd be back in the game and hanging on to him like I should.

I know its a bit silly but am I missing something here..?

 

[stevesey]

It's the torque curve you should be looking at - stay within the flat topped area of that. Of course gear ratios and change points mean you might want to go a bit higher at times.

 

[kth286]

do not let it go any higher than 3300 revs for max power and performance.

 

[Adam230K]

So I don't think its torque that matters, i think its power delivered. I think the more power being delivered the greater acceleration achieved per unit time. Can anyone prove me wrong or reinforce what my thoughts?

 

[stevesey]

So I don't think its torque that matters, i think its power delivered. I think the more power being delivered the greater acceleration achieved per unit time. Can anyone prove me wrong or reinforce what my thoughts?

You need to read this then.

In particular this bit:

"Any given car, in any given gear, will accelerate at a rate that *exactly* matches its torque curve (allowing for increased air and rolling resistance as speeds climb). Another way of saying this is that a car will accelerate hardest at its torque peak in any given gear, and will not accelerate as hard below that peak, or above it. Torque is the only thing that a driver feels, and horsepower is just sort of an esoteric measurement in that context."

So keep on the top of the torque curve as much as you can.

 

[Number_Cruncher]

With a gearbox with fixed ratios, in any particular gear, you get maximum acceleration when the engine is producing maximum torque.

With a continuously variable gearbox, you get maximum acceleration when the engine is held at maximum power.

These two statements abvove, despite seeming contradictory, are both correct.

With the gearbox with fixed ratios, if the engine were running at maximum torque, [i.e., doing the best possible in that gear] you would get better acceleration by changing down and making the revs nearer the maximum power speed.

 

[stevesey]

Adam - interpreting your question a little it's really what rpm should I change gear at to make sure I get maximum acceleration isn't it.

Applying NC's rule - change at max power so 3300rpm on your graphs (using the at the wheels figure) which kth would agree with.

Seems a long way below the redline? I thought'd I try some maths to illustrate when to change gear.

So accelleration is proportional to torque at the wheels. And the gearbox acts as a torque multipler. So for an illustration lets takes a 3rd to 4th gearchange and assume a 4th ratio of 1:1 and 3rd of 1.35:1 (N.B. using the at the wheels graph figures).

4000rpm in 3rd = 100 lb.ft = * 1:1.35 gear ratio gives us a figure of 135.
Change gear and you have:
3000rpm in 4th = 145 lb.ft = * 1:1 gear ratio gives us a figure of 145

Higher figure after changing gear - should have changed earlier - lets try 3500rpm

3300rpm in 3rd = 135 lb.ft = * 1:1.35 gear ratio gives us a figure of 175.
Change gear and you have:
2450rpm in 4th = 160 lb.ft = * 1:1 gear ratio gives us a figure of 160.

Higher figures at both points than the first example - Clio is now way back in your rear view mirror.

Basically if you change as high as 4000rpm you are spending all your time on the downward side of the torque curve - your changes should bridge the peak to maximise the area under the graph.

 

[Number_Cruncher]

The zip file attached is a spreadsheet which allows some basic performance data to be estimated.

I'm not saying it's the last word in accuracy - there are some tweaks I would like to make to it, and there's certainly no warranty!

However, if you enter your vehicle's details in the cells marked out in bold on the tractive effort page, it should calculate some interesting numbers.

One large weakness is in the estimation of engine performance - it falls down a bit with modern tuned manifolds where the power and torque curves are more "bumpy" than older engines.

The cells in red are the rolling resistance values of the tyres - A, the speed independent term, and B the speed dependent term. It's a bit difficult to find good published figures for these, but, usually, rolling resistance isn't a large contributor to a vehicle's drag, especially at speed.

The change up engine speed can be changed, and typically, you'll find that revving beyond the maximum power speed will give a predicted improvement in acceleration.

One interesting (to me at least!) feature of the spreadsheet is that it calculates performance in two ways - one a torque based method, and the other a power based method - they both agree, as they absolutely should.

I've uploaded the spreadsheet in the spirit of co-operation rather than in any pretence that the spreadsheet is absolutely correct - if anyone has any suggestions for its improvement, or can spot any obvious howlers!, I would be very pleased to hear about them.

 

[Dieselman]

Use an inclinometer or accelerometer to find the G force in each gear across the rev range, then plot graphs and see where they intersect. once acceleration in one gear is lower than in the next gear up, you know the revs to change up at.

 

[Adam230K]

Interesting stuff.

I always thought that where the car is delivering the most power thats where it will also accelerate the most. In this case my dyno sheet said that max power was attained at 3460rpm, max torque at ~3k.

I might do a 10-80mph tests (to alleviate take-off variation) and see which one takes longer. Judging by what I saw on that day and what I definitely feel in the car, I'm quite sure that i'll be following the torque curve of the car as suggested. Will report back results soon.

 

[Adam230K]

With a gearbox with fixed ratios, in any particular gear, you get maximum acceleration when the engine is producing maximum torque.

With a continuously variable gearbox, you get maximum acceleration when the engine is held at maximum power.

These two statements abvove, despite seeming contradictory, are both correct.

With the gearbox with fixed ratios, if the engine were running at maximum torque, [i.e., doing the best possible in that gear] you would get better acceleration by changing down and making the revs nearer the maximum power speed.

These 2 statements seem to contradict each other. Which is it, better acceleration at max power or max torque?

 

[stevesey]

These 2 statements seem to contradict each other. Which is it, better acceleration at max power or max torque?

Once in a fixed gear - max torque.

 

[stevesey]

The change up engine speed can be changed, and typically, you'll find that revving beyond the maximum power speed will give a predicted improvement in acceleration.

Haven't had a look yet - but that will depend on the shape of the torque curve won't it. May well work out that way for a normally aspirated petrol engine, but for a TD if you rev beyond max power - when you change up you may well find you are on the downward side of the torque curve (becuase there is so much torque lower dwon).

What you need is to select the gear change points to maximise the area under the torque curve isn't it?

 

[Dieselman]

Yet more cheap as chips testing...no fancy Dyno needed....Again... !!

Browser Warning

 

[stevesey]

The zip file attached is a spreadsheet which allows some basic performance data to be estimated.

Is it just me that can't extract for spreadsheet - or has it been corrupted?

 

[Dieselman]

Is it just me that can't extract for spreadsheet - or has it been corrupted?

It won't open. What zip compression is being used?

 

[Number_Cruncher]

To squeeze under the size limit of this site, it uses the ultra compression option in 7-zip. Sorry, I should have mentioned that.

 

[wemorgan]

Could you try hosting it on Google Documents?

 

[stevesey]

Had a bash at generating torque at wheels graph - based on Adam's figures

Using the flywheel figures I end up with (speed = X and torque = Y - although these aren't accurately scaled).


The lines don't cross which suggests that Adam should take it to the red line in each gear to maximise the torque at the wheels:

However if I use the lower figures which include transmission losses I get:


Here the lines cross and give the gear change points to obtian max acceleration. The cross over points are generally below 4000rpm - which ties in with Adam's feeling that the car doesn't pull as well if he pushes if above 4K.

N.B. As i don't know the exact gears on Adams polo the axis on boths graphs shouldn't be taken as absolute figures - just illustrative.

When I first looked at Adam's figures I was quite suprised at the way the transmission losses increase above 2500rpm. Losses seem more than they should be to me.

 

[Dieselman]

When I first looked at Adam's figures I was quite suprised at the way the transmission losses increase above 2500rpm. Losses seem more than they should be to me.

Possibly a result of the increased torque being fed through the transmission. Losses are exponential to the power supplied.

How did you calculate the transmission losses, I thought the graphs were before and after remap?

If the losses really are showing as large then this may be how the much improved outputs have been shown...a bit of fiddling with the losses goes a long way to showing increased output.