Power Range
© Frits Overmars
Peak power alone does not say much about the usefulness of an engine. The combination of power curve
and transmission, the sort of application that the engine will be used for, the abilities of the rider and the
condition of the track, together define which is the optimum engine character.
An example: a 125 cc road race engine can always be kept between 10.000 and 14.000 rpm thanks to its
six-speed gearbox. Whether this engine produces 2 hp or 20 hp at 6000 rpm, is irrelevant.
But a kart engine with direct drive without a gearbox has trouble staying above 5000 rpm in slow corners;
it can only manage to do so by means of a very short gearing that forces the engine to rev over 17.000 rpm
on the straightaway. For such a kart it is imperative that acceleration out of slow corners does not cost too
much time. There may be only one such corner on the whole circuit, but time lost there cannot be made up
by a higher top speed on the straight.
Therefore a sensible tuner will not concentrate on peak power; he will make sure that the power is never
really bad in the whole range from 5000 to 17.000 rpm (in this kart-example).
I prefer to work not with a power band but with a power range, which I define as the highest rpm of a power
curve, divided by its lowest rpm.
Experience has taught what kind of power range is needed for a certain application. Road racing calls for a
range of about 1,4 . When a Continuously Variable Transmission is involved, a range of 1,2 or even less
might be enough. Motocross calls for something like 2. A touring bike needs at least 3 to be comfortable.
And the direct-drive kart in the above-mentioned example needs about 3,4 (17.000/5000).
Let us assume that we have a measured power curve from 7000 to 14000 rpm. That gives a power range of
2.
Within this power curve all possible range values are investigated. From range=1,00; range=1,01;
range=1,02; range=1,03 , etc, up to range=2.
For each of these range values the whole power curve is examined in order to find which lower and upper
rpm values yield the highest average power.
For range=1,5 for instance, we start with calculating the average power between 7000 and 10.500 rpm. Then
we proceed with 7010--10.515; then with 7020--10.530; and so on, until the final possibility of 9333--14.000
rpm. And the highest value found is stored as the average power for range 1,5 , together with its
corresponding lower and upper rpm limits.
All stored values for average power are displayed in range graphs. So when you are preparing an engine
that will need a range of 1,6 , you can see at a single glance which engine delivers the best average power
at range 1,6 . Furthermore you can see between which rpm values this engine is most effective.
And you don't need to fall into the trap that all engine builders have fallen into at some time: attaching too
much importance to peak power.
Maybe you are used to comparing the acceleration times of different engines on an inertial dyno. But that
only makes sense if these engines are all run between the same initial rpm and the same final rpm,
and all with the same gearing.
The range concept does not suffer from any of these limitations; it functions under all circumstances.
And because you can compare the range graphs of all your engines, you will be able to distinguish much
quicker which range is best for a specific application. That is an experience you would otherwise only gather
after years of trial and error.
Suppose you are preparing an engine for an application that requires a power range of 1,5. Then which of
the two power curves in the picture below left should you pick? Hard to say, isn't it? But when you look at the
power range curves in the picture on the right, it is clear as day: the yellow curve wins.
Two more power curves. The yellow curve does not look very useful. But maybe it could work with a
Continuously Variable Transmission....
The power range curves tell us what we need to know. If the CVT can keep the revs within a 1,2 rpm range,
the yellow curve is OK. And if the CVT can keep them within a 1,1 rpm range, the yellow curve is a winner!
Another advantage of the power range curve: it can show you the rpm limits you should operate between.
At range 1,2 the yellow power curve works best between 9970 rpm (the blue curve) and 11.964 rpm
(the white curve).
Average power
In the above text I mentioned average power. Average power is not the area under the power curve.
A claim like this calls for an explanation, so here goes.
Take a look at the above A_F_E_G_B power curve. It is a peculiar curve where the power totally collapses
at point E, but then comes back. The reason I drew it like this will become clear in a minute.
We can look at the red area under the curve and convert that to a rectangle. The yellow rectangle in the
picture below has the same area as the red areas under the original power curve. So you could say the
average power between rpm-point A and rpm-point B has a magnitude equal to the height A--D.
Or could you?
Let us put the engine with the peculiar red power curve in a vehicle and start accelerating, from rpm-point A
to rpm-point B. How long will that take? It will take forever because the power at rpm-point E is zero,
so acceleration at point E is zero; we will never get past that point.
Acceleration from A to B will take an eternity which means that average power between A and B is zero!
Admittedly you won't encounter a power curve like the one in the first red picture very often. But the
acceleration versus area argument holds for any power curve. Let's take a look at a 'normal' curve without a
zero-power point. The curve below is split up into 9 equal rectangles, each with a horizontal dimension A--B
that represents rpm, and a vertical dimension A--G that represents power.
When you convert the sum area of those 9 rectangles into a single rectangle with the same rpm spread A--F,
it will have a height A--H that is 1,8 * A--G. Assuming A--G equals 1 HP, then the average power A--H would
be 1,8 HP (picture below).
Now let us look at the same power curve and determine the average power via the acceleration approach
(picture below). We will assume that acceleration from A to B with the available power A--G takes 1 second.
Between B and C the power is twice as much, so acceleration from B to C will take ½ second. And so on.
Total acceleration from A to F takes 20/6 s. If this acceleration were to have a constant value, it would take
1/5 * 20/6 = 4/6 s from A to B. This would require a constant power of 6/4 * A--G. Assuming again that A--G
equals 1 HP, the average power would be 1,5 HP instead of the 1,8 HP that came out of the area-approach.
By the way, if you really want to grasp the difference between the 'area under the curve'-approach and the
acceleration approach, put another 97 yellow rectangles on top of the 3 rectangles between C and D in the
above picture, and repeat the area-calculation and the acceleration-calculation of average power.