Nissan Patrol 4x4 (Africa)

In Carlo’s thread I stated that torque cannot be measured on a dyno.
In essence the statement is correct.
The dyno employs a load-cell to measure torque as it is impossible to measure torque at the rollers.
"Do you want your engine to make HORSEPOWER or TORQUE?"
This question is usually asked in a tone which strongly suggests that "experts" believe power and torque are somehow mutually exclusive.
In fact, the opposite is true, and you should be clear on these facts:

POWER (the rate of doing WORK) is dependent on TORQUE and RPM.
TORQUE and RPM are the MEASURED quantities of engine output.
POWER is CALCULATED from torque and RPM, by the following equation
HP = Torque x RPM ÷ 5252
An interesting fun fact is because of the mathematical constant

All vehicles, no matter which, will therefor produce their same torque figure and power figure at 5252 RPM.
Get your head around that!
An engine produces POWER by providing a ROTATING SHAFT which can exert a given amount of TORQUE on a load at a given RPM. The amount of TORQUE the engine can exert usually varies with RPM.

So using the sketch below with respect to an engine
The simplest explanation is thus
Torque is the tangential force to turn the shaft
That shafts ability to drive something else is the power
And that’s why the dyno measure the power rather than anything else.

To expand somewhat:

A chassis dynomometer has some element (normally the rollers under the wheels, but there are ones that bolt directly to the hubs) which is driven by the car.

These driven elements (rollers from now on) are braked somehow to create a load. There are many different types of braking systems, from eddy-current (apparently now the most common) to hydraulic (very similar to a torque converter). However it works, the braking is controlled. The brake mechanism is mounted such that the torque it reacts can be measured. This is where the above-mentioned load-cell comes in. The roller's speed is also measured.

Torque is rotational force, and the formula is force applied multiplied by the length of the lever used to apply it.

Power is, as already stated, torque multiplied by the speed of rotation, with some constant added depending on what units of measure you are working in.

The thing is that the gear ratio (which includes not only the ratio of the gear the car is being tested in, but also the final drive ratio, the tire size, and the roller's diameter) affects the torque, but it does not (if you neglect the effects of friction, which of course you can't) affect the power output.

Now comes the tricky bits:
1) the dyno does not know the gear ratio or the tire size.
2) The dyno does not know how much the tires are slipping on the rollers (which they do)
3) Most dynos do not know how much friction there is in your drive-train. However, some dynos can measure it with some accuracy on manual cars. They do this by measuring how long it takes for the system to slow down when the clutch is depressed. But these are very un-common, and it's not perfectly accurate anywayas it can't measure the slippage.

Therefore, any dyno torque curve must include some sort of calculation to determine the gear ratio. Whether this is by the operator revving the engine to a certain RPM indicated on the rev counter, and then pressing a button, or by telling the dyno what car it is, and selecting a gear ratio from a list of loaded values, either is going to be wrong to some extent, in the first case because the operator WILL press the button at slightly the wrong time, and the rev counter ids probably wrong anyway, and in the second place because no two sets of tires are ever going to be exactly the same diameter, due to wear, tire pressures etc.

So, to sum up:

As stated before, because power is a function of torque and speed, any two engines that produce the same torque at the same RPM will also produce the same power at that same speed (RPM).

But that only applies at that specefic RPM.
Also, there are many factors that affect the accuracy of the result. Air pressure, temperature, tire grip and pressure,etc.
Thus, chassis dynos are as far as I am concerned of limited use for comparing different cars, but of great use for comparing one car before and after modifications, as long as those tests are close together in time, so that the tires are still the same... Tests done six months apart are of limited value.

Thanks Peter for expanding and making it even more clear, in a more thorough technical manner.

My experience has been consistent, same car, same dyno, months apart.

Obviously regular calibration of equipment should be carried out.

''Therefore, any dyno torque curve must include some sort of calculation to determine the gear ratio.''

The reasoning for pegging the torque to factory spec.

Very thoroughly explained Peter! Just the way I have it.

Thanks Julian and Peter - I am certainly left more informed but unfortunately also more skeptical of the vast majority of dyno operators out there, as well of the dyno figures themselves. I'd rather be skeptical than ignorant.