subject line says it all, really…
From howstuffworks you hook the engine up to a dynamometer and see how much torque is generated at various rpm levels. You then multply torque by rpm/5252.
All of this sounds rather complicated, so I wonder if this is how it is actually done for every application.
oops, forgot to add;
If you’re wondering where the 5252 comes from, check here
Mechanical horsepower = torque (in foot-pounds) * RPM / 5252.
The torque output of an internal combustion engine varies with RPM, being zero at very low RPM, rising to a peak level, then falling to zero again at the engine’s maximum speed. Some engines produce high torque only in a narrow band of output speeds, others have a wider power band. Maximum horsepower is determined by looking at the point on the speed-torque chart where torque*rpm is greatest. This gives you an idea of the maximum power the engine can produce, but to get the full picture you need to look at the power/speed curve to see how the engine will behave in reality.
Originally horsepower was considered to be one unit when a horse could raise, by single pulley, one hundred fifty pounds at a rate of 2.5 mph, or 550 foot pounds per second.
This in energy terms turns out to be 746NM per second, or in electrical terms is 746Watts.
If you mean, how do they arrive at the bhp rating for engines, well there are a couple of problems that have to be overcome.
Piston engines are unusual in that they do not provide maximum torque at low revs, unlike say an electric motor.
What is done is to run the engine up to some speed, and then use the output to drive a brake, the idea is that you can easily measure the force on the brake by using a friction band, it can be as simple as a spring connected to a gauge.(there are very much more sophisticated ways of laoding up an engine)
Next you increase the pressure on the brake, whilst increasing the fuel input to the engine, and the speed of the engine should stay the same, when an increase in engine power cannot keep up with the increase in brake pressure, the brake gauge is checked for the load and the engine speed is recorded.
You probably know that
Power = Torque *revs, and since you have the torque on the brake gauge and you have an engine tachograph, its then easy to work out.
This is why piston engines are rated in Brake Horsepower(bhp)
I feel the need to point out (for those who don’t follow jk1245’s link) that 5252 is purely an artifact of units.
It’s a unit conversion constant, and requires: power be measured in horsepower; torque be measured in foot-pounds; and engine speed be measured in RPM.
Measure any of those quantities in different units, and the constant changes. (e.g., power in watts, torque in newton-meters, and/or engine speed in radians/second).
I’ve actually run a dynamometer once myself when I was in college. We had a Saturn straight four hooked up to a device that was sort of like a large generator, only the current it produced didn’t really do much of anything. The generator (actually, the correct term is an inductive brake) was rigged so that a load cell on its mount could tell you how much torque the engine was making at a particular RPM, and you calculate the horsepower from that. Running the device was pretty simple, no more complicated than driving a car. The setup has adjustments for throttle and how much load the inductive brake applied, and readouts for the torque, RPM, and since the engine is in a lab instead of a car a lot of other things about the engine - fuel consumption, coolant temperatures, cylinder pressure, and a lot more. Making a dyno pull doesn’t take very long; the pull itself could often be done in less than a quarter of an hour.
I’ve also been to a few “dyno days” where someone rents out a version of a dynamometer that connects to a car’s drive wheels. This gives a lower rating, since you lose some of the power from the gears and all in between the motor and the wheels. The people there had the car on the machine for about an hour each, during which they could fiddle with things under the hood and see if a few adjustments could give them more power.
As for whether the original companies go through this for every application: You’d better believe it. The automotive manufacturers have pretty deep pockets, so the cost is only a small fraction of what it takes to develop an engine. They test every type of motor they’re using in a production car if they have made any changes to it, not just for horsepower but also for emissions and durability. Not only do they test examples of every motor they put into cars, they also test a large number of engine designs that never wind up appearing in anything as part of their research and development. They use these dynamometers to determine the best programming for the fuel injection, also.