There are two types of dynamometers - chassis and engine.
The main advantage of a chassis dyno is the relative ease of access. You just drive the car's drive wheels onto the rollers and begin measuring - simple. This makes it ideal for basic fault-finding that can't be performed on the road, quantifying mods and basic experimenting. It can also be used for engine management mapping, but it isn't as suited to this as an engine dyno.
One often-overlooked feature of chassis dynos is their ability to be used to calibrate vehicle speedometers.
When a car is put on a chassis dyno, a large cooling fan is directed onto the radiator, the car is nestled into position and the safety tie-downs are connected. Rear wheel drive cars are generally the easiest to dyno, but the more powerful ones (over about 250hp at the wheels) can create traction problems. Depending largely on the car's suspension design, the car might want to wriggle about and wheel-spin on the rollers. This gives small irregular spikes along the dyno graph, but extra tie-downs can usually solve this problem.
Front wheel drive cars try to torque-steer on the rollers, and small guide wheels are usually secured into position next to the car's front wheels to prevent the car from walking about. Permanent 4WD cars and commercials can only be dyno'd on a 4WD dyno. These expensive machines have adjustable wheelbases to accommodate different size vehicles. And obviously, the car must be very well tied down...
Common measurements taken (other than power and tractive effort) are air/fuel ratio at the tailpipe, intake air temp and sometimes a full 5-gas tail-pipe gas analysis.
When chassis dyno-ing a car, be wary of the build-up of excess heat - especially in highly modified turbo engines. The cooling provided by the front fan helps, but often it isn't capable of providing the same cooling airflow that you encounter on the road. Chassis dynos give power "at the wheels". A power run on a chassis dyno should cost around A$60-100, while a full EFI system mapping from scratch should come in at under $1000 - regardless of engine type.
Note that when doing comparative dyno figures (ie 'before' and 'after' a mod), ensure the power pulls are conducted in an identical manner. An air temp sensor in a different location, a closed bonnet, varying ramp speeds or a revised correction factor might all conspire to fudge your results. And always make sure the final power curve is one that is repeatable, not a one-off best.
To use an engine dyno, the engine must be removed from the car. Because of this, it is not practical to re-dyno an engine after only relatively small mods have been made - a chassis dyno is the more appropriate alternative. Engine dynos are, however, more suited to programming management systems (or carb/ignitions) on an "unknown" engine or one with major mods. Any workshop equipped with an engine dyno is likely to have invested in comprehensive instrumentation, full computer control and data logging. This type of dyno also has the advantage of giving extremely precise control over engine load and speed.
A major benefit of an engine dyno is that it allows the use a quick-response exhaust gas oxygen (EGO) meter or an exhaust gas temperature (EGT) sensor positioned directly after the exhaust collector - or even one after each exhaust port. Extensive use of gas analysis is also common. Some modern engine dynos have electronic throttle control to keep the engine 100% on track, and there is also the advantage of significantly better engine cooling - depending on the capacity of the water tank or cooling system employed. Engine dynos give their power reading "at the engine", or "at the flywheel".
Having decided to opt for an engine dyno - what should you pay and what should you expect in return?
Well, because of the time involved in setting up an engine on an engine dyno, the price difference between a power run only and a full EFI mapping isn't that great. Around A$450 will get you a power run, but from approximately A$50 more you can get a full mapping job done! Depending on the facilities and the engine though, up to around A$4000 could be spent mapping a twin-turbo small block with sophisticated aftermarket injection. Note that these prices can vary for "oddball" engines that might require fabrication to hook them up to the dyno. Note that some workshops charge a flat rate of around A$600 per day, regardless of what is being done.
In summary, using an engine dyno is the most accurate way to dyno an engine - after all, that's how engine manufacturers do it.
All dynos work on the principle of measuring torque. It helps to imagine torque as twisting effort at the crank (or wheels). The unit for this measurement is Newton Meters (Nm), kg-m or ft-lb.
The other aspect that a dyno needs to measure to determine engine power is speed in revolutions per minute (rpm). This is due to the fact power is equal to torque multiplied by revs.
However, for a dyno to attain a proper and steady reading, it must be able to apply and then hold the appropriate load on the test engine. This will allow the engine to develop power. The way an engine or chassis dyno achieves this is with a brake - the origin of the term bhp (ie measured power at the engine brake).
Modern dynos achieve this loading in one of five ways:
Used in both engine and chassis dynos. These use a dyno head with a rotor that spins in a large watertight housing. As it spins, the rotor pushes the water outward into water cups mounted on the inside of the casing. The engine's power is dissipated into the water, which also has the additional benefit of being used as a cooling medium. This approach is quite good at holding a constant engine load.
Used mainly in chassis dynos. A pair of discs spin close to large electric coils, which magnetically resist the rotation of the discs as more current is applied to them. Because current can be controlled more easily than the flow of water, an Eddy Current dyno is more finely adjustable. Excellent for creating a precise and constant load.
In this application, an oil pump is driven directly off the engine and the output of the pump is variably restricted to increase the applied load on the engine. Rarely used today, but can hold quite a steady load.
Here, two load sensors are located along a shaft between the engine and the dyno. With the metal characteristics of the shaft known, the twist of the shaft under load can be measured and used to determine the applied torque.
This is the technique used in the US Dynojets. The engine is used to spin up a large external flywheel with the control software able to measure acceleration of the flywheel. With the inertia of the flywheel known, torque can then be calculated. This type of dyno can calculate dynamic load only, not hold a constant load.
Because all of these types of dynos need to dissipate the engine's total power, an efficient cooling system is important to maintain consistency. In the case of engine dynos, this is usually done with a large water tank or cooling tower. The water passes through the hot dyno head and then back into the tank or tower. If a tank system is being used, mapping an engine from scratch might consume the entire capacity of the tank, in which case hot water will start to be re-circulated. On the other hand, most chassis dynos use internal fan cooling to transfer heat to the air. These are often incorporated into the rotating assembly and can be quite effective depending on the amount of power being dissipated over a given duration.
To make compensation changes for varying atmospheric conditions, there are usually some +/- adjustment facilities available to the operator of a dyno.
All dynos have an air temperature probe that is designed to be located in the engine's induction stream. There are also usually inputs for barometric pressure and humidity. Note that it is easy for dyno operators to "create" more power by fiddling with these controls, or by suddenly dumping high load onto the engine at any given rpm. Under these conditions, the measured power output might vary by up to 20 per cent - so it's very important to have faith in the dyno operator.
Chassis Dyno Tractive Effort - What Is It?
Chassis dyno graphs display both power and tractive effort at the wheels - not power and torque like an engine dyno. Tractive effort is the force being applied at the tread of the tyres and is proportional to engine torque. It differs from torque because it takes into account the gearing of the drivetrain (gearbox, final drive and tyre diameter). Power runs are usually conducted in second or third gear, and the tractive effort graph can be plotted against either engine rpm or road speed at the wheels.
Engine dynos are great for mapping a fully modified or relatively unknown engine for the best results. Chassis dynos are ideal for quantifying changes, fault-finding, tuning and more cost-effective basic engine mapping.
As a guide, all race teams thoroughly engine dyno their motors since they have the financial resources, but we who aren't as fortunate can still achieve excellent results with a chassis dyno. But in any scenario, it is important to have an experienced dyno operator that you trust. Ask some of the workshop's previous customers of their impressions and make an informed decision before entrusting your car or engine to a dyno workshop.