[Cuda340]

What's the difference between running an engine on an engine-dyno, or, as installed in the car, on a chassis-dyno?

Most chassis dynos extrapolate an engine's power and torque based on vehicle inertia. Many variables can affect chassis dyno results. It's a great convenient tuning aid, but you can't directly equate the reported torque and power numbers back to engine dyno results, or even the results obtained from two different chassis dynos. Operator experience is critical for obtaining consistent results from small incremental changes.

If you want to know how the engine will run in the real world as-installed in your vehicle and/or determine overall drivetrain performance and efficiency, then the chassis-dyno is the better choice. However, it doesn't directly measure torque. Instead it extrapolates output according to the chassis dyno manufacturer's software logic. Also it's harder to get back-to-back repeatability on even the same chassis-dyno, so expect at least a 5% variance in otherwise-identical back-to-back runs. Simply put, there's greater variance in results from different chassis-dynos then there are on two different, but properly calibrated and corrected, engine dynos.

However, if you want big, accurate numbers, in a carefully controlled environment, the engine dyno is the way to go. Engine dynos directly measures true torque at the flywheel, so it will tell you how much power the engine makes in a perfect world. A properly controlled engine dyno cell, a carefully calibrated engine dyno, and a good operator is capable of test repeatability within 1%.

Here are some “take-aways” when you compare a chassis to an engine dyno:

• Engine dynos are best for accurate and repeatable measurement of an engine's actual torque and horsepower numbers at the flywheel.
• Chassis dynos are a great in-car tuning tool but can't be relied on for highly accurate and repeatable power and torque measurements.
• With an automatic transmission, experts claim a 25-27% horsepower loss on a chassis-dyno versus an engine dyno.
• With a manual transmission, experts claim a 15-17% horsepower loss on a chassis-dyno versus an engine-dyno.
• Many variables affect chassis dyno accuracy, but all these "little" things affect the results and the amount of power-drop compared to flywheel numbers.
• Engine-driven mechanical fans cost power.
• Wheel weight, tire size and weight, and how fast you are accelerating that tire/wheel mass can drastically change a chassis-dyno's recorded results. Then there's airflow through the engine compartment, oil and coolant temperatures, inlet air temps, air cleaner size, the effect of engine-driven accessories, airflow quality through the engine compartment with the hood shut, how tight the dyno's tie down restraints are, etc.

 

[Frank Julian]

Engine Dyno, great if you are building an Engine and want numbers BEFORE you drop it in the car.
Chassis Dyno, Great if you are modding an engine that is already in a car and you don't want to pull the engine.

 

[VQ4NFO]

After building the engine I use both. I had the engine dyno'd before installed. Let it warm up, light throttle to around 4500 rpms and then full throttle, repeated several times to make sure the components were good and rings seated. Really not interested in the numbers because drivetrain losses affect the real world power to the wheels.

After installed, the fine tuning and baseline dyno results give me an ideal of the power to the pavement. After, with each additional modification, I use the same dyno and compare the results to the baseline. I even try to dyno in the same weather conditions, same operator and always ask if the dyno has had a computer upgrade that would skew the results.