Diesel Power Summer 2005
photography Courtesy of Gale Banks Engineering
How to get factual horsepower and torque numbers using dynamometers—and why inertia dynos are dyno jokes!
How do you get factual horsepower and torque? There are two ways. One is constant-speed, full-load engine dynamometer testing. Worldwide, this is the accepted method of developing an engine’s power and torque ratings and most closely duplicates diesel’s historic use–hauling a heavy load. The power curve is developed by running the engine until all the iron, aluminum, steel, plastic, coolant, lubricants, turbocharger, and charge-air cooler are at stable operating temperatures. Then, the horsepower and torque curves are developed by running full-throttle at several constant rpm points from idle to redline. These points are used to plot the horsepower and torque curves.Gearhead vehicle discussions often go something like this: “How fast?” “What’s the e.t.?” “How much boost?” “What’s the horsepower and torque?” Whether bragging or fact, the answers are usually the subject of pride. Personally, I take no pride in bragging.
The second way is to accelerate the engine as it would when installed in the vehicle and take measurements as rpm increases, producing the curves.
A properly engineered, fully instrumented dyno cell costs at least one million dollars. An engine dyno cell with full emissions equipment costs around $2.5 million. Dyno test cells at this level of accuracy are found at GM, Ford, Daimler-Chrysler, CAT, Cummins, and International to name a few. Aftermarket companies rarely have this level of equipment.
For the past 40 years, we’ve checked the accuracy of our dyno cells by obtaining lab test engines from these Detroit manufacturers. Then, we compared their test data to ours on the same engines. One recent test engine from Cummins “Skunk Works” produced power just at the limit of its dyno cell capability of 400 hp. That engine, when tested at Banks, produced within 2 hp of Cummins numbers. This comparison ensures that we’re dealing in dyno “fact.”
We also use chassis dynos. They’re great for drivetrain development, but you really have to know what you’re doing to extract correct information. Wheelslip, clutch or transmission slip, any changes in fluid temperatures, tire temps, inflation pressure, and the overall mechanical condition of the drivetrain will all impact the ability to compare the results with previous runs or with other dynos. As a result, we measure these items on each run, ensuring they are proper and have not changed. We also ensure that road-speed air velocity and volume are provided to the engine radiator and charge-air cooler on each test. It takes at least a 15hp fan to do this correctly.
Many aftermarket companies use chassis rather than engine dynos. Chassis dynos measure wheel horsepower or the power delivered to the ground. While not as accurate as engine dynos, some chassis dynos are more accurate than others.
There are two types of chassis dynos, the first is a continuous-load-capable “real” dyno that loads the wheels with a power absorption unit that uses either water-brake or eddy-current power absorbers. These load dynos reasonably duplicate the work environment of the diesel engine in a truck. Load dynos are also capable of acceleration tests, accurately duplicating the inertial resistance a truck has to overcome to accelerate.
Load dynos can also accurately duplicate aerodynamic load, which becomes tremendous at highway or dragstrip speeds. In other words, the possibility for accuracy when testing with these dynos is very high. In the world of chassis dynos, these are dyno jewels.
The next type of chassis dyno that has come upon the scene is the inertia dyno. These low-buck units have no power absorber whatsoever and are completely incapable of running an engine at sustained full load. So, if you want to evaluate the power output of your engine the way the factory guys do it, you can forget these dynos.
The inertia dyno’s problem is inertia–more specifically, the inertia of the dyno versus the inertia of your truck. An inertia dyno can only do one thing: It can measure the time it takes to accelerate or spin a weight (usually a large steel drum) from one rpm to another. Now, I won’t get into calculations here, but realize that weight is only part of the inertial value. If the inertial value of the dyno is 2,600 pounds and the inertial value of the truck is 11,000-plus pounds, you begin to see the problem.
Now, some guys don’t care about all of this; they just want to know how well their truck works at the dragstrip. You know, you idle up to the line and the coolant, oil, block, heads, turbocharger, and charge-air cooler are not up to full-operating temperature. Then, you blast down the quarter-mile and get real-world results. Unfortunately, inertia dynos won’t duplicate this. Why? Because dragstrips are flat. To achieve a flat land evaluation of quarter-mile power, the dyno would have to have an inertial value equal to the inertial value of the truck.
Due to this mismatch of inertia, the rate of change of engine rpm in any gear will occur much faster than in the real world. The engine itself and all its guts, from the front pulley to the back of the clutch or torque converter as well as the rest of the drivetrain, also have inertia. The quicker the engine revs, the more power it uses to accelerate itself. And since the engine’s rpm is increasing quicker than in the real world, the turbo boost at any rpm would also be lower because the turbo is lagging behind due to its inertia. So, at any rpm, there is less power at the flywheel. In other words, the numbers are incorrect.
Finally, an inertia dyno doesn’t take aerodynamic drag into consideration. Even if the inertia did match, the aero-drag would not be there, so the numbers would still be incorrect. Due to their inertia gap, these dynos are like racing downhill–and a steep hill at that. Who would want to know his quarter-mile time and power output down a steep hill?
Some guys fix these low numbers with ungodly, correction factors, so these fantasy results fit the ego needs of the vehicle owner. Bottom line: These pretenders to the dyno world are nothing more than dyno jokes. Yes, I know, there are a lot of them out there, and there are a lot of guys who think they’re wonderful. In my opinion, someone needs to wake up the aftermarket before it standardizes the worst method of dyno testing, ever.
If you would like more information about dynos, check out: www.bankspower.com/dyno_testing.
Also in this issue of Diesel Power:
Riding the Big Hoss, by Steve Temple
Banks Goes Camping