I found this article on the InterNet and thought it was worth sharing.
Date: 05/18/2006
Author: Brian Ebert
Title: Dyno Testing - Facts and Myths
Details: This is the first of several articles regarding the building and tuning of high performance automotive engines. Hopefully, it will help you to differentiate the facts from the fictions that are far too prevelant in our information rich society today (i.e. - the internet). As with anything in life, good knowledge is of utmost importance in making decisions about your high performance automobile and it is my goal to try help you in this area. Our first subject.... dyno's.
A dyno (dynamometer) is a device for measuring the output of an engine. It is used for comparison TESTING..... I emphasize testing because far too many people have come to believe that a particular dyno output number is a goal in and of itself and then use them to compare with other numbers from other cars and dynamometers. Doing this generally leads to nothing but misconceptions and disappointment. The reason for this is in the procedure itself. Most of you have probably heard of the term "scientific method". It refers to the procedure of testing something by establishing a baseline in a controlled environment and then changing one thing at a time (to negate the effects of other influences) and noting the resulting of the change, or lack thereof. When we apply this to engine testing, it means we have to have control of all the things that can influence the power output of the engine being tested. This is no small task, as the list of things affecting an engine's power is quite large, such as - atmospheric conditions (temp., barometer, humidity), coolant temp, oil viscosity, oil temp, oil level in the oil pan, accessory friction losses (water pump, alternator, etc.), fuel type, fuel temp, and so on. When you run an engine in the car, you add a whole new set of variables such as trans type, trans oil type and temp, condition of u-joints, u-joint angles, rear end friction from ring and pinion, oil weight, oil temp and level, wieght of wheels and tires, tire inflation, tire rolling resistance and drag from the brakes. Therefore, trying to use power numbers from one shop's dyno and vehicle and comparing it to another shops dyno and vehicle without knowledge of all these variables is wasted effort. A dyno's true purpose is to allow an operator the ability to compare and optimize the POTENTIAL of a particular combination of parts. How that potential is used is a completely different matter (more on that in another article).
The next thing to understand is the difference in the way power is measured. We express power in two different ways: torque and horsepower. Torque is the measurement of static FORCE. We commonly express it in lb/ft. It is determined by applying twisting force to a shaft and then measuring the linear force it creates tangental to a bar one foot long radiating from the center of the shaft. Horsepower is the measurement of WORK performed, meaning the application of force over a period of time. It is computed by the formula 550 lb/ft per second = 1 horsepower. In our automotive use, working with rotational torque numbers, we get the well known algebraic equation HP= (TQ x RPM)/5252. There are two common ways dyno's measure power. The first is by using a water brake (like used on a SuperFlow dyno) or electrical retarder (such as on a Mustang dyno) to apply a load to a running engine, and then measuring the force required to hold the housing of the brake or retarder stationary. This determines the torque output of the engine, and by filling in the variables in the HP equation (tq and the RPM of the engine at which it produces said tq), we compute the horsepower to engine produces. The second is by inertial load testing. An object of known inertial mass (such as the drum on a DynoJet 248) is accelerated by an engine under power. Through a mathematical formula, the rate of acceleration of the drum tells us how much horsepower the engine is producing. Then, by filling in the equation again, we can determine the engine's torque output. Note that since the load bearing dyno changes the load on the engine to control it's rate of acceleration and an inertial dyno's rate of acceleration changes with horsepower input, it is impossible to compare results from the two different methods of measurement without knowing the rotational mass and frictional losses of the crank, flywheel, transmission, tires, etc. One method of power measurement is not necessarily better or worse than the other, but they ARE different.
The last thing I need to address may be the most important.... the variables. This is where you will really see a difference from dyno center to dyno center. As I stated earlier, controlling the variables is VERY important in engine testing. Let's face it, if you allow more than one thing to change from test to test, how can you tell what caused the change? Some variables are easy to monitor and control such as oil temps and levels, tire pressures, etc. One thing we can't control however, is the weather. It is constantly changing, even from minute to minute. Fortunately, we have a solution for this problem......IF it is used correctly! That solution is called the correction factor. It's purpose is to cancel out the changes in an engines power output based on atmospheric changes. Simply put, it is a mathematical formula that corrects the observed power output of an engine to what that engine WOULD produce given a standardized atmosphere. This way, a dyno operator can compare results taken during two or more different periods of time and different weather conditions, and if all other variables are kept the same, the corrected results should all be the same even though the daily results were different. To this end, two sets of standards are commonly used: Standard Correction and SAE standard J1349 rev. June '90. Standard correction uses 29.92 inches on mercury (Hg) at 60 degrees F with no humidity as the base atmosphere. The SAE standard uses 29.23 inches Hg at 77 degrees with no humidity. (Please note that these formulas are for naturally aspirated engines - forced induction requires a different formula although rarely actually used). Naturally, the engine corrected to the higher barometer and cooler temp will show more power - it doesn't really matter which correction factor is used, as long as the same one is always used when comparing results. This leads us to what I believe is the most overlooked factor in dyno facility configuration - proper ventillation. Since any changes in atmosphere can have drastic effects on power output (by changes in temp, humidity, baro, or contamination of intake air from exhaust gasses), a stable environment is critical in dyno testing. This point is emphasized by dyno mfgs. such as SuperFlow, Land & Sea, DTS, etc. in their recommendations for properly ventillating a dyno cell. It should be noted that these recommendations are for testing an engine on a stand AND in the car on a chassis dyno. Yet, while many engine dyno facilities have good ventillation systems, it is VERY rare to see a chassis dyno cell that is properly ventillated. What is proper ventillation? According to a Land & Sea article, http://www.land-and-sea.com/dyno-tec...ell_design.htm, (and similar from SuperFlow), WITHOUT dealing with radiator, driveline and exhaust system heat, a dyno room's air should be changed at the rate of 2,000 cfm per 100 HP! That means that when testing an engine making 500 HP on a chassis dyno (WITH the radiator and exhaust system heat to deal with), the cell should be vented at a rate of approx. 15,000 cfm! A simple fan or two in front of the car does NOT accomplish this task. Since the weather readings for the correction are seldom taken directly in the path of the engines intake air, but somewhere else in the room, the only way to keep the readings accurate is to keep the entire room fresh by venting it at the massive rates just mentioned. Anything short of this diminishes the possibility of getting repeatable results. At our shop here in Minnesota, we have a 13,000 cfm fan in our current test cell, and will double that at our new shop this fall. This is why we can make back-to-back tests with no changes and have the same results within 1 HP. Remember, if you can't repeat within 5 HP, how can you be sure that timing change you just made was responsible for the 3 HP gain on the next test?
To sum it all up, it is time and money well spent to find a properly equipped dyno facility and an experienced operator to get the most out of your combination. Also, try not to spend TOO much time worrying about other people's 'numbers' when you don't know all the facts about how they were derived. A 'number' does not necessarilly equate to a certain on-track performace. More on that in another article. As always, please feel free to contact us if you have any questions regarding this article or anything else about your vehicle.