 |
Unfortunately, seeing a dyno sheet of a particular modification on a vehicle is not necessarily a representation of what it does in reality. Dyno data can be intentionally or unintentionally manipulated in many ways.
Temperature: engine temp *greatly* affects power output. Dyno runs done during different levels of coolant, engine and oil temps will have drastically different results, with the cooler runs typically producing more power. Some engines, such as the forced induction vehicles, can lose up to 15% of total hp and torque as temps rise. Thus a colder run done after a particular mod will exaggerate gains, if any. Redirecting external cooling fan air can have tangible effects on data, especially when air is concentrated on intercoolers, open air filters, etc. The best dyno operators will use high velocity fans (not just high volume fans) directed at critical areas of the engine and drivetrain in order to mimic real world conditions as much as possible.
Gearing: a chassis dyno is susceptible to gearing bias. Recorded power can be affected on a dyno graph by simply running the test car in too low or too high a gear. Dyno operators should pick the trans gear that is closest to a 1:1 ratio to avoid gearing bias. This ratio is 4th gear in most cars. In a 1:1 ratio, the two gears that make up a gear set are the same size. When one is larger than the other, either in a lower gear than 1:1 or a higher gear than 1:1, there are frictional losses that translates to less power put to the wheels. Also, too high of a gear that causes the test car to go to a top speed far greater than the available fan cooling capacity will produce invalid results.
Atmospheric conditions: all dyno runs should be corrected for atmospheric conditions, typically humidity, barometric pressure, and ambient temperature using the SAE correction factors. Neglecting to correct for these variables can skew data, sometimes purposely. However, forced inducted cars should only correct for humidity and ambient temperature as the pressure inside the intake manifold is much higher than the ambient barometric pressure. The SAE correction factor for barometric pressure assumes a naturally aspirated engine.
Fuel composition: most modern vehicles have multiple ignition timing maps to allow the fuel injection ECU to compensate for the varying qualities of fuel found in their intended markets. For example, the US Spec Porsche 997S has four ignition timing maps to choose from depending on the amount of measured knock activity present. Thus, on 91 octane RFG (Reformulated Gasoline) fuel found in California and other parts of the US, the fuel injection ECU jumps to a less aggressive ignition timing map, resulting in a lower power output than on the 93 octane fuel found in other parts of the US. On some highly modified cars, fuel octane and composition can become even more important, and the positive effects of high octane racing fuel can make a tremendous difference. When viewing a dyno sheet, it is best to determine what fuel the test car was run on.
Manual graphing: in reality, dyno data can be totally fabricated. There are many software programs that will generate graphs with any data inputted, such as Microsoft Excel. Try to request actual dyno sheets, bearing some sort of external run data.
Our Mustang AWD-500-SE 4WD chassis dyno is the second chassis dyno that we have owned, and it is rated at a maximum hp load of 650(800 in 2WD mode.) However, we can load it from 0hp (simply spinning the attached flywheel and rollers like the Dynojet) up to 650hp and anywhere in between. We can specify a time period over which we want a load to be applied. We can simulate a series of hills, as the car goes up and down in elevation, which obviously varies the load on the engine. We can set a specific rpm at which the dyno can apply a specific load.
In other words, the load application on our dyno is practically infinitely variable, and can be as punishing or gentle on the engine as we want. All controlled by the Windows based software on the controller PC, with hand held operator controls. For development use, the load ability is crucial, to produce a product that will perform correctly in the real world.
This is all in contrast to the Dynojet inertia type testing, which simply requires the car to run on the rollers from standstill or low rpms, to redline or anywhere in between (in one or several gears). The rate of acceleration is measured, and power is mathematically calculated using the constant of roller weight. The time it takes to accelerate a given mass to a particular speed can be used to calculate hp. It is relatively painless to the car, and quickly measured. Plus, the Dynojet software allows automatic graphing, which produces an attractive consumer product.
However, the downside to this measuring form is that, in the real world, the vehicle is working harder than what is required in accelerating the rollers on a Dynojet. Between the weight of the car itself, elevation changes, and surface friction, the actual environment the engine works in during use is quite different from its experience on an inertia type dyno. This is an important factor to consider when interpreting the hp gain of a particular modification or performance product. For example, lots of ignition timing can be thrown at an engine to produce big hp numbers on an inertia dyno, but the same ignition setting can result in less hp when the engine is actually required to work on the street. Under load, (on the street or on a load dyno) pre-ignition or detonation can occur, tripping the knock sensors and causing the ECU to roll back the timing to less than stock. This is not good for power. Another example is fueling. Leaner fuel maps may work well numbers-wise when subjected to the relatively gentle inertia rollers, but can be disastrous when under load on the street.
Another popular dyno that has appeared on the market in recent years is the Dynapak. Its attractive attribute is that it has individual power absorbers that bolt directly to the hubs of the test vehicle. This is attractive to shop owners in the fact that when the absorbers are not being used, they can be rolled away into a corner. This design takes up a lot less floor space than a traditional chassis dyno. However, the major downside of this design is that it is in fact not measuring power at the wheels, as the wheels are not even on the car at the time of the test! Thus, the extra weight and the load that the wheels and tires produce, especially as wheel speeds increase, is conspicuously absent. This will have a very significant affect on how hard the engine is working, as well as presenting the question on how to interpret the data. Power to the wheels means power to the tires, which is representative of the power that propels the car down the road. Power to the crank is representative of the power than the engine makes with no other drivetrain components bolted to it. Power to the wheel hubs is representative of what?
A final note to consider is that comparing wheel power numbers across different dyno brands and even across the same brand of dyno at different locations should be avoided. Comparing across dyno brands should be avoided for the very fact that different brands of chassis dynos actually measure power in different ways, as explained above. Comparing across the same brand at different locations should be avoided unless it has been determined that the dynos in question have been calibrated in the same manner. The software controls of chassis dynos today allow key variables to be changed such as the calculated weight of the dyno rollers or the way that the control software filters electrical noise from the incoming signal of the dyno. Changing these variables can have a significant impact on recorded power numbers.
We hope that you find this informative and useful!
|
