Dyno Accuracy Testing - Dyno Scruples

It's not often that you reveal the punch line in the opening paragraph of a story, but this is one of those rare occasions. While there is still no substitute for real-world track testing, chassis dynos are great tuning tools that allow dialing-in a vehicle in a safe and controlled environment. Unfortunately, instead of being utilized to help maximize on-track performance, chassis dynos are replacing tracks altogether. These days, many enthusiasts would rather compare dyno printouts than timeslips. However, there is very little parity when it comes to different types of dynos and how they measure horsepower, so comparing readings between them is futile. Furthermore, although dynos themselves are remarkably precise pieces of machinery, accuracy is limited to how well an operator can control a multitude of test variables. Simply put, unless you're willing to go to great lengths to level the playing field, like we do for the Jeg's Engine Masters Challenge, dynos are for tuning, not racing.

Ironically, to find out just how much dyno numbers can be manipulated, you need a really accurate dyno to spot the swing in power. For that, we sought out the highly accurate Mustang Chassis Dyno at Performance Turbo Motorsports (Belton, Texas). Our victim vehicle was a stock '99 Corvette six-speed with a high-flow air intake and an after-cat exhaust system. That's a good thing, since a relatively low-horsepower, naturally aspirated, internally stock motor should produce rather consistent power output from pull to pull. After establishing some baseline numbers, variables such as tire pressure, location of the external cooling fan, tie-down strap tension, and engine cooling were altered first individually, then cumulatively.

BaselineWithout solid baseline numbers, it's impossible to gauge changes in horsepower in any of the subsequent pulls during a dyno session. PTM's Mustang MD-600-DE eddy-current dyno loads the rollers to simulate vehicle weight and wind resistance. PTM's normal test procedure calls for raising the hood, and positioning the external cooling fan roughly 6 feet in front of the bumper. All pulls throughout the entire test were made with coolant temperature between 205 to 210 with 10 to 5 minutes of cool down time between each run. Two pulls were performed after changing each variable, and the published results are an average of those two figures. A testament to the MD-600's repeatability, each of the three baseline dyno runs were within one-half hp of each other for a final average figure of 273.5 hp at 5,600 rpm and 284 lb-ft of torque at 4,300 rpm.

Hood DownSimply by closing the hood, peak horsepower and torque plummeted to 257.9 and 265.5, respectively. Coolant temperature also soared to 246 after the second pull. This really isn't a shocking revelation considering how much closing the hood traps heat and restricts airflow. Granted, you don't drive down the road with the hood up, but a quaint external cooling fan simply can't simulate the airflow a car experiences on the road. For example, at 6,000 rpm in Fourth gear, our test mule would normally have 130 mph worth of air flowing through its cooling system and air intake. On the other hand, the external fans used in most shops would be lucky to generate 20 mph air speed. "Most fans are undersized and only move 2,500 to 5,000 cfm of air," says Michael Caldwell of Mustang Dynamometer. This, in turn, prompts the PCM to richen up the air/fuel mixture and retard timing to prevent detonation in response to extreme coolant temperature and a heat-soaked inlet air temperature (IAT) sensor. "It doesn't take long for things to get too hot with the hood down, which is why we recommend testing with the hood up." Although a carbureted motor may experience less of a horsepower loss, it would also be much more likely to detonate.

Cooling FansThere is no universal rule for where a dyno facility's external cooling fan must be positioned in relation to the test vehicle. The question is, can it really impact horsepower output? For this test, the fan was moved within 2 feet of the front bumper. In addition, a second fan was pointed right at the air filter. As marginal as the increase in airflow may have been, power did increase to 274.8. "Fans are used for cooling and can also replicate the ram-air effect a car experiences while moving down the road," says Mustang's Caldwell.

Tire PressureIt's not hard to conceptualize how decreasing rolling resistance can increase horsepower. To test this theory, we inflated the tires from 32 psi to the recommended maximum of 50 psi. Surprisingly, horsepower dropped to 268.5, perhaps due to tire slippage. While torque peaked at 4,300 rpm in the baseline pulls, it dropped down to 3,700 rpm during this test. Conceivably, tire pressure variations do not impact street radials all that much at this power level, and any decrease in rolling resistance was offset by reducing the contact patch. On a side note, that's been proven at the dragstrip time and time again, and now (at least in this test) on the dyno as well. "Every tire's different and you can't make blanket statements, but air pressure and the load placed on a tire can vary results quite a bit," says Caldwell.

Strap TensionFor safety purposes, vehicles must always be tied down with straps to hold them securely in place throughout the duration of a dyno test. However, there is a difference between "extra tight" and "just tight enough." Less tension equates to lower rolling resistance, which should increase power. For this test, strap tension was decreased and horsepower increased to 275.6. Nonetheless, it is inadvisable to compromise safety by loosening tie-down straps in search of a few extra horsepower. Leave stupid antics like that to us.

The NumbersAt the end of the day, dyno tricks were worth a total gain of 7.5 hp over the baseline numbers, which equates to a change of 2.7 percent. With such a marginal change in power, is this test legit? According to Michael Caldwell, "With horsepower changes of 1 percent or less, it's really inconclusive as to whether or not that reflects an actual change in power." Granted, each individual trick that resulted in power gains fell within that 1 percent window, but the cumulative effect of combining them into a single run yielded gains too large to attribute to normal run-to-run variation. Furthermore, if the same 2.7 percent increase in power was applied to a far more potent motor, the difference in horsepower would probably be much more significant. As unscientific as this test may be, the bottom line is that even the best dyno equipment is only as accurate as its operator permits, and if shops want to fudge dyno numbers to promote the perceived effectiveness of its parts, it most certainly can.

Hero RunTo analyze the cumulative effect of each of the variables tested in this story, two final pulls were made with every dyno trick employed. Since increasing tire pressure decreased power, it was left at 32 psi for this test. To recap, our hero runs were performed with the hood up, cooling fans positioned close to the car, and some tension removed from the retaining straps. As expected, these changes yielded the highest peak hp figure of the day at 281 hp.

Engine DynosThe best way to eliminate the variables involved in chassis dyno testing is to test on an engine dyno instead. While chassis dynos don't require pulling the motor and measuring the actual hp that gets to the tires, when it comes to precisely measuring incremental horsepower changes, an engine dyno is still king. "Instead of measuring power once it goes through the driveline, an engine dyno measures right off the flywheel which is why it's so accurate," says Dan Roberts of DTS. Likewise, variables such as tire pressure, strap tension, fan location, and engine heat soak are eliminated as well. "There are some variables as far as the climate that the engine runs in, and there needs to be a sufficient flow of air to shed radiant heat from the motor and to feed the induction system. All of these issues are easily controlled with the ventilation systems and high-power fans in place in a dyno room." The only variables left, then, are coolant and oil temperature, which can be controlled much more easily with the engine in a cradle than with it in a car. Engine dynos also have the flexibility to feed the induction system its own separate supply of air. "To reduce the effects of correction factors in extremely hot or cold climates, you can direct air into the induction system from a climate-controlled room or turn the dyno room itself into a climate-controlled environment." Again, horsepower variations aren't so much a mechanical shortcoming of chassis dynos, but with fewer variables that can impact performance, it's much harder to cheat on an engine dyno.