How We Test - Vehicle Performance TestingPart 1: Power, Acceleration And Braking
Photography by SCC Staff
writer: Jay Chen






We are the only magazine in our segment that provides full vehicle performance testing. We think it's important to call out vehicle manufacturers and tuners alike and print real numbers, so our readers can make informed purchase and modification decisions. However you use this information, be it for fun or for guidance, our testing methodology should be crystal clear. As engineering editor, I will attempt to make it so...
Theory and PracticeTesting is a lot hotter and less sexy than you might imagine. It takes at least two people all day at the race track or stadium parking lot to set up cones and timing gear and test and test and test until we have repeatable results we're satisfied with. That's just for one test on one car.
Our testing mindset and method is different from mainstream books like Motor Trend and Automobile. Not only are the cars we test different (we don't count test minivans or count cupholders), our aim is to gather real-world results and data that's relatable-and repeatable. Indeed, it's more likely an SCC reader will dyno or quarter-mile their own project car to see how it stacks up against one of ours.
Now that test equipment like data acquisition systems, GPS-based telemetry recorders and even dynos have become cheaper and accessible to the average guy, quantitative performance measurements by readers are possible. But testing still requires time, financial resources, and most of all, experience that few people outside of race teams or engineers possess. Simply having the right equipment isn't enough-you need to have the right track, driver and geek to generate useful data.




Things get even more hairy with aftermarket tuning. There are countless variables that testers of stock vehicles can ignore. But anything that can be tweaked usually is. Things like boost curves, fuel, tire pressure, spark plug gapping, suspension settings and so on add to the long list of things we have to put into the equation.
Tuners can spend hours making minor changes at the dyno or track, playing with tire pressures, alignments or even cool-down times trying to squeeze out that last 10th. We test each car we receive from a tuner as-is. We expect them to have enough common sense to bring a car that's dialed-in and ready. That rarely happens, but our job is to provide the fairest conditions we can as we try to generate repeatable numbers from usually inconsistent tuner cars.
Our core tests and procedures have not changed since the day SCC got its first radar gun and timing lights, even though we constantly explore new ways of making our tests faster, more accurate and (most importantly) relatable. Here are some standard tests SCC performs:
Dyno TestingThe fact we measure power on a chassis dyno should tell you that this is a flawed means of measurement already. But drivetrain loss variables aside, we stick with Dynojet's 248 series two- or four-wheel inertial dyno because it is the simplest, non calibration-dependent and abundant dynamometer around. All our dyno graphs show wheel-horsepower numbers generated on these dynos, even though other eddy current dynos-like the Maha-may be more sophisticated in how they acquire an accurate power figure.




For the uninitiated, the Dynojet is just a large drum roller of known properties that the car has to spin like a giant gerbil wheel. The faster a car can accelerate this heavy drum, the more power it makes. It's simple and leaves little guesswork or variability.
But there are problems. One we frequently run into is airflow. The fact that the car is stationary means it's not getting enough cold air for cooling or the intake charge. Few dyno facilities have fans that come close to replicating true airflow at 80mph. So to cope with this unrealistic scenario, we perform dyno pulls with the hood open and spray down intercoolers, then wait a minimum of five minutes between runs. It's a compromise, but it helps even out the numbers.
We've also never used SAE (Society of Automotive Engineers) horsepower correction. Long story short, when we started working with Dynojet's Winpep software, it didn't offer SAE atmospheric corrections as an option. Since then, we've stuck with it to keep our dyno charts consistent, even though it corrects to a higher ambient pressure compared to SAE.

This just means our power figures are something like 0.3 percent higher on the Standard (STD) correction compared to SAE. And for all the carping about proper horsepower figures (if such things exist on a chassis dyno), keep in mind most of our dynos here in Southern California are within 200 feet of sea level and our dyno conditions don't wander too far from standard temperature and pressure. We might move to SAE corrections if we change the format of our dyno charts in the future (see Appendix J, page 24), but even with SAE's power correction factor applied to our numbers, the results will still be off, since cars today are tuned to run differently based on ambient conditions.




For those familiar with Winpep, we also use a smoothing factor of three out of five to filter out noise and spikes in the data. The more smoothing used, the lower the peak horsepower appears on the graph. As with most of our tests, we have to achieve the same numbers twice before results are published. This means a lot of pulls for some temperature-fickle turbo cars.
Acceleration and BrakingEven with fancy GPS or dash-mounted accelerometers available, we stick with the Stalker ATS radar gun, like the ones used by the cops. We've avoided using accelerometers because they require added calibration time per vehicle and are susceptible to data drift. GPS systems also have similar problems because many integrate accelerometer data to calculate the results. They are also ridiculously expensive and time-consuming to process just simple acceleration data. Although these systems are great for consistent road course telemetry use, the shock loads and huge changes in motion from braking and acceleration runs make these systems inaccurate from dead starts or coming to a complete stop.
The Stalker radar gun measures and records a vehicle's speed once every 20th of a second. The software it comes with takes all this data and figures out how long it took to move a set distance, or to reach a certain speed. There are several options for calculating the quarter-mile ETs and trap speeds. We use NHRA standards for quarter-mile trap speeds and ETs, but do not consider reaction time or use a one-foot rollout. We're testing the car, not the driver. The quarter-mile time starts when the car starts rolling (we don't use a light tree). Each car is tested with different combinations of throttle and clutch modulation and shift points until we see trap times plateau.Then we round the ET to the fastest repeatable 10th of a second to pick out the best run.




More importantly, we don't apply altitude or atmospheric correction factors, even though we test at the same track as most other magazines. Since driver and condition inconsistencies will easily outweigh most correction factors-especially for aftermarket cars-we print what we measure. We do record test conditions and take notes on each car's behavior, as that gives us an idea why a number might look slow. Ultimately, what you run is what you get, even on a hot dusty day in a 20mph headwind. This is why our test results for new cars are typically slower than those of other mags.
Typically we perform a braking test from 80mph with the same Stalker radar where braking is applied at 85mph (so that response time and transient effects are not pertinent). Although braking distances from 70mph and 60mph can be extrapolated from the same run, an exponentially larger amount of energy gets stored in the brakes when you start the braking at 80, thus making the number inaccurate. So we've thrown out the 60mph braking test altogether. Depending on the car and braking system (ABS or non-ABS) we look for the same repeatability within several feet before we accept a result.
For big brake upgrades on project cars, we perform a completely different set of repeated tests to measure distances and rotor temperatures. Here we look for consistency in distances and how the front rotor temperatures change (when compared to stock brakes we've baselined on a previous occasion). Both the stock brakes and big brake kits are tested on the same tire at the same location under similar conditions.
Hopefully you have some idea of what we do on a day-to-day basis and the effort put in to provide accurate, real-world results. In the next segment, we'll get into our dynamic handling tests and track shootouts. Until then, keep our methods in mind next time you read about horsepower or quarter-mile times published in other magazines. If they can't tell you how they arrived at those numbers, you can bet they can't vouch for their accuracy, either.