Variables to control when doing G-Tech runs

Spook, bless his heart, tried to evaluate the effect of ambient temperature on the performance of his SSR, and posted the results on this board earlier today or yesterday.

Spook, you then commented to me in one of your posts that you thought you had isolated that effect. I'm not so sure.

The problems I see in what you posted are

1. The difference is really big between the 2 temperatures - too big I think

2. Both results are too good for a bone stock SSR with a 200 lb driver and fuel, even with a pro driver

3. There are LOTS of variables that need to be controlled to make a G-Tech run:
(a) accurate
(b) consistent (which goes a long way towards proving the accuracy)
(c) comparable to an earlier run on the same vehicle before latest mods
(d) comparable to another SSR

The variables that need to be controlled include at least (pardon me if I miss some - I'm pretty tired tonight):

- Verifiably level road. You CANNOT do this by eyeball - use a laser level or a LARGE ball at a minimum, and take measurements at MANY points along the run. The human eye looking down a road positively cannot detect gradients even in the 1 to 3% range, and those are large enough to blow the test results badly

- Straight road (a curve activates the differential and adds road friction)

- NO wind or at worst 3 mph, no more. Cannot have headwind, tailwind, or crosswind

- Temperature as close as possible to an agreed upon standard. 70 degrees is a good one. Even small differences in temperature skew the results enough to hide all but the most successful mods

- Humidity as close as possible to the norm in your test area (If you live in the desert southwest, it's going to be low most of the time, and if it's Atlanta, it's going to be high virtually all the time, so you can't standardize across the country, but you NEED to standardize in your own testing

- Barometric pressure needs to be as close as possible to the norm in your area for same reasons.

- You can't have gravel or lots of dust or leached pavement oils on the road. It makes consistency impossible. Anywhere near construction is a disaster because of the fine dust generated

- Exact same amount of fuel in the fuel tank each time, and since I am interested in REAL world performance, to me that means 1/2 tank

- No extra "stuff" in th car. Consistent weight is crucial

- Obviosuly, no passenger! You think I am kidding. I've seen people try to do runs with a buddy watching the instrumentation. That extra weight puts the performance on a different planet

- The convertible top must be UP! The aerodynamics of an open convertible are horrid.

- The AC and fan should both be off. Yes, I know that under full throttle, the PCM is supposed to shut off the AC, but there is a lag, right at the worst time: the launch

- The engine should be at the same temperature just before the launch, each time. Otherwise, the results will vary a lot

- The transmission should be at the same temperature just before the launch, each time.

- The number of options present or absent on the vehicle has a profund effect on curb (NOT just "dry") weight. To be a meaningful set of data for comparison, your vehicle needs to be reasonably typical (Comparing a loaded SSR to a barebones one is futile)

- You HAVE to get an accurate weight, using either 4-point track scales (best) or at a minimum, a truck stop scale reading, WITH the actual test driver and everyhting (and ONLY everything) that will be aboard for EVERY test done going forward

- The rollout distance should be set to the same value each time, and since the default for a G-Tech is 12 inches, and since that is consistent with a shallow stage at a dragracing track, that is a good standard

Once you have these ALL consistently under control, you have a good START. Then, you have to make multiple, not single runs, for each scenario. By doing this, you either verify that you HAVE good control, or you prove to yourself that you do not. If my repeat runs, when graphed by the G-Tech, do not cover each other virtually identically, I consider them unreliable and discard them less I be tempted to try to use them as valid data.

Obviously, the best way to do this is at a track, BUT you need to go on days when the turnout will be poor due to a competing event elsewhere, etc, or you won't have control over the starting engine temp, tranny temp, or even possibly the ambient temperature.

If you use a different "unofficial" location, you should always use the SAME location, for each set of runs, and for each run in each set, and just to be sure, in the same direction each time (in case there is a tiny slope too small for your crude tools to reliably detect). Obviously pick a spot where you will FOR SURE be ALONE (except for buddies NOT in your vehicle) so that if you screw up, no one but you gets hurt. And, don't do like I did in Atlanta and accidentally set up your lab 1/4 mile from a police shift change station.

THese are just some of the SETUP basics.

Then, there is technique, which I am too tired to write about right now, and don't know enough about anyway.

Then you get to recordkeeping, which has to be DETAILED (yuk) and absolutely METICULOUS in format and content and precision of the metrics. Otherwise, you are wasting yuor time and worse, you are deceiving others with bad data.

When I look at other peoples' data I look at the number of repeated runs, the consistency of each of those runs, the attention paid to each of the variables listed above, and the track record of the poster on detailing, accuracy, reliability, and trustworthiness. You would be amazed at how many national magazines fail sometimes (and some of them more often than not). I look FIRST at the terminal 1/4 mile velocity. If it varies by even 0.5 mph from run to run, there is a SERIOUS uncontrolled variable, as that mph number should be identical even if bad technique or other issues disturb the elapsed time from run to run.

One good stating point that tells you something is wrong is when a heavier than average test driver was used, the vehicle used for testing was heavy on options, the tank was full, and the test results obtained are BETTER than the manufacturer was able to obtain with a lightweiht driver, a PRO driver, driving the lightest variant of the vehicle, with minimized fuel and other fluids, on the absolute best day, best track, and best other conditions that the manufacturer was able to find or produce. Since the manufacturer has the absolute highest motivation to get the absolute best results possible, evenif it means abusing the vehicle (nothing to them - they can replace anythign they need to on it at no real cost!), if someone with all the DISadvantages listed does better than the manufacturer, I conclude that there were indeed uncontrolled variables, and look elsewhere for usable data.

GM claims, for the 03 and 04 SSR, something like 0 to 60 in 7.5 seconds, and the 1/4 mile in 15.9 at 89 mph. You can be certain that they moved Heaven and Earth to get that set of numbers.

For an illustration of poor control of variables, look at the graphs I publsihed for the Bad Boy loaner Trailblazer I did the SSR fanatics first road test on. I am on vacation. I don't have a laser level with me, I had to take the best road sections I could find in the mountains of all places, I had to use 2 different sites because of the intense police patrol activity there, and I did NOT have the luxury of waiting for the exact ideal ambient conditions I wanted. Plus, at 7000 feet elevation, the results should be a little low anyway (but not as low as they turned out!!)To you, the two runs shown there might look acceptable. To me, they are marginally usable, but then I didn't need really precise results to prove what my seat of the pants dyno had already told me: I got issued a slug of a vehicle that simply did not perform anywhere clsoe to how a 275 hp vehicle should. If I needed comparable and quotable results, I would have passed on the testing. But, I was merely doing an entertainment piece so it did not matter.

Anyway, just a few thoughts.

Jim G