Topic: Uncorrected Dyno Graphs

[Jitenshakun]

By Patr....

i dont normally post here (pro lurker) but since someone was asking I thought I would give a detailed answer.
 
we dont publish or compare corrected numbers. Thats why tora posted that those were corrected numbers. just FYI there are different kinds of correction that dyno operators use. Some use SAE correction, which is totally invalid for turbo cars. Some run a base correction for their ambient barometric pressure (i.e. altitude). Some run a base correction for the ambient temperature. Some run a base correction for the ambient humidity. Some run a base correction on all of the above. None of it matters if you are doing a tuning session and normalize all of the above (see below), but to compare apples to oranges as people like to do, its all worth mentioning.
 
Now, if a customer asks for a corrected graph to a certain other type of dyno at sea level, we can provide it for them. We know what the differences are because we've had the same car on a whole slew of different dynos (all at sea level). When we do all of our internal comparisons, or put dyno plots on websites, and when we talk about cars, we dont put up any corrected numbers at all. If someone else wants to put up whatever graph they want, we cant control that. Most people do it because there are a whole bunch of Dynojet posts on a partcular internet forum and people like to brag about this or that. But those in the know understand that xxx is good on yyy dyno.
 
A bone stock WRX will pull NORMALIZED 165-175 on our dyno. The variance is due to cars as much as the three mentioned things above. A bone stock STi will pull NORMALIZED between 210 and 220 onour dyno. Same reasons for the variance. When I say NORMALIZED I mean stable ambient temp, stable intake air temp (measured on the car via ODB-II as well as on the dyno via thermocouple), and stable coolant temp, stable ignition advance tables, stable air fuel correction, etc. and a CONSISTENT ramp/acceleration rate, and a CONSISTENT WOT hold period from a CONSISTENT rpm, and a CONSISTENT tire pressure, gearing, and method of operation. If one wanted to, one could VERY EASILY make a stock STi pull 200, or a stock WRX pull 150. The point being, if you dont normalize stuff, the tool is basically a useless toy. Even on an accurate dyno, with all of the above normalized, depending on the tuning strategy employed on the car, you can often see a variance of +/- 20 ft lbs if you do the pulls in a certain way (i.e. hot, or long pauses at full boost before ramping, etc. etc.)
 
Having used Mustangs, Dynapacks, Mahles, Dynojets and DDs with the same car, I can say for sure that at least the Mustangs in seattle and portland read about 18% higher than our dyno (they are both at sea level, in any case), and a sea level Dynapack and Dynojet is off the chart (35%+). I think Blaine's dyno should read similar to ours for stock cars, and probably closer to ours on higher boost. 220 for an STi would be right assuming its similar to the Mustangs in Seattle and Portland, with the higher altitude factored in. But there is a lot at play.
 
Here is just one example (there is more to it than this, but this is just one example). If you drive on the street and have the ability to read OBD-II, take a look at your intake temps while driving in whatever gear you plan to do your pulls in. Then take a look at them before and during your pulls. If a highly boosted car is getting hot air, it can pull timing (it can be dangerous too but that is another story). Furthermore, not only will heat increase intake temp charge, but it will also result in less dense air, thus to make xx psi the turbo has to work harder than it did in colder air, and so its out of its efficiency range. On a highly tuned car, or a car tuned to the edge, this can result in a significant difference pull to pull. Just as an example, when I do a final "full pull" test on a car (a hot worst case test), if its highly strung you WILL see a drop of up to 20ft. lbs. YOU WANT THAT. Otherwise, you are leaving something on the table. Its all ok as long as the car does its job, and it adds power then it can and pulls it when it cant without any side effects.
 
But NONE OF THAT MATTERS if a car pulls say 20 psi on the street, then only pulls 18psi on the dyno. This can usually be addressed by using a higher gear or slowing down the acceleration ramp rate (if the dyno allows you to), but both of those solutions come with the added side effect of significantly increased heat dissipation requirements. If you aren't simluating what you are doing on the road, the tool is only useful as an excersize in curiosity.
 
We will normally do our STi pulls in 4th, but on all cars tuned I always "pull them out" in 5th and sixth and then run 4th at a super slow ramp rate. This is to test the 'worst case' heat scenarios. And yes, the numbers are lower when you do that. If we made that the NORMALIZED basis, then of course we could do that to all cars and have an "even lower" basis, so then when you publish a nice "cold pull" you could get even more props. But that would be against the premise of it all. We try to publish pulls taken at coolant temp of 92-95, intake temp of 25-29, ambient temp of 70F (or higher), with a 4 second full WOT hold at 2500rpm then a 8-12s ramp, with an ECU not "tricked", and at least 2 minutes running at 2000-2500 rpm prior to the pull. With the right cooling setup this is very close to what you'd get on the road, depending on which gear.
 
Those numbers are published uncorrected. We dont use correction for anything else. If a customer wants a graph, and we see it published, we are the first to say its corrected (as was done here).
 
Long response to a short question, eh ?
 
-Pat