Simple Performance Check

[Siestarider]

With all the complicated ways we can evaluate performance and how it is increased or decreased with modifications to our cars, the first one I used and still use is also the simplest. I use the accessport logging function and calculate how long a full throttle pull takes.

Just log a pull in one gear and show the data in a spreadsheet. Pick an rpm range that you reliably have 100% throttle, I use 2500-6500 rpm. I.e. I always make sure I am at full throttle before 2500 rpm, and don't let off till past 6500 rpm. Interpolation between times is required, they never record exactly on 2500 or on 6500.

Take your 6500 rpm time, subtract the 2500 rpm time, and voila. The time tells the truth.

For example, when I was stock and had added a Mountune air box and 2 J downpipe, I logged and found no difference in delta t. What? I had improved breathing, but the data did not measure it.

Duh. Our cars will not respond to breathing improvements without revised ECU programming to take advantage of them. Hence loading a suitable Cobb tune was required to drop the time interval between 2500 to 6500 rpm.

My stock delta t was 8.6 seconds. Once I got lighter tires and wheels, rear seat delete and hard parts required for running Cobb S3 tune, I got down to 7.6 sec. Could not improve over that with 3" exhaust vs stock exhaust, so put the 3" aside.

I tried to improve by using Panda pro-tuning, which after two revisions failed to show any improvement in this simple measure of performance. Still fighting with them over the money I paid. I had the impression they promised improvement over Cobb S3, but the data just did not show any improvements, even with their E30 tune. I guess Panda does not believe in simple math.

Subsequent to that fiasco, I went back to Cobb S3. I also remounted my old 3" exhaust and fooled with it enough to make it barely streetable for me. Meaning 3" straight through Vibrant resonator and Vibrant muffler, and found a way to dampen resonation.

My latest Cobb S3 V220 logs consistently show 7.15 sec for 2500-6500 full throttle pulls. I honestly do not know how much of the improvement is due to 3" straight through exhaust vs Cobb's revised tune, but its a good combo for my car.

As one of my first posts on this forum says, F=MA. If M is unchanged between pulls, and A does not improve (delta t), then you did not get more F.

Obviously there is a whole lot of utility to real dyno's and VD comparison of curves for tuning, this technique is more like running a timed 1/4. You either make the car quicker, or you don't. For example SS's car is obviously quicker than mine in a 1/4 mile.

I calculated the SS 253 whp log time just for fun, 2500-6500 is about 7.0 seconds. This tells me is this particular simple measure of performance does not transfer very well car to car.

But for your own car, its a simple and useful method of determining whether you added F by mods.

 

[Hijinx]

I feel like this is way over simplified.

 

[Siestarider]

I feel like this is way over simplified.

Which part of F=MA is too simple?

 

[Hijinx]

Which part of F=MA is too simple?

Nice.

 

[FearTheSnail]

On the MazdaSpeed's they always used 60mph-100mph because it was a perfect pull in fourth. Now, for those of us with big turbos, this 2500-6500 pull that you are talking about will be very skewed compared to someone with a stock turbo.

I will take a virtual dyno chart any day over how long it took me to go from 2500-6500. As long as you are not using a sloped road, and you have no bumps on said road the virtual dyno can be very accurate. If you are worried about weight go to any truck stop, for $15 or so they will weigh your car.

 

[Joker328]

I feel like this is way over simplified.

It is. It is a reasonably useful way to show if you are gaining any performance than it is an indicator of how much performance you are gaining or where you are gaining it. It aggregates variation across the rev range, which may or may not be valuable depending on the type of driving you are doing and the type of mod. For example, a higher flowing down pipe or exhaust might show gains in the middle of the rev range (with a tune), but ultimately be limited by the small turbo in the high end of the range. Thus, a pull from 2500 to 6500 may show some gains, but in a racing situation where you are mostly in the high end, you may not see much performance difference. In daily driving, you may feel it more in the butt dyno though. So oversimplified, yes, but sometimes simple is all you need.

 

[Siestarider]

My example is obviously skewed to stock turbo. I assume anyone who wants to collect good comparison data would use the same piece of road, log run both ways, and use whatever RPM range they want for their purposes. Plus no change to vehicle weight. My point was, if you do those things, then any increase in performance will show up in reduced time.

Since my main interest is road/tracking, I use the larger RPM range. It will be interesting to see how the Cyborg Santa is bringing me compares with this measure on my car.

I agree VD and regular dyno tuning are great tools, far more revealing as to where performance is gained across RPM range, etc than simple elapsed time.

On the other hand, simple elapsed time is very honest for what it is.

 

[InsTwin]

What's the weight of the wheels you are using and what gear are you pulling in?

Seems like that was what gave you your biggest gain.

 

[Siestarider]

What's the weight of the wheels you are using and what gear are you pulling in?

Seems like that was what gave you your biggest gain.

OZ Ultras, 16x7, 14.7lbs? Tires also a lb or two lighter. I typically run 3rd gear pulls, seems a good compromise gear.

Loosing rear seats and spare tire/jack helped too. FTS is right, I need to have car weighted, nearest scale is 20 miles away in a direction I never go.

I would say the biggest gain in F, ie reduction in delta t, was from Cobb S3 tune. But by the time I learned that adding hard parts would not add HP without a tune, I had the stuff and went straight to S3.

Now that I am messing around with aero, have to consider most "improvements" add drag to gain downforce, so this simple method will help me understand how much I give up. For example, my air dam tests indicate less than 1% reduction in delta t on pull, but added 5.5% to coast down time, with frontal area increased and 1" clearance between pavement and air dam.

So now I need to use a manometer to see where the pressure changes are due to larger air dam. Its all fun.

 

[Siestarider]

On way home from picking up Cyborg from SS yesterday, I went by a truck stop and weight ticket 2660 lbs with 5/8 tank of fuel + Cyborg on bed + maybe 10 lbs of junk in car.

I have been using 2900 lbs aggregate for VD comparisons in 2015. Looks like 2830 is more accurate with me in car. My goal from beginning was to break 2600 lbs. Remove rear seat delete bed = -30 lbs. Battery swap = -15 lbs. May have to bite my left arm off for that last 15. Already jacked my own thread. I love this car.

 

[Siestarider]

Playing with datalogs posted by others, Joker328's point is clear. You have to pick the rpm range of interest based on modification intent. Preferably you already know what the rpm range of interest is, and log that range. Since my car is a DD and stock turbo, I stayed with standard tuning log pulls until recently.

Big turbo cars don't pull well from 2500 rpm in 3rd gear. While I found using tuning logs over wide rpm range good for quick F=MA results by elapsed time, mods that push torque and HP peaks higher in RPM range may not show up clearly or at all in a wide rpm range pull.

Looking at my tracking logs, I see most of my time is spent between 3500 and 6000. Now that I know shifting at redline offers a little bit of time advantage, next track logs will be shifting at redline going up through gears on straights. Should be a few 10th's per lap there.

Another way to look at it, via Datazap, is check shape of rpm curve under a WOT pull, if its linear over a given rpm range, you are accelerating at a constant rate. If its concave, you are accelerating faster with increasing rpm.

My stock datazap graphics show rpm curve is becoming convex at 6000+, turbo is "running out of air", but the gear advantage remains better than next gear up until redline.

In the interests of full disclosure, thought I better revise my opinion before putting the Cyborg on.