Dynoing using Data log?

FastDriver

My dad had a bra
SN Certified Technician
Sep 5, 2001
4,538
1,244
214
Seaside, CA
I have been geeking out on my best friend's 2016 mustang. He's got an OBD2 port bluetooth scanner that can datalog using his phone in real time.

I wrote some script in python and R to try and calculate HP & Torque. Unfortunately, the numbers I'm getting are considerably lower than I expected. The Mustang GT with PowerPack3 should be making ~430 rwhp. In the graph below, the red line is RPM, Blue is Power. I calculated the power based on the weight, velocity, and rate of acceleration. Happy to get into the formulas if you want. It's slightly smoothed (averaged data over 0.2 seconds). Each recorded point is ~0.036 seconds.

Anyways, 4th gear looks really weak at ~300 rwhp. We'd kind of expect to see more numbers like what we see in 2nd gear. I was just wondering if anyone here has experience comparing what they see on the dyno, to what they're getting from datalogs? Is it usually significantly lower in the real world than on the dyno?
1569220461979.png
 
  • Sponsors(?)


a91what

SendMeUrDataLog
Mod Dude
Apr 6, 2011
9,761
5,808
204
31
Hillsborough county
I have used injector pw and bsfc to calculate HP and TQ from fuel flow.. it's usually within 10% unless the car is under or over fueled.. if you can post a copy of the datalog I can try it on mlvhd.
 

FastDriver

My dad had a bra
SN Certified Technician
Sep 5, 2001
4,538
1,244
214
Seaside, CA
That's a pretty cool approach, but injector pulse width isn't captured in this datalog.

It hit me last night... air resistance. It probably costs on the order of 50+ hp at 100 mph. I'll be damned if I can find the coefficient drag or frontal area of an S550, though.
 

FastDriver

My dad had a bra
SN Certified Technician
Sep 5, 2001
4,538
1,244
214
Seaside, CA
Booyah! After accounting for air resistance & smoothing out the datapoints:

1569357200348.png


Many "power pack 3" S550 Mustangs have dynoed on youtube at ~430 rwhp on the dyno. So, for the real world, this looks pretty believable, but it will still inevitably be off by the cumulative errors in the DA correction, my best guess in the Cd & frontal area of an S550 Mustang (.35 & 24 ft^2) error & imprecision in the datalog, spikes due to kinetic energy transfer between the rotating assembly and tires. I'm thrilled to see expected ballpark numbers!

That data was pulled from https://airdensityonline.com/track-results/Spokane_County_Raceway/ and then the relative effect on engine power was calculated here: http://www.csgnetwork.com/relhumhpcalc.html. I got my buddy's to scale his car to get the 3980 lbs race weight.
 
Last edited:

FastDriver

My dad had a bra
SN Certified Technician
Sep 5, 2001
4,538
1,244
214
Seaside, CA
Damn, I thought this was cooler than it apparently is. Let's try this another way: $50 OBD2 scanner + this = $10k dyno, except better because it's in real-world conditions on the track or street!
 

FastDriver

My dad had a bra
SN Certified Technician
Sep 5, 2001
4,538
1,244
214
Seaside, CA
Did not update the legend in this new pull with the Dark Blue AFR curve toward the bottom of the graph.
upload_2019-9-28_1-38-47-png.png

Graph from a 3rd gear street pull today. This one shows a blue and green line from a pull in a single gear from 1500 to 7500 RPM (hitting redline). The blue and green curves look a lot more like a what you'd expect to see on a dyno since it's one continuous pull without shifting.. A peak of around 395/350 rwhp/rwtq. This is lower than most dynos in the 435 range, but in the real world, you're spinning those front wheels and tires, and dealing with other issues, as well.