Page Options
You are here : Magazine Articles
  Minimize
09

 

Physics Lesson - Calculating Torque and Power From Datalogs

By Khiem Dinh

Khiem Dinh is an engineer for Honeywell Turbo Technologies at the time of this writing.  All statements and opinions expressed by Khiem Dinh are solely those of Khiem Dinh and not reflective of Honeywell Turbo Technologies.

The motion of cars is ruled by physics, so using some basic physics and data enables us to calculate cool things.  Everyone likes cool things right?  The modern automobile has electronics that gives us the ability to gather useful data letting us calculate useful things.  In this case, we're going to calculate torque and horsepower making ourselves a virtual dyno.

I recently had a Hondata KPro installed allowing me to get the necessary data.  To calculate torque and horsepower, only engine speed and time need to be datalogged.  With just those two parameters, we can apply good ole F=ma (Force = mass x acceleration).  We do need to know other parameters specific to the car in order to perform the calculations: tire size, vehicle weight, gear ratios, final drive, coefficient of drag, frontal area, and air density.

First, we need to look at the forces acting on the car.  We have the force at the wheels pushing the car forward and the force of aerodynamic drag pushing on the car in the opposite direction.  The difference between the two is the remaining force providing acceleration.  Terminal velocity is the situation where the drag force equals the force at the wheels resulting in zero force for acceleration.  I called the force at the tire a net force as there are also frictional losses; there's not really a good way to decouple the frictional losses from the force put down at the tires (which I called the gross force), so I lumped them together into a net force.  This is the equivalent of measuring torque at the wheels on a chassis dyno where the car is strapped down to rollers.

The basic equation we need to calculate torque and power is F=ma.  We know the mass of the car, so we need to calculate the acceleration of the car which gives us the force.  Remember the engine speed and time we datalogged?  We're going to use that information to calculate velocity giving us acceleration and drag force.  Knowing those two things gives us the net force at the tires.

 

To get velocity, we need to know how fast the tire is spinning and also the tire's circumference.  Calculating the circumference of the tire is a simple matter of plugging and chugging the tire width, aspect ratio, and wheel diameter.  The rotational speed of the tire is calculated by dividing the engine speed by the gearing.  In the case of the S2000, it has a primary gear reduction in addition to each individual gear ratio and the final drive.

Now that we know velocity, it's a simple matter of calculating acceleration.  From acceleration, we get the net force at the tire.  Using the value of the net force and the radius of the tire, we can calculate the torque at the tire.  To convert to torque at the engine (as measured at the wheels/tires such as most chassis dynos), the value of the torque at the tire is divided by the gearing.  Once you know the engine torque, horsepower is just an equation away. 

 

Pages: 1 of 4 Next Page
Posted in: Magazine, Tech

Comments

Tech@EPR
# Tech@EPR
Friday, March 09, 2012 1:22 AM
quick easy way to give you BHP is simple.

We use this when we flow bench heads and utilize the engine dynoe.

CFM X .25 X No of cylinders. This will give you an approximation to within 5-10% variance. Simple quick way to determine what the engine will make without having to go into math problems that some don't want to take the time in getting involved with or just don't know how.

Figured I'd share this with you.

Laters
8695Beaters
# 8695Beaters
Friday, March 09, 2012 8:27 AM
If you want a quick and dirty horsepower calculator, this one (http://www.virtualengine2000.com/Calculator.htm) uses 1/4 mile trap speed and the weight of your car to do the same thing. It only gives you a peak number and it's not as accurate, but it is a start.

Speaking of Virtual Dyno, I've been asking Brad about VD's compatibility with Tuner Pro. Anyone have a Tuner Pro datalog to send him so we can figure this out? I was hoping to use VD to help me clean up my S13's tune similar to how Project EVO IX has been tuned. And before anyone asks, yes I know it's not ideal, I'm not looking to break records, I just want to lean out my car a bit and fill in some gaps. I also plan to test out his software against a real dyno with our SAE car...just to put the naysayers to rest. Anyway, if you have a Tuner Pro log, please shoot me a PM, nobody has answered my requests (though I did get a nice rant from some FFF about how JWT doesn't know how to tune cars because he uses a Dynojet).
Jeff
# Jeff
Friday, March 09, 2012 11:24 AM
Been using this for 5 years (not free): http://www.dataloglab.com/
spdracerut
# spdracerut
Friday, March 09, 2012 12:52 PM
@Tech, the head flow measurement relation to HP makes sense as HP is a function of airflow. The more airflow, the more fuel you can spray, the more power. That’s a good little correlation.

@Beaters, just make your own spreadsheet. It’s really a pretty basic physics problem, much easier than your typical college engineering physics class problems. The only tricky part will be cleaning up the data, i.e. the data processing as far as accuracy goes. I bet you have a high resolution crank trigger wheel though on your SAE car. On my S2000, I think there’s only like 8 teeth or so on the crank wheel and I think only 4 on the camshaft wheel. I did a quick search and you can get trigger wheels with 36 to 60 teeth, so that really improves your resolution which should greatly reduce the scatter and error in the data.

You can back out your coefficient of drag and friction by doing a coast down test with your Formula car. And making your own spreadsheet, you can throw in any temperature correction you’d like.
Tech@EPR
# Tech@EPR
Friday, March 09, 2012 7:16 PM
Correct why its the easiest and most functional formula we've come up with that gets you relatively close enough to give a semi accurate figure. obviously we can't configure this for a FI application but if you want raw engine power based on head flow dynamics...this is the best and easiest way.
Steve
# Steve
Saturday, March 10, 2012 4:01 PM
Is that relative dip at 5500 rpm VTEC-ness kicking in? Sorry if that's a noob question, idk Honda engines at all really. Cool write-up!

Post Comment

Only registered users may post comments.

MotoIQ Proudly Presents Our Partners:


Copyright 2012 by MotoIQ.com
Privacy StatementTerms Of UseAdvertise