This is going to be a long post so apologies to people with short attention spans.

Here's one from way out of left field that I have been considering for a while. There are a lot of threads on this site about how much hp=how fast, & how much better was the upgrade etc, but not many of us can justify heading for the dyno shop or drag strip every time we make a change to the car..... so here's what I'm thinking:

What you need:
1. Flat, straight road
2. Windless day
3. video camera, preferably digital.
4. A car with a tacho and the specs on its gear ratio's, diff ratio's & tyres, preferably a datto 1200 but anything will do.


Attach camera so it gets clear view of your tacho.

Drive car from 1500 rpm to redline in gears 1,2, 3 & possibly 4 if you have enough road (slowing back down to 1500 before each run)

Do a couple of replicates in each direction up the road to keep it scientific.

Possibly retard timing by 5 deg and do another run to test sensitivity of method. (that should drop a 10-20 hp depending on your car)


Back at the computer.

Time rpm in increments of 500 for each pull

Create graph of the pulls (gears 1,2,3 &4) and compare the shape of the curve. This is the torque curve at the rear wheels + the wind resistance for a given speed + the rolling resistance. (At this stage there are no units for torque)


Normalise the data by dividing the torque + resistance curve by the gear ratio.

After this step, the remaining difference between the torque curves is equal to the rolling resistance +the wind resistance


Create a graph of the combined rolling resistance & wind resistance and fit a curve.

Use the equation from that curve to remove roll+ wind from the data.

All three curves should look the same now.


Fom this, create a torque curve (at the rear wheels) using physics principles.

Calculate a power curve from that.


I am working on the basis that rolling resistance and wind resistance are the only 2 factors working against the power of the car, therefore you dont need to measure each individually, you can lump them together in a common equation.

would anyone be interested in a speradsheet that can do this if I put in the effort to work it all out?

Are there any maths or physics guru's out there who can check up on the work?

comments appreciated

Yep,
It will work but is a very hard way to do it.

I use a new series Gtech pro RR and they are so close its not funny.

I did some runs on a friends car 3 weekends ago while doing some base level tuning and then we did 37 power runs at Melbourne performance centres dyno last weeknd and the difference was 2% between Gtech estimate and dyno RW Kw's.

Its an interesting and challenging mental excercise if you have the time :)

Gosh thats good...I might consider getting one now!!

More time than money

How does the Gtech account for wind & rolling resistance? do you have to to estimate parameters for it?

Edit: That's a pretty impressive bit of gear

http://www.gtechpro.com/rr.html

your theory sounds pretty good. however. you will ofcourse need sevral basline horspower ratings for the same vehicle that have already been dynoed for this method to work. if you take to of the same vehicles with the same gear box and tire size and engine. (engine mods not relevant on eigther) and have them dynoed, then do the time recorded in each gear recording on each vehicle you will have a basline and therefore a formula.

Quote:

Yeah, a mate of mine is dynoing his XR6T on the weekend so I was going to grab his data for verification.

Answering my own Q's here-from the G-tech site:

"The Gtech measures NET horsepower, which takes aero drag and the rolling resistance of the road into account. A dyno does not have to contend with aero drag but the Gtech does. This often accounts for a large part of the difference between a dyno slip and the Gtech results."

Feral- so kind of surprising that it was still within 2%??

That reminds me ... I have an original Gtech in the closet. I should pull it out and test my 0-60 mph (0-100kph) times.

when working all this out would you have to take into account the friction from the road and tire pressure? something you need not worry about on a dyno obviously...just a thought

I can kind of get around that- either by just taking the Gforce approach and measuring net torque/hp or by calculating the difference between each gear AFTER the curve has been normalised for torque multiplication from the gearbox.

Just got some data to work with @ 15 frames/sec so that should be precise enough to get the numbers flowing.

And the best thing is if I run the vid at 1.3X i get to hear what it will sound like with a blower..... and at 2x it sounds just like a hypo 4cyl bike