Calculate HP base on OBD II data?

**jfitzpat** · 03-09-2010, 01:57 PM

Originally Posted by nixit

OK so to accurately calculate hp you would need...

If your concern is *precision*, you would probably want to look into directly reading torque. If the point is consistent readings, so you can, say, tune with confidence, then VSS, RPM, and good acc readings can be good enough.

Originally Posted by nixit

Why cant you just record deceleration coasting in neutral with no wind? wouldn't this just be the sum of all force acting against the vehicle? I don't understand why it wouldn't work. I know its not linear but you could build a look up table with the data.

For a rough estimate, it would work, but think about it this way, even with the clutch engaged, is the mechanical friction zero? The problem (or the purposes of precision) is that you are looking at a bunch of different factors and they do not behave identically in response to changes in velocity.

-jjf

**jfitzpat** · 03-09-2010, 01:59 PM

Originally Posted by Vitaliy

Sorry for an off-topic comment.. are you a pilot? I've just resumed ground school after a long break and recently had gone over induced/parasitic drag, maximum lift/drag ratio, etc. The Jeppersen book has a sidebar where they talk about WWII pilots waxing their planes before each flight to reduce parasitic drag and squeeze a few extra knots out of them.

FWIW, I've been a CFI-I for many years, though I don't teach as much as I used to (it's always been a hobby for me).

-jjf

**Vitaliy** · 03-09-2010, 02:51 PM

Originally Posted by jfitzpat

FWIW, I've been a CFI-I for many years, though I don't teach as much as I used to (it's always been a hobby for me).

-jjf

Nice.

**nixit** · 03-09-2010, 04:39 PM

Originally Posted by Vitaliy

For better accuracy, I would also suggest adding attitude indicator, a stall warning horn, and a windsock.

LOL. I'm a little OCD I like to see lots of data and for it to be as accurate a possible.(thats the whole point of the carPC for me). I know its overkill and a waist of time to most but "it's what i do".

The direct torque aproch is way to pricey

The g-tech performance meters claim +/- 1%. For $200 + all the other features

(still not as cool as making your own)

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

**jfitzpat** · 03-10-2010, 11:08 AM

Originally Posted by nixit

The g-tech performance meters claim +/- 1%. For $200 + all the other features

(still not as cool as making your own)

I tend to take vendor specs in motorsports with a grain of salt. But, like I said, in my experience VSS + RPM + acc can give pretty consistant results. Then it is a matter of tweaking weight, drag coefficient, or whatever to give some semblence of accuracy.

Building your own is educational, that's for sure.

-jjf

**MX5_Carputer** · 03-27-2010, 03:50 PM

If I were you I would do the following assuming petrol car!:

- Take a note of my cars peak horespower and at what RPM (this will give you an idea of whether you are in the right ball park!)

- Using MAF or speed density (depending on available PIDS) obtain mass airflow g/s
- Calculate fuel flow g/s assuming lambda = 1 (14.7 parts air to one part fuel by mass)
- Assume for now that your vehicle does not run rich under high load, this is entirely possible- I know some ECUs are mapped to avoid running rich at all costs (maintain emissions...)
- multiply the chemical energy stored in the fuel by 0.3 (approx ICE efficiency)
- I have done this and it worked nicely for me!

If anyone is interested in taking this approach I am happy to do a bit of legwork and provide a walkthrough on how to do the above?

The other approach mentioned a few times in the post was essentially F=ma assume you know mass calc acceleration you get force this equates to useful power at the wheels. Assuming you do multiple runs for wind on a flat road this is a more true figure of your vehicle's power.

**split63** · 02-18-2011, 05:18 PM

Originally Posted by MX5_Carputer

....Calculate fuel flow g/s ....... multiply the chemical energy stored in the fuel by 0.3 (approx ICE efficiency)
- I have done this and it worked nicely for me!

I believe what you are calculating is the energy or horsepower potential in the fuel. Since engines vary in efficiency from each other and as a function of RPM, I believe this may be a very gross estimate. Also, during Wide open Throttle, the system is typically open loop and 14.7 AFR is not happening.

**MX5_Carputer** · 02-19-2011, 09:14 AM

Originally Posted by split63

I believe what you are calculating is the energy or horsepower potential in the fuel. Since engines vary in efficiency from each other and as a function of RPM, I believe this may be a very gross estimate. Also, during Wide open Throttle, the system is typically open loop and 14.7 AFR is not happening.

Ok so you tried it and failed?

I don't want people to be put off this technique based on your speculative comments. As I already said in my post, I have tried it and found it provided a reasonable estimate of manufacture quoted figures (on more than one vehicle). I would say within 2-3% error.

**Nitroking** · 02-20-2011, 09:44 AM

Originally Posted by MX5_Carputer

- multiply the chemical energy stored in the fuel by 0.3 (approx ICE efficiency)

How do I det the chemical energy? If I have 2 gallons/hour for example, how to I get the chemical energy for that?

**MX5_Carputer** · 02-20-2011, 01:08 PM

I just wrote a post, giving an overview of the maths involved, even for those without a MAF signal. Chemical energy of petrol is approx 43MJ/kg. Check out the post here

---------- Post added at 07:08 PM ---------- Previous post was at 06:59 PM ----------

Originally Posted by Nitroking

How do I det the chemical energy? If I have 2 gallons/hour for example, how to I get the chemical energy for that?

So for you;
2 gallons / hr = 7.571 litres/hr
density of petrol = 749.5 g/l
mass of 2 gallon/hour = 5674.46 g/hr
fuel flow = 1.576 g/s
Power = 0.33 * 1.576 * 43 [kW]
P = 22.36 kW
P = 30.4 hp