Realtime Horsepower?

[CybrStuff]

So I know you all love the abstract technical questions, so here's a doozy for you. What I've been tossing around in my mind for the past few weeks is whether it would be possible to measure a car's power output in realtime. I have a few ideas and I was wondering if I could get some input from the experts around here. Here's the basic idea:

 

Base horsepower can be derived from the rear wheel torque, which is a function of acceleration (F = ma), wheel diameter, and engine rpm.

From the base horsepower, we add a correction factor for wind resistance at the relative wind speed around the car (using an anemometer or some other wind speed measurement).

We also include a factor to account for the angle of the car (it will be accelerating naturally down hills, etc.), which will be measured by a tilt sensor of some sort.

Finally, we add a factor to account for frictional losses (that is, the fact that a car rolling in neutral on a flat surface will eventually stop).

 

Now, assuming I didn't leave something out (which is possible), there are still a couple of tricky parts. First, we have to determine the rolling resistance, which might be possible by using a roll-down test from a relatively low speed (to avoid wind resistance). Next, we have to figure out a rough approximation of wind resistance as a function of speed, which should be possible by making a dyno run to get a feeling for horsepower, and then making a similar run on the street (probably not in 4th though ) and, using the accelerometer, figuring out the difference in power at a given wind speed. Since we now know the rolling resistance, we can account for environmental variables (temperature, humidity, altitude, all that stuff that affects mass air flow) and come up with a reasonable approximation of horsepower loss to wind. Once we do these weeks of painstaking tests and add probably a dozen instruments to the vehicle, I think it should be possible to get some kind of realtime horsepower display inside the vehicle. Any thoughts?

[rudypoochris]

It is called Gtech..... gets everything but the wind, which is a big deal at high speed, but if you were itno it enough im sure you could write an equation to fix that.

[CybrStuff]

It is called Gtech..... gets everything but the wind, which is a big deal at high speed, but if you were itno it enough im sure you could write an equation to fix that.

 

Yeah, I know about Gtech, but does that work when you're just cruising, say at highway speeds, when all of the power of then engine is going into maintaining speed against rolling resistance and wind drag? From all the data I can find on their site, it's more for doing dyno pull type runs on the street, and it doesn't figure out how much power your engine is really making to the rear wheels because it doesn't care about aero drag or rolling resistance.

[rudypoochris]

Yeah, I know about Gtech, but does that work when you're just cruising, say at highway speeds, when all of the power of then engine is going into maintaining speed against rolling resistance and wind drag?

 

Sorry i think i misunderstood, but not sure good question. I know the newer models can calc the torque because they can read the rpms throw the cig lighter. It is an intresting question, but i would doubt it can read your cruising hp.

[CybrStuff]

Sorry i think i misunderstood, but not sure good question. I know the newer models can calc the torque because they can read the rpms throw the cig lighter. It is an intresting question, but i would doubt it can read your cruising hp.

 

Yeah, it's no problem . That's the reason behind my initial idea. Since the Gtech can't do this, I was wondering if it would be feasible to develop a system that could.

[HeatRaveR]

I drove a Chevy SSR recently and I noticed it had Torque gauge built in. I'd never seen/heard of such a thing, let alone on a stock vehicle. I have no idea what it's hooked up to or how it works though.

[Mikelly]

...Those Gtechs are only as accurate as the data input.

 

There was recently a comparison in road and track or Car & driver that did a shootout on performance meeters. It was a decent read and talked about some of the issues and solutions for measuring HP and torque.

 

Mike

[CybrStuff]

There was recently a comparison in road and track or Car & driver that did a shootout on performance meeters. It was a decent read and talked about some of the issues and solutions for measuring HP and torque.

 

Link to the shootout

An interesting sidebar on weather correction

 

I read through the article, but unfortunately, they don't really talk about horsepower and torque numbers. Their testing is based more on quarter-mile times, 0-60 times, lateral g's, and braking. They mention that one of the meters can give you a road horsepower reading by analyzing some acceleration data, but I get the feeling that it's the same thing as the Gtech, where it just represents your net horsepower, not your actual horsepower to the wheels, and also that it is not a realtime measurement (e.g. you're cruising down the highway at 70 mph and the display reads '57 hp'). Thanks for the heads up on the article, though.

[rudypoochris]

Delete

[mom'sZ]

CybrStuff: Just curious, what exactly are you trying to ascertain? How much HP your car is making when cruising down the road? Are you just curious or is there some other reason you want to know? I'm just wondering. My understanding about performance meters is that they can calculate HP from acceleration using the weight of the car. So long as you accuratly input the weight of the car, they are fairly accurate from what I've heard. When these things first came out, I was excited. I haven't gotten one yet but intend to. Until recently your average person wasn't allowed to even own a accelerometer, apparently they are a key component in rocket guidance systems. With an accelerometer and an accurate vehicle weight, calculating HP figures would be very similar to how a momentum dyno works, such as a DynoJet. And no one seems to question their accuracy. Anyhow, just for another data point, years ago I read some stuff that was talking about how much power it took to keep a vehicle moving down the road once it was up to speed. This was years ago when gas mileage first became an issue. If I remember correctly, it was a miniscule amount. Something along the lines of a two ton car moving 60 MPH took less the 10 HP to keep it up speed. I think 7 HP was the number.

 

So... does anybody have one of these performance meters. If so, what do ya think? Are they cool? Does it work. Can you tune with it?

[CybrStuff]

CybrStuff: Just curious, what exactly are you trying to ascertain? How much HP your car is making when cruising down the road?

Exactly.

Are you just curious or is there some other reason you want to know?

Really, it's just sheer morbid curiosity. I'm somewhat of a data freak and would love to see how much horsepower you actually need in most situations. Plus, correlate that with a fuel computer and you could get some interesting mileage vs. power graphs. Mostly though, it would just be a fun thing to try to implement using my digital circuit knowledge.

My understanding about performance meters is that they can calculate HP from acceleration using the weight of the car. So long as you accuratly input the weight of the car, they are fairly accurate from what I've heard.

You're right. As far as measuring effective horsepower (actual horsepower - rolling resistance - drag), such systems are probably quite good. The main reasons I'm looking into something more are that:

1) They don't give you a realtime reading.

2) They don't let you see gross wheel horsepower (the number before parasitic losses).

Anyhow, just for another data point, years ago I read some stuff that was talking about how much power it took to keep a vehicle moving down the road once it was up to speed. This was years ago when gas mileage first became an issue. If I remember correctly, it was a miniscule amount. Something along the lines of a two ton car moving 60 MPH took less the 10 HP to keep it up speed. I think 7 HP was the number.

This is a tricky subject because it depends a lot on the shape of the car and the local atmospheric conditions. One that I know for sure is that wind resistance goes up dramatically as a function of speed (on the order of the speed cubed?), so your drag at 70 mph is considerably more than your drag at 60 (and who among us never goes over the speed limit?).

 

Anyway, I know it's not a gigantic amount (unless you're looking at going over 100 mph or so, when the drag can require hundreds of hp to overcome), but the geek in me (that is to say, all of me ) is interested in knowing how much power I'm developing at any given point in time.

[mom'sZ]

very cool..

 

They don't give you a realtime reading.

 

So you set the thing, you punch it, youi haul *** a bit, and then you stop and it tells you your horsepower? sorry for being an idiot.

 

any how, you don't have to tell me about the whole wind resistence vs power thingy. I learned that years ago motorcycle racing. Raced heavy weights - 750 cc and up, they went up to 170 mph. At that speed, aero was important. Also raced a lightweight - around 400cc. Well one time I'm racing the light weight and I crash the damn thing. Bugger up the fairing pretty good and didn't have a spare. I was bummed out but found some number plates and went out there without it. Guess what, went faster, seems that pretty fairing didn't help, it created drag and slowed me down. Left it off after that.

[CybrStuff]

So you set the thing, you punch it, you haul *** a bit, and then you stop and it tells you your horsepower?

As far as I can tell, this is how all of the main commercial ones work. Of course, I'm willing to admit I'm wrong if anyone can show me one that does what I'm looking for. The way I see it, if I wanted a dyno run, I would go to a dyno (prices notwithstanding) and get much more accurate data.

[Guest Z Draci]

The Bentley Arnage T I was in had a "% power available" meter.

Instead of telling you the direct output, the meter told you how much more of the 6.75 liter twin turbo V8 is available to use.

[Mikelly]

We had a Gtech I sold to a member here... My stepson bought it and he eventually ended up owing me money and gave me the thing in leu of cash... I'd have prefered the $400!

 

Anyway, I took the unit and my trusty old Corvette out to an isolated country road and did repeated runs with the unit. I then tweaked the numbers in the unit, trying to better dial it in to the car, and every variable I plugged in really skewed the results to the point that it just didn't seem worth the trouble. Furthermore, Unless you another method to verify the data (Dyno or going to the track with a timing device/ starting tree), you're just assuming that the thing is working. For me, I had numbers on the car from previous dyno runs and I still couldn't get the Gtech to register the same power. The unit constantly registered more power than we knew the car to make on the dyno... Significantly more. So I didn't trust it.

 

It could have been me. But I boxed it up, sold it and called it a day. I can do dyno runs for about $70 for three pulls, or pay $25 and go to the track for a test and tune night, so why bother.

 

Mike

[CybrStuff]

For me' date=' I had numbers on the car from previous dyno runs and I still couldn't get the Gtech to register the same power.

[. . .']

I can do dyno runs for about $70 for three pulls, or pay $25 and go to the track for a test and tune night, so why bother.

I feel like the reason you had such a hard time trying to get the Gtech lined up with "the real world" is that it doesn't take enough variables into account. For example, when you're going 60 mph, the difference between a 20 mph headwind and the same tailwind is going to be probably tens of horsepower loading on the engine. There are no instruments that come stock on a vehicle that can measure wind speed, so naturally any measurement based on stock instruments are going to be incorrect. I think by supplementing those stock instruments with a few others, it should be possible to get reasonably close to measuring real horsepower.

[mom'sZ]

I'd have prefered the $400!

that's funny!

Anyhow, interesting.... you are the first person I've heard who wasn't impressed with these things. Although most guys would be thrilled with HP numbers that were to high and not bother to try to verify. How about using it as a tuning tool? Even if they are 'overly optimistic' numbers is it still accurate enough to tell if you've made a gain? And how about for lateral Gs, you can go to the drag strip or dyno for HP or straight line acceleration testing but there is no easy way to verify handling improvements. The one model also seems to have some features aimed at helping a road raceer or autoXer with lap times. Hell, I was ready to order one of these things. Then again I love techy stuff... ha ha ha