can you measure real horsepower?
 

Hi everyone,
Recently I have come across a very interesting article in C&D "Measuring your horsepower depends on whose yardstick you use".
The conclusion is that cars are becoming too sophisticated for a dyno test. Contemporary cars are using many sensors at different points to measure variables like airflow speed, inlet air temp, water temp, etc. If the conditions are not optimal the computers react by backing off the spark and increasing the air-fuel richness. The example of BMW tuner Dinan shows different power outputs depending on the conditions the car was dynoed in. For example 470hp Dinan's M5 (415 expected at the wheels) puts out only 334hp at the wheels on a dyno without blowing air and closed hood. Then they open the hood and turn on a small shop fan blowing air at 10mph. This time bmw shows 371hp at wheels. Still not there. So they start a $7000 electric fan that blasts air at 75mph. This time it seems the car is producing 411 hp at the wheels. So basically without a wind tunnel or an efficient blower there is no way to measure real hp on a dyno. And even if you take the engine out, the computer won't be fooled and without blowing air it wont show you its full potential.

Well this puts a different perspective on dynoing rx8. I am sure its electronic system is as complicated as those of bmw and dynoing does not really make sense unless you know what you're doing. That would explain low rx8's power output on dyno. I was wondering how it affects tuning, since we dont really have all the info from the computer.

What ya think?

youre right...dyno's are inaccurate. even engine dynos are, but they're more consistant.

in this day and age, it's really no longer necessary to tune a car on a dyno. canzoomer doesnt do it. all you really need is a wideband and some open roads to do some pulls. get the AF ratio smooth and at a safe level, and good power will follow.

Well, at least there is some sort of study going on trying to control the variables. I think it makes a pretty damn strong case as to why the '8s are dynoing at about 180 HP instead of 200+.

If you want a real number at the flywheel you need to remove the engine and put it on a stand.

However, if you want a real wheel hp number then your going to need to drive the car and do a g-tech style run in real life with a data logger... except unlike a g-tech do things like correct for things like drag, weather, ect. It can be done, and i predict soon in the future there will be tools out to do it... ;)

If no one is pulling 5.9's to 6.3's (0-60), and everyone is dyno'ing 170's, (sorry not even 180's), we are all fooling ourselves to think anything different....

I believe the RX8 is making 220-225hp stock.

I also believe dynos have become more than useles with new cars no question. Little things like how large of a fan the shop is using can make a difference. Of all the claims I have seen made on this board I tend to only believe Racing Beat.

A wideband and datalogging in real time on the road is the way to go and actually cheaper in the long run for tuning when you consider the cost and headache of always having to go to the dyno. I ran into this conclusion with my Eclipse turbo and the various cars we worked on.

A drag strip is a good way to measure improvements once you become consistent.

THe most recent issue of Sport Compact Car has an article on various dyno's used in industry. Conclusion: variation between different dynos (even between same brands of dyno) was very high. Their suggestions, stick with the same dyno and measure percent changes. Comparison between different dynos is basically useless.

Also they stated that smart ECU's in modern cars may learn during dyno runs, and will adapt accordingly in successive dyno runs. This adds variability that make comparison even between successive runs difficult.

Horses.....
 

How about some physics.....let the quarter mile times tell YOU how many horsepower you have -

Quarter Mile to HP convertor....

Obviously, it's a little inaccurate, but it will give good ballpark figures if you know your vehicle weight and a good elapsed time.
.
.
.
doc

I had my 8 on a Dyno Dynamics this afternoon. It's bone stock and is the base model so the TCS and DSC don't come into play. Magic8 had mentioned the article in Sport Car Compact. In that article they wrote that although it was one of the lowest reading dynos, it was the most consistent and repeatable. I witnessed that first hand as they ran my car several times with practically no change in readings. That's what is needed for tuning.

I just got the M flash done a couple of weeks ago, so I got the dyno run today for a baseline on Canzoomers mods. I put down 162.9 at the wheel. They calculated (its a guess at best) that I was seeing 213.8 at the flywheel. Since that dyno is supposed to be low compared to Dynojets and Dynapacks, I feel pretty good about my little 8. The proof will be in the follow up run after the Canzoomer goodies. I just ordered my mods a week ago though, so don't anybody stay up late to see when I post the next dyno. It's going to be a wait for sure.

How to reliably test an engine for performance :

Eddy current / AC engine dynamometer (more stable than water brakes), recently calibrated.
Engine with full working FEAD and clutch, full vehicle intake and exhaust systems. Auto spark / auto fuelling. Reference fuel.
Steady state testing (wait a few minutes at every operating points for temperatures to stabilise).
Average of several runs (3 minimum, 10 best).
Corrected to ISO 1585 or local equivalent.

Anything else is just guesswork.

The above dyno problems are related to the engine control system. There are the basic dyno problems also.

Also, I think alot peaple don't understand the difference between horsepower measured from acceleration based dyno measurements and an engine brake horsepower measurement (it measures actual power from the engine). Often the dyno is called a rear wheel measurement but it also subtracts engine inertial losses. Inertial losses are proportional to the rate of engine acceleration. (It is a good measure power for acceleration at lower speeds). Power is also used to overcome wind resistance at higher speeds so the dyno's artifically low values are not accurate.

Some of the more knowlegable understand there is an inertial loss but think it is straight percentage which is completely based on the engine inertia. It is also highly related to the torque increase thru gearing.

My post above was a slightly off subject but I am somewhat tired of some the post about RX-8 horsepower vs the power in car B. Also I am annoyed with post about the RX-8 being a low torque car (All cars have low engine torque - you cannot spin wheels even 500 ft-lbs of torque if gearing is not used to increase it). Most of these type post are only related to drag racing. I am more of road racer type where the RX-8 shines.

Most of these peaple making such post do not understand how these items are measured and calculated.

If you are interested there are some good definitions of this stuff on the web (The physics of Racing series is good-> http://www.miata.net/sport/Physics/) and if you have some extra money some of the books by Karl Smith (Tune to Win)
have good discussion of power, torque and how to set gearing.