Chassis Dyno Testing

Thanks to Gary Morrell for yet another informative tidbit.

 
 		 Dynamometer test results, especially with respect to the horsepower increases claimed as a result of making modifications to the car, should always be taken with a grain of salt. Here's why: 

The Yamaha engine is rated 220HP (at Sea level) at the crankshaft, the driveline losses are rated at about 17%, that means 183HP at the wheels with a stock engine. The difference that I have personally witnessed, with my own car, on a chassis dyno in Colorado Springs, by removing a brand new Motorcraft paper filter and substituting it with NOTHING, was a 4HP increase at the wheels, a measurement that took several runs to verify due to equipment repeatability. These tests were performed on a Clayton, 200HP water brake dyno that had been recently calibrated. The repeatability of this machine, with a VERY good operator can approach +/- 3HP. 

I have performed and observed quite a bit of chassis dyno testing over the years and have learned several things: 

  1. The end results are highly variable due to operator technique, equipment variation, and environmental conditions. I've seen peak numbers change by 20HP just by changing the guy who's sitting behind the wheel.
  2. DO NOT TELL the dyno techs what HP levels you are expecting, they will be less likely to tell you what you want to hear. Act surprised when you see the numbers.
The famous SHO Borla exhaust dyno test is a good example: I have a copy of the article from Super Ford that was run by performance Dyno Shops. ( now out of business, seems that no one can locate their records, hmmmm... ) They quoted +32HP from installing the Borla, which somehow got inflated to +38HP in some popular literature. How come everybody else who has before-and-after-dynoed the Borla, on different dynos, all across the country, only gets +6 to +10HP, figures that are just slightly larger then the repeatability numbers quoted a few paragraphs ago? 

Just because engines are now electronically managed doesn't mean that the laws of physics are somehow circumvented. To make more horsepower, you need to increase the rate at which an engine consumes fuel and air, period. This is accomplished by using the tried and tested methods of fluid mechanics to up the rates at which the air/fuel mix and exhaust are input and extracted from the engine. The means by which this is accomplished may be lowering intake and exhaust flow restrictions, forced induction, opening valves longer and/or farther, increasing valve overlap, or raising compression, to name a few. There are no magic potions or bullets, just physics. The auto manufacturers (I work for one...) are constantly striving to increase efficiency and output while reducing exhaust emissions. If the automakers could get the 'claimed' increases in HP and efficiency from aftermarket products like oil additives, split electrode spark plugs, magnet thingee's on the fuel line, etc., the cars would leave the factory with them.

Example

 The famous Pogue 100MPG carburator was not quashed by the auto companies, the damn thing simply wasn't practical. The theory behind it was that even a moderately aerodynamic car could cruise at highway speed with only 10 to 15HP being expended. Hell, a one-lunger Briggs & Stratton could supply that! The problem was how to get this little engine to accelerate said 3000lb vehicle to highway speed between now and next Christmas. The Pogue was invented in the 1930's, there's probably a copy of the patent in your local public library. Look it up. 

If you look at a modern engine with respect to specific output vs fuel consumed vs waste heat produced, you'll find that its thermal efficiency is quite close to the theoretical maximums predicted by the Otto cycle pressure/volume and temperature/entrophy planes. It is physically impossible to do any better than this unless you design an engine around a more efficient thermal cycle, such as the Carnot. Problem is, Carnot cycle engines are nearly impossible to build.

Dynojet dynamometers have a good reputation (among the various chassis dynos in existance) for repeatable, accurate results. Here's a site that'll help you locate the nearest Dynojet facility if you want to buck up and find out what kind of ponies your car is really kicking out.

Some interesting information from Ted Breaux (who, it should be noted here does (or did) sell LPMs - computer chips for the SHO) :

I am writing this note simply as an FYI of some ongoing observations which have become increasingly apparent over the months, which I recently had an opportunity to investigate thoroughly.

We have all been made aware of the Dynojet dyno, which is a fine piece of equipment which offers nice features, convenience, and repeatability.  While I have always been a proponent of dyno testing, it is important to realize that while dyno testing is a great indicator of power levels, it is not infallible, and therefore the numbers given by the device are not to be 'set in stone'.  Unfortunately, no dyno can precisely mimic street conditions or vehicle dynamics, both of which play a significant role in vehicle performance.

There have been several recent instances in which some SHOs which were dynoed seemed to give numbers which were different than expected with respect to different mods to the car.  The most recent was with Darin's car, in which I asked him to do some very detailed street testing following his recent dyno episode.  Although the dyno indicated no appreciable power differences with his LPM, the street testing proved otherwise, showing significant and consistently repeatable performance improvements with the
LPM.

Why would dyno data conflict street data?

Well, the only explanation I can give is point which Mike Wesley (Ford Engineer, Dynojet Rep) made me aware of many months ago.  He told me with unwavering confidence that there may be discrepencies in the data on vehicles which were heavier than the load which could be applied by the dyno (i.e heavier than 3200lb).  This observations was verified when dynoing Mustangs.  To compound the problem, when you throw in air resistance, etc. it makes it impossible to duplicate real world load.  Now, with conventional vehicles (i.e. drag cars, etc.), this is not going to be as apparent as with those computer controlled vehicles which are programmed to constantly sense vehicle load.  The load sensed by the computer is used to determine just how much fuel and spark the computer gives the engine.  The way LPMs are programmed, although your car will be noticeably snappier and torquier in normal driving, the most dramatic deviation from the factory programming occurs at 100% load, which is where you really want it.  The trouble is that unless you can put the car under full load, the computer will never get to the top of the load tables.  This may explain the conflicting information, and at the present, it is the best of several explanations I can give.  It does make perfect sense to me however.  Of course, an LPM is only as infallible as the programmer, and well......I ain't perfect, but I'm pretty good!  Hell, I've had enough experience with this to be about as good as you can be, and most SHOs respond noticeably well.

Ok, What does this mean?

What it means is that if you want to go out and dyno your car, by all means, do so, and have a good time doing it.  The cost is reasonable, and if I had the kind of time I used to, I'd probably be out playing with it as well.  You should remember however that the dyno cannot necessarily tell the whole story as it happens on the street, and because of this, the information it gives should be used as an indicator and not an absolute.

Which brought up this response from Don Mallinson, who owns/operates a Dynojet :

       My dyno has the "high inertia" drums which have a total weight of 4200 #  This would make the load on my dyno more than stock for an SHO.  This can effect slightly the performance, but when properly used, the dyno is to test back-to-back several modifications.  Used this way it is about as accurate as you can get, and for sure better than stop watches used on the street.  Even high quality equipment like the Vericom is not that repeatable on the street.  I have tried and with changing wind, road slope and weather even 20-60 runs are very hard to get repeat numbers on.
        About the only thing missing from the dyno run is air through the intake system, but that is only applicable if you have a true ram-air system, and it's only good for higher speeds.  I put a fan on most cars that can give about 40 mph wind to the front, I don't do it to put air into the intake, but to keep air moving across the motor and help keep things cooler.  Of course the electric fan is still better at cooling since it's right ON the radiator.
        We tried several times to get many of your chips to work that day the SHOcago group came down, and when the car won't even run, its surely not the fault of the dyno dynamics.  You know we tried many times to get one to work in my car and never did succeed, although I really wanted it to work.
        I admit the Dynojet is not perfect, but it's about as good as it gets today for the value and accuracy.  If the chip works on the street, it should work on the dyno.