Chassis Dynos

[Topspin]

Before I put last months Hot Rod mag to bed, I thought I'd mention this, as I know some here have swapped dyno #s before. Myself, I've never been much of a believer in chassis dynos and certainly wouldn't spend much to visit one. From simple 1/4 mile top speed and some other data you can extrapolate a hp number that's probably as accurate as the average dyno. Anyway Hot Rod took a new Shelby Super Snake with 750 flywheel hp and visited 5 local dyno shops in a 2 day period. They made multiple pulls at each place without touching the engine at all.

1st stop had a Dynojet dyno and the 3 pulls were; 651, 662, and 656.

2nd stop, a few hours later had a Mustang dyno and multiple pulls were made all between 577 and 579 hp. 75 hp less than dyno 1??!!

3rd stop, next morning at another dynojet the 3 pulls were; 638, 653, and 664. Each pull more hp than the last?!

4th stop, Dynapac dyno and the 4 pulls were; 602, 580, 598, and 561.

5th stop, Superflow dyno and the 5 pulls were; 624, 632, 632, 638, and 624.

Moral to the story, if any, is comparing your #s to someone's #s from another time and place is virtually meaningless. This car went from 577 to 664 hp untouched - such is the world of chassis dynos!

[bgetz]

Great post. I went back and read that article.

QUOTE=Topspin;956244]Before I put last months Hot Rod mag to bed, I thought I'd mention this, as I know some here have swapped dyno #s before. Myself, I've never been much of a believer in chassis dynos and certainly wouldn't spend much to visit one. From simple 1/4 mile top speed and some other data you can extrapolate a hp number that's probably as accurate as the average dyno. Anyway Hot Rod took a new Shelby Super Snake with 750 flywheel hp and visited 5 local dyno shops in a 2 day period. They made multiple pulls at each place without touching the engine at all.

1st stop had a Dynojet dyno and the 3 pulls were; 651, 662, and 656.

2nd stop, a few hours later had a Mustang dyno and multiple pulls were made all between 577 and 579 hp. 75 hp less than dyno 1??!!

3rd stop, next morning at another dynojet the 3 pulls were; 638, 653, and 664. Each pull more hp than the last?!

4th stop, Dynapac dyno and the 4 pulls were; 602, 580, 598, and 561.

5th stop, Superflow dyno and the 5 pulls were; 624, 632, 632, 638, and 624.

Moral to the story, if any, is comparing your #s to someone's #s from another time and place is virtually meaningless. This car went from 577 to 664 hp untouched - such is the world of chassis dynos![/QUOTE]

[poppa]

The only place that really counts is who is ahead at the end of 1320 ft of pavement just like the old days ;o)

[xoxoxoBruce]

Just too many variables to count, from air temperature to the friction of the rollers, Jim G mentioned that in his testing. I think the only honest use for a chassis dyno, is when a tuner/laptop is hooked up setting tune points, to show the result of changes.

[Achias]

The mech who installed Ilsa's blower mentioned that the dudes who run the dynos have a lot of latitude available with the results. Some mess with it to make the results more favorable.

His advice (here in the midwest) was to trust nobody except Speed, Inc. in Schaumburg, IL. They charge over $500 for the job, which is about the same that the locals charge.

[ssrtruck]

Interesting to see the two DynoJet Dynos were very close in numbers.
Getting more power from each pull is common, engine temp and ambient air temp has a lot to do with the outcome

[Topspin]

Tire slippage is certainly a problem on chassis dynos. The Dynapac dyno is unique in the fact that the tires are removed and connectors are actually bolted directly to the axle shafts, to eliminate any variable with wheel spin. Still that unit produced 41 hp variance between the pulls! As far as ambient air being an influence, that was realized long ago and all raw data is corrected to an air temp., humidity, and barometric pressure standard. There are two standards used, so it's important to know which one was used on your dyno session. They're pretty similar with SAE STD using 60 deg. F. instead of 77 deg. for air temp. and 29.92 bar. press. instead of 29.23. You'll get around 4% more power using STD corrections.

Bruce, I guess my real point here is, if you can't produce an accurate, repeatable baseline reading, how can you really "tune" for more power on a chassis dyno?! It's funny the shop with the lowest hp #s, (by the way that was Granatelli Motorsports - a pretty big name in the business) was the most consistent. With a 2 hp variance, they're probably the only one that I'd let "tune" my ride!

[xoxoxoBruce]

I defer to your vastly superior knowledge on engine building/testing, but I'm still skeptical of dyno results.
If you're set up and running on a dyno, any dyno, making changes to the tune on the fly, you have the best chance of eliminating as many variables as possible. If you do a pull, and go home to make changes, then bring it back in a day or two, there are too many things that can't be controlled. I wouldn't trust those numbers.

The SAE STD correction calculates the theoretical deviation from temp/baro/hum changes, but those variations may not be accurate for your engine/tune. As you well know, two "identical" engines will not perform the same.

Oh, and don't forget the phase of the Moon, and the weight of the jockey.

[Topspin]

Hey Bruce, reread my opening post. I don't in any fashion promote using a chassis dyno for anything. Concerning the corrections, they are the very foundation that allows engineers to do long term engine development and testing on engine dynos.

[Mike in AZ]

Hey Topspin,

I've worked for an engine manufacturer for 34 years. Most of that was spent testing engines in the test cells and in the air. I have never believed in transient measurements for performance data.

We use (huge) water brake dynos in our test lab for testing turboshaft and turboprop engines. There is real science in the measurements we make and the reading of torque is done with a load cell (strain gage bridges) and very, very precise measurement. Engine torque is extremely difficult to accurately measure in transient condition. To get accurate readings, we hold the engine at various fuel flow settings, up to max power (~5,000 shp on the T-55 engines for the Chinook) for 3 minutes steady BEFORE taking data. At that standardized point, the thermals on the engine have settled to about 99% of their potential and we call it "close enough" and capture hundreds of data points in three snapshots that are averaged. The measurements we make are accurate to within 0.1% of the absolute numbers and have to be traceable to NBS. It takes a LOT to do that.

We use similar testing methods for our turbofan engines, but instead of a water brake, we have calibrated sleds that the engine hangs from. We can measure thrust to +/- 2 pounds on a 7,000 pound thrust engine. Fotr the turbofans, we really need to stabilize on test point to get the thermals to settle. The engine is about 6 feet long and grows over an inch in overall length between idle thrust and max power.

Transient testing on chassis dynamometers is full of uncontrolled variables. I would hesitate to call anything accurate to more than +/-10% of reality. The only thing they are good for is back-to-back testing.... and the accuracy of that is up to the operator...... even with pulls that are less than 10 minutes apart.

If I use some of my standard seat of the pants data "filtering"...... I'd say you have a solid case for claiming 630 to 640.....

Regards,

Mike

[xoxoxoBruce]

Originally Posted by Topspin

Hey Bruce, reread my opening post. I don't in any fashion promote using a chassis dyno for anything. Concerning the corrections, they are the very foundation that allows engineers to do long term engine development and testing on engine dynos.

Yeah I know, but I would be skeptical of even Granatelli's accuracy, despite their apparent repeatability. Timeslips are king.
Like I said, the corrections are theoretical, and can only be useful when all the other variables are controlled.

Mike I was the lead tech for the installation and calibration of the instrumentation, at Westinghouse's steam turbine test facility. I hate water brakes.

[ssrmaddox]

Gentlemen: Enjoy the ride! It's an SSR.....why worry about how much measured HP you are getting out of your LS2 when it's in a 5,000 lb vehicle!
If you really want a drag racer..build one! Or spend more time in making the SSR about a ton lighter!

[Topspin]

Mike, Thanks for your input. Your "job" sounds REAL interesting to me! I guess what surprised me, in this article, was not the variance from shop to shop, but the variance from run to run. Some runs were made back to back, to heat saturate the engine, so a lower hp # was predictable. Other runs were made with cooling time in between, to try to duplicate the previous conditions, and still the #s would be randomly up or down.

I don't suppose you have a picture of the hanging turbofan engine laying around? That is one brave test operator, though I'm sure he's well protected!

Ssrmaddox, I don't think you really "get" hotrodding. Man, we even "race" flashlights here - see who can turn off a yard light from the farthest distance.

I'm with you Bruce, I'd just as soon spend $20 and go to the track!!

[Double-L]

Originally Posted by ssrmaddox

Gentlemen: Enjoy the ride! It's an SSR.....why worry about how much measured HP you are getting out of your LS2 when it's in a 5,000 lb vehicle!
If you really want a drag racer..build one! Or spend more time in making the SSR about a ton lighter!

No that is the fun of it putting all out race car on their trailers with your 5,000 lb 10 sec. SSR and then driving home, On the other I will stay clear but I do like having a dyno pull anytime I can get it and it is great for tunning in my opinion