Before calling BS on dynos...

[spdu4ea]

Over the last few weeks I've seen a number of posts expressing skepticism over various SC dynos. Two recent examples:

Person A dynos their stage 1 SC and makes 500rwhp
Person B dynos their stage 2 SC and makes 520rwhp

People claim person A's dynos must be wrong (too high)


Person C dynos their stage 2 SC and makes 460rwhp
Person D dynos their stage 2++ SC and makes 600rwhp

People claim person D's dynos must be wrong (too high)



At face value, those claims might seem reasonable (although one might ask why they don't suspect B & C of dynoing "too low"). But there is A LOT more variance in dynos than most people realize. PG was kind enough to open his dynodb to me and using only stock vehicles here's what I discovered:

Stock dynos ranged from 303 to 363 rwhp. Those weren't just single outliers either: there were 3 over 360 and 2 under 305. Even just amongst the dynojet runs, results varied by over 40 horsepower. Stock! So arguing simply because a dyno is 20-40hp away from expected on a modded car is just silly (not saying you shouldn't question results, but chill with the accusations until proven).


Other interesting tidbits:

Average stock dyno: 334rwhp

Broken down by:
6MT: 339rwhp (18% drivetrain loss)
DCT: 326rwhp (21% drivetrain loss)
E92: 338rwhp
E90: 330rwhp
91oct: 331rwhp
93oct: 337rwhp

[Eugene-TAIWAN]

wow great information!

thanks a lot for taking the time to explain and figure this out for us.

[DLSJ5]

Good post.

[sciaenops]

Originally Posted by TrackRat

How the dyno is operated, what correction factors are used for weather and drivetrain loss, etc. all enter the picture. When properly conducted a properly maintained Dynojet and Mustang dyno showed the same power within 3 hp from the same vehicle. Few dynos are properly operated however in my experience. That's why most chassis dyno testing is a waste of time if you're looking for accurate torque/HP data. Car companies use engine dynos for power and chassis dynos for emissions work. In addition without controlled test conditions none of the data is accurate.

I think you could say this 20,000 times and some people still wouldn't get it...

excellent post.

[Bob MG]

So true!
Each car runs slightly different power as you state. Some cars are just stronger than others since all engine internals are build to be within a variance range, some being mire in sync than others as far as balance is concerned.

The way you run the motor in should also make a difference. Will these variable, it is logical each factory motor will perform within a Hp range.

[Small Yellow]

I hope mine is among those over 360 rwhp hhaha.

[MspiredM3]

Very nice post!

[disruptv]

Ok with all that being said, what's the best way to dyno a dct car? I 4got the exact process but I remember there was a way to trigger all the lights versus the car not letting u shift all the way to redline...

Thanks

[///Monster]

NICE!!

Originally Posted by PencilGeek

In the past few days, I've added more than two dozen dyno charts to the DynoDB. I've added a "transmission type" field to the Dyno Database -- to help correlate differences between the two types of vehicles.

Then I performed all of the research necessary to track down the transmission type of each of these cars -- literally dozens and dozens of DynoDB entries. I even had to resort to finding where the OP sold his car on ebay -- just to find what transmission he had; listen to his videos to hear the shifting -- to know if it's DCT or 6MT; or find his thread about putting in a short shift kit (or 4.10 final drive) to determine that he's a 6MT. I also tracked down the premium gasoline octane ratings in at least one half-dozen states -- just to make sure the DynoDB entries were correct. And during this process, I found at least 6-8 more baseline dynos that were never entered into the DynoDB -- most all of them from California -- and some of them on the same exact dynos as the baselines from other cars.

As I pointed out to the OP through PMs, his analysis did not take into account any results that were known to be on the same exact dyno, nor differentiate between any dyno brands, nor compare the same type of dyno correction factors, nor try to categorize the results on regional basis. Without taking most of these factors into account, there was no meaningful results that could have been correlated from those methods. Basically, the results needed to be scrapped and restarted.

After entering in all of these new entries and tracking down the transmission types of almost every single car, I looked over all of the Dynojet, 91 octane entries with SAE correction. I didn't take regional differences into account -- because almost all of the results were already from California and Arizona -- where the gasoline is arguably equally as bad as each other. This is what I found.

6MT: 91 Octane; Dynojet, SAE Corrected, 6-Entries. 312-337 whp. Average: 324.5 whp.
DCT: 91 Octane; Dynojet, SAE Corrected, 7-Entries. 320-342 whp. Average: 334.0 whp.

(Click the 6MT or DCT links above to see the entries.)

I began to look at 93 octane results also -- but there just weren't enough. There were only four 93 octane entries from only three different vehicles and all of them in three different states. Comparing those results would be meaningless.

The results seem to show that not only are the lowest baseline dynos all 6MT's -- they're lower by a substantial amount over their DCT counterparts. There's enough Dynojet entries from California and Arizona alone to come up with some pretty meaningful correlation.

EDIT: 18-May-2010, 8:01 AM
For the 6MT, the average was 324.5hp and the stastical midpoint was also 324.5hp. There were three entries below the midpoint, one entry at the midpoint, and two entries above the midpoint. This means the majority of etries were either at the midpoint or lower than the midpoint -- and the two entries above it were the outliers. (Yes I realize this is not the statistical definition of an "outlier.") Just the opposite was true for the DCT. The average was 334hp, but the midpoint was 331hp. There were five entries above the midpoint, and two entries far below the midpoint. Those two entries dragged down the DCT average. Clearly for DCT, the "norm" were the entries above the midpoint and the entries below the midpoint were the outliers. This is very damning evidence for anybody who wants to say the DCT has higher drivetrain losses than the 6MT -- when in fact the data shows that not only is the opposite true -- but the DCT is more efficient than the 6MT by a rather large margin. But if that wasn't enough, consider this...
END EDIT

There were four entires that came from the same exact dyno: two DCT and two 6MT. The two DCTs measured within one horsepower of each other -- and the two 6MT's measured within 3hp of each other. Yet both DCT's measured 15+ horsepower higher than the their two 6MT counterparts.

In the next few weeks, I'll be getting my hands on a bunch of Maha dyno charts -- all baseline runs from the same Maha dyno. Some results were measured in crank horsepower, and others as wheel horsepower. I plan to get as many entries of both varieties as possible. I'm told there are multiple 6MT and DCT entries. Since this is the exact dyno BMW uses at the factory to publish their results, anybody who wants to argue with the Maha and say it can't be trusted -- please be my guest. I've been told that all of the Maha results are all within one to three horsepower of each other regardless of 6MT and DCT. Once I get the results, I'll add them to the DynoDB -- and then as far as I'm concerned, that will be the final nail in the coffin for this "DCT is weaker than 6MT" or "DCT has higher drivetrain losses than 6MT" myth.

[Meff]

Originally Posted by PencilGeek

6MT: 91 Octane; Dynojet, SAE Corrected, 6-Entries. 312-337 whp. Average: 324.5 whp.
DCT: 91 Octane; Dynojet, SAE Corrected, 7-Entries. 320-342 whp. Average: 334.0 whp.

So why the 22 -25 whp variances on stock cars?

1. dyno operator error ?

2. factory engine variances ?

3. how the motor reacts to proper/improper initial break in ?

4. all of the above ?

[Sales@Evolve]

Number 1 mainly + fuel quality used.

The variances from engine to engine are extremely small.

Will post up a few graphs of stock 6MT and DCT's using same fuel to show how consistent a dyno can actually read if used correctly.

We will show our correction factors also.

[aka_phox]

Nice Thread!

[CanAutM3]

As I have posted on a previous thread:

The whole discussion on dynos also got me very interested. I am an aerospace engineer and have been working on aircraft engines for the past 18 years, mostly in the engine test department. So I am quite used to working around dynos (we also do turboprops and helicopter engines, in case some of you wonder how we would dyno a jet engine ).

I have to agree with swamp on the dyno discusion. Inertia based dynos are not very accurate when producing absolute numbers. They essentially estimate the horsepower at the wheel by calculating the potential to accelerate a rotational mass. These dynos not only take into account drivetrain losses (real losses, which are mostly friction) but also take into account the intertia of the whole drivetrain (flywheel, gearbox, drive shafts, wheels and tires, etc...). For example, installing a lighter wheel-tire combo will yield better results for the same car on an inertia dyno, when in fact the engine power is un-altered.

I quickly went through the www.rri.se site. Their test methodology and equipment (brake dynos) is much closer to what we use in the aerospace industry and what car manufacturers use to establish engine output. These methods, IMO, are much more accurate and repeatable than any inertia dyno. You can actually test the same engine on very different days and still get very repeatable corrected results (well within 1%).

Steady state measurement on a brake dyno is the most accurate and repeatable way to measure engine output. However, intertia dynos do provide a good approximation of a cars ability to accelerate in the real world and are a good tool for tuners.

My 2 cents...

[kmarei]

"... and then as far as I'm concerned, that will be the final nail in the coffin for this "DCT is weaker than 6MT" or "DCT has higher drivetrain losses than 6MT" myth"

Or that the DCTs make more power to account for the higher drivetrain losses?

[tibra1]

Originally Posted by kmarei

"... and then as far as I'm concerned, that will be the final nail in the coffin for this "DCT is weaker than 6MT" or "DCT has higher drivetrain losses than 6MT" myth"

Or that the DCTs make more power to account for the higher drivetrain losses?

Surely you can agree that DCT is at least ALOT heavier than an MT..

Plus heat issues down the line would be a concern.

[kmarei]

Originally Posted by tibra1

Surely you can agree that DCT is at least ALOT heavier than an MT..

Plus heat issues down the line would be a concern.

You're preaching to the choir
DCT is heavier
Not sure how pencil geek reached that it has lower power train losses
And I've never heard of a 6mt overheating
But in his statement he was overlooking an obvious conclusion

[tibra1]

Originally Posted by kmarei

You're preaching to the choir
DCT is heavier
Not sure how pencil geek reached that it has lower power train losses
And I've never heard of a 6mt overheating
But in his statement he was overlooking an obvious conclusion

Honestly..this is a two year old thread so its quite possible some of these conclusions have been revisted and re-evaluated..

[Krozi]

Good post.


LOL just saw its like a 2 year old thread haha

[CanAutM3]

Originally Posted by Krozi

Good post.


LOL just saw its like a 2 year old thread haha

Just noticed this too .

Helluva

[m3an]

Originally Posted by CanAutM3

As I have posted on a previous thread:

The whole discussion on dynos also got me very interested. I am an aerospace engineer and have been working on aircraft engines for the past 18 years, mostly in the engine test department. So I am quite used to working around dynos (we also do turboprops and helicopter engines, in case some of you wonder how we would dyno a jet engine ).

I have to agree with swamp on the dyno discusion. Inertia based dynos are not very accurate when producing absolute numbers. They essentially estimate the horsepower at the wheel by calculating the potential to accelerate a rotational mass. These dynos not only take into account drivetrain losses (real losses, which are mostly friction) but also take into account the intertia of the whole drivetrain (flywheel, gearbox, drive shafts, wheels and tires, etc...). For example, installing a lighter wheel-tire combo will yield better results for the same car on an inertia dyno, when in fact the engine power is un-altered.

I quickly went through the www.rri.se site. Their test methodology and equipment (brake dynos) is much closer to what we use in the aerospace industry and what car manufacturers use to establish engine output. These methods, IMO, are much more accurate and repeatable than any inertia dyno. You can actually test the same engine on very different days and still get very repeatable corrected results (well within 1%).

Steady state measurement on a brake dyno is the most accurate and repeatable way to measure engine output. However, intertia dynos do provide a good approximation of a cars ability to accelerate in the real world and are a good tool for tuners.

My 2 cents...

and that's why I apprecitate/use Mustang Dynos. The ability to place load on the car imitates the actual road and also allows for a more accurate/reliable tune IMO as a Dynojet is not using the load. I don'y know about Dynapack though.
But to the OP I find your statement on Mustang vs. Dynojet reading very close weird. I have never seen a mustang dyno that read w/in 10hp of a dynojet (same car).

[CanAutM3]

Originally Posted by kmarei

Not sure how pencil geek reached that it has lower power train losses

I don't understand it either...

My instincts tell me that DCT should have higher losses. First due to the pumping losses of the clutches spinning in oil rather than air. Second due the power required to run the oil pump for the actuators.

Interesting how a big chunk of the thread was deleted...

[tibra1]

Originally Posted by CanAutM3

Interesting how a big chunk of the thread was deleted...

Agree..two years ago we could have seen the unedited version..