TB993tt

In the interests of geekiness I wanted to have my car "power tested" to compare with my RS Tuning engine dyno sheets.
I chose Manthey Racing who have one of the best reputations for race car prep in Europe and have a state of the art Maha LPS3000 chassis dyno. I met up with RLers Stummel and Phelix.....

Preparing the car for the dyno takes about an hour with them decoupling the 4WD , chocking the springs at the back, putting on some "dyno wheels" and snugly fitting a shroud over the intercooler so they can blow air through during the run

The car is postioned on the rollers and all the other air blowers are positioned. There is a huge one at the front which takes care of the oil cooler then two big ones at the back, one blowing through the sealed intercooler shroud and the other blowing directly into the air box.

A series of warm up tests are done where the dyno would put a load on the engine, a 1 minute clip is here :http://www.youtube.com/watch?v=vPqQe6qEuRA
The clip is boring but gives you an idea, the round gauge on the far left is the load being applied to the rollers, the next one along is the power in KW and the next is the revs. The warm up goes on for about 30 minutes but all the time the air flow is maintained so the car does not get any hotter than on the road so the conditions are very realistic.

The dyno is then set into power run mode, it seemed to have very detailed settings and was set for "aircooled turbo motor".
The power run is just like other chassis dynos with the BIG difference being that the Maha loads with resistance against the oncoming torque at the wheels -this is where the problem hits....

Bottom line is it seems that around 700-720NM (515-530lb/ft) is the limit of grip for the rear wheels - I know Jean's car had the same issue....
The operator tried about 5 times but the torque was too much he resorted to using part throttle through peak torque then nailing it at higher revs -eventually he managed to get me a peak power number, the part throttle run took 23 seconds

I have studied the torque diagram and compared it with my engine dyno sheets and it appears that the machine only managed to record all the torque at the last couple of hundred revs with the number at 6775rpm being virtually spot on (actually 2PS more !)

Just to confirm my "head geek" credentials I insisted on sitting in the car and watching the Intake air temperature (on the Andial/davtron gauge) whilst the operator did a few more maximum rev runs - to my amazement it only went 20degC above ambient showing what a great air flow set up they have there and also that the Secan was indeed working as advertised....

So the question is, how does the best chassis dyno available (Maha LPS3000) measure torque of more than ~720NM and just how accurate do we think these big torque numbers are coming from cheaper chassis dynos ?

This made me think of a dyno sheet I saw from Gemballa who also operate this type of dyno - they managed to measure 841NM and 699PS on a 996GT2 on their Maha
Here is the sheet:

I have exactly the same sheet as this for the part throttle run and it makes for some fascinating reading....

The RWHP (the blue line) on my part throttle run at 5000rpm is almost the same as on the Gemballa run, so does this mean that my car also has 841NM @5000rpm ?

The RWHP on my run at the peak power is within 4KW of the Gemballa peak power - so does my car also have 699.5PS ?

The answer lies in the Ansaugluft Temperatur - this is the air intake temperature. The Gemballa has an IAT of 75 degC. My IAT was 10.5degC....

Because of the very high Gemballa number that engine was given another 88hp as a correction for the inlet temp and it got 136 hp from coast down losses compared to my (ultra efficient 993tt ) 90hp.

Now that 10degC intake temp measured for my car was courtesy of the big pipe blowing cold air (from outside) directly into the airbox which is where the temp sensor was sitting - where did Gemballa put their sensor ? In the post intercooler intake tract ? that seems most likely....

The question is, is this "cheating" - I would say probably not (unless the Maha handbook specifically states where that temp should be read from) but isn't it amazing how it can effect the results and how Gemballa can show a 700hp engine which would probably read more like 570hp if measured a different way - and of course (presuming the Gemballa had grip on the dyno) if my car could have been run at full throttle it would certainly have pulled much higher torque number at the wheels than the Gemballa......

Now about those 8 second dynojet runs.........

Stummel

good to see you back home.

the dyno session was endless pleasure although one of those beers the night before must have been bad.

It is funny that the "Schleppleistung" (drag loss?) on the Gemballa reads 50% higher than on your runs... If the operator puts his foot slightly on the brake at the 2nd part of the run one can get whatever power figures one wants or needs.

For those interested in a RS-Tuning custom tune. TBs car runs as smooth as a stocker. Difference is that it accelerates stronger in 5th gear than a stocker in 2nd gear. Also has lots more power off boost at low rpm than stock one.

911/Q45

The massive supplemental air flow used should be a clue as to the usefulness of the wimpy oil cooler fan that some wire to switch on while the car is moving.

JasonAndreas

TB993TT,
Have you tried any of the hub mounted dynamometers?

Basal Skull

Great information! Thanx

vs
So when I read about IAT or intake air temperature, depending on the set up it can mean before or after the intercooler?
So, with your Secan, your IAT after the intercooler is <30.5degC vs 75 degC of the Gemballa?
or am I getting terms mixed up - is there a specific name for the temp before or after the intercooler?

The IAT the I can get from my OBD2 port is after the intercooler?

Ryojo

TB993tt

I do seem to recall mentioning that one beer was a little warmer than the previous 16 - that could be the culprit
Good point - there is no reason why a 996GT2 should have so much more drag/frictional losses is there ?
911/Q45
I noticed that the oil gauge did not budge all session, stayed right at the bottom - it is amazing watching your car seemingly being "tortured" but in reality, with all the temperatures being representative of the road it is no harder on the car.....
Jason
That should get rid of the slippage issue shouldn't it ? I don't know enough about chassis dynos but I guess Manthey bought what they thought was the best for testing real power - maybe the hub ones cannot measure as accurately ? Certainly when one considers the frankly pretty measily 7 second power runs which reveal 750hp and 700lb/ft for just whizzing a drum up to speed there is a lot of snake oil swooshing around out there.....
Just what are the dynojets doing when confronted with the torque, they don't attempt to brake it and measure what force is needed to overcome it, they are just working on the rate of acceleration of the drum I guess which seems prone to all sorts of errors, but fortunately those errors are what bring smiles to their customers faces

TB993tt

You're welcome, thanks for saying thanks

No you are not getting confused, you are correct. In the above case my IAT was actually 28degC and (I am guessing subject to the arguments above) the Gemballa is 75DegC

There is no specific name for post intercooler temp - For turbo cars we are only interested in the temp AFTER the intercooler since this is what will determine the output of the engine -but as the Gemballa example above shows you have to really check the detail to get the correct picture....

I have emailed Manthey to ask them what Maha specify to use as "ansaugluft temperature" on these turbo engines....

Jean

TB

Great post, thanks for sharing your experiences. I wish I had looked into 50% of the detail that you did but I couldn't wait to jump into the 'Ring.

Phelix you seem to have gotten the addicition to dynos, I guess you are an expert MAHA operator by now

I am not sure I have digested yet all the info but it does look like very interesting learnings. I have been trying to find my dyno sheet but I can't I will call them on Monday and hope they have it on file, otherwise will go through the video. It will be interetsing to compare data and curve, since it is on the same dyno and the same sadistic operator.

I recorded close to 750NM if I recall correctly, but that was after several loaded tries on the dyno for over half an hour and a lot of slippage, they finally gave up and just handed the info showing us the door...

Glad that you all made it well and safely, I bet you had a blast. Stummel when is your turn ?

Stummel

Jean, they will no longer have it on file after 2(?) yrs.

My turn seems to be somewhere in 2009 or so, when we also have general speed limit in Germany :-(
At least now I am quite sure about the power level I neeeed.

Stummel

TB, you could give the technical support at MAHA a call:
+49 (0)8374-585-90

TB993tt

Jean
Thanks for your inspiration....

The highest torque it managed to "grab" was actually ~766NM, you can just make it out on the screen shot below, it is 353.9KW @4410rpm

I suspect the dyno sheet they gave you was very much like mine, completely useless really

eclou

Do they use special tires for the testing? I would consider using R compounds perhaps and get them up to operating temps prior to making the dyno runs if you are experiencing slippage. If they cannot set-up the dyno to get enough traction, it really doesn't matter how great the Maha is eh? Impressive output despite the slippage anyway!

TB993tt

The slippage comes because the dyno loads up and tries to match the torque - other dynos don't do this so don't get this kind of slippage -whilst obviously this is a limitation of the Maha and it seems measuring over 750NM is challenging..

I wonder just how inaccurate other chassis dynos are in measuring high torque, we have seen many instances of silly torque numbers (remember the Weltmiester debacle, the 9M Bosch dyno showing 800NM on K24 cars with just 1 bar ?) -is it that non loading (and non slipping) dynos cannot measure peak torque accurately at all on these turbo motors ?

eclou

I think it was Jean who mentioned that .9g is about the limit of what street tires can provide for traction in acceleration, and race tires closer to 1.x. I would think that this could help in getting a MAHA run with better resistance to slippage.

Aslo, if your 800+ NM car can get traction in real world acceleration runs on street tires vs the Maha being unable to read 800+ NM without losing traction then I would argue that the Maha empirically cannot duplicate real world measurements - does that sound logical?

TB993tt

I think the underlying point may be how many cars are there with 800NM REAL torque ?
Sure I don't doubt that other tuners can claim 8XXNM as shown on their non loading chassis dynos or done like Gemballa above but get the feeling that Manthey are quite used to handing people a dose of reality when they test their cars -when they did my first complete run and converted KW to PS using a calculator, the operator nodded as he thought the number was too high and they had made an error somewhere (point being they do not see too many 993tts with those sorts of numbers, the last was probably Jean's)

The traction issue road vs dyno is a bit of a red herring since the dyno is done in 5th gear so one would have thought traction should be possible...

Who knows the inside story about the Maha at AMS in Cali, someone said it shut down because people didn't like the numbers they got and business dried up ? urban myth ?