Mystery Solved: Complete correlation between S55 Dyno's, Simulations, and Real World

[swamp2]

Originally Posted by P1 Motorcars

That is not the correct way to do this. CarTest has separate tire and drive train loss calculations. You're proposing to add tire losses to drive train losses. That would add a double correction and would be incorrect. You can see what I mean in the following graph. This graph from CarTest shows separate wheel and drive train (transmission) losses.

It is not at all cool to bash my work as so flawed and then turn right around on post graphs from my analysis as examples to support minor points of your arguments. Wow, talk about inconsistency and hypocrisy. You should minimally cite those results as my work.

[e39>all]

The champ has arrived with opening statements! I'm not sure they'll broadcast this trial on cnn, but I'm anxious to follow the action. I'm not sure swamp is on the right side of this argument, but he's the best litigator on this forum and my money is on him to pull out a stunning comeback victory of attrition nonetheless. 12 round split decision for swamp.

[M3guy3]

So the M4 is indeed underrated. makes sense

[solstice]

Originally Posted by Boss330

Could very well be, but SAE corrections has been proven to be a quite big factor (in one instance for example 40hp). A correction factor of say 40hp for a given atmospheric condition is a lot of noise...

Trap speeds of 114-119MPH (so far) indicates anywhere from 424 to 480hp in a 3528lbs vehicle. And that's before considering that trap speed calculations are based on engines without a plateau power curve, which gives a substantially higher average hp in each gear than the traditional engines trap speed calculators are calibrated for. Just try to put the M550d in a trap speed calculator and you will find that the trap speed indicates a 100hp under stating from BMW.

I'm not saying that I have the final answer or know all the answers. But I am quite certain that the trap speeds we have seen so far really doesn't indicate a under stating from BMW (or at least not more than the 5% margin they are allowed by EU regulations).

It's also interesting how the OP is critical of the shortcomings of the Insoric and dismisses the results that gave (which didn't fit in with the OP's other "findings"), but seems to forget the shortcomings of dyno's and the inaccuracies shown again and again on dynos... The OP accused the other "author" (swamp2) of just using data that fit in with his views, seems to me there was a bit of that going on here instead

The M550d point out a potential weakness in simulation used in isolation which seems to be core of what the OP's work here is trying to highlight, to use different sources to detect this type of potential errors in the individual methods. When applying dyno data to the M550d which seem to correlates well to stated hp ( I know little of this car and just did a quick search and fund some results around 340 hp ) you can see the problem in the simulation. Here the OP detects no such inconsistency, the dyno correlates with the simulation results. The OP goes further and add more data sources as VBOX and find further correlation.

For trap speed it's seem more logical that errors are in the lower trap speed recordings due to the driver being part of the formula. Here you have two out of four "professional" runs, two I assume independent runs at different facilities by different sources recording 119 mph trap speed. I would think this is closer to where the cars power capability lies than the 114 runs and that we will see well executed trap speeds gathering more round 120 mph than 114 mph.

[Boss330]

Originally Posted by P1 Motorcars

We did this to follow proper CarTest procedures. CarTest documentation instructs us to use SAE corrected numbers.

HORSEPOWER: Enter the peak SAE standard J1349 measured net horsepower and the engine rpm at which it is developed.

HORSEPOWER CURVE VALUES: If you are using car-specific parameters, you can override the idealized engine power curve shape that CarTest prepares with the exact engine power curve if it is known. ... All data must be SAE corrected.

CarTest is trying to simulate performance based on real world conditions. Therefore it applies its own (SAE) weather correction to the dyno charts we input, from the new weather we input into the program. This is why we think CarTest makes this requirement.



This isn't our subject matter or expertise. But I still might be able to help. There are four bone stock dyno's in the Dyno Database. One of them is our own. If the S55 ECU can correct for weather to ensure it always has the same rated power regardless of weather conditions, then the UNCORRECTED dyno results should always be nearly the same regardless of conditions.

Here's the graph. You be the judge if these all look the same. Even though our results are in the Dyno Database, we excluded our own results because they are clearly outliers. Our car had less than 100 miles on it and we didn't even take it to redline. But if you want to download and look, they're in the db.





Interesting to see what your opinion is of the uncorrected dyno's above. SAE correction is selected as per CarTest instructions.



The SAE correction conditions are documented in the Dyno Database, and the dyno files are available for download. Here's a recap of what they show. I'm not sure if this is what you consider significant or not.

SAE correction for stock conditions:
Temp: 77.88
Pressure: 29.07 inHg
Humidity: 33%
Correction: 1.021
Uncorrected horsepower: 411 whp
Corrected horsepower: 420 whp

SAE correction for JB4 Tuned conditions:
Temp: 79.52 deg-F
Pressure: 29.10 inHg
Humidity: 30%
Correction: 1.021
Uncorrected horsepower: 461 whp
Corrected horsepower: 470 whp

Since we documented all of our input parameters to CarTest (except two), you are free to use our values. We welcome people checking our work, or pointing out any errors we made with our input values. If the previous CarTest article published this data, we believe we would have been able to point out what was wrong with it. You are free to verify our results and download the files from the Dyno Database just as we did, change to uncorrected horsepower and see what you get.



That is not the correct way to do this. CarTest has separate tire and drive train loss calculations. You're proposing to add tire losses to drive train losses. That would add a double correction and would be incorrect. You can see what I mean in the following graph. This graph from CarTest shows separate wheel and drive train (transmission) losses.





The BT tool doesn't have F8x support yet (we tried), and the Actual Moment isn't accurate anyways. Here's a graph from that shows this inaccuracy. The first graph shows the actual dyno. The second graph shows a comparison between what the actual dyno saw vs. what the ECU "Actual Moment" saw at the same time.







Nope, we haven't tried it. It looks pretty cool. I'll look over the web site technical data and give an update if you're curious what we think. It's worth noting that the math they use to calculate horsepower should be the same thing the vBox Dyno does. The formulas are well established physics formulas. So if you believe in the Dash Dyno, you should believe the vBox Dyno as well.



Interesting to see what you think of those uncorrected dyno's as well.



Dinan uses a Motec ECU on the engine dyno. You can't look at their engine dyno results and tell anything about under or overrating. It is not running the same software as your car.



Here's how to put that into perspective. SAE correction is only valid for +/- 7% correction value. At the maximum 7% correction, the minimum whp required to cause a 40 whp swing would be approximately 612 whp. But more to reality, I just checked the S65 dyno database and see the maximum recorded SAE correction (that doesn't violate the +/- 7% spec) was 3.4%. It would take approximately 1175 whp at 3.4% correction to cause a 40 whp swing. Curious if you have any links to the 40whp correction swing you mentioned?



I don't know about EU regulations and that's not our area of expertise. But I am curious what you think of the vBox Dyno calculations. At the end of the day, we're pretty sure all horsepower calculator uses the same physics formulas as used by the vBox Dyno, and you can see from the vBox Dyno description it does a helluva lot of calculations and has very fine control of the input parameters.



I'm not sure I follow you here. I think you're talking about comparing Dyno Brand-D to Dyno Brand-B. We've only modeled CarTest based on Dynojet results, so that's the only thing I can really vouch for our methods, accuracy, and results. If you have other verified examples of dyno's + vBox results from the same car, we will be happy to perform the same test and see if we get the same results. PM us for details if this data is available.



No, Terry Burger is based in California and uses 91 octane, that car (if it's the OP's car) is based in MD and used 93 octane. I also don't think Terry does street racing.



Also curious of your opinion of the uncorrected dyno results. BTW if anybody has any more bone stock F8x Dynojet files, I'll be happy to add them to the uncorrected graph and see how they compare. And if anybody has any dyno's + vBox results, I'll be happy to run the same tests and present the results no matter what they show.

Thanks for the response and explanations.

My questions/comments are:

• That CarTest asks for SAE corrected numbers is just as expected, if not the software simulations would be inconsistent. However, if the S55 already has done the SAE correction on the dyno, then this needs to be considered and understood before blindly using SAE correction in this case. I agree that a 9hp correction isn't exactly a large number. But it's 3 times larger than the variation between the uncorrected stock dyno runs (3,31hp), and it represents a 2% increase in HP. When the SAE std has a tolerance on factory hp of 1% and EU allows a 5% tolerance, then 2% is actually quite a bit...
  
  
• The uncorrected dyno's show a variation of just 0,8% in peak power values. They are also quite similar through the rpm range, with the EAS having the largest deviation. Considering how many factors there are that affect chassis dyno readings (tire pressure, friction, operator, and dyno variation etc), this is astonishingly close and certainly seems to indicate that the S55 tries to achieve the same power output under different circumstances (and seems to succeed judging by those dyno results). However, chassis dynos have historically shown a large deviation between different dynos (of the same type and manufacturer) between shops/operators. Even though these dynos seems very consistent (and also could be considered evidence towards the S55's ability to "SAE adjust"), the same caution should apply when comparing these different dyno results.
  
  
• In your reply on the Actual Moment from the ECU vs Dyno, it seems you imply that the chassis dyno is a 100% correct tool, and use that as evidence to support why Actual Moment is wrong... How do you then explain the large variation we see on dyno numbers (same dyno brand and model)? swamp2 compiled a database of various dyno graphs for the S65, and found as much as a 15% variation (again same brand and model dyno). And if the answer is that all of the variation is down to operator and use of the dyno, then how do we know which dyno result to trust?
  
  
• Do we have verification that Dinan doesn't use BMW DME and software on the dyno? Why would they run a Motec ECU on the dyno when the purpose also is to develop software for the BMW DME?
  
  
• Regardless of Dinan, for type approval purposes in EU they have to run the engine on a engine dyno at a independent technical service. They need to have the std exhaust system as well as intake system present when doing this test. The manufacturers claimed hp/Tq is only accepted if the power readings are within 5% of what BMW claims it to be. If we are to believe that there is a significant under rating, we will also have to believe in a conspiracy, by BMW, to mislead European authorities. Such a conspiracy would need to involve quite a number of employees from BMW and/or test personell at TÜV. A lot of people that BMW needs to keep silent about this deception...
  
  
• In Norway there was a civil case against Nissan that lasted 5 years, where it was claimed that the Nissan X-Trail had a engine that was over rated by 20hp. This claim was made on the basis of chassis dyno testing by a consumer organisation. They adviced their members to get a refund of approximately USD 4-5.000. In the end Nissan won the trial because of the uncertatinty of drivetrain losses and chassis dyno accuracy. The court used the type approval documentation and their engine dyno numbers as those where the only reliable source...
  
  
• And if BMW has under stated the crank hp and the engine in reality makes 480hp, then that would mean a new car tax evation in Norway of over $ 11.000,- per car... Bad for reputation and possible heavy fines if someone finds out (and people in Government and EU Commission also read forums and might start asking questions to BMW...)
  

[Boss330]

Originally Posted by solstice

The M550d point out a potential weakness in simulation used in isolation which seems to be core of what the OP's work here is trying to highlight, to use different sources to detect this type of potential errors in the individual methods. When applying dyno data to the M550d which seem to correlates well to stated hp ( I know little of this car and just did a quick search and fund some results around 340 hp ) you can see the problem in the simulation. Here the OP detects no such inconsistency, the dyno correlates with the simulation results. The OP goes further and add more data sources as VBOX and find further correlation.

For trap speed it's seem more logical that errors are in the lower trap speed recordings due to the driver being part of the formula. Here you have two out of four "professional" runs, two I assume independent runs at different facilities by different sources recording 119 mph trap speed. I would think this is closer to where the cars power capability lies than the 114 runs and that we will see well executed trap speeds gathering more round 120 mph than 114 mph.

My understanding is that what throws most of us off is the "strange" hp curve of the S55. The power plateau from 5500-6300rpm, where the engine constantly makes 425hp, is bound to make the F8x perform way better than a car with a engine that only makes 425hp from, say, 6200-6300rpm (like in a traditional NA engine).

On a 1/4mile run in the F8x, every time you shift gear the revs drop to around 5500rpm, but you still have 425hp on tap from the engine. In a NA engine like the S65 (or any engine with a traditional "non plateau" power curve) you have max power at your shiftpoint/redline, but when you change gear and drop down in the rev range you don't start with max power. You might be 50hp down from what you had at redline in the previous gear...

What matters here is the average hp in each gear. In the S55 that is 425hp (because we are always at max power under acceleration), but in the S65 it might be as "low" as 380-390hp (because, in each gear, we start off at a low hp level and end up at max hp only at redline). In effect this means that the S55 performs like a 420whp engine (with a traditional power curve) since it probably has the same average HP in each gear as a more powerful engine with a "normal" hp curve has.


For instance if we just use as an example a engine that in each gear starts with 380hp @ 6000rpm and ends up at 480hp @ 7800rpm. Let's also assume, for ease of calculation, that the power delivery is perfectly linear between 6000rpm to 7800rpm.

With this engine we would start at 380hp in each gear under acceleration (apart from in 1st at launch) and the engine would gradually build power towards the shiftpoint at 7800rpm where we have max power of 480hp.

The interesting thing here would of course be what average power this engine develops in each gear (because that would tell us about the total acceleration. Peak power means little since the engine only sits there for a very brief moment).

Average power, in gear from 6000-7800rpm, for this hypothetical engine would be:

480+380/2 = 430hp

Here we have a 480hp engine that makes an average power, in gear, of 430hp. Not that far off the average power of the S55, which only is a 425hp engine...

I believe my example shows how the S55 can perform like a much larger "traditional" engine under acceleration. Because it's max power is also it's average power, in each gear, under acceleration.

Hope this makes sense

[solstice]

If the car for all practical purposes acts, performances, dynos and feel closer to 500 hp than 425 who really care about the academic question if a lab mounted engine is within specs in a scenario it will never be used in? It seems like a question only interesting to officials and not car buyers. It's not even interesting for simulation if the resulting academic crank hp is no longer reflective of real life performance. I still suspect the engine is underrated vs. actual power delivered when the S55 is working in a production car even if I don't for a minute think it's a conspiracy. Some tuning of parameters or ( legal ) lack of or normalization of some DME inputs maybe but not a conspiracy involving officials as in lying of measured results.

[Boss330]

Originally Posted by solstice

If the car for all practical purposes acts, performances, dynos and feel closer to 500 hp than 425 who really care about the academic question if a lab mounted engine is within specs in a scenario it will never be used in? It seems like a question only interesting to officials and not car buyers. It's not even interesting for simulation if the resulting academic crank hp is no longer reflective of real life performance. I still suspect the engine is underrated vs. actual power delivered when the S55 is working in a production car even if I don't for a minute think it's a conspiracy. Some tuning of parameters or ( legal ) lack of or normalization of some DME inputs maybe but not a conspiracy involving officials as in lying of measured results.

The "scenario" is to as exactly as possible replicate how the engine will run in the car. If not it wouldn't make any sense to dyno it... The unrealistic dyno sessions was the way they dyno'd the good old muscle cars with gross power ratings. When they started with net ratings, power dropped by 10-20% because they suddenly had to replicate a realistic "in car" situation. This means same intake system including air boxes and air filter box. Same as std exhaust system, and ancilliaries etc.

There is a reason race engine builders use engine dynos to test and develop their engines. And they also want to make sure that the engine performs the same way when it's in the race car...

My point is that the "academic" crank hp actually represents very well how the S55 performs in the car and why it performs like a much more powerful engine, due to it's high average hp.

I believe, both because of the trap speeds and high average hp of the S55 AND the rigid and thourough EU-type approval framework (both legal and technically) that there is a VERY small possibility that the S55 is under rated by anything more than 5% (which is the allowed EU tolerance).

BTW, any "tuning" of parametres or "normalization" of DME parametres would all imply misleading and misrepresenting the engine when tested for type approval authorities. The dyno room conditions are set up according to strict parametres and any "special software calibrations" for the dyno run would certainly have to involve a lot of people that knowingly program those parametres to achieve a lower hp result...

[lfelunden]

OT but, how do you measure whp on the vBox??

[gthal]

I'm still confused

[solstice]

The power plateau is nothing really unusual. Look at the N63TU for example. It is quite interesting since it has similar power levels and power delivery. It is rated 20 hp above the S55 but seem to dyno about 20 hp below the S55. Furthermore when looking at the shape of the power curves it should average similar power from 4500 rpm to 6500 rpm as the S55 between 5500 and 7500. I.e it should get higher to match it's 445 hp rating to make up for it's drop off towards the redline. I.e it's power plateau and area under the graph should provide similar effect in power between shift points.

Unfortunately there is no car of the M3s' weight with an N63TU, however I'm not sure how much the weight difference really impact trap speed in cars of this size and engine type. The track day I spent with all the N63TU cars I managed the highest trap speed of the day at 109.5 mph in the 750 Li... I ran the others as well as the 650 coupe but it was not faster...

Looking at other trap speeds for these cars it seems like the N63TU cars trap at about 10 mph down on the S55 cars. That seems a lot, more than the lower weight and stated less hp should allow. It just seems that whatever reasonable theory is raised for no under rating as double correction and power plateau they just don't hold to explain real life data, relative or absolute.

[P1 Motorcars]

Originally Posted by Boss330

That CarTest asks for SAE corrected numbers is just as expected, if not the software simulations would be inconsistent. However, if the S55 already has done the SAE correction on the dyno, then this needs to be considered and understood before blindly using SAE correction in this case. I agree that a 9hp correction isn't exactly a large number. But it's 3 times larger than the variation between the uncorrected stock dyno runs (3,31hp), and it represents a 2% increase in HP. When the SAE std has a tolerance on factory hp of 1% and EU allows a 5% tolerance, then 2% is actually quite a bit...

I don't think you mean to say the DME does SAE correction, but I think you mean to say it regulates and even limits power based on ambient conditions. It may be a very fine nuance, but the two are very different.

In your reply on the Actual Moment from the ECU vs Dyno, it seems you imply that the chassis dyno is a 100% correct tool, and use that as evidence to support why Actual Moment is wrong... How do you then explain the large variation we see on dyno numbers (same dyno brand and model)? swamp2 compiled a database of various dyno graphs for the S65, and found as much as a 15% variation (again same brand and model dyno). And if the answer is that all of the variation is down to operator and use of the dyno, then how do we know which dyno result to trust?

We've done a lot of work in this area. Yes we can say with absolute certainty that Actual Moment is not accurate. In the example we gave, the shape of the curves don't look remotely similar and the Actual Moment curve doesn't look like anything you've ever seen on an S65. There's also a ~140 whp difference between dyno and Actual Moment.

Do we have verification that Dinan doesn't use BMW DME and software on the dyno? Why would they run a Motec ECU on the dyno when the purpose also is to develop software for the BMW DME?

Yes, 100%. You can see it in the photos of their engine dyno's of the S65, and you can call (or in your case email) and that's what they will tell you. Dinan has a second dyno they use for DME testing. They test DME software on their Dynapack.

I can't comment on the EU regulations. We don't know anything about them. But you certainly make a good point.

Originally Posted by lfelunden

OT but, how do you measure whp on the vBox??

I believe it's done through a series of well known physics equations. You would need to contact the admin at vboxtools.com for more info.

[FogCityM3]

Thanks for your response, appreciate it.

Yes, I would be curious to see your opinion of DashDyno, its strengths/deficiencies, etc.

Also, as far as the actual moment from the ECU not being accurate, none of my own logging corroborates this. If anything, it tends to run higher than I would think (not lower and then I do the dash dyno run and apply the 13%-15% drivetrain losses and seems about right). However, if that dyno was from a tuned M3, that could account for the difference/inaccuracies. I own two tunes, one of the tuners changes the minimum torque tables (which result in extremely high TQ numbers), one of them does not (results in in line TQ/hp numbers for a tune), and I have logged the stock ECU, and again seems to produce around 410-420 bhp. I believe that others who have done this in the past also have corroborated that on an untuned engine, the actual moment from the ECU is fairly accurate..this is actually what got me started looking at logging this parameter.

Since support is not there for the F80, maybe doing a stock actual moment logging and vbox (or dash dyno) logging would be great to establish a baseline and then use the same procedure for an untuned F80 once the support is there.

Thank you in advance.

Originally Posted by P1 Motorcars

We did this to follow proper CarTest procedures. CarTest documentation instructs us to use SAE corrected numbers.

HORSEPOWER: Enter the peak SAE standard J1349 measured net horsepower and the engine rpm at which it is developed.

HORSEPOWER CURVE VALUES: If you are using car-specific parameters, you can override the idealized engine power curve shape that CarTest prepares with the exact engine power curve if it is known. ... All data must be SAE corrected.

CarTest is trying to simulate performance based on real world conditions. Therefore it applies its own (SAE) weather correction to the dyno charts we input, from the new weather we input into the program. This is why we think CarTest makes this requirement.



This isn't our subject matter or expertise. But I still might be able to help. There are four bone stock dyno's in the Dyno Database. One of them is our own. If the S55 ECU can correct for weather to ensure it always has the same rated power regardless of weather conditions, then the UNCORRECTED dyno results should always be nearly the same regardless of conditions.

Here's the graph. You be the judge if these all look the same. Even though our results are in the Dyno Database, we excluded our own results because they are clearly outliers. Our car had less than 100 miles on it and we didn't even take it to redline. But if you want to download and look, they're in the db.





Interesting to see what your opinion is of the uncorrected dyno's above. SAE correction is selected as per CarTest instructions.



The SAE correction conditions are documented in the Dyno Database, and the dyno files are available for download. Here's a recap of what they show. I'm not sure if this is what you consider significant or not.

SAE correction for stock conditions:
Temp: 77.88
Pressure: 29.07 inHg
Humidity: 33%
Correction: 1.021
Uncorrected horsepower: 411 whp
Corrected horsepower: 420 whp

SAE correction for JB4 Tuned conditions:
Temp: 79.52 deg-F
Pressure: 29.10 inHg
Humidity: 30%
Correction: 1.021
Uncorrected horsepower: 461 whp
Corrected horsepower: 470 whp

Since we documented all of our input parameters to CarTest (except two), you are free to use our values. We welcome people checking our work, or pointing out any errors we made with our input values. If the previous CarTest article published this data, we believe we would have been able to point out what was wrong with it. You are free to verify our results and download the files from the Dyno Database just as we did, change to uncorrected horsepower and see what you get.



That is not the correct way to do this. CarTest has separate tire and drive train loss calculations. You're proposing to add tire losses to drive train losses. That would add a double correction and would be incorrect. You can see what I mean in the following graph. This graph from CarTest shows separate wheel and drive train (transmission) losses.





The BT tool doesn't have F8x support yet (we tried), and the Actual Moment isn't accurate anyways. Here's a graph from that shows this inaccuracy. The first graph shows the actual dyno. The second graph shows a comparison between what the actual dyno saw vs. what the ECU "Actual Moment" saw at the same time.







Nope, we haven't tried it. It looks pretty cool. I'll look over the web site technical data and give an update if you're curious what we think. It's worth noting that the math they use to calculate horsepower should be the same thing the vBox Dyno does. The formulas are well established physics formulas. So if you believe in the Dash Dyno, you should believe the vBox Dyno as well.



Interesting to see what you think of those uncorrected dyno's as well.



Dinan uses a Motec ECU on the engine dyno. You can't look at their engine dyno results and tell anything about under or overrating. It is not running the same software as your car.



Here's how to put that into perspective. SAE correction is only valid for +/- 7% correction value. At the maximum 7% correction, the minimum whp required to cause a 40 whp swing would be approximately 612 whp. But more to reality, I just checked the S65 dyno database and see the maximum recorded SAE correction (that doesn't violate the +/- 7% spec) was 3.4%. It would take approximately 1175 whp at 3.4% correction to cause a 40 whp swing. Curious if you have any links to the 40whp correction swing you mentioned?



I don't know about EU regulations and that's not our area of expertise. But I am curious what you think of the vBox Dyno calculations. At the end of the day, we're pretty sure all horsepower calculator uses the same physics formulas as used by the vBox Dyno, and you can see from the vBox Dyno description it does a helluva lot of calculations and has very fine control of the input parameters.



I'm not sure I follow you here. I think you're talking about comparing Dyno Brand-D to Dyno Brand-B. We've only modeled CarTest based on Dynojet results, so that's the only thing I can really vouch for our methods, accuracy, and results. If you have other verified examples of dyno's + vBox results from the same car, we will be happy to perform the same test and see if we get the same results. PM us for details if this data is available.



No, Terry Burger is based in California and uses 91 octane, that car (if it's the OP's car) is based in MD and used 93 octane. I also don't think Terry does street racing.



Also curious of your opinion of the uncorrected dyno results. BTW if anybody has any more bone stock F8x Dynojet files, I'll be happy to add them to the uncorrected graph and see how they compare. And if anybody has any dyno's + vBox results, I'll be happy to run the same tests and present the results no matter what they show.

[ASAP]

I have a simple / unscientific test that could be telling.

Why not dyno a stock M4 on this dyno and get a result of 425 whp or whatever the number everyone has been claiming is. Then throw an FBO and e85 N54 that almost always dyno's around that power on Dynojets. If the power results on the dyno are similar, then race the two cars or take them to the track and compare trap speeds. Wouldn't this immediately resolve any inaccuracies in adjustements for ambient conditions if the cars are far off from each other in terms of real world performance?

[Black Gold]

Originally Posted by solstice

The power plateau is nothing really unusual. Look at the N63TU for example. It is quite interesting since it has similar power levels and power delivery. It is rated 20 hp above the S55 but seem to dyno about 20 hp below the S55. Furthermore when looking at the shape of the power curves it should average similar power from 4500 rpm to 6500 rpm as the S55 between 5500 and 7500. I.e it should get higher to match it's 445 hp rating to make up for it's drop off towards the redline. I.e it's power plateau and area under the graph should provide similar effect in power between shift points.

Unfortunately there is no car of the M3s' weight with an N63TU, however I'm not sure how much the weight difference really impact trap speed in cars of this size and engine type. The track day I spent with all the N63TU cars I managed the highest trap speed of the day at 109.5 mph in the 750 Li... I ran the others as well as the 650 coupe but it was not faster...

Looking at other trap speeds for these cars it seems like the N63TU cars trap at about 10 mph down on the S55 cars. That seems a lot, more than the lower weight and stated less hp should allow. It just seems that whatever reasonable theory is raised for no under rating as double correction and power plateau they just don't hold to explain real life data, relative or absolute.

good post

I think its also appropriate to consider the fact that this is not the only engine that seems to be underrated

the new 5.5L turbo from merc, the 3.0SC from audi, the 4.0TT from audi to name a few have all been dyno'd at significantly higher values than the manufacturers claimed values

they also perform in line with those numbers.

[Boss330]

Originally Posted by P1 Motorcars

I don't think you mean to say the DME does SAE correction, but I think you mean to say it regulates and even limits power based on ambient conditions. It may be a very fine nuance, but the two are very different.

A very valid correction

And, yes that's how I understand the DME operates. As you say (and has we have discussed at length in the other thread), the understanding is that the DME regulates for ambient/altitude (probably also including limiting/lowering boost in cold weather at low altitude).

[SOM3]

So bottom line is: M3/4 > you

Right?

[fusionchicken]

Read the whole thread

Just one question

Uwotm8

[Brosef]

this thread is hopelessly bereft of data. where's Swamp??

[P1 Motorcars]

Originally Posted by swamp2

I sincerely appreciate additional investigation into what is indeed a somewhat complex and confusing situation. However, there are significant problems with this analysis some of which have already been pointed out. There are also completely inaccurate characterizations of my work and post of the topic. I will provide some substantial additional comments when I get a moment to "surface".

We posted all of our input data, and we welcome you to do the same. If it will help, we will post screen shots of our input screens and block out the two pieces of data we want to protect. We welcome you to do the same and then we can all compare and contrast and have a purely technical discussion about it. We are really curious why our simulations worked the first time and yours were so far off when you used the WHP method. We’d really like to see what values you used; so please post them.

Originally Posted by swamp2

It is not at all cool to bash my work as so flawed and then turn right around on post graphs from my analysis as examples to support minor points of your arguments. Wow, talk about inconsistency and hypocrisy. You should minimally cite those results as my work.

Today we learned the word "douchbaggery" was officially added to the English dictionary. I think you're trying to say we conducted some douchbaggery. Last night it was late on the east coast when I was trying to answer people’s questions and yes I grabbed the easiest graph I could. I have replaced the graph with one we generated ourselves. It still shows the same thing regardless of who generated it.

Originally Posted by FogCityM3

Also, as far as the actual moment from the ECU not being accurate, none of my own logging corroborates this. If anything, it tends to run higher than I would think (not lower and then I do the dash dyno run and apply the 13%-15% drivetrain losses and seems about right). However, if that dyno was from a tuned M3, that could account for the difference/inaccuracies. I own two tunes, one of the tuners changes the minimum torque tables (which result in extremely high TQ numbers), one of them does not (results in in line TQ/hp numbers for a tune), and I have logged the stock ECU, and again seems to produce around 410-420 bhp. I believe that others who have done this in the past also have corroborated that on an untuned engine, the actual moment from the ECU is fairly accurate..this is actually what got me started looking at logging this parameter.

Here's what I get from your own analysis. The ECU is not generating Actual Moment by pure calculations -- but aided by data tables. That alone proves it’s not accurate. If the data were generated by pure calculations (such as F=M*A), then the results would be rather immune to engine modifications. BMW could even add wind resistance, wheel losses, friction losses to the equation because they know exactly the weight distributions of the wheels, tires, etc. But we can see they didn't do that, and your own analysis shows that the values you get out are dependent on the data programmed into the ECU torque tables. The graph I posted shows just how far off this method can be.

BTW, I’ll look at the Dash Dyno a little later. I’m really curious where they’re getting velocity. But we’ll look into it and post an update. We might even buy it and try it on our own F80. It would be a really good comparison to the vBox and vBox Dyno.

[FogCityM3]

My point is that it appears that some sort of minimum torque setting that can be overridden with certain tunes and influence the actual moment, otherwise it is a strict calculation. But on a stock car and with certain tunes the calculations appear to be directly affected by various parameters, unless something is overridden (based on my discussions with tuners and my own logging of dozens of parameters, where it can be clearly seen). It is my understanding the ECU looks for a minimum torque value that is calculated (or set so low in the table that it won't matter unless things are going wrong..I'll have to look back at my logs but I believe I get different miminum torque ratings based on environmental factors, showing that untuned, it is a calculated value from various parameters) or alternatively with different tunes, this calculation be overridden. I just know what I've gotten with stock tune, and two different tunes using actual moment with dozens of logs. The person who maintained the dyno DB also did some actual moment logging and seemed to work fine with stock cars, but not with all tuned cars. You can talk to the various tuners as to what they modify/don't modify. Also, when I put in better fuel, the timing targets are hit (vs timing pulled on 91) and there is a direct change in power as calculated by actual moment, even at equivalent temperature/ pressure. Also, I have observed that changes in mass airflow, altitude, atmospheric pressure, IATs vs ambients all influence power/actual moment produced. I have the benefit of road logging on flat roads at sea level, where it is very easy to control for environmental factors and established enough baselines to see how changes in various parameters influences power. So to me is shows that various inputs change the calculation, unless things in certain tables get overridden.

For example, I just logged an E85 mixture, and I got the highest Actual Moment value I have ever logged, under any temperature or pressure conditions.

Since we're talking stock, from all my logging, there does not appear to be evidence I can see from my dozens of logs that actual moment isn't accurate on a stock car, as it matches the factory stated bhp (with 93 octane) and my dash dyno logging.

Also, one of the points I was trying to make earlier, but noticed I was missing a few words that made it really unclear is that would be curious to try Vbox/Dash Dyno and compare to BT actual moment on an E90 M3, on the open road. This would establish a baseline for the accuracy of open road logging per wheel hp calculations and what the ECU says the engine is producing (per BT actual moment) for the S65, then test it with the S55 once BT offers support for the S55.

Thx for looking into the dash dyno.. I'm curious about your thoughts.

Originally Posted by P1 Motorcars

Here's what I get from your own analysis. The ECU is not generating Actual Moment by pure calculations -- but aided by data tables. That alone proves it’s not accurate. If the data were generated by pure calculations (such as F=M*A), then the results would be rather immune to engine modifications. BMW could even add wind resistance, wheel losses, friction losses to the equation because they know exactly the weight distributions of the wheels, tires, etc. But we can see they didn't do that, and your own analysis shows that the values you get out are dependent on the data programmed into the ECU torque tables. The graph I posted shows just how far off this method can be.

BTW, I’ll look at the Dash Dyno a little later. I’m really curious where they’re getting velocity. But we’ll look into it and post an update. We might even buy it and try it on our own F80. It would be a really good comparison to the vBox and vBox Dyno.

[e39>all]

Originally Posted by Got Trails

Cliff notes?

Lol, this whole thread is probably just the cliff notes of the cliff notes of swamps thread