GetBMWParts
BMW Garage BMW Meets Register Search Today's Posts Mark Forums Read

Go Back   BMW M3 and BMW M4 Forum > BMW F80 M3 / F82 M4 Technical Topics > Engine / Drivetrain / Exhaust / Bolt-ons / Tuning

Closed Thread
 
Thread Tools Search this Thread
      08-17-2014, 10:48 AM   #89
Boss330
Major General
Boss330's Avatar
No_Country
1712
Rep
5,108
Posts

Drives: BMW
Join Date: Mar 2012
Location: Earth

iTrader: (0)

Quote:
Originally Posted by CanAutM3 View Post
Yet the Maha dyno showed 465ps at the crank. That is 34ps more than the official rating, or 7.8%.

As I posted more than once, the WHP, or P-rad, number obtained on a Maha dyno cannot be used for comparison. Only the crank power number is really valid for comparison.

That being said, I still believe that there is no "underrating" per say. I am still convinced that the S55 (and other FI engines) behave differently on chassis dynos compared to the official test protocol on bench dynos.
I know, but it's still hard to understand the MAHA crank numbers, as they are "all over the place" in the Sport Auto tests...

The whp number of 375ps the MAHA measured, is actually what a 431PS engine with 13% drivetrain losses would get...
Appreciate 0
      08-17-2014, 11:00 AM   #90
CanAutM3
General
CanAutM3's Avatar
Canada
21105
Rep
20,741
Posts

Drives: 2021 911 turbo
Join Date: Jan 2012
Location: Montreal

iTrader: (1)

Garage List
Quote:
Originally Posted by Boss330 View Post
I know, but it's still hard to understand the MAHA crank numbers, as they are "all over the place" in the Sport Auto tests...

The whp number of 375ps the MAHA measured, is actually what a 431PS engine with 13% drivetrain losses would get...
Why do you say they are all over the place?

Further, where do you get that 13% number from ? To me it does not seem to have any foundation.

As I pointed to you many times before, the shape of the P-rad curve does not
correlate to anything we have seen. You cannot only consider the peak power numbers to establish losses, at which RPM they are obtained is also important. By your logic, the F8X would have 25% drivetrain loss at 7300rpm since the Maha dyno WHP is about 325ps at that RPM. Think about it a little...

Last edited by CanAutM3; 08-19-2014 at 01:25 AM..
Appreciate 0
      08-17-2014, 12:08 PM   #91
Boss330
Major General
Boss330's Avatar
No_Country
1712
Rep
5,108
Posts

Drives: BMW
Join Date: Mar 2012
Location: Earth

iTrader: (0)

Quote:
Originally Posted by CanAutM3 View Post
Why do you say they are all over the place?

Further, where do you get that 13% number from ? To me it does not seem to have any foundation.

As I pointed to you many times before, the shape P-rad curve does correlate to anything we have seen. You cannot only consider the peak power number to establish losses, at which RPM they are obtained is also important. By your logic, the F8X would have 25% drivetrain loss at 7300rpm since the Maha dyno WHP is about 325ps at that RPM. Think about it a little...
I agree that the shape of the whp curve also doesn't add up... And agree that that would mean a 25% loss...

It's all over the place because the M4 was 34PS over the factory claimed number, the AM Vantage V12 was 34PS under the factory claimed number. And other dyno results in their Supertests have also shown a inconsistency/variation from stated factory numbers.

The MAHA chp numbers seems to be just as reliable as Dynojets and the other chassis dynos out there...
Appreciate 0
      08-17-2014, 01:37 PM   #92
solstice
Major General
5457
Rep
7,037
Posts

Drives: 2015 M3 6MT
Join Date: May 2013
Location: Seattle

iTrader: (0)

Quote:
Originally Posted by Boss330 View Post
I agree that the shape of the whp curve also doesn't add up... And agree that that would mean a 25% loss...

It's all over the place because the M4 was 34PS over the factory claimed number, the AM Vantage V12 was 34PS under the factory claimed number. And other dyno results in their Supertests have also shown a inconsistency/variation from stated factory numbers.

The MAHA chp numbers seems to be just as reliable as Dynojets and the other chassis dynos out there...
Why do you by default and conviction assume that the MAHA dyno is inconsistent? Why can't it just as well be the stated hp that is "inconsistent" vs a potentially true measured/calculated MAHA hp? In a way I wish I still shared your full and complete trust in the precision and perfection of authorities and regulators...but I don't.

Last edited by solstice; 08-17-2014 at 01:57 PM..
Appreciate 0
      08-17-2014, 02:31 PM   #93
Boss330
Major General
Boss330's Avatar
No_Country
1712
Rep
5,108
Posts

Drives: BMW
Join Date: Mar 2012
Location: Earth

iTrader: (0)

Quote:
Originally Posted by solstice View Post
Why do you by default and conviction assume that the MAHA dyno is inconsistent? Why can't it just as well be the stated hp that is "inconsistent" vs a potentially true measured/calculated MAHA hp? In a way I wish I still shared your full and complete trust in the precision and perfection of authorities and regulators...but I don't.
It's not about the "precision and perfection" of authorities and regulators. They don't make the cars... It's whether the manufacturers and independent technical services are complying with the legislation they are giving their legally binding assurance that they comply with!

And the factory stated HP is measured on a multi million engine dyno certified to do that. Seems like many by default and conviction assume those dyno tests are inconsistent. NO chassis dyno is certified to measure crank hp...

We can either choose to believe a chassis dyno that is not certified to measure crank hp and which we we have plenty of evidence has a variation of 10-15% between them. Or, we can choose to believe the independently measured hp numbers taken on a engine dyno at a technical service.

For me, the choice is really easy Especially when trap speeds/performance also match the factory stated hp number!

Even the OP (P1) of this thread admitted that the S55's average power of 425HP "quite possibly" could explain the trap speed of the F8x... Shouldn't that be a hint that there might be questions about the OP's simulations, when he hadn't given any thought to the difference in average power of the S55 vs S65? And him being honest enough to agree that this "quite possibly" could account for the difference?


I know you don't put much trust in regulators and the manufacturers ability or willingness to comply , but for information for the rest of us I will just post what Directive 2007/46/EC says about conformity:

http://eur-lex.europa.eu/legal-conte...140101&from=EN" rel="" target="_blank">http://<a href="http://eur-lex.europ...01&from=EN</a>

Quote:
In order to ensure that the procedure for monitoring conformity
of production, which is one of the cornerstones of the
Community type-approval system
, has been correctly implemented
and functions properly, manufacturers should be
regularly checked by the competent authority or by an appropriately
qualified technical service appointed for that purpose
.
Quote:
1. The manufacturer is responsible to the approval authority for all aspects of the approval process and for ensuring conformity of production, whether or not the manufacturer is directly involved in all stages of the construction of a vehicle, system, component or separate technical unit.
Quote:
1. The Member State which grants an EC type-approval shall take the necessary measures in accordance with Annex X to verify, if need be in cooperation with the approval authorities of the other Member States, that adequate arrangements have been made to ensure that production vehicles, systems, components or separate technical units, as the case may be, conform to the approved type.

2. The Member State which has granted an EC type-approval shall take the necessary measures in accordance with Annex X in relation to that approval to verify, if need be in cooperation with the approval authorities of the other Member States, that the arrangements referred to in paragraph 1 continue to be adequate and that production vehicles, systems, components or separate technical units, as the case may be, continue to conform to the approved type. Verification to ensure that products conform to the approved type shall be limited to the procedures set out in Annex X and in those regulatory acts that contain specific requirements. To that end, the approval authority of the Member State which has granted the EC type- approval may carry out any of the checks or tests prescribed in any of the regulatory acts listed in Annex IV or Annex XI on samples taken in the premises of the manufacturer, including production facilities.

3. When a Member State which has granted an EC type-approval establishes that the arrangements referred to in paragraph 1 are not being applied, deviate significantly from the arrangements and control plans agreed, or have ceased to be applied, although production is not discontinued, that Member State shall take the necessary measures, including the withdrawal of the type-approval, to ensure that the conformity of production procedure is followed correctly.

I might be naive, but I find it very unlikely that a European manufacturer deliberately and knowingly deceives and misleads EU type approval authorities. There is a regime to verify conformity, including spot checks at random. So the risk of being caught cheating definitely is present! And with all the forum activity about under rating, the good people responsible for those spot checks might find it relevant to do a spot check on engine power... (Lot's of those working in the auto industry, on both sides of the "fence", are car enthusiasts reading forums like these (posting as well perhaps )...). That is something BMW obviously will need to have considered as a possibility if they have indeed under rated their engines to the extent it's claimed here.

And, no I also don't believe that there is a mysterious factor inside, or outside, the dyno room explaining a "legal" difference of 60-70hp. It would also be strange that that factor only should be present on the S55... If anything, the dyno room is a ideal environment and should not yield lower power ratings.

Let me also add that this isn't proof of no underrating, merely evidence based on a regime of testing, compliance and spot checks to verify that compliance.

And, another (unrelated to this thread) interesting legal requirement:

Quote:
It is also important for equipment manufacturers to have access to
certain information that is available only from the vehicle manufacturer,
that is to say, the technical information, including
drawings, required for the development of parts for the aftermarket.

Last edited by Boss330; 08-17-2014 at 03:53 PM..
Appreciate 0
      08-17-2014, 05:37 PM   #94
solstice
Major General
5457
Rep
7,037
Posts

Drives: 2015 M3 6MT
Join Date: May 2013
Location: Seattle

iTrader: (0)

Looking at some MAHA dynos for some other BMW FI engines they don't seem to vary wildly vs rated hp instead they are pretty much within the regulation allowance:
N63: 393.7 vs 400 hp rated.
http://www.benz**********/content.ph...d&snpAttempt=1

S63TU: 573.7 vs 560 hp rated
http://www.aston**********/content.p...2&snpAttempt=1

N55: 307.1 vs 300 hp rated
http://www.***********.com/content.p...atted-downpipe

The S55 seem to be the exception by generating far more power vs stated on the MAHA than other BMW FI engines at 465 vs 431 stated.

Edit: seems like copy of the links doesn't work. Just google the engine and maha dyno and check the results.

And is an underperforming British car really a surprise
Furthermore the AM's engine is hand-built. Precise conformity must be even more tricky between species.

Last edited by solstice; 08-17-2014 at 06:04 PM..
Appreciate 0
      08-17-2014, 09:46 PM   #95
P1 Motorcars
Private First Class
26
Rep
181
Posts

Drives: Performance Shop
Join Date: Jun 2014
Location: Stamford, CT

iTrader: (0)

Quote:
Originally Posted by swamp2
Rebuttal...
I appreciate the thorough and detailed response. I noticed a few things below I wanted to address. I wanted to say right up front that some of the educational corrections seem wrong themselves. So I wanted to add our knowledge to the topic and hopefully help bring more understanding to this thread.


Quote:
Originally Posted by swamp2
And P1 Motorcars does tuning and have a Dynojet and thus have an inherent conflict of interest to "prove" that dynos and in particular Dynojets are accurate. Your entire argument here begins with this proposition that chassis dynos are accurate and that is a big problem. More on that in a bit.
Our posts tell of our experiences and expertise and what we brought to the table to this topic. Nothing more, nothing less.

Quote:
vBox analysis...
Correction: vBox hardware and vBox software (by RaceLogic) doesn't calculate NHRA trap speeds. vboxtools.com software used in our article has full NHRA settings, 1-ft rollout, and trap speed calculations. That's why both vMax and trap speeds are called out separately in our article.

Quote:
Maha dyno's and crank horsepower...
Correction: We have a partner shop that has a Maha. I saw a few of the Maha comments in your thread and noted the people didn't understand what they measure, what they report, and how they work. I didn't read any further at that point. All of our prior analysis and success using CarTest to match real-world results has always been with Dynojet results. We added the Maha comparison as an afterthought. Even though we stand by those results and note they seem to be just as accurate as our Dynojet results, we'd prefer to ignore them and instead concentrate on the Dynojet results.

Quote:
Dynojets...
Correction: We know the differences between most different dyno brands. Some use F=M*A, others use strain gauges, others use eddy current + strain gauges, you get the idea. The different measuring techniques report different results. In our prior works, we used Dynojets and all simulations matched the real-world.

Correction: The 13% power changes you're talking about on the S65 may be more understandable the more you study their sources. I know you've been around long enough to see this analysis before. 91 Octane vs. 93 Octane and 6MT vs. DCT both account for some of those differences. Then sometime around late 2010 or early 2011, a software update was introduced that bumped base power significantly. By memory, I think it was about 10 whp bump. I've seen multiple independent vendors report the same increase in power at about the same time.

Quote:
Cherry picking data...
Correction: We cut-paste those all from your thread (also without attribution...sorry). We threw out the items you called outliers and non-DCT entries. The car in this test is an M3, not M4. If there are any missing items or cherry-picks, as of today, I still don't see them added to your own list.

Quote:
Our input data...
  • Good idea on the tire size calculations.
  • We changed tire pressure because we measured it.
  • We used 175 pound driver because that's what we were told it was.
  • DCT shift times. I will confirm that we use 50ms.

Quote:
SAE Corrections...
Correction: There are white papers on the accuracy of SAE corrections. The paper's I've read showed them to be much more accurate than most people think. It's ironic reading people arguing against the accuracy of the correction, but also arguing the EEC regulations are enforceable law (which depend on the accuracy of these calculations).

Correction: Yes we've read the SAE documents. Without this knowledge, we wouldn't have been able to comment about the lacking humidity input in the Insoric section of our article. BTW, you forgot to mention the SAE correction also accounts for mechanical efficiency.

Quote:
Wind...
Agreed, it's a micro climate. But we always add it when we know it.

Quote:
Drive train losses...
Correction: CarTest doesn't use them on wheel power simulations. Since we're using Dynojet wheel power simulations, we can skip this discussion as it doesn't affect any of our work or any of our results.

Quote:
Insoric...
Correction:
  • It sounds like they shared slightly more information with us about their hardware than they did with you. A vBox would still be more accurate.
  • We pointed out the flaws in their various techniques by analyzing what they allow for input parameters. This isn't speculation, this is very obvious.
  • Regarding Crr calculations. Have you tried changing tire pressure in CarTest and see how it changes the results? BTW, we used this formula for Crr: http://www.engineeringtoolbox.com/ro...ce-d_1303.html

Quote:
Conclusions...
I didn't see anything blaring that we supposedly did wrong. The most I got out of it was you don't play with wind and tire pressure, but we do. Our method is simple, yet time consuming. We use the actual results for each car and specific run. Our simulations always seem to match the real-world within a much tighter margin of error than 0.7 MPH using your approach. Yes it's philosophical approach and there’s a benefit to both. We didn't want to answer a theoretical question with theoretical data and draw iron-clad conclusions. We'd prefer to help draw conclusions based on actual data, and hopefully that's what we brought to the table.

One thing about the Dynojet models in CarTest. We can't speak for other dyno brands and we don't make any claims about them. All we're trying to say is that we have a track record of simulations that match real world data, and that track record is based on using Dynojet SAE corrected results. If it's not broken, don't fix it.

1. Chassis dynos, explained, we use Dynojet for CarTest simulations.
2. In-situ, wheel based, Insoric dyno, flawed,
3. Maha whp results, completely misunderstood
4. Trap speeds (and other performance metrics), consistent with simulations
5. BMW stated crank hp, BFD
6. Simulation, matches real-world results
7. Legal requirements, very interesting. SAE isn't a legal requirement, it's a standards body. Makes me wonder if EEC is the same thing. Now that I've read 3 or 4 of those documents, they don't appear to have any teeth in them whatsoever. Like the SAE, they sound like a standards body that if you want to put their stamp on something, you need to comply with their procedures. The later documents posted in this thread all but say it's voluntary. There are ample quotes by Boss330 from the EEC documents themselves that indicate that submission is purely voluntary; but once you submit, you are bound to the agreement. Again, sounds like it has absolutely no teeth in it whatsoever. But I'm also going to hedge my bets by saying I'm not an expert in anything regarding the EU.
Appreciate 0
      08-18-2014, 12:11 AM   #96
Boss330
Major General
Boss330's Avatar
No_Country
1712
Rep
5,108
Posts

Drives: BMW
Join Date: Mar 2012
Location: Earth

iTrader: (0)

[QUOTE=P1 Motorcars]
Quote:
Originally Posted by swamp2
1. Chassis dynos, explained, we use Dynojet for CarTest simulations.
2. In-situ, wheel based, Insoric dyno, flawed,
3. Maha whp results, completely misunderstood
4. Trap speeds (and other performance metrics), consistent with simulations
5. BMW stated crank hp, BFD
6. Simulation, matches real-world results
7. Legal requirements, very interesting. SAE isn't a legal requirement, it's a standards body. Makes me wonder if EEC is the same thing. Now that I've read 3 or 4 of those documents, they don't appear to have any teeth in them whatsoever. Like the SAE, they sound like a standards body that if you want to put their stamp on something, you need to comply with their procedures. The later documents posted in this thread all but say it's voluntary. There are ample quotes by Boss330 from the EEC documents themselves that indicate that submission is purely voluntary; but once you submit, you are bound to the agreement. Again, sounds like it has absolutely no teeth in it whatsoever. But I'm also going to hedge my bets by saying I'm not an expert in anything regarding the EU.
I can categorically say that you are 100% wrong in your assumption that EU type approval standards are voluntary. Those are the legal requirements you need to adhere to if you want to market and sell cars in the EU!

You can of course say that it's voluntary to sell cars in the EU market... But as a mass manufacturer (or even low volume, but with less rigid regs) you HAVE to get a EU type approval to be allowed to sell your vehicles throughout the EU market.

I know, I work with this every day of the week...

Article 4, #3 of Directive 2007/46/EC (the framework Directive), clearly states that this isn't voluntary. The notion that it hasn't got any "teeth in them whatsoever" is clearly wrong:

Quote:
3. Member States shall register or permit the sale or entry into service only of such vehicles, components and separate technical units as satisfy the requirements of this Directive.
And, this really doesn't matter anyway. The F8x HAS EU-type approval and thereby has to adhere to the various technical Directives required for type approval.

The EU type approval information document for the M3/4 lists that engine power is in accordance with:

Quote:
40 MOTORLEISTUNG
ENGINE POWER
e24*80/1269*1999/99*0111*00 12.12.2013 alle / all


1.4. Verifications according to Annex V, no. 1.(d), 2007/46/EC
with VO (EU) 65/2012
:
The test vehicles comply with the relevant data in the information folders of the approvals of the separate directives named in part III of the information document. The selection of vehicles allowed a proper assessment of the different combinations to be approved, with regard to the criteria according Annex V No.2 2007/46/EC.

1.8. Verifications according to Annex V, no. 2., 2007/46/EC
with VO (EU) 65/2012
:
Inspection have been carried out to ensure the proper control of the various combinations of engines, gearboxes, powered axles, steered axles, body styles, number of doors, hand of drive, number of seats and level of equipment to be approved.

US legislation differs from EU in that it relies on self certification from the manufacturer.

From Wikipedia:

Quote:
Self-certification

Rather than a UN-style system of type approvals, the US and Canadian auto safety regulations operate on the principle of self-certification, wherein the manufacturer or importer of a vehicle or item of motor vehicle equipment certifies — i.e., asserts and promises — that the vehicle or equipment complies with all applicable Federal or Canada Motor Vehicle Safety, bumper and antitheft standards. No prior verification is required by a governmental agency or authorised testing entity before the vehicle or equipment can be imported, sold, or used. If reason develops to believe the certification was false or improper — i.e., that the vehicle or equipment does not in fact comply — then authorities may conduct tests and, if a noncompliance is found, order a recall and/or other corrective and/or punitive measures. Vehicle and equipment makers are permitted to appeal such penalties by filing petitions for finding of noncompliance inconsequential to safety.


Please also explain what we all got wrong in our understanding of the MAHA whp (this genuinely is something I'd like to know more about, and since you claimed we have gotten it wrong I also assume you know the correct answer here).

The INSORIC at least seems to get very consistent and good results, flawed as you might claim it is... Engine power within the EU allowed tolerances.

Last edited by Boss330; 08-18-2014 at 06:05 AM..
Appreciate 0
      08-18-2014, 04:58 AM   #97
Boss330
Major General
Boss330's Avatar
No_Country
1712
Rep
5,108
Posts

Drives: BMW
Join Date: Mar 2012
Location: Earth

iTrader: (0)

Quote:
Originally Posted by solstice View Post
Looking at some MAHA dynos for some other BMW FI engines they don't seem to vary wildly vs rated hp instead they are pretty much within the regulation allowance:
N63: 393.7 vs 400 hp rated.
http://www.benz**********/content.ph...d&snpAttempt=1

S63TU: 573.7 vs 560 hp rated
http://www.aston**********/content.p...2&snpAttempt=1

N55: 307.1 vs 300 hp rated
http://www.***********.com/content.p...atted-downpipe

The S55 seem to be the exception by generating far more power vs stated on the MAHA than other BMW FI engines at 465 vs 431 stated.

Edit: seems like copy of the links doesn't work. Just google the engine and maha dyno and check the results.

And is an underperforming British car really a surprise
Furthermore the AM's engine is hand-built. Precise conformity must be even more tricky between species.
Good find on MAHA dyno results Agree that those results seems more accurate.

As regards hand built engines. Please think over what you said one more time... In engine building there is something called "blueprinting". That is when a engine is hand built to be as close as possible to the specs and tolerances in the blueprints of the engine's design. Hand built engines are hand built in order to be more close to spec and has tighter tolerances. Bespoke racing engines are hand built, both because of cost of tooling but because with mass production there is always bound to be variations in the different components. Meaning that there needs to be larger tolerances in order for all the components to fit together (like if you have a crank journal diameter that is on the large end of the manufacturing tolerance and a bearing liner in the block on the small side of tolerances). In a hand built engine, dimensions are usually hand measured and the proper thickness bearing is chosen so that the tolerance is much closer to the optimal designed tolerance.
Appreciate 0
      08-18-2014, 10:34 AM   #98
solstice
Major General
5457
Rep
7,037
Posts

Drives: 2015 M3 6MT
Join Date: May 2013
Location: Seattle

iTrader: (0)

Quote:
Originally Posted by Boss330 View Post
Good find on MAHA dyno results Agree that those results seems more accurate.

As regards hand built engines. Please think over what you said one more time... In engine building there is something called "blueprinting". That is when a engine is hand built to be as close as possible to the specs and tolerances in the blueprints of the engine's design. Hand built engines are hand built in order to be more close to spec and has tighter tolerances. Bespoke racing engines are hand built, both because of cost of tooling but because with mass production there is always bound to be variations in the different components. Meaning that there needs to be larger tolerances in order for all the components to fit together (like if you have a crank journal diameter that is on the large end of the manufacturing tolerance and a bearing liner in the block on the small side of tolerances). In a hand built engine, dimensions are usually hand measured and the proper thickness bearing is chosen so that the tolerance is much closer to the optimal designed tolerance.
I guess it depends to some degree at what extent the hands are involved but I'm pretty sure that if you let a modern high quality industrial robot machine 100 engines they will be pretty much identical while if you let 10 different people hand build 100 engines these engines will have more variations to each other.

Last edited by solstice; 08-18-2014 at 10:49 AM..
Appreciate 0
      08-18-2014, 10:47 AM   #99
JoeFromPA
Colonel
1791
Rep
2,995
Posts

Drives: '15 AW M3 6MT Stripper
Join Date: Aug 2013
Location: SE PA

iTrader: (0)

Quote:
Originally Posted by solstice View Post
I guess it depends to some degree at what extent the hands are involved but I'm pretty sure that if you let a modern high quality industrial robot machine 100 engines they will be pretty much identical while if you let 10 different people hand build 100 engines these engines will have more variations.
But that would mean thousands of internet posters with "strong" factory engines were WRONG. And it would mean that if you dyno two vehicles with the same drivetrains on the same day/dyno and running the same fuel, they should be pretty much identical.
__________________
AW/Carbonstructure 6MT 2015 M3 picked up 8/22/2014. Stripper except for adaptive suspension. Weighed at 3,450 pounds with 1/4 fuel. 70,000 miles as of February 2020.
Appreciate 0
      08-18-2014, 10:54 AM   #100
solstice
Major General
5457
Rep
7,037
Posts

Drives: 2015 M3 6MT
Join Date: May 2013
Location: Seattle

iTrader: (0)

Quote:
Originally Posted by JoeFromPA View Post
But that would mean thousands of internet posters with "strong" factory engines were WRONG. And it would mean that if you dyno two vehicles with the same drivetrains on the same day/dyno and running the same fuel, they should be pretty much identical.
Well there's the factor of use and abuse as well prior to getting each car on a dyno. I also think there are a fair amount of human touches on each complete car. Then there is material and environmental differences during production. But between hand built and robot built entities I would say the chance of conformity is stacked in favour of the robots with a large amount. Last but not least are SW updates and configuration tuning of ECU,DME and robots.

Last edited by solstice; 08-18-2014 at 11:00 AM..
Appreciate 0
      08-18-2014, 11:32 AM   #101
P1 Motorcars
Private First Class
26
Rep
181
Posts

Drives: Performance Shop
Join Date: Jun 2014
Location: Stamford, CT

iTrader: (0)

Quote:
Originally Posted by Boss330
We can either choose to believe a chassis dyno that is not certified to measure crank hp and which we we have plenty of evidence has a variation of 10-15% between them. Or, we can choose to believe the independently measured hp numbers taken on a engine dyno at a technical service.
See my previous response to swamp2. Much of the 13% variance in the Dyno Database can be explained. Of course there are some outliers. But what do you do with outliers? You throw them away. After you throw them away, you should see groups of 6MT vs. DCT and 91 Octane vs. 93 Octane. You should also see an additional spike across the board after sometime late 2010 or early 2011 but I'm not sure if the database can show this or not.

Quote:
Even the OP (P1) of this thread admitted that the S55's average power of 425HP "quite possibly" could explain the trap speed of the F8x... Shouldn't that be a hint that there might be questions about the OP's simulations, when he hadn't given any thought to the difference in average power of the S55 vs S65? And him being honest enough to agree that this "quite possibly" could account for the difference?
Our simulations models use actual dyno results placed into the simulation model. Our simulation models aren't based on any assumptions of average or peak power. Therefore we wouldn't need to think about this difference because the answer wouldn't change our simulation model. In the end, our simulations matched real-world results.

Quote:
I know you don't put much trust in regulators and the manufacturers ability or willingness to comply , but for information for the rest of us I will just post what Directive 2007/46/EC says about conformity: ...
I know what you're saying and I saw your responses to my other comments. Here's what got my curiosity. It was comments like these:
  • "1. The Member State which grants an EC type-approval shall take the necessary measures in accordance with Annex X to verify..."
  • "When a Member State which has granted an EC type-approval ... "

Is there such a thing as a member state which doesn't grant EC type approval? What happens when a vehicle fails to gain certification because it doesn't pass section 8.1 during testing? Does that mean that car can't be sold in the EU? I still have my doubts.

Quote:
Originally Posted by Boss330
Quote:
Originally Posted by P1 Motorcars
Here's what the compliance section says:
8. TOLERANCES FOR MEASURING THE NET POWER
8.1. The net power indicated by the manufacturer for the type of engine shall be accepted if it does not differ by more than ± 2 % for maximum power and more than ± 4 % at, the other measurement points on the curve with a tolerance of± 1,5 % for engine speed, from the values measured by the technical service on the engine submitted for testing.
8.2. During the tests to verify conformity of production the power shall be measured at two engine speeds S1 and S2 corresponding respectively to the measurement points ofmaximum power and maximum torque accepted for type approval. At these two engine speeds, which are subject to a tolerance of± 5 %, the net power measured at at least one point within the ranges S1 ± 5 % and S2 ± 5 % shall not differ by more than ± 5 % from the approval figure.
Section 8.2 is the compliance section that I'm pretty sure you and the other guys are talking about. But look very closely at what it says. It says power will be measured at two points (maxTQ and maxHP), and it must be +/- 5% at ONE of those points...not BOTH! That means BMW could comply with this simply be getting maxTQ within +/- 5% and still jack up the HP to whatever they want.
Agree. Section 8.1 and 8.2 clearly says that both HP and TQ is to be measured and checked for conformity of production.
When I made that post, I almost highlighted the section now appearing in RED. But I thought it was obvious enough, that I didn't think I would have to. When I read this, here's what it meant to me. Section 8.1 discusses the variance allowed for the specific engine submitted for test. Section 8.2 defines the variance allowed to ensure conformity of production. Yes both are required (who thought I said they weren't?), but 8.1 is the specific engine submitted for test, and 8.2 is to ensure conformity of production.

Once again, I'm going to mention the irony of others arguing against SAE/EEC weather correction formulas while at the same time arguing for the inviolate law of EEC regulations whose enforcement depends on the accuracy of these same weather correction formulas.

Quote:
Originally Posted by Boss330
Please also explain what we all got wrong in our understanding of the MAHA whp (this genuinely is something I'd like to know more about, and since you claimed we have gotten it wrong I also assume you know the correct answer here).
I don't remember if they were your comments or not. I wasn't singling out any individual just making a general comment and admitting I only saw two posts. If it all got corrected in the end then it all worked out.
Appreciate 0
      08-18-2014, 02:19 PM   #102
Boss330
Major General
Boss330's Avatar
No_Country
1712
Rep
5,108
Posts

Drives: BMW
Join Date: Mar 2012
Location: Earth

iTrader: (0)

Quote:
Originally Posted by P1 Motorcars View Post
I know what you're saying and I saw your responses to my other comments. Here's what got my curiosity. It was comments like these:
  • "1. The Member State which grants an EC type-approval shall take the necessary measures in accordance with Annex X to verify..."
  • "When a Member State which has granted an EC type-approval ... "

Is there such a thing as a member state which doesn't grant EC type approval? What happens when a vehicle fails to gain certification because it doesn't pass section 8.1 during testing? Does that mean that car can't be sold in the EU? I still have my doubts.



When I made that post, I almost highlighted the section now appearing in RED. But I thought it was obvious enough, that I didn't think I would have to. When I read this, here's what it meant to me. Section 8.1 discusses the variance allowed for the specific engine submitted for test. Section 8.2 defines the variance allowed to ensure conformity of production. Yes both are required (who thought I said they weren't?), but 8.1 is the specific engine submitted for test, and 8.2 is to ensure conformity of production.

Once again, I'm going to mention the irony of others arguing against SAE/EEC weather correction formulas while at the same time arguing for the inviolate law of EEC regulations whose enforcement depends on the accuracy of these same weather correction formulas.



I don't remember if they were your comments or not. I wasn't singling out any individual just making a general comment and admitting I only saw two posts. If it all got corrected in the end then it all worked out.

Many member states don't issue type approvals. Germany, Italy, France, England are the main countries issuing type approvals. That's because that's where manufacturers mostly are placed and where there are technical services (test laboratories) placed to do this. If a car model has received type approval, that means it can be sold and registered in any EU country (just like a US car that has the "This vehicle conforms with all applicable FMVSS standards in place at the date of manufacture" can be sold in every US state. A car can't have more than one EU type approval. If it has a type approval number starting with e1, that means it was issued in Germany.

I see you have your doubts as regards type approval in EU. But, yes I can assure you that a mass produced passenger vehicle in category M1 (a car with a maximum of 9 seats in total) that fails to meet the requirements of the technical directives, can not be registered in member states. There are certain allowances for small volume type approval, one off vehicles and amateur built vehicles. A one off vehicle in category M1 can instead undergo a IVA procedure (Individual Vehicle Approval), but this means that EACH individual car has to be documented according to each of the technical Directives in Annex IV of 2007/46/EC. And that includes engine power as well...

I will once again quote what Directive 2007/46/EC says about the need to comply with the required standards:

Quote:
3. Member States shall register or permit the sale or entry into service only of such vehicles, components and separate technical units as satisfy the requirements of this Directive.
EU type approval is the cornerstone of vehicle registration within EU and was introduced as a way of harmonization and free trade of motor vehicles between member states.

As I said previously, type approvals and EU legislation is something I work with on a daily basis. Getting the M3/4 type approved for sale in EU isn't something BMW can opt out of!


As regards the engine power testing, I'm not sure where the confusion is?

The manufacturer has to submit a engine for testing. This has to be measured according to the standards and the entire power curve has to be verified. If the max power is within 2% of the stated hp and if the remaining test points are within 4% of what the manufacturer states, then the engine's factory claimed power can be accepted.

This is 8.1 and concerns the engine submitted for testing. That has to be within 2% of factory claimed max hp.

And then we have 8.2 which concerns conformity of production. This means that when the authority, or technical service, does spot checks during production (which could be today, next month or in two years), the engine they pick for conformity of production (a random engine) has to be within 5% of the approval figure.

For type approval testing, the requirement is a tolerance of max 2% on max power. For the subsequent testing for conformity of production the tolerance is higher at 5%.

Which is why we repeatedly have said that if the S55 indeed is under rated, then that most likely won't be more than by the EU tolerance of 5%.

Since this both IS a legal requirement that BMW needs to adhere to, AND a requirement BMW has adhered to in acquiring EU type approvals for the M3/4, this should lead to the following conclusions:

1. BMW presented a engine (with DME, intake and exhaust systems as per production spec) for type approval testing that was within 2% of the claimed HP

2. The engine submitted for testing was either:
-Completely in conformity with requirements and with the engines subsequently used in production vehicles
-Not in conformity with the engines later used in production vehicles

If BMW have chosen the latter solution, then they are playing a dangerous game as they risk getting caught cheating as soon as conformity of production spot checks take place... To me this seems unlikely and such a policy would result in hugely bad PR for BMW in EU if they are caught cheating (heads would roll at BMW M HQ)!

And I would say that any claim of under rating in the magnitude that is claimed (60-70hp) certainly also HAS to imply wrongdoing by BMW with regards to type approval testing. Unless we are to believe that the engine dyno test is so flawed that it misses the engine's performance in the car by 60-70hp (or 14-16% in this case)... I don't believe that there is anything "magical" happening when the engine is mounted in the car that suddenly increases power by 60-70hp over what it had on the engine dyno! That means that any claim of under rating also needs to consider the likelyhood of BMW knowingly deceiving the governing bodies of all EU countries these cars are marketed in. Yes, that might be the case. Maybe BMW couldn't care less about possible consequences and loosing type approval for the F8x (and thereby be unable to sell the cars in the EU).


And, it doesn't matter who made the comments about MAHA whp. You said that this was "completely misunderstood". I would REALLY like to have your explanation on what was completely misunderstood and what the correct understanding is.

The MAHA dyno is something we genuinely have debated here and are curious to learn more about! Since you claimed that we had completely misunderstood the whp from a MAHA, I'm sure you can explain where we went wrong (at least in the two posts you read)!

Last edited by Boss330; 08-19-2014 at 05:16 AM..
Appreciate 0
      08-18-2014, 05:18 PM   #103
Boss330
Major General
Boss330's Avatar
No_Country
1712
Rep
5,108
Posts

Drives: BMW
Join Date: Mar 2012
Location: Earth

iTrader: (0)

Quote:
Originally Posted by solstice
Quote:
Originally Posted by Boss330 View Post
Good find on MAHA dyno results Agree that those results seems more accurate.

As regards hand built engines. Please think over what you said one more time... In engine building there is something called "blueprinting". That is when a engine is hand built to be as close as possible to the specs and tolerances in the blueprints of the engine's design. Hand built engines are hand built in order to be more close to spec and has tighter tolerances. Bespoke racing engines are hand built, both because of cost of tooling but because with mass production there is always bound to be variations in the different components. Meaning that there needs to be larger tolerances in order for all the components to fit together (like if you have a crank journal diameter that is on the large end of the manufacturing tolerance and a bearing liner in the block on the small side of tolerances). In a hand built engine, dimensions are usually hand measured and the proper thickness bearing is chosen so that the tolerance is much closer to the optimal designed tolerance.
I guess it depends to some degree at what extent the hands are involved but I'm pretty sure that if you let a modern high quality industrial robot machine 100 engines they will be pretty much identical while if you let 10 different people hand build 100 engines these engines will have more variations to each other.
Why are F1 engines hand built then?

By hand built we don't talk about a person making the components from scratch. The components are manufactured by CNC operated machines. But with any such machining there are variations in manufacturing. As tools are worn, there becomes differences (small but present) as well as numerous other factors.

On a "robot" assembled engine all engines are put together using the same std components. Meaning that there needs to be a certain amount of tolerances present in order to allow for components on the low or high end of tolerances.

On a hand built engine, all the components aren't only assembled by hand, but measured individually and the assembler makes sure all the tolerances are within a much smaller margin than mass production can allow for.

The difference is becoming smaller, but a hand built (assembled) engine is hand assembled in order to have lower tolerances and being closer to the optimal tolerances for each component.

I've been to several engine builders, including F1 engine supplier Cosworth and witnessed this meticulous process of hand built engine. BTW, a little known fact is that Cosworth hand assembled some of the special versions of the AM V12 engine (I was there when some of that took place). The regular V12's are assembled by Mahle in Germany...

Oh, and the AM V12 is basically two Ford V6 engines put together... Developed when AM was owned by Ford and based on the then current Ford 2.5 and 3l 60deg V6 engines.
Appreciate 0
      08-18-2014, 05:34 PM   #104
solstice
Major General
5457
Rep
7,037
Posts

Drives: 2015 M3 6MT
Join Date: May 2013
Location: Seattle

iTrader: (0)

You don't build several thousand F1 engines. Further more these engines do not need to be exactly the same, they just need to be good and follow the F1 framework. You don't hand build when you manufacture large series and replication is not a main goal of hand building. Another reason for hand building is cost of setting up a fully automated production. It make sense to hand build at low volumes of exotic units that can carry high cost. Automation is likely preferable of many reasons, consistency is one of them but if you build few engines the investment might not make sense.

If you think hand built engines of a larger volume is more identical to each other than robot production, think again.

Last edited by solstice; 08-18-2014 at 05:41 PM..
Appreciate 0
      08-19-2014, 01:38 AM   #105
CanAutM3
General
CanAutM3's Avatar
Canada
21105
Rep
20,741
Posts

Drives: 2021 911 turbo
Join Date: Jan 2012
Location: Montreal

iTrader: (1)

Garage List
Quote:
Originally Posted by solstice View Post
You don't build several thousand F1 engines. Further more these engines do not need to be exactly the same, they just need to be good and follow the F1 framework. You don't hand build when you manufacture large series and replication is not a main goal of hand building. Another reason for hand building is cost of setting up a fully automated production. It make sense to hand build at low volumes of exotic units that can carry high cost. Automation is likely preferable of many reasons, consistency is one of them but if you build few engines the investment might not make sense.

If you think hand built engines of a larger volume is more identical to each other than robot production, think again.
Robot built engines will be more identical to each other, albeit with larger tolerances. Compromises are made to achieve high throughput and ease of manufacture for the robots.

Properly "hand built" engines usually include more steps during the assembly process to ensure optimal fits and gaps, which allows for tighter tolerances. They will be less identical because each engine is sort of customized which yields a more optimized engine.
Appreciate 0
      08-19-2014, 02:09 AM   #106
solstice
Major General
5457
Rep
7,037
Posts

Drives: 2015 M3 6MT
Join Date: May 2013
Location: Seattle

iTrader: (0)

Quote:
Originally Posted by CanAutM3 View Post
Robot built engines will be more identical to each other, albeit with larger tolerances. Compromises are made to achieve high throughput and ease of manufacture for the robots.

Properly "hand built" engines usually include more steps during the assembly process to ensure optimal fits and gaps, which allows for tighter tolerances. They will be less identical because each engine is sort of customized which yields a more optimized engine.
Agreed.
Appreciate 0
      08-19-2014, 03:48 AM   #107
Boss330
Major General
Boss330's Avatar
No_Country
1712
Rep
5,108
Posts

Drives: BMW
Join Date: Mar 2012
Location: Earth

iTrader: (0)

Quote:
Originally Posted by solstice View Post
You don't build several thousand F1 engines. Further more these engines do not need to be exactly the same, they just need to be good and follow the F1 framework. You don't hand build when you manufacture large series and replication is not a main goal of hand building. Another reason for hand building is cost of setting up a fully automated production. It make sense to hand build at low volumes of exotic units that can carry high cost. Automation is likely preferable of many reasons, consistency is one of them but if you build few engines the investment might not make sense.

If you think hand built engines of a larger volume is more identical to each other than robot production, think again.
I see you agree with what CanAutM3 wrote, which is the same as I have been saying all along...

I will also add that a F1 engine doesn't merely need to be good, they need to be assembled to a much higher level of precision than a production car engine. In 2006, the Cosworth V8 raced at 20.000rpm. And before there was a limitation on how many F1 engines you could use each year, the teams brough around 10 engines per weekend. Those engines where dismantled and reassembled after each GP weekend.

That means that back then each team went through as many as 100-140 engines each (not always a brand new engine, but one that was disassembled and reassembled for every race weekend).

There was 10-12 teams, which means that roughly 1000-1500 engines were built each year (by 3-6 engine suppliers).

But, F1 engines aren't the best example of higher volume hand built engines. They are however a prime example of hand assembly in order to make sure that each engine is as optimally put together as possible. That means you just can't grab a set of bearings and throw in as a replacement. You need to measure and find the optimally sized bearing for the combination of tolerances on the parts you assemble (as well as with the other components of the engine). It's kind of like a VERY high end watch. Hand assembly in order to make the most presice and accurate clockwork possible.

An article explaining the principles of engine blueprinting:

http://www.precisionenginetech.com/t...inting-part-1/

And here is a example of modern, high volume, hand assembly of engines with the AMG concept of "one man, one engine":

http://www.mercedes-benz.ca/content/...mg/engine.html

https://www.mercedes-amg.com//video/...erview_eng.mp4

https://www.mercedes-amg.com/engineering.php?lang=eng





Here is a typical mass production robotic assembly line. Would you rather have your engine assmbled by the workers in this video or in the AMG video

Appreciate 0
      08-19-2014, 06:20 AM   #108
RingMeister01
Croatian
RingMeister01's Avatar
No_Country
880
Rep
3,613
Posts

Drives: PORSCHE
Join Date: Dec 2012
Location: NYC

iTrader: (0)

Garage List
Very smart post P1 and without the BS
__________________
NARDO GREY '18 PANAMERA TURBO
Ducati V4 S Corse (track only)
Husqvarna FS450 (track only)
Looking for an SUV
Appreciate 0
      08-19-2014, 09:42 AM   #109
Boss330
Major General
Boss330's Avatar
No_Country
1712
Rep
5,108
Posts

Drives: BMW
Join Date: Mar 2012
Location: Earth

iTrader: (0)

Quote:
Originally Posted by P1 Motorcars View Post
Is there such a thing as a member state which doesn't grant EC type approval? What happens when a vehicle fails to gain certification because it doesn't pass section 8.1 during testing? Does that mean that car can't be sold in the EU? I still have my doubts.


Once again, I'm going to mention the irony of others arguing against SAE/EEC weather correction formulas while at the same time arguing for the inviolate law of EEC regulations whose enforcement depends on the accuracy of these same weather correction formulas.
As of july 2014, the engine power is to be measured according to ECE regulation 85 (UN). Only minor differences, but basiacally even the same wording as in the EU directive on most paragraphs.

http://www.unece.org/fileadmin/DAM/t...regs/r085e.pdf


This is how evaluation of results under conformity of production testing is described here:

Quote:
5. EVALUATION OF RESULTS
5.1. If the net power of the engine tested pursuant to paragraph 2 above
fulfils the requirement of paragraph 4 above, the production is
considered to conform to the type approval.

5.2. If the requirements of paragraph 4 above are not fulfilled two more
engines are tested in the· same way.

5.3. If the net power figure of the second and/or third engine of
paragraph 5.2 does not fulfil the requirements of paragraph 4 above, the
production shall be considered not to conform to the requirernents of
this Regulation and the provisions of paragraph 7.1 shall be put into
effect.
ECE Reg 85 has been amended 5 times, here is the last amendment:

http://www.unece.org/fileadmin/DAM/t...gs/r085a5e.pdf

Pretty detailed info on the engine to be tested...

And, in the EU regulation (136/2014) that now refers engine power to be according to ECE Reg 85 there is also a provision for engines like the S55:

http://eur-lex.europa.eu/legal-conte...4R0136&from=EN

Quote:
2.3 Power correction factors
By way of derogation from paragraph 5.1 of Annex V to UNECE Regulation No 85, when a turbo-charged engine is fitted with a system which allows compensating the ambient conditions temperature and altitude, at the request of the manufacturer, the correction factors α a or α d shall be set to the value of 1.
This basically confirms that such a engine should have a correction factor of 1 (if requested by the manufacturer). As we suspected, a SAE/ECE correction on this kind of engine is a "double dipping". The engine self corrects and shouldn't have a correction factor applied.

This is also now acknowledged in legislation on engine power and correction factors. So the regulators don't necessarily use correction factors anymore. Taking one more inaccuracy out of the equation...

Last edited by Boss330; 08-19-2014 at 10:09 AM..
Appreciate 0
      08-19-2014, 11:16 AM   #110
solstice
Major General
5457
Rep
7,037
Posts

Drives: 2015 M3 6MT
Join Date: May 2013
Location: Seattle

iTrader: (0)

Quote:
Originally Posted by Boss330 View Post
I see you agree with what CanAutM3 wrote, which is the same as I have been saying all along...
Not even close. CanAutM3 is saying that robot built engines have less variations between them than hand-built engines which was my original point with the AM Vantage engine vs the S55 and other BMW FI engines. You argued the opposite.

Tolerances is totally unrelated to variances between engines. You can have two totally different engines with 1000hp difference and they can have the same tolerances. This is impossible with two new near identical robot built engines independent on what tolerance they are built at. That hand-built engines can be built to higher tolerances does in no way mean that they are built to higher conformity as a series since they aren't. Robots are simply superior in repeating the same procedure with the same parameters time after time than any human.

I can very easily see how one hand-built Vantage V12 engine can be built with 34 hp difference to another ( while still being built to equal tolerances ). It's a bit harder to see with robot built S55s. If it happens it's more likely due to the parameters I mention in another response than production line variations.

It's becoming slightly OT but my main point is that the AM is not the best evidence to discredit the precision of the MAHA dyno's crank hp calculation. I would think the precision against other BMW FI engines like the S63TU, N63 and N55 are better indicators when evaluating the S55's result. If you don't agree let's leave it at that to not make this slight sidetrack dominate this thread and risk overshadow all good work done by the OP.

Last edited by solstice; 08-19-2014 at 12:06 PM..
Appreciate 0
Closed Thread

Bookmarks

Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off



All times are GMT -5. The time now is 12:08 PM.




f80post
Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2024, vBulletin Solutions Inc.
1Addicts.com, BIMMERPOST.com, E90Post.com, F30Post.com, M3Post.com, ZPost.com, 5Post.com, 6Post.com, 7Post.com, XBimmers.com logo and trademark are properties of BIMMERPOST