|
|
08-09-2009, 09:27 AM | #45 |
Major General
1291
Rep 7,389
Posts |
Great stuff mixja.
I think that 10% to 15% is too large a change. Of course, autocrosses are one and lost in .001 second, sometimes, so the first 20 mph/kmh is very important. Was this done at a track, or is this calculated? Also, did you happen to do this with the 6MT going from 3.85 to 4.10? Dave
__________________
|
Appreciate
0
|
08-09-2009, 02:41 PM | #47 |
Automotive Industry Insider
462
Rep 1,948
Posts |
Mixja's numbers make a lot more sense...
These statistical results more accurately mirror what I have seen by changing out the differentials in several of my cars. (BMW and non-BMW cars alike) The accelerating metrics from a standing start are closer to reality in my experience. These numbers show a clear separation between the various FD ratios which swamp2's numbers did not. I'm never wanted swamp2 to think I was just dismissing the calculated numbers the Car Test software provided. I was just skeptical of certain aspects, due to the whitewashing in some areas. (virtually no difference in key metrics) A shorter FD ratio change wll give you an advantage in torque multiplication. That is a fact, not my opinion. It's a mechanical torque multiplier. I could care less what the software says, if it's trying to tell me there is absolutely no difference in acceleration with one short geared FD while an even shorter FD has no problem? Please...that's just silly. That distinction was not clearly illustrated in swamp2's numbers, which is why I knew something was wrong in the assumption factors that you must type into the software algorithm. Assuming the same car, transmission, tires, surface, and conditions...the shorter geared car will always be at least a little faster out of the hole if you don't loose traction. If traction is maintained, then the car will be faster in 1st and 2nd gear than the longer geared car. (every time) I grew up in a family of mechanics (my uncle was a ASE certified mechanic and avid drag racer). My father would race his Mustang at the track when I was a kid. I've seen Mopars, Chevy's and Fords raced with every conceivable differential ratio you can possibly imagine. And I can't recall a single instance where the shorter gear swap made no difference in 60ft time or even an 1/8 mile track. (unless you smoked the tires at the starting line of course) If it hooked...then the shorter gearing would give you an advantage every time out. (without exception) Now the advantage may vary from car to car, but there will be an advantage. It won't be 0.00%... The software should have validated that fact, yet it did not. That is why I kept insisting the metrics were wrong. It makes no sense when applied to the real world performance of a 10% improvement in trq multiplication (3.45:1 ring and pinion ratio), that I have witnessed over the past 30 years. FYI: The Car Test software can not tell you something very important... How often you will lose traction (grip) due the the various surfaces you will encounter on the street during daily driving. (smooth and flat vs. bumpy and uneven) Wet roads in particular will be a little more problematic with slick road conditions and or worn tires.
__________________
The best is yet to come...
|
Appreciate
0
|
08-09-2009, 04:14 PM | #48 | |
Dog Listener
703
Rep 7,850
Posts |
Quote:
I have a couple of questions that I hope don't showcase my ignorance: Is the 5% increase in power HP, TQ, or both?I'm guessing the model used a constant increase of 5%. Adding some kind of power/rpm/performance in each gear is going to be pretty complicated to model. The FD change comparisons all assume a constant in terms of engine performance while the power increase adds in other variables (where the power increase "is"; what is meant by power) that may make that comparison a little bit more tricky. No worries there, I'm not trying to nitpick or find fault. I'm trying to make sure I apply the knowledge I take away from this in a reasonable manner by knowing as much as possible about the assumptions. There's no way I could build a model to account for different levels of power increase in RPM ranges I don't expect you to either! Then again, there may be no issue here to consider at all (all comparisons are valid; 5% constant is fine; power range variables are accounted for) and my questions may simply be naive. |
|
Appreciate
0
|
08-09-2009, 07:27 PM | #49 |
Captain
50
Rep 780
Posts |
The increase was 5% peak tq and 5% peak power. The software automatically creates its power/tq curve automatically based upon these figures and the curve that is generated looks reasonably close. So the gain is 5% across the entire curve...
|
Appreciate
0
|
08-09-2009, 07:53 PM | #50 |
Dog Listener
703
Rep 7,850
Posts |
Thanks again for posting the data! This is great stuff and I've got a clearer handle on where to focus performance mods in terms of bang for the buck!
|
Appreciate
0
|
08-09-2009, 08:15 PM | #51 | |
Lieutenant General
611
Rep 10,407
Posts |
Quote:
An item that really got a bee in Lemans bonnet was the identical times to a certain speed. Lemans: Do note these predictions offer a similar state of affairs - for instance 0-100 km/hr, 3.45 vs. 3.62, and a few others as well. I do expect some other differences as I have adjusted many of the default parameters, both with justifications and in some regards to adjust to observed figures. Specifically I have used a non standard clutch engage time and shift time (obvious for DCT). Along with that change you should adjust the primary integration variable the step time downward (I used .03 s with good results opposed to .05 as the default which can not even capture a short DCT shift!) Next most of the mechanical losses are slightly altered from the default values. Lastly, I did find an error as well in my inputs, despite copying and pasting to create these vehicles the 3.62 run I initially posted had 2% more parasitic losses than the other two vehicles. Here are my revision with these corrections. I trust you also used the optimize launch feature for each car? Again #1 is stock DCT, #2 is 3.45, #3 is 3.62. The last of the three images are the results not allowing a shift, which I have referred to before.
__________________
E92 M3 | Space Gray on Fox Red | M-DCT | CF Roof | RAC RG63 Wheels | Brembo 380mm BBK | | Vorsteiner Ti Exhaust | Matte Black Grilles/Side Gills/Rear Emblem/Mirrors | | Alekshop Back up Camera | GP Thunders | BMW Aluminum Pedals | Elite Angels | | XPEL Full Front Wrap | Hardwired V1 | Interior Xenon Light Kit | |
|
Appreciate
0
|
08-09-2009, 08:26 PM | #52 | |||||
Lieutenant General
611
Rep 10,407
Posts |
Quote:
Quote:
Quote:
Quote:
Quote:
__________________
E92 M3 | Space Gray on Fox Red | M-DCT | CF Roof | RAC RG63 Wheels | Brembo 380mm BBK | | Vorsteiner Ti Exhaust | Matte Black Grilles/Side Gills/Rear Emblem/Mirrors | | Alekshop Back up Camera | GP Thunders | BMW Aluminum Pedals | Elite Angels | | XPEL Full Front Wrap | Hardwired V1 | Interior Xenon Light Kit | |
|||||
Appreciate
0
|
08-09-2009, 10:20 PM | #53 |
Automotive Industry Insider
462
Rep 1,948
Posts |
Swamp2,
Listen, I really admire how you like your artificially derived number in a computer simulation, but these number will never bare out in real life. You are hanging your entire hopes on your ability and the software's ability to accurately scale the performance gain of a mechanical device that improves your low end torque performance. I'm impressed that these programs can simulate what should happen (in a perfect world), but your continued insistence that these numbers are somehow infallible are preposterous. Of course they can be wrong...a human being designed them based on a number of scientific theories that are highly subjective to say the least. Forgive me if I tend to side with the reality that I have seen on the track for the past 30 years. Call me crazy I guess? Your metrics are dead wrong in the acceleration rates between the 3 possible FD ratios. (especially the 3.15 vs. 3.45) There is no way in hell they will record exactly the same time 0-50, 0-60 mph. That will never happen in real life. There is also no way they run within .01 in the 1/4 mile. Seriously, do you actually believe that JUST because the program tells you this? Any good engineer worth his salt will always question a calculation that seems odd or implausible. I take it that you think none of the results look odd to you? Apparently not, because you seem to keep defending the results of your simulation without exception. Again, it's a software program based on dozens of assumptions that may or may not be 100% accurate. That means the results could be wrong...but you seem to resist the notion that this is even possible. (or so it seems) I get the distinct impression that you are so devoted to these numbers, because you lack the real world experience gained from seeing the results of these mechanical devices IN ACTION. (on the track) No software. No simulations. Just bolting these differentials into the car (with your own two hands) and running back-to-back tests in the real world. (and logging the results you get) That's the real measuring stick. And it removes all doubt. I've yet to see what your numbers suggest with my own two eyes (0.00 improvement with nearly 10% improvement in trq multiplication), so of course I'm not going to buy these numbers. Why would I? They make no sense in the real world. If that doesn't cause a light bulb to go off in your head...then I guess that's the real problem. If you were a rabid drag racer, and grew around gear heads like myself, there would be no doubt that some of these numbers are just wrong. If I were you, I'd be carefully dissecting the program itself to find out why the numbers are not showing any difference in some of these key accelerating metrics. And why the 1/4 mile times show a statistical dead heat. (which is also rediculous) Again, if you had any substantial track experience racing cars at the strip, you would understand this better. That alone would probably cause you to look more closely at how these numbers are being calculated. In all the FD ratio changes that I have done personally, and the dozens of other FD I've seen others make, I have a clear understanding of what works and what doesn't. That has a tangible VALUE to me in regards to real world performance. The stopwatch is the final arbiters in these matters. That's the kind of consistency that I can believe in. Those numbers can not be disputed. I really wish you would stop being so condescending with your replies as well. I guess we will have to agree to disagree. Cheers.
__________________
The best is yet to come...
|
Appreciate
0
|
08-10-2009, 01:26 AM | #54 | |
Captain
50
Rep 780
Posts |
Quote:
Here are the comparisons between MT 3.85 (Car 1) and MT 4.10 (Car 2)... |
|
Appreciate
0
|
08-10-2009, 02:14 AM | #55 | |
Captain
50
Rep 780
Posts |
Quote:
I do think it will change the dynamic of the car - in that it will feel quite different purely because it will get through the gears a lot quicker. The main reasons I'm staying away from this modification for now are: - 7 Speed DCT has close ratios anyway. The step from a 6 speed to a 7 speed transmission has much more bearing on performance than a change in FD. - Resale value - this is a "non-reversible" modification, unless you are willing to shell out some extra $$$ to keep your original FD ratio. Where I come from a non-reversible modification will dramatically affect resale value (bearing in mind in NZ an M3 costs around 2.5 x average income, so buyers tend to be older and more conservative). - The performance gains are negligible and very dependent on the scenario. Other performance upgrades generally only give you improvement and don't have any possible negative scenarios. Last edited by mixja; 08-10-2009 at 03:07 AM.. |
|
Appreciate
0
|
08-10-2009, 08:12 AM | #56 | |
Major General
1291
Rep 7,389
Posts |
Quote:
Thanks again. How much time do you insert for the shifts? Dave
__________________
|
|
Appreciate
0
|
08-10-2009, 08:14 AM | #57 | |
Major General
1291
Rep 7,389
Posts |
Quote:
Important point.
__________________
|
|
Appreciate
0
|
08-10-2009, 03:21 PM | #58 |
Captain
50
Rep 780
Posts |
|
Appreciate
0
|
08-10-2009, 04:03 PM | #59 |
Major General
1291
Rep 7,389
Posts |
swamp2's speed in gear data corresponds well to my on-the-street experience. I'll try to control the rolling start so that I'm at 4000 rpm in 2d gear and then give the signal. This gives me a big jump on almost anything (I've got the 4.10 in a 6MT) and I've got an 8600 redline and my 2d-3d shift is quicker than .3 seconds (that is about right for the 1-2 shift, even with my UUC SSK).
Also, many people often fail to realize that a tenth or two is a lot when you translate it to car lengths. Dave
__________________
|
Appreciate
0
|
08-18-2009, 06:38 AM | #60 |
Captain
50
Rep 780
Posts |
I found a cool feature in CarTest that allows you to graph "parameter sensitivity". Essentially it shows you what happens when you change a single parameter but leave all other parameters constant for a given performance metric.
For the graphs below, the performance metric is standing quarter mile time in seconds (shown on the Y axis). Here is graph for changing DCT Final Drive Ratio: Here is graph for changing MT Final Drive Ratio: So you can see in the DCT case, there is a small benefit of 0.07s going from 3.15 to 3.45, but a negligible benefit in going from 3.45 to 3.62. In the case of the MT, the software suggests that 3.85 is close to optimal for FD ratio and going to 4.10 gives no benefit at all in terms of standing quarter mile times... Out of interest here are the graphs for changing horsepower and weight on the DCT Coupe, which as you can see have much more impact on quarter mile times (and interestingly suggest you need almost 500HP or a reduction of ~600lbs in weight to get into the 11s)... |
Appreciate
0
|
08-18-2009, 09:02 AM | #61 |
Major General
1291
Rep 7,389
Posts |
Mixja, that is fantastic information.
It caused me to think more in-depth about my own particular situation. At 5400 feet above sea level, I start out losing around 65 rwhp to the guys at see level. Of course, I also lose torque, but I've forgotten then rather large number. When I went from a 3.85 FD to a 4.10, my car's sensitivity to that parameter was considerably greater than in the car's orginal design parameter at sea level. I was sitting here thinking, "This is a clear win, how can these guys not see it." Now I'm thinking, "Oh, it may not be near so clear at sea level." Thanks for your great contributions. Dave
__________________
|
Appreciate
0
|
08-18-2009, 02:17 PM | #62 | |
Captain
50
Rep 780
Posts |
Quote:
The effect of increasing to 4.10 is greater, which is inline with the thinking that FD ratio has more effect the less power you have... |
|
Appreciate
0
|
08-18-2009, 02:35 PM | #63 | |
Major General
1291
Rep 7,389
Posts |
Quote:
Dave
__________________
|
|
Appreciate
0
|
08-24-2009, 05:01 PM | #64 | ||||||||
Lieutenant General
611
Rep 10,407
Posts |
Sorry for the delay. I've been relaxing on vacation for a couple of weeks. I'm keen on continuing this debate.
Quote:
Quote:
Quote:
Quote:
The immense value of such simulations is exactly the ability to limit ONE SINGLE parameter to variation. You can not do this in the field with real test, it simply CAN NOT be done. Lastly you mention "any engineer...". Do you happen to have any formal background in science or engineering? Quote:
Quote:
Why do you continue to discount (ignore is probably a more accurate term) the effects of more time spent in gears with lower accelerations (higher numbered gears) when increasing FD ratio? The average in gear accleration in 2nd gear for the M3 is about half of what it is in 1st gear. THIS DIRECTLY counteracts the benefits of torque multiplication. Please do answer my question as to why you can not simply multiply and multiply using an infinite or simply really large FD ratio to get a correspondingly large torque multiplication and thus performance improvements? I have asked this key question over and over and you simply won't reply. Quote:
Quote:
I really wish you would start engaging in a real debate with some real numbers or data rather then pounding your fists on the table over and over that the M3 will behave exactly as prior vehicles or 30 year old hot rods. Each case and each performance metric are positively unique and your refusal to admit this is really quite silly.
__________________
E92 M3 | Space Gray on Fox Red | M-DCT | CF Roof | RAC RG63 Wheels | Brembo 380mm BBK | | Vorsteiner Ti Exhaust | Matte Black Grilles/Side Gills/Rear Emblem/Mirrors | | Alekshop Back up Camera | GP Thunders | BMW Aluminum Pedals | Elite Angels | | XPEL Full Front Wrap | Hardwired V1 | Interior Xenon Light Kit | |
||||||||
Appreciate
0
|
08-24-2009, 05:05 PM | #65 |
Lieutenant General
611
Rep 10,407
Posts |
Thanks for the feedback. Of course getting shift points nailed is only an algebraic type of math problem in CarTest nailing the time to speed numbers will be subject to slightly more discrepany with test as this requires the discretization and numerical integration and uses many more of the input parameters for each car.
__________________
E92 M3 | Space Gray on Fox Red | M-DCT | CF Roof | RAC RG63 Wheels | Brembo 380mm BBK | | Vorsteiner Ti Exhaust | Matte Black Grilles/Side Gills/Rear Emblem/Mirrors | | Alekshop Back up Camera | GP Thunders | BMW Aluminum Pedals | Elite Angels | | XPEL Full Front Wrap | Hardwired V1 | Interior Xenon Light Kit | |
Appreciate
0
|
08-24-2009, 07:27 PM | #66 |
Automotive Industry Insider
462
Rep 1,948
Posts |
smamp2,
What's the deal here? I have tried several times to explain why I think your numbers are off. (on the timed 0-50, 0-60, 1/4 mile acceleration results you posted) That's my opinion. You have yours. Why must you continue this silly charade? My opinion is based on the results I have actually seen over the past 3 decades of drag racing. I'm sorry if that 'evidence' contradicts what your software simulation program is telling you. You have have made it perfectly clear, that you do not respect that opinion, and would rather use simulation numbers only. That's your right. Nobody is trying to say you can't. How about showing some respect for the individuals who don't bow at the alter of the almighty Car Tech software? I certainly hope you don't expect me to abandon my belief system...just to appease you? I have met a great deal of Engineers over the course of life, and I have the utmost respect for many of them. They can calculate anything you can possibly imagine...on paper. I have also seen these engineers out on the track. (after running endless simulations on 'generic' software programs). One size does NOT fit all. When some of their calculations don't TRANSLATE well to the real world performance on the track, there is a moment where they appear to look like a deer caught in the headlights. I have seen this look many times. Regardless of how many degrees you have, or how much book knowledge you posses...that has to be backed up with some basic common sense. I have met a number of Engineers who had an abundance of the former, and virtually none of the latter. (which turned out to be their Achilles heal) I don't doubt the fact that these software simulations can point you in the right direction, but I think it's very foolish to put all your eggs in that basket. Your risk is actually greater by choosing that path, than actually choosing your mod upgrades from a trusted individual(s) who have vast experience making those hardware changes. (again, real world track experience is worth it's weight in gold) Seeking the knowledge of a trained professional (who has literally seen it all), is significantly more valuable to me than a set of artificially derived numbers on a software simulation. Sorry to see that you think so little of using this methodology. I'm sure all the racing organizations I speak with, don't factor in all the experience they have gained over the years at all. It's all about what the computer simulation tells them to do. I'm sure that's exactly how it works, when they need to upgrade their hardware componants... No need to 'test' a number of different hardware packages in the real world, let's just buy the option the software is suggesting. The results can't be wrong after all, it's based in infallible metrics data. On a more serious note... I tend to put more stock in the tried and true method of upgrading parts, which is based on real world experience and knowledge. The software 'simulations' are indeed a useful tool. And I am not discounting that point at all. There is certainly no doubt that it can help you in some aspects. But I do not believe for one minute, that it should be the determining factor in choosing one path over another. If you make even the slightest error in your input values, the results can (and will) look very different. That can change your entire perspective on which path to take. You never qualified your numbers until I dared to challenge the results. Do you realize that? All the simulation numbers in this thread are 'guesstimates'. Regardless of how many mathematical formulas were used, that's what they are... The numbers are derived using theories and assumptions. They are not infallible, and your refusal to admit that is telling. When it comes down to making a hard (and expensive decision), I always error on the side of real world experience. (from an actual human beings) That method has prevented me from making any colossal mistakes over the years, so why in the world would I abandon that process? Because the results of some random software simulation? If those are my two choices, guess which path I'm going to take. I see now, that's it's truly impossible for you to see things from that real world perspective. (after trying in vain to get that across to you several times) You wrap yourself up in a bunch of arbitrary numbers, and you resist any attempt to discuss things in a rational real world manner. You seem to discount anything else out of hand, which is incredibly arrogant. I'm not sure if you even realize that you do this on a regular basis? :confused I guess we will have to agree to disagree, since you cannot possibly bring yourself down from your lofty perch...and admit that you precious calculations could indeed be flat out wrong. (because of a input error you may have made) The hubris that you project to the members of this board is breathtaking. In that respect, you are unrivaled. You are right, and everyone else is always wrong. That's the image you project from what I have seen in a number of threads. Not only that, but you seem to go out of your way to type most most condescending posts I have ever seen on an internet message board. (if someone challenges your opinion) That feeling is shared by many of the fellow m3post members here. I go out of my way to help the members of this board on a variety of automotive related topics. I post the information they need to make an informed decision on a specific mod. My history clearly indicates that intent. I personally respect the knowledge you possess. But unfortunately, your aptitude in debating any issue, with anyone on this board (sans the condescending tone)...is an unmitigated disaster. You do this every time someone disagrees with your opinion. There can be no debate without at least the tacit acceptance, that maybe, just maybe, your simulation numbers won't accurately reflect the real world performance of this mod. Don't worry though...I won't hold my breath waiting to hear that from you.
__________________
The best is yet to come...
|
Appreciate
0
|
Post Reply |
Bookmarks |
|
|