how do I improve torque?

[klammer]

Originally Posted by BMRLVR

If you want more torque you bought the wrong car. The S65 begs you to wring it's neck and shift gears. It is an NA sports car engine with a high redline and a nice progressive power band to match, big lazy torquey engines are boring and don't force the driver to work and therefore are unrewarding in my opinion.

Exactly! It's like people who ride Harley's bitching that a 600 sport bike is slow

[M2withACC]

S65 is a very balanced engine. My old C63 had too much torque. Oh well

[swamp2]

Where to begin...

Originally Posted by pbonsalb

Interesting that an E9X M3 will be equally fast with a 3.15 or a 4.10, or any other contrast in gear ratios for that matter. We have now debunked the "performance axle ratios" offered by the factory for many cars as a fraud. And Dinan must not have been using the right simulation software when testing its aftermarket differentials. Maybe it was actually testing them in the real world instead.

Testing is not more accurate than simulation. Anyone who believes this does not understand either. PERIOD. Have a look at the fairly large performance database of M3's performance here on the forum. For any given contest there is a very large range of values obtained. Variables contributing to test (i.e. "REAL" world) testing include but are not limited to: driver, weather, road condition, road slope, tires, tire wear state, tire inflation, level of vehicle break in, fuel variability and on and on and on. Unfortunately with almost all "REAL" testing that you've done yourself and even that is reported by the media too many of these factors are random and uncontrolled leading to large variations. With simulation you remove EVERY variation expect the one or ones the user wishes to control. This is true cause and effect isolation along with control of the size of the input change/difference. Basically not even practically possible in the "real" world.

This was a 6MT to 6MT simulation comparison. The stock FD ratio is 3.846.

Oh and do you know how OEMs figure out how fast or efficient any given car will be long long before prototypes are built? Physics based simulation. Do you know how all mechanical systems are tested virtually for a wide range of loads like structural thermal, aero, acoustic, etc. long before the prototypes are built - physics based simulation. Do you know how cars are successfully manufactured today taking into account things like stamping, welding, casting, and other manufacturing processes - physics based simulation. Get with the 1980's dude!

Originally Posted by pbonsalb

With the DCT, the shift speed factor is nearly eliminated. However, an extra shift adding time is only an issue when the race ends at a speed that requires the lower geared car to do an extra shift and the higher geared car not to do one.

Although M-DCT is fantastic and offers a much much larger performance increase compared to a FD of any ratio I was not discussing it and it was not relevant to the simulations I posted.

Originally Posted by pbonsalb

You can choose a target speed that will favor the higher geared car, but you can also choose a target speed that will favor the lower geared car. Or you can choose a speed range in one gear or in a couple of gears that favors one car or the other.

...
I do agree that diff ratio changes come with compromises. Highway rpm may be higher, MPG may be less, traction may be worse off the line, acceleration will be slower or faster to certain arbitrary target speeds or through certain arbitrary speed ranges. The factory's compromise is usually good for the average driver.[/quot]

More or less correct as the simulations indicate. Although prior work done my another M3post.com member showed that the existing M3 FD ratio were pretty close to optimal for some common performance metrics (may have been ET, don't quite recall).

Originally Posted by pbonsalb

But having done diff ratio changes in the real world (3.15, 3.23, 3.38 and 3.64 on my 99M3), I like them and believe that reasonable ones make a performance improvement that the performance driver appreciates. When my 99M3 was centrifugal supercharged and relatively torqueless, making only 368 lbs rwtq, I liked the 3.64. Now that it is turbocharged and making 517 lbs rwtq, I like the 3.15 and for traction reasons, I may go to a 2.93. However, I am not basing my like or dislike on arbitrary target speeds or arbitrary speed ranges, because I have found based on experience that there is much more to the real world than computer simulations with limited inputs.

Blah, blah, blah. I like the "feel" blah, blah, blah. Post up multiple runs with completely controlled parameters mentioned above and show a concrete, statistical advantage for your 99 M3 FD mod. I could bother to run some simulations but they would only show it does not really make a difference. Stop the refuting physics by feel BS. Numbers or go home. I tire of such BS.

Originally Posted by talontid

WOW! You guys are debunking "the myth" of gears on the internet based on an app that shows you who's faster... Where are all the real world tests?? I would bet alot of $$ that says the car w/ gears if faster in the 1/4. Your app is also missing 60ft times as well. I dunno, if you guys wanna believe it ok, but I have had 3:27, 3:73 and 4:10 (although diff cars, the results are independent of each other and proved gearing makes a big diff).. Now to the OP original question, yes gears will make city driving feel that much better, and will keep the car in its' powerband

Uhhh, yes. See above. This is not a typical "app". This is physics based simulation. Please re-read comments about real world testing above. Too many variations for any single test to be that useful.

I'll bet that a pre and post "geared" MT E9X M3 car will not show STATISTICALLY significant 1/4 improvements with a 3.85 -> 4.10 FD mod. In fact I'd bet pinks (if you car has a comparable value...). Money where your mouth is.

The app does provide 60 foot times.

Originally Posted by talontid

Oh and BTW based on the app screen shot you posted, that et does not correlate to the trap speed; that's a LOW 12 sec pass easily.

Times are traps both correlate well with the range of data from the performance database of stock M3s posted by various magazines/journalists. That being said I have noted that traps are somewhat less accurately predicted compared to times. However, in this case we only care about the A-B comparison as all losses such as aero and drivetrain and held EXACTLY fixed between the two hypothetical cars.

Dinan and other FD folks are selling FEEL and peak and average in gear acceleration (which is what you feel). Any sophisticated seller will realize that for most cars FD mods do not appreciably improve numbers across the board.

There is no such thing as a free lunch. Without more power or less weight (or perhaps less losses or improved aero) you simply DO NOT, CAN NOT get an overall improvement in performance. Those that say otherwise are not better than snake oil peddlers.

I really tire of this debate somewhat but feel strongly enough about it to continue to try to promote a wider understanding of it.

Next? ...

[m3an]

Originally Posted by swamp2

Where to begin...



Testing is not more accurate than simulation. Anyone who believes this does not understand either. PERIOD. Have a look at the fairly large performance database of M3's performance here on the forum. For any given contest there is a very large range of values obtained. Variables contributing to test (i.e. "REAL" world) testing include but are not limited to: driver, weather, road condition, road slope, tires, tire wear state, tire inflation, level of vehicle break in, fuel variability and on and on and on. Unfortunately with almost all "REAL" testing that you've done yourself and even that is reported by the media too many of these factors are random and uncontrolled leading to large variations. With simulation you remove EVERY variation expect the one or ones the user wishes to control. This is true cause and effect isolation along with control of the size of the input change/difference. Basically not even practically possible in the "real" world.

This was a 6MT to 6MT simulation comparison. The stock FD ratio is 3.846.

Oh and do you know how OEMs figure out how fast or efficient any given car will be long long before prototypes are built? Physics based simulation. Do you know how all mechanical systems are tested virtually for a wide range of loads like structural thermal, aero, acoustic, etc. long before the prototypes are built - physics based simulation. Do you know how cars are successfully manufactured today taking into account things like stamping, welding, casting, and other manufacturing processes - physics based simulation. Get with the 1980's dude!



Although M-DCT is fantastic and offers a much much larger performance increase compared to a FD of any ratio I was not discussing it and it was not relevant to the simulations I posted.


...
I do agree that diff ratio changes come with compromises. Highway rpm may be higher, MPG may be less, traction may be worse off the line, acceleration will be slower or faster to certain arbitrary target speeds or through certain arbitrary speed ranges. The factory's compromise is usually good for the average driver.[/quot]

More or less correct as the simulations indicate. Although prior work done my another M3post.com member showed that the existing M3 FD ratio were pretty close to optimal for some common performance metrics (may have been ET, don't quite recall).



Blah, blah, blah. I like the "feel" blah, blah, blah. Post up multiple runs with completely controlled parameters mentioned above and show a concrete, statistical advantage for your 99 M3 FD mod. I could bother to run some simulations but they would only show it does not really make a difference. Stop the refuting physics by feel BS. Numbers or go home. I tire of such BS.



Uhhh, yes. See above. This is not a typical "app". This is physics based simulation. Please re-read comments about real world testing above. Too many variations for any single test to be that useful.

I'll bet that a pre and post "geared" MT E9X M3 car will not show STATISTICALLY significant 1/4 improvements with a 3.85 -> 4.10 FD mod. In fact I'd bet pinks (if you car has a comparable value...). Money where your mouth is.

The app does provide 60 foot times.



Times are traps both correlate well with the range of data from the performance database of stock M3s posted by various magazines/journalists. That being said I have noted that traps are somewhat less accurately predicted compared to times. However, in this case we only care about the A-B comparison as all losses such as aero and drivetrain and held EXACTLY fixed between the two hypothetical cars.

Dinan and other FD folks are selling FEEL and peak and average in gear acceleration (which is what you feel). Any sophisticated seller will realize that for most cars FD mods do not appreciably improve numbers across the board.

There is no such thing as a free lunch. Without more power or less weight (or perhaps less losses or improved aero) you simply DO NOT, CAN NOT get an overall improvement in performance. Those that say otherwise are not better than snake oil peddlers.

I really tire of this debate somewhat but feel strongly enough about it to continue to try to promote a wider understanding of it.

Next? ...

stay in the lab buddy, "real-life" data will always prevail.. Any yes, car manufacturers are still acutally racing cars, 1/4 mile, road racing, etc.

I used to drag race ALOT, and I can't listen to someone who is posting statistics versus real world information..
One last Q... Have you ever even drag raced a car?? Or do you just like to read? Schnell has gears and first hand info.. My last car picked up like 2/10 from gears. Keep in mind every .1 in the 60 ft equates to about .2 upstairs.. There's so many factors that a simulator cannot depict. Hence why people start w/ sims and move to the real thing.

[m3an]

A neat little article on gears that if read correctly helps to understand why BMW chose the rear gears they did..


http://www.carcraft.com/techarticles...rks/index.html

[swamp2]

Originally Posted by SehrSchnell

Well, I know swamp didn't quote me, but I feel I should give my two cents (again) regarding the "feel" of the gears...

Although as stated many times, I agree with swamp regarding the numbers, but I find it absolutely laughable when you say that the "feel" should be dismissed entirely. At the end of the day - the feel is what matters! I don't care about simulations and numbers... I do that for a living, and I highly appreciate them, and again, numbers don't lie, so there is no point to argue about them.

But... driving the car is more than just clinical numbers, it's all about how it feels, and I simply like it. I'm not trying to convince anybody regarding the 4.10 gears, nor do I try to justify my purchase (I could switch them out and sell them the next day), and I wasted tons of money on other mods, and cars, so I'm the first one to admit if I wasted cash.

You can run simulations all day long and crunch numbers, but you buy a car after a test drive - to see how it feels, and not after looking over all the test results.

Agree! There is nothing wrong with feel. One should just be FULLY aware of what they are doing and you are pretty well in the minority in this regard.

One other last point to make here is that a larger FD will get you rowing through the gears faster. This aspect of the change will almost for sure make one feel that the car is better/faster. However, the effect is actually positively harming performance. A bit like a nice throaty/racy exhaust, it will absolutely sound faster but wont be. Be cautious of your ability to really feel the true result!

Originally Posted by talontid

stay in the lab buddy, "real-life" data will always prevail.. Any yes, car manufacturers are still acutally racing cars, 1/4 mile, road racing, etc.

I used to drag race ALOT, and I can't listen to someone who is posting statistics versus real world information..
One last Q... Have you ever even drag raced a car?? Or do you just like to read? Schnell has gears and first hand info.. My last car picked up like 2/10 from gears. Keep in mind every .1 in the 60 ft equates to about .2 upstairs.. There's so many factors that a simulator cannot depict. Hence why people start w/ sims and move to the real thing.

Well then you stay in the 1970's. I'll stick with both. Clearly you missed the fact that I referred above to simulations and a large-ish database or real performance numbers right here from this site. Have a look here or at the last/most current post/update. What is "THE" 1/4 mi time for an E9X M3?

Just to humor you I have been the drag's both to watch and to drive my own car. I've also been to the track many times and autox as well.

If you do not understand the need for statistics for concrete proof when measurement error and/or random error are larger than the effects of many modifications themselves then the entire conversation is pretty well lost on you.

Although there are some factors not adequately captured by simulation those factors for street cars, even exotics at the 1/4 mi track are very few and can be almost entirely eliminated since we are doing an A-B comparison. Again since you don't understand this the point is pretty well lost.

[m3an]

Originally Posted by swamp2

Agree! There is nothing wrong with feel. One should just be FULLY aware of what they are doing and you are pretty well in the minority in this regard.

One other last point to make here is that a larger FD will get you rowing through the gears faster. This aspect of the change will almost for sure make one feel that the car is better/faster. However, the effect is actually positively harming performance. A bit like a nice throaty/racy exhaust, it will absolutely sound faster but wont be. Be cautious of your ability to really feel the true result!



Well then you stay in the 1970's. I'll stick with both. Clearly you missed the fact that I referred above to simulations and a large-ish database or real performance numbers right here from this site. Have a look here or at the last/most current post/update. What is "THE" 1/4 mi time for an E9X M3?

Just to humor you I have been the drag's both to watch and to drive my own car. I've also been to the track many times and autox as well.

If you do not understand the need for statistics for concrete proof when measurement error and/or random error are larger than the effects of many modifications themselves then the entire conversation is pretty well lost on you.

Although there are some factors not adequately captured by simulation those factors for street cars, even exotics at the 1/4 mi track are very few and can be almost entirely eliminated since we are doing an A-B comparison. Again since you don't understand this the point is pretty well lost.

ok, I guess as you say I don't get it. Keep resorting to "statistics" and I will keep testing in the "real world" racing.. Good day!

[GIdriver]

Originally Posted by SehrSchnell

Another point regarding "feel", and why it can't be dismissed: Take f@cking for example, technically it's for making babies, but we all do it because it feels good.

LMFAO!!!

I agree that how you "feel" your car should count as well.

[pbonsalb]

Again, BMW chose the 3.85 from a range of reasonable choices, as the best choice for the average driver given performance and fuel economy targets. The choice was a compromise. It was not the choice that made the car the fastest or that gave it the best fuel economy. There are good reasons why people may want to accent one or the other more, and change the ratio to achieve that.

Each car will have to make 5 shifts to reach its top speed, so shift time ends up being equal. The only difference is that the shorter geared car will have a slightly lower top speed and will have a slightly reduced maximum speed in each gear. It will go through the speed range of that gear faster, however, because torque is being multiplied better.

Sure, pick some speed range that requires the shorter geared car to have to change more gears than the longer geared car, and the longer geared car will likely be faster (though DCT dramatically reduces any gear change time advantage).

Having run 4 different diffs on my E36M3, I have experienced what ratio changes do in the real world. Even your own simulation in Post #48 shows that one knocks a second off one's 60-130 time with the shorter diff, and people routinely do other mods that you probably agree make a performance difference in order to knock that same second off their 60-130 time.

[swamp2]

Originally Posted by talontid

ok, I guess as you say I don't get it. Keep resorting to "statistics" and I will keep testing in the "real world" racing.. Good day!

Wow you can't even get a hang of the vocabulary...

I am relying on physics based simulation.

I contend that far too many make rash and even incorrect conclusion from "real world" test data due to lack of statistics and uncontrolled influences.

[swamp2]

Originally Posted by pbonsalb

Again, BMW chose the 3.85 from a range of reasonable choices, as the best choice for the average driver given performance and fuel economy targets. The choice was a compromise. It was not the choice that made the car the fastest or that gave it the best fuel economy. There are good reasons why people may want to accent one or the other more, and change the ratio to achieve that.

No.

The diff ratio was chosen almost entirely for performance. It is very close to the best ratio to maximize 1/4 mi times. A slightly LOWER ratio would actually produce a better 1/4 mi time. Generally in gear results are improved by going with a large ratio. CarTest allows a cool parameter exploration/optimization tool. Obviously neither in simulation nor in the real world can you optimize for diverse performance metrics simultaneously. See chart below.

Originally Posted by pbonsalb

Even your own simulation in Post #48 shows that one knocks a second off one's 60-130 time with the shorter diff, and people routinely do other mods that you probably agree make a performance difference in order to knock that same second off their 60-130 time.

No.

You misread the data. By changing the starting gear the 4.10 geared car can gain a maximum of 0.4 seconds improvement for 60-130. Make an arbitrary change though to perhaps 50-120 or 70-140 and there would probably be a loss of 0.4 seconds. Now if all you care about is ONE particular contest then indeed some very minor benefits can be gained.

It seems abundantly clear here you are painfully proud of all of you diff swapping exercises and will go to any measure to prove they have large benefits. Almost any other mod that simply provides some speck of broadband power gains (pulley, tune, high flow cat, etc.) is a far better mod than ANY final drive mod. Why, it adds power and that improves results ACROSS the board.

[lippi]

Originally Posted by SehrSchnell

Another point regarding "feel", and why it can't be dismissed: Take f@cking for example, technically it's for making babies, but we all do it because it feels good.

Best example ever! Hahaha

[duk]

SC it

[pbonsalb]

Originally Posted by swamp2

You misread the data. By changing the starting gear the 4.10 geared car can gain a maximum of 0.4 seconds improvement for 60-130.

Maybe I should have majored in simulations rather than English, but the way I read your simulation of 60-130 mph times in Post #48, the 4.10 is from 0.25 seconds to 1.5 seconds faster depending on what gear the run starts in.

On the street in daily driving, we are not always doing our computer simulation. Sometimes we accelerate without having the computer simulate the best possible gear changes. We actually drive the cars. I have run from 60-130 starting in 4th, for example. And the 4.10 would be a second faster. Not being in the computer chosen perfect gear making computer perfect shifts is the real world and the torque multiplication shows up there.

Even in your computer perfect world, the 4.10 diff is faster from 60-130 regardless of what gear you start in. Why is that? Because torque is multiplied better. And where do you see the most dramatic difference? Where there is a long pull in the same gear, like running 60-130 iin 6th. You see a 1.5 second improvement because the torque multiplication is better.

Why don't you run some simulations to debunk the underdrive pulley? Let's say it adds 10 rwhp. Go run your computer. You will come back and tell us not to do it because on the computer the gains show up as little or nothing in the quarter mile.

You could play this game for any minor mod. Do it for a tune that adds 10-15 rwhp. Then tell people that tunes are all in your head and that there is no point because your computer simulation of maximum acceleration only with perfect and ideal shifts only shows minimal gains. The 4.10 diff is multiplying torque better by 5% in every gear at every rpm. The tune that adds 10-15 rwhp is adding most of that power from 5000 rpm and next to nothing from 750-5000 rpm where I drive 99% of the time.

I'll stay in the real world. Simulations have their value, but the real world has much more value to me.

[m3an]

Originally Posted by swamp2

Wow you can't even get a hang of the vocabulary...

I am relying on physics based simulation.

I contend that far too many make rash and even incorrect conclusion from "real world" test data due to lack of statistics and uncontrolled influences.

so you debunk facts founded in the "real world" unless your physics based sim software agrees?.. Read that back to yourself and see if it makes sense.

Originally Posted by swamp2

No.

The diff ratio was chosen almost entirely for performance. It is very close to the best ratio to maximize 1/4 mi times. A slightly LOWER ratio would actually produce a better 1/4 mi time. Generally in gear results are improved by going with a large ratio. CarTest allows a cool parameter exploration/optimization tool. Obviously neither in simulation nor in the real world can you optimize for diverse performance metrics simultaneously. See chart below.



No.

You misread the data. By changing the starting gear the 4.10 geared car can gain a maximum of 0.4 seconds improvement for 60-130. Make an arbitrary change though to perhaps 50-120 or 70-140 and there would probably be a loss of 0.4 seconds. Now if all you care about is ONE particular contest then indeed some very minor benefits can be gained.

It seems abundantly clear here you are painfully proud of all of you diff swapping exercises and will go to any measure to prove they have large benefits. Almost any other mod that simply provides some speck of broadband power gains (pulley, tune, high flow cat, etc.) is a far better mod than ANY final drive mod. Why, it adds power and that improves results ACROSS the board.

you just confirmed what I/we have said about a diff cutting 1/4 mile times, where you originally argued it would not make a diff at all... Again every .1 downstairs is worth about .2 upstairs. So yes, a diff change will help 1/4 mile times, which last I checked, is pretty much the benchmark for testing. I don't see car companies doing many other tests for speed reference.

to your second point, no again.. The OP asked for a good mod for torque. For the best bang for buck, rear gears are a great mod. You are not gaining power, we know that.

Originally Posted by pbonsalb

Maybe I should have majored in simulations rather than English, but the way I read your simulation of 60-130 mph times in Post #48, the 4.10 is from 0.25 seconds to 1.5 seconds faster depending on what gear the run starts in.

On the street in daily driving, we are not always doing our computer simulation. Sometimes we accelerate without having the computer simulate the best possible gear changes. We actually drive the cars. I have run from 60-130 starting in 4th, for example. And the 4.10 would be a second faster. Not being in the computer chosen perfect gear making computer perfect shifts is the real world and the torque multiplication shows up there.

Even in your computer perfect world, the 4.10 diff is faster from 60-130 regardless of what gear you start in. Why is that? Because torque is multiplied better. And where do you see the most dramatic difference? Where there is a long pull in the same gear, like running 60-130 iin 6th. You see a 1.5 second improvement because the torque multiplication is better.

Why don't you run some simulations to debunk the underdrive pulley? Let's say it adds 10 rwhp. Go run your computer. You will come back and tell us not to do it because on the computer the gains show up as little or nothing in the quarter mile.

You could play this game for any minor mod. Do it for a tune that adds 10-15 rwhp. Then tell people that tunes are all in your head and that there is no point because your computer simulation of maximum acceleration only with perfect and ideal shifts only shows minimal gains. The 4.10 diff is multiplying torque better by 5% in every gear at every rpm. The tune that adds 10-15 rwhp is adding most of that power from 5000 rpm and next to nothing from 750-5000 rpm where I drive 99% of the time.

I'll stay in the real world. Simulations have their value, but the real world has much more value to me.

BTW, people use real world data, just for that reason. Any race team will actually race the car (drag, road race, etc.) b/c as you stated, your simulators are for perfect situations and does not allow for variables or "uncontrolled instances" as you said which are not "realistic" Keep playing w/ apps; the rest of us will stick to trial and error!

[swamp2]

Originally Posted by pbonsalb

Maybe I should have majored in simulations rather than English, but the way I read your simulation of 60-130 mph times in Post #48, the 4.10 is from 0.25 seconds to 1.5 seconds faster depending on what gear the run starts in.

On the street in daily driving, we are not always doing our computer simulation. Sometimes we accelerate without having the computer simulate the best possible gear changes. We actually drive the cars. I have run from 60-130 starting in 4th, for example. And the 4.10 would be a second faster. Not being in the computer chosen perfect gear making computer perfect shifts is the real world and the torque multiplication shows up there.

Even in your computer perfect world, the 4.10 diff is faster from 60-130 regardless of what gear you start in. Why is that? Because torque is multiplied better. And where do you see the most dramatic difference? Where there is a long pull in the same gear, like running 60-130 iin 6th. You see a 1.5 second improvement because the torque multiplication is better.

60-130 is a pretty competitive endevor for those that actually time their runs. In light of that it only makes sense to report the gain from best without mod to best with mod and that is about 0.4 seconds here. Not bad I guess. However, if you drive lazy (i.e. in the wrong gear) why bother timing your run anyway?? Here a FD mod obviously helps more. The worse gear you choose the more benefit. If having the best benefit under the worst cicuymstance (where any driver actually drying would blow you away) then knock yourself out. You should probably buy a 335 over an M3 as well. Again change the contest slightly from 50-120 or something similar and you could find a harming effect not a benefit.

Originally Posted by pbonsalb

Why don't you run some simulations to debunk the underdrive pulley? Let's say it adds 10 rwhp. Go run your computer. You will come back and tell us not to do it because on the computer the gains show up as little or nothing in the quarter mile.

You could play this game for any minor mod. Do it for a tune that adds 10-15 rwhp. Then tell people that tunes are all in your head and that there is no point because your computer simulation of maximum acceleration only with perfect and ideal shifts only shows minimal gains. The 4.10 diff is multiplying torque better by 5% in every gear at every rpm. The tune that adds 10-15 rwhp is adding most of that power from 5000 rpm and next to nothing from 750-5000 rpm where I drive 99% of the time.

I'll stay in the real world. Simulations have their value, but the real world has much more value to me.

Glad you mentioned it. Here are the sims. This is the power of simulation. Curious - just run it (of course being mindful of garbage in garbage out which is VERY possible with such tools). AA pulley +12 hp vs. 4.10 gears vs. stock. Other than a very small advantage to distance off the line up to about 100' the car with the $400 pulley has a clear across the board advantage. How much are Dinan 4.10 gears, $3,000. If that does not seal the deal I don't know what will. Spend (waste) your money how you like and keep it "real" with no real numbers to back up your claims (I'm still waiting on those numbers from you by the way...). I'd do a pulley at three times the price before 4.10 gears.

The best band for the buck is actually M-DCT. Depending on the contest and shift speed of a MT driver is can be good for an equivalent 20-50 hp over the 6MT. That's a whole other discussion though... Dig around for those threads if interested.

Lat but not least if you spend 99% of your time driving a high reving car like an M3 below 5000 you clearly have no interest in actually making the car go and the entire discussion is pretty well moot.

[swamp2]

Originally Posted by talontid

so you debunk facts founded in the "real world" unless your physics based sim software agrees?.. Read that back to yourself and see if it makes sense.

BTW, people use real world data, just for that reason. Any race team will actually race the car (drag, road race, etc.) b/c as you stated, your simulators are for perfect situations and does not allow for variables or "uncontrolled instances" as you said which are not "realistic" Keep playing w/ apps; the rest of us will stick to trial and error!

You just don't get it...

Engineering and science are measurement based activities. No disagreement whatsoever on that. However, the devil is in the details and race team and professionals (say the OEMs). Have the time and money for very high quality testing under controlled circumstances. What do they do BEFORE and AFTER they test - THEY SIMULATE!!!

What is THE 1/4 mi time for the E9X M3? If you consult the"database" link I provided prior you will find not a number but a pretty wide range of numbers:

12.6 s - 13.3 s

Similar ranges exist for all contests.

Due to uncontrolled circumstances natural variations in results will occur. For each time reported by these sources how many runs do you think they took to get their best times? I'd guess around 10. Each of those runs probably showed a couple tenths of variation. Let's imagine a "12.7" second M3 showed runs ranging from 12.7 up to 13.1 (we always report bests). If you then turned around and immediately added a pulley and did one run at 12.8 would you imagine the pulley did not add power? You would probably have to take ANOTHER 5-10 runs and average them to see the actual difference. Heck maybe from heat soak even with the pulley you could not best the 12.7 time. "Real world" testing sort of tells us that the pulley IS NOT MAKING MORE POWER (well less losses technically but basically the same thing). That is absolutely, positively THE WRONG CONCLUSION.

With simulation ALL of these natural variations are AUTOMATICALLY equalized and with ONE run you can ISOLATE the effect you are looking for.

Are we making any progress here?

It does not take statistics to verify significant mods say such as adding a blower to an M3. It most likely WILL take statistics and some effort at environmental control to see the effect of low power modifications (or zero power modifications such as the FD as being discussed here). Physics based simulation can effectively shortcut the arduous and perhaps expensive testing process through natural and perfect control of potentially confounding or uncontrollable inputs.

Get it?

[pbonsalb]

Originally Posted by swamp2

Lat but not least if you spend 99% of your time driving a high reving car like an M3 below 5000 you clearly have no interest in actually making the car go and the entire discussion is pretty well moot.

You must be one of the fast and furious crowd that lives on a race track. Why don't you datalog your driving since you like computers so much? You will find that about 99% of the time your car is running, it is below 5000 rpm.

My sole enjoyment from my car is not dragstrip runs through the quarter mile. The vast majority of the 30-50 mph, 50-70 mph and 60-130 mph runs in your simulation show the 4.10 gears are best. What do you do most in daily driving? 1/4 mile runs, 30-50 mph accelerations, 50-70 mph accelerations, 60-130 mph accelerations? I do a lot more 30-50s and 50-70s than I do 1/4 mile runs in my daily driving. And life is not a perfect computer simulation. I am not always in the perfect gear.

[GIdriver]

Well if this is going to continue, I'm gonna grab some

Intresting discussion, though. Nice to see that this data supports that a pulley works that well. Probably the best bang for the buck in our cars, IMO.

[m3an]

I've seen enough. I mean there's a reason dynos (specifically mustang) replicate real world data w/ load bearing rollers; there's a reason why street tuning is still great; there's a reason b/c countless variables not observed from a sim program. For 1/4 mile I will still bet, two cars, stock for stock, one w/ 4: 10 or lower gear (shorter gear (higher #) ) will be faster in the 1/4 mile. Shit, the OP was looking for best bang for the buck to increase torque?? And you still would spend 3x amount for a pulley?? OK, good advice.


One last question though chief; you use a sex simulator before you do the real thing?

[GIdriver]

Originally Posted by talontid

I've seen enough. I mean there's a reason dynos (specifically mustang) replicate real world data; there's a reason why street tuning is still great; there's a reason for countless variables not observed from a sim program. For 1/4 mile I will still bet, two cars, stock for stock, one w/ 4: 10 or lower gear (shorter gear (higher #) ) will be faster in the 1/4 mile. Shit, the OP was looking for best bang for the buck to increase torque?? And you still would spend 3x amount for a pulley?? OK, good advice.


One last question though chief; you use a sex simulator before you do the real thing?

[m3an]

Originally Posted by GIdriver

lol it makes sense now doesn't it??