E55 vs M5 rolling runs: Torque vs Gearing
#126
Senior Member
Join Date: Oct 2002
Location: Norway
Posts: 404
Likes: 0
Received 0 Likes
on
0 Posts
ALPINA B12 5,7 Coupe #22/57
Originally Posted by Improviz
Unfortunately for your argument, the E55's drag coefficient is better than the M5's by 13% (0.27/0.31 = 0.87). Hence, at high speeds the E55, not the M5, has the aerodymanic advantage. My previous post to Wolverine has links to both BMW and Mercedes sites giving the Cd for both cars. Feel free to click them.
I had no idea about the Cw value on the M5.
Originally Posted by Improviz
So you still lost on the point, even after shifting the argument away from Gustav's central contention, which was that weight has NO effect at higher speeds. Clearly, it does, and clearly, you were in error in defending his incorrect assertion.
Originally Posted by Improviz
I.e., the E55's drag should be lower than the M5's, as its drag coefficient is 13% lower than the M's
![thumbs](https://mbworld.org/forums/images/smilies/thumbsup.gif)
But it wold be interessting to hear what Wolverine had to say about the relationship and importance of the frontal area in connection with the factor called Cw. Is the result of those to what you guys call drag ?
I would guess the Frontal area of the E is larger than the 5 but probably enought to be an important factor. But now I am only guessing.
Last edited by Erik; 04-30-2005 at 02:55 AM.
#127
MBWorld Fanatic!
Originally Posted by Erik
But it wold be interessting to hear what Wolverine had to say about the relationship and importance of the frontal area in connection with the factor called Cw. Is the result of those to what you guys call drag ?
I would guess the Frontal area of the E is larger than the 5 but probably enought to be an important factor. But now I am only guessing.
I would guess the Frontal area of the E is larger than the 5 but probably enought to be an important factor. But now I am only guessing.
Wolverine - thanks for the interesting lesson.
Is the frontal area of the vehicle already incorporated into the calculation that results in the drag coef.?
#128
MBWorld Fanatic!
Originally Posted by Improviz
Interesting. So, let's say that the Cd of one car in question, namely the E55, is 0.27 , while for the second car in question, namely the M5, it is 0.31 .
Which car would be more affected by drag at high speeds? Would it not be the car with the lower Cd? I believe it would...given that the Cd in the equation you provided is the drag coefficient.![Smilie](https://mbworld.org/forums/images/smilies/smile.gif)
Which car would be more affected by drag at high speeds? Would it not be the car with the lower Cd? I believe it would...given that the Cd in the equation you provided is the drag coefficient.
![Smilie](https://mbworld.org/forums/images/smilies/smile.gif)
#129
Originally Posted by Erik
Fine, I did not have the M5 in mind and my comments was only based on you claim that Gustav said was wrong. It came up when you discussed the SL and the E.
Originally Posted by gustav
Weight has an insignificant advantage in higher speeds. Aerodynamic advantage is alot mroe important where I beleive the SL55 has its advantages over the E55.
Lastly, nobody here has stated that f=ma => a = f/m doesn't work at high speeds. The drag enters into the picture, but we are not discussing a mack truck versus a Porsche here, Erik, we are discussing two cars with very similar drag coefficients.
Originally Posted by Erik
I had no idea about the Cw value on the M5.
![thumbs](https://mbworld.org/forums/images/smilies/thumbsup.gif)
Originally Posted by Erik
What ? I think not, but thats OK.
Originally Posted by Erik
I know, and still it is able to move a way from the E55 at high speed. ![thumbs](https://mbworld.org/forums/images/smilies/thumbsup.gif)
![thumbs](https://mbworld.org/forums/images/smilies/thumbsup.gif)
So, from 30 mph up to 140+, in this particular run, with two extra gears, over 30 more rated horsepower, and a bit less weight, the brand new M pulled one carlength against the three year old E55.
Wow. And this has you M guys excited?
![Roll Eyes (Sarcastic)](https://mbworld.org/forums/images/smilies/rolleyes.gif)
![thumbs](https://mbworld.org/forums/images/smilies/thumbsup.gif)
Originally Posted by Erik
But it wold be interessting to hear what Wolverine had to say about the relationship and importance of the frontal area in connection with the factor called Cw. Is the result of those to what you guys call drag ?
I would guess the Frontal area of the E is larger than the 5 but probably enought to be an important factor. But now I am only guessing.
I would guess the Frontal area of the E is larger than the 5 but probably enought to be an important factor. But now I am only guessing.
![thumbs](https://mbworld.org/forums/images/smilies/thumbsup.gif)
#130
Pretty evident who the pople are here to get upset and slander people... Anyway, Impr you ignored that the E5 got the jump in the video as well. If the M5 had a similar jump Im sure the E55 would never be infront of the M5. It would be carlenghts behind as I experienced in other races
![Smilie](https://mbworld.org/forums/images/smilies/smile.gif)
#131
Oh yeah, Impro-man, in case you didn't notuce the E55 got a massive jump on the rolling run. Looks like over a carlength. And there were 3 people in the M5.
You are really sad 'cos I have a feeling even if you get exacty what you ask for, you will still dispute. I can't wait to hear you excuses when you see with your own eyes what an M5 will do to an E55.
You are really sad 'cos I have a feeling even if you get exacty what you ask for, you will still dispute. I can't wait to hear you excuses when you see with your own eyes what an M5 will do to an E55.
#132
Originally Posted by Gustav
Pretty evident who the pople are here to get upset and slander people... Anyway, Impr you ignored that the E5 got the jump in the video as well. If the M5 had a similar jump Im sure the E55 would never be infront of the M5. It would be carlenghts behind as I experienced in other races ![Smilie](https://mbworld.org/forums/images/smilies/smile.gif)
![Smilie](https://mbworld.org/forums/images/smilies/smile.gif)
In the meantime, incredibly gullible people can take the unscientific test results of paid BMW spokesperson like you with the grain of salt it so richly deserves.
![Roll Eyes (Sarcastic)](https://mbworld.org/forums/images/smilies/rolleyes.gif)
#133
Originally Posted by Monkey&Moron
![bs](https://mbworld.org/forums/images/smilies/bs.gif)
![bs](https://mbworld.org/forums/images/smilies/bs.gif)
![bs](https://mbworld.org/forums/images/smilies/bs.gif)
![bs](https://mbworld.org/forums/images/smilies/bs.gif)
![bs](https://mbworld.org/forums/images/smilies/bs.gif)
![bs](https://mbworld.org/forums/images/smilies/bs.gif)
Btw, you still haven't produced the data I asked for. Third time now: show me one test, from anywhere, where two cars tested 0.8 apart at 125 tested 0.3 to 0.4 apart in the 1/4 mile as you're claiming.
![crazy](https://mbworld.org/forums/images/smilies/crazy.gif)
You can't, and the results will prove it when they come out, and I'll be here with your quotes to show how full of **** you are.
Results, please. Stop dodging and start producing.
#134
Originally Posted by Gustav
Pretty evident who the pople are here to get upset and slander people... Anyway, Impr you ignored that the E5 got the jump in the video as well. If the M5 had a similar jump Im sure the E55 would never be infront of the M5. It would be carlenghts behind as I experienced in other races ![Smilie](https://mbworld.org/forums/images/smilies/smile.gif)
![Smilie](https://mbworld.org/forums/images/smilies/smile.gif)
I dont get it ----if you did the "test" and you 100% confident in its results, why do you come here and argue with the type of people that "get upset
and slander people"?? Why do you bend over trying to prove something to these "people"?
If you feeel that E55 got a jump on you-----post another video where the race was honest , no jumps, no pulls, whatever, or do you base your whole BS here on one race? Surly you have video where E55 does not "jump" and is compleetly "humiliated" as a result?
I did not read in the rules of M5board, that moderators from that board are requiered to go on other boards and prove something to angry, upset people. I dont see the difference between the "upset" people and you, honestly.
#135
Senior Member
Join Date: Oct 2002
Location: Norway
Posts: 404
Likes: 0
Received 0 Likes
on
0 Posts
ALPINA B12 5,7 Coupe #22/57
Originally Posted by wolverine
Excellent points. I'm surprised that the M5 is that much quicker than the E55 in the 0-200 test - 1.8 seconds is about a 10 CL margin.
SportAuto has the well earned reputation as producing the most consistent test results in the world in terms of their test methods. They are typically German and precise in identifying and controlling as many variables as possible. If you're going to use anything as a benchmark, this is it.
SportAuto has the well earned reputation as producing the most consistent test results in the world in terms of their test methods. They are typically German and precise in identifying and controlling as many variables as possible. If you're going to use anything as a benchmark, this is it.
Could you please tell my how you come to the conclution of 10 CL ?
Just curious
![thumbs](https://mbworld.org/forums/images/smilies/thumbsup.gif)
#136
Super Member
Originally Posted by Improviz
Interesting. So, let's say that the Cd of one car in question, namely the E55, is 0.27 , while for the second car in question, namely the M5, it is 0.31 .
Which car would be more affected by drag at high speeds? Would it not be the car with the lower Cd? I believe it would...given that the Cd in the equation you provided is the drag coefficient.![Smilie](https://mbworld.org/forums/images/smilies/smile.gif)
Which car would be more affected by drag at high speeds? Would it not be the car with the lower Cd? I believe it would...given that the Cd in the equation you provided is the drag coefficient.
![Smilie](https://mbworld.org/forums/images/smilies/smile.gif)
On drag caused air, you are correct the M5 should have more adverse effect.
However, you will also need to look at the mass of each car. Yes, M5 might have less horses to accelerate after taking into drag. However, I believe M5 is lighter. In addition, gearing also matters in how the car accelerate and how the power is transfered.
It's not simply looking a cof. of drag.
If your car is not geared well, no matter how much HP/Torque you have you will lose simple as that.
As far as 30+HP. We all know that E55's engine is not rated as mercedes claimed 469 HP. It was simply a marketing figure. The truth is that W211 engine puts the same amount of HP at the crank as any other 55 (close to 500). So you are talking around 10 HP difference. Not to mention that E55 has way more torque than M5 (which is espeically important for acceleration runs in automatic).
As far as who is faster. That E55 in the 2 trailer had one car length jump on the M5. The M5 was able to overtake the E55 with 3 people on board (probably extra 300 lb easily- i.e. take away the M5's weight advantage). It pulled over a car length ahead in rolling start.
This is a rolling start. E55 should have even a stronger advantage due to its massive torque.
But at least the trailer did prove without a doubt that the new M5 is faster than E55 in stock form.
And this thread should just die.
#137
Super Member
Originally Posted by reggid
i've looked at some test data and believe the test by Autocar claiming a 119mph trapspeed is wrong my numbers show 117mph.
the test data is
0-110 11.5
0-119 12.8
0-120 13.3
just by looking at the numbers they seem unrealistic becasue 110-119 (difference of 9mph) in 1.3sec but 119-120 (difference of 1mph) in 0.5sec and this is with a gear change at 125mph so that isn't the cause.
Other data
Sportauto
0-100km/h 4.5sec
0-200km/h 13.8sec
my interpolated values
60ft 2.2sec
0-60mph 4.35s
1/4mi 12.55@116.7mph
Autozeitang
0-100km/h 4.4sec
0-200km/h 13.9sec
my interpolated values
60ft 2.15sec
0-60mph 4.15s
1/4mi 12.45@116.8mph
Autocar
0-60mph 4.6sec
0-120mph 13.3sec
my interpolated values
60ft 2.45sec
0-100kph 4.9s
1/4mi 12.9@116.9mph
i fit an equation to the time to speed data and integrated the equation to get time to distance. Based on how well my eqation matches the original data points from which the equation was derived i believe my trapspeeds are +/- 0.5mph and time to distance accurcay about +/-0.1s.
I predict about 117mph trapspeeds at 12.4-12.5 ET.
i compared the test of the preproduction m5 and cls55 from automotor and sport
m5 cls55
0-100 4.7 4.7
0-200 14.8 15.6
my calcs show both did approx 12.9s the m5@115mph and cls55@112mph its worth noting that this m5 was 1 sec slower from 100-200kph than any other test i have since found.
the test data is
0-110 11.5
0-119 12.8
0-120 13.3
just by looking at the numbers they seem unrealistic becasue 110-119 (difference of 9mph) in 1.3sec but 119-120 (difference of 1mph) in 0.5sec and this is with a gear change at 125mph so that isn't the cause.
Other data
Sportauto
0-100km/h 4.5sec
0-200km/h 13.8sec
my interpolated values
60ft 2.2sec
0-60mph 4.35s
1/4mi 12.55@116.7mph
Autozeitang
0-100km/h 4.4sec
0-200km/h 13.9sec
my interpolated values
60ft 2.15sec
0-60mph 4.15s
1/4mi 12.45@116.8mph
Autocar
0-60mph 4.6sec
0-120mph 13.3sec
my interpolated values
60ft 2.45sec
0-100kph 4.9s
1/4mi 12.9@116.9mph
i fit an equation to the time to speed data and integrated the equation to get time to distance. Based on how well my eqation matches the original data points from which the equation was derived i believe my trapspeeds are +/- 0.5mph and time to distance accurcay about +/-0.1s.
I predict about 117mph trapspeeds at 12.4-12.5 ET.
i compared the test of the preproduction m5 and cls55 from automotor and sport
m5 cls55
0-100 4.7 4.7
0-200 14.8 15.6
my calcs show both did approx 12.9s the m5@115mph and cls55@112mph its worth noting that this m5 was 1 sec slower from 100-200kph than any other test i have since found.
Being a engineer, 1st thing that I was taught at school is that different data taken with same samples in same condition will show variation. Not to mention that the test was done by different people, different sample, and in different conditions.
The variation along in the 0-60 and 0-100 kph is already big.
To fit only 4 data points with huge variation to claim that 119 mph trap speed by autocar is wrong is beyond me.
I don't want to drag numerical analysis method into discussion. But the variation along in 0-60 time is already 10%. To fit a equation based on this than claim a less than 2% difference on trap speed is beyond stupid.
In addition, we have no idea how many test were run to obtain the results. Perhaps autocar ran the 0-100 0-119 0-120 at different time. There are so many variation that can be introduced. And yet, we have someone based on 4 data points (which we have no clue how it was obtained) to make a judgement call.
In addition, it is impossible to compare one set of test to another test. Conditions, the car itself are all variation.
The only thing that can be conclusive from the video is that on that day. The M5 with 3 people out gunned the E55 easily. ( For car at this HP range, to be able to do what M5 did to the E55 is significant.).
Last edited by krispykrme; 05-02-2005 at 08:21 PM.
#138
MBWorld Fanatic!
Originally Posted by krispykrme
But at least the trailer did prove without a doubt that the new M5 is faster than E55 in stock form.
And this thread should just die.
Krispy - your next post (lambasting somebody for hypothesizing about possible quarter mile times) says that there are variations in test data. So I ask you - How does one "test" of a car prove an absolute "without a doubt"??
![nix](https://mbworld.org/forums/images/smilies/nixweiss.gif)
I really am ready to believe it is generally faster - as it was designed to be. But not based on a paid spokesperson's test of one car that happened to run zero to 100 mph (WITH TWO PASSENGERS AND A DRIVER) in the same time it took Motor Trend to get the new Ferrari F430 to achieve that same speed. And, I might add, much faster (again with two passengers and a driver) than anything else that has been recorded for that car. It is great to reach a conclusion, but how one gets there is often just as important.
When the US cars get here, with or without the Euro spec launch control and SMG, we'll see what it does at the strip and on the highways. In fact, please take yours out to let your brothers (and sisters) on this board know what it can do.
![thumbs](https://mbworld.org/forums/images/smilies/thumbsup.gif)
![wwf](https://mbworld.org/forums/images/smilies/chairshot.gif)
#139
Originally Posted by krispykrme
I find it amazing that you can fit a equation based on couple data points.
Being a engineer, 1st thing that I was taught at school is that different data taken with same samples in same condition will show variation. Not to mention that the test was done by different people, different sample, and in different conditions.
The variation along in the 0-60 and 0-100 kph is already big.
To fit only 4 data points with huge variation to claim that 119 mph trap speed by autocar is wrong is beyond me.
I don't want to drag numerical analysis method into discussion. But the variation along in 0-60 time is already 10%. To fit a equation based on this than claim a less than 2% difference on trap speed is beyond stupid.
In addition, we have no idea how many test were run to obtain the results. Perhaps autocar ran the 0-100 0-119 0-120 at different time. There are so many variation that can be introduced. And yet, we have someone based on 4 data points (which we have no clue how it was obtained) to make a judgement call.
In addition, it is impossible to compare one set of test to another test. Conditions, the car itself are all variation.
The only thing that can be conclusive from the video is that on that day. The M5 with 3 people out gunned the E55 easily. ( For car at this HP range, to be able to do what M5 did to the E55 is significant.).
Being a engineer, 1st thing that I was taught at school is that different data taken with same samples in same condition will show variation. Not to mention that the test was done by different people, different sample, and in different conditions.
The variation along in the 0-60 and 0-100 kph is already big.
To fit only 4 data points with huge variation to claim that 119 mph trap speed by autocar is wrong is beyond me.
I don't want to drag numerical analysis method into discussion. But the variation along in 0-60 time is already 10%. To fit a equation based on this than claim a less than 2% difference on trap speed is beyond stupid.
In addition, we have no idea how many test were run to obtain the results. Perhaps autocar ran the 0-100 0-119 0-120 at different time. There are so many variation that can be introduced. And yet, we have someone based on 4 data points (which we have no clue how it was obtained) to make a judgement call.
In addition, it is impossible to compare one set of test to another test. Conditions, the car itself are all variation.
The only thing that can be conclusive from the video is that on that day. The M5 with 3 people out gunned the E55 easily. ( For car at this HP range, to be able to do what M5 did to the E55 is significant.).
look at the first paragagraph of the post where its gives 0-110,0-119 & 0-120 and tell me its not suspicous (119-120mph in 0.5sec at the very least it may be 118.5-120 in 0.5 sec due to rounding but still i think there is a typo).
Typically 0-100kph takes 2 tenths longer than 0-60mph (96kph) which is not a whole lot different from what i found.
I never claimed the M5 won't trap at 119 just in that particular test it would be more like 117 +/-
#140
Super Member
Originally Posted by reggid
there was actually about 10 data points used (not all shown) and the resulting equation(s) were checked against the data points. i did not claim 100% accurcay but its better than alot of other claims around here. The equation (a 5th order polynomial) is continous and the actual situation due to gear changes is not, so the point of my comments was to give a bit of insight into the time to distance and time to speed debate.
look at the first paragagraph of the post where its gives 0-110,0-119 & 0-120 and tell me its not suspicous (119-120mph in 0.5sec at the very least it may be 118.5-120 in 0.5 sec due to rounding but still i think there is a typo).
Typically 0-100kph takes 2 tenths longer than 0-60mph (96kph) which is not a whole lot different from what i found.
I never claimed the M5 won't trap at 119 just in that particular test it would be more like 117 +/-
look at the first paragagraph of the post where its gives 0-110,0-119 & 0-120 and tell me its not suspicous (119-120mph in 0.5sec at the very least it may be 118.5-120 in 0.5 sec due to rounding but still i think there is a typo).
Typically 0-100kph takes 2 tenths longer than 0-60mph (96kph) which is not a whole lot different from what i found.
I never claimed the M5 won't trap at 119 just in that particular test it would be more like 117 +/-
Typicall car test are not done based on a single run. It may appear that the test result is inconsistent. But it could be that one run is done when driver shifted a better point, the tire had better grip, may be wind was a factor. Or simply the measuring equipment also made a variation in test result.
No matter how many order of polynomial is used. When you are fitting data to a equation, there are always error. Simple as that. Equation is good as it give us an good insight to an objects behavior in a comman language we can understand. But by nature it is an approximination.
When magazine set up to test, they usually test different objective one at a time. Not all at once. Hence, each result i.e. 0-100, 0-110, 0-120, 0-119 could be results of seperate test. Hence each results are indep and can't be duplicate as I have stated earlier the control in this experiment is not perfect.
In that particular test, I find it's highly likely that the result would be a 119 not 117. Simply because you and i can't 100% certain that 0-119 and 0-120 test results are from the same test.
This is a car driven by human. Even in a machine controlled testing environment you are going to have variation.
In this case, the variation in data points that you have used in your equation is more than the result you are claiming. Hence, in statistical term, your conclusion can't be supported.
Unless you have thousands of data points, in which you have identical control in the experiment, you really can't use data points from different test with different car samples, different control (wind, driver.....) to make arrive to a conclusion. Especially the variation in data point inputed is greater then the result.
I am not saying the concept behind your post is wrong. It's just the way the data are inputed in your model is wrong.
#141
Super Member
Originally Posted by enzom
Krispy - your next post (lambasting somebody for hypothesizing about possible quarter mile times) says that there are variations in test data. So I ask you - How does one "test" of a car prove an absolute "without a doubt"??
Hang out at a drag strip during a Corvette challenge and watch how the Z06 times and speeds are all over the place. Just like drivers, cars have variations, too.
I really am ready to believe it is generally faster - as it was designed to be. But not based on a paid spokesperson's test of one car that happened to run zero to 100 mph (WITH TWO PASSENGERS AND A DRIVER) in the same time it took Motor Trend to get the new Ferrari F430 to achieve that same speed. And, I might add, much faster (again with two passengers and a driver) than anything else that has been recorded for that car. It is great to reach a conclusion, but how one gets there is often just as important.
When the US cars get here, with or without the Euro spec launch control and SMG, we'll see what it does at the strip and on the highways. In fact, please take yours out to let your brothers (and sisters) on this board know what it can do.
You will be in the best position to kill this thread once and for all. Please kill it. ![wwf](https://mbworld.org/forums/images/smilies/chairshot.gif)
![nix](https://mbworld.org/forums/images/smilies/nixweiss.gif)
I really am ready to believe it is generally faster - as it was designed to be. But not based on a paid spokesperson's test of one car that happened to run zero to 100 mph (WITH TWO PASSENGERS AND A DRIVER) in the same time it took Motor Trend to get the new Ferrari F430 to achieve that same speed. And, I might add, much faster (again with two passengers and a driver) than anything else that has been recorded for that car. It is great to reach a conclusion, but how one gets there is often just as important.
When the US cars get here, with or without the Euro spec launch control and SMG, we'll see what it does at the strip and on the highways. In fact, please take yours out to let your brothers (and sisters) on this board know what it can do.
![thumbs](https://mbworld.org/forums/images/smilies/thumbsup.gif)
![wwf](https://mbworld.org/forums/images/smilies/chairshot.gif)
In that day, the M5 with extra weight was faster than the E55. The car is at least running at same time (less variation in weather, surface...). The fact of matter is that E55 clearly had a jump over 1 car length (judging from the engine sound and the movement of the car). For the M5 to overcome such a disadvantage at the power output range we are discussing is quiet significant.
And probably can't be explained by variation caused by real world environment.
If the car were much closer in the video, then you can safely argue that variation in car itself could lead to different result. Moreover, E55 is an automatic, the chance of making error in driving it vs the M5 is much smaller. (I will always take a computer controlled response over a human response). Judging from the video and its result, it's very clear that M5 Gustav was driving is significantly faster than that E55.
Like i said, no one knows what US M5 will perform. I am willing to put both car for test and kill the thread once for all. But no takers.
I will probably ended up doing this on my own and eat all the cost, since my interest has been aroused. (actually not by the test result, I am more interested to see what kind of action I will get on this forum).
To me which car is faster is not really of significance to me. But I am always intrigued by level of maturity on this board. After spending all my life dealing with numbers, electrical, and computer systems. I am getting very interested in human behavior. I just find it fascinating that people on this forum take so much pride/ego in a piece of machinery.
#142
Krispy is correct in how the data inputting is incorrect. The concept is fine, but the execution is flawed. It's just like a post I made earlier, the sample sizes are too small.
Here's a breakdown of what we're looking for:
1) A SRS (simple random sample) of tests. == What that means is that the tests we pick out to compare against whatever we are comparing (in this case, creating an equation to model future tests agianst) must be randomly selected, and every run/test/result must have an equal chance of getting selected as EVERY other run. No if's or buts.
2) n > 30 so that the sampling distribution of means is normal. == What his means is we need our sample size to be at least composed of 30 samples. The second part of the assumption means that the reason we need a sample size of that much, is so we can model if we took every single possiblity of combinations of tests, the averages would be able to be plotted as follows. As per the empirical rule, 68% of all of the averages must be within one deviation of the sampling distribution mean. 95 within 2, and 99.7 within 3 deviations. This accounts for outliers and such and will try to take care of such cars such as a prepped press car.
Then last but not least
3) n must be no greater than 5% of the population. == What this means is what it says. If I have 30 tests in my sample, N (population) must be equal to or greater than 600.
Without these basic assumptions being met, it's hard to conduct almost any kind of test.. well.. correctly.
Here's a breakdown of what we're looking for:
1) A SRS (simple random sample) of tests. == What that means is that the tests we pick out to compare against whatever we are comparing (in this case, creating an equation to model future tests agianst) must be randomly selected, and every run/test/result must have an equal chance of getting selected as EVERY other run. No if's or buts.
2) n > 30 so that the sampling distribution of means is normal. == What his means is we need our sample size to be at least composed of 30 samples. The second part of the assumption means that the reason we need a sample size of that much, is so we can model if we took every single possiblity of combinations of tests, the averages would be able to be plotted as follows. As per the empirical rule, 68% of all of the averages must be within one deviation of the sampling distribution mean. 95 within 2, and 99.7 within 3 deviations. This accounts for outliers and such and will try to take care of such cars such as a prepped press car.
Then last but not least
3) n must be no greater than 5% of the population. == What this means is what it says. If I have 30 tests in my sample, N (population) must be equal to or greater than 600.
Without these basic assumptions being met, it's hard to conduct almost any kind of test.. well.. correctly.
#143
Originally Posted by krispykrme
It's not simply looking a cof. of drag.
Originally Posted by Krispykrme
If your car is not geared well, no matter how much HP/Torque you have you will lose simple as that.
Read the new Car & Driver for a discussion about how in the new SLK55, which weighs more than the previous one and has only six more rated horsepower, still managed to pull over 1/2 second faster in the 1/4, and was significantly faster in all 0-xxx times. Why? The magazine concluded the extra two gears did it, along with the extra torque.
No mystery here...like all BMWs, the M5 is geared far more agressively than Mercedes products. And having two extra gears allows the engine to be kept in the peak power zone more efficiently, as there is not as big of a drop off in gear ratios in the upper reaches. Once the E55 et al shift to a seven speed, this advantage will be negated.
Originally Posted by krispykrme
As far as 30+HP. We all know that E55's engine is not rated as mercedes claimed 469 HP. It was simply a marketing figure. The truth is that W211 engine puts the same amount of HP at the crank as any other 55 (close to 500). So you are talking around 10 HP difference. Not to mention that E55 has way more torque than M5 (which is espeically important for acceleration runs in automatic).
Originally Posted by krispykrme
As far as who is faster. That E55 in the 2 trailer had one car length jump on the M5. The M5 was able to overtake the E55 with 3 people on board (probably extra 300 lb easily- i.e. take away the M5's weight advantage). It pulled over a car length ahead in rolling start.
This is a rolling start. E55 should have even a stronger advantage due to its massive torque.
This is a rolling start. E55 should have even a stronger advantage due to its massive torque.
BMW 5 series M5 (E60 (05-)) 3.99/5.5 mph 2.65/8.3 mph 1.81/12.2 mph 1.39/15.9 mph 1.16/19.1 mph 1.00/22.1 mph
Mercedes E class E55 AMG (211 (02 - )) 3.59/7.9 mph 2.19/13.0 mph 1.41/20.1 mph 1.00/28.4 mph 0.83/34.2 mph
I used 8,000 redline for M5, and 6,500 for E55. Please correct me if I'm wrong, don't have either one in front of me and am going from memory, but they're close.
M5:
mph/1000 5.5 8.3 12.2 15.9 19.1 22.1
max spd 44.0 66.4 97.6 127.2 152.8 176.8
E55:
mph/1000 7.9 13.0 20.1 28.4 34.2
max spd 51.4 84.5 130.7 184.6 222.3
So, look at what happens: first is close, second is a bit further, but third and up is quite dramatic. Ratio-wise (M5/E55, giving a percentage of torque multiplication per gear), we have:
1st: 0.69
2nd: 0.64
3rd: 0.61
4th: 0.56 <-- looky here!
5th: 0.56
So, The discrepency in gearing does grow larger at higher speeds.
Interestingly enough, dividing the M5's max torque of 370 (memory again) by E55's max torque of (memory again) 550 lb-ft gives 67%. So, as long as the gearing is such that the E55's gearing has 67%, acceleration should be pretty equal. Once this drops off, surprise surprise surprise, physics still works, and the M5 will start to enjoy a torque multiplication advantage. This doesn't become pronounced until fourth gear, which does gel with what we see on the video.
But guess what: look at the spacings on the seven speed Benz auto and see how they compare. Advantage = gone. Enjoy it while it lasts.
Originally Posted by krispykrme
But at least the trailer did prove without a doubt that the new M5 is faster than E55 in stock form.
But yes, based upon its gearing, the M has an advantage in torque multiplication at all speeds, but particularly above 100 where percentage tips to one greater than engine torque discrepancy can cover. But it also has a disadvantage aerodynamically. I do suspect that at triple digit speeds above 130, it will enjoy an advantage, but would be surprised if actual test results show a large gap.
Finally, the tested 0-125 times for these cars are only 0.8 apart. Per my earlier posts, I have not seen one instance, even in comparisons of cars faster than these, wherein a car with an 0.8 advantage in 0-125 was more than 0.1 or 0.2 faster in the 1/4 mile, and in many cases not even that. But again, we will see when tests come out....which hopefully won't be long, but BMW seems in no hurry to release the M stateside.
![nix](https://mbworld.org/forums/images/smilies/nixweiss.gif)
The reason the thread will not die is because certain people insist in seeing this test not as an unscientific test run with unknown conditions inside the E55 (and M5), but as The Absolute Final Last Word on the subject. It is not, and I and many others will be awaiting instrumented test results on time to distance testing, as the time to speed testing posted so far simply does not point to a huge time-distance advantage for the M.
We shall see. Like you, I am an Engineer, and I prefer objective instrumented data generated under controlled conditions. If the M comes out faster under such conditions, fine, but for the aforementioned reasons I am unwilling to make a sweeping general statement for all data points based upon a sample size of one.
Obviously, the M's time-speed data are irrefutable at high speeds, but I'm doubting that the time-distance will be significant, and am betting that 1/4 mile speeds will be a tenth or two apart max, if that.
We shall see...
#144
Super Member
Originally Posted by Improviz
But guess what: look at the spacings on the seven speed Benz auto and see how they compare. Advantage = gone. Enjoy it while it lasts.
Originally Posted by Improviz
The reason the thread will not die is because certain people insist in seeing this test not as an unscientific test run with unknown conditions inside the E55 (and M5), but as The Absolute Final Last Word on the subject. It is not, and I and many others will be awaiting instrumented test results on time to distance testing, as the time to speed testing posted so far simply does not point to a huge time-distance advantage for the M.
We shall see. Like you, I am an Engineer, and I prefer objective instrumented data generated under controlled conditions. If the M comes out faster under such conditions, fine, but for the aforementioned reasons I am unwilling to make a sweeping general statement for all data points based upon a sample size of one.
Obviously, the M's time-speed data are irrefutable at high speeds, but I'm doubting that the time-distance will be significant, and am betting that 1/4 mile speeds will be a tenth or two apart max, if that.
We shall see...
We shall see. Like you, I am an Engineer, and I prefer objective instrumented data generated under controlled conditions. If the M comes out faster under such conditions, fine, but for the aforementioned reasons I am unwilling to make a sweeping general statement for all data points based upon a sample size of one.
Obviously, the M's time-speed data are irrefutable at high speeds, but I'm doubting that the time-distance will be significant, and am betting that 1/4 mile speeds will be a tenth or two apart max, if that.
We shall see...
There are just not enough data for anyone to make a conclusion other than Gustav's drive in the M5 was very fast as compare to the E55 that day.
#145
Senior Member
Join Date: Oct 2002
Location: Norway
Posts: 404
Likes: 0
Received 0 Likes
on
0 Posts
ALPINA B12 5,7 Coupe #22/57
Originally Posted by Improviz
Then perhaps you were ill advised to argue about aerodynamics helping the vehicle out against the E55 at higher speeds.
![thumbs](https://mbworld.org/forums/images/smilies/thumbsup.gif)
Again: the only people arguing that mass is insignificant here are you and Gustav. The equations show otherwise.
The only thing I said about this was
"So in low speed like the 0-60 sprint the air resistance in of no practical importance, only weight to power ratio ( and gearing, traction...)
but from high speed like 155 mph it is off huge importance and weight are getting a relativly lesser and lessser important factor.
Cant see what was wrong with that.
I as I said, I did not use this as an argument to defend or support the contest E55 vs M5. It was only a general observation....
#146
Originally Posted by Erik
The only thing I said about this was
"So in low speed like the 0-60 sprint the air resistance in of no practical importance, only weight to power ratio ( and gearing, traction...)
but from high speed like 155 mph it is off huge importance and weight are getting a relativly lesser and lessser important factor.
"So in low speed like the 0-60 sprint the air resistance in of no practical importance, only weight to power ratio ( and gearing, traction...)
but from high speed like 155 mph it is off huge importance and weight are getting a relativly lesser and lessser important factor.
Originally Posted by Erik
Well that is off course correct, at least if its done in a vacum, NO air.
At increasing speed the mass is getting less and less important since the major recisting factor is the air resistance, but since you obviously are an expert in the area you already know that.
Another thing is, the Cw value is only a factor telling us how slippery the body/shape is, to get the air resistance you have to multply with the Area to puch thrue the air.
So in low speed like the 0-60 sprint the air resistance in of no practical importance, only weight to power ratio ( and gearing, traction...)
but from high speed like 155 mph it is off huge importance and weight are getting a realivly lesser and lessser important factor.
And why dont you stop the stupide Gustav bashing, it does not give you much credit.
At increasing speed the mass is getting less and less important since the major recisting factor is the air resistance, but since you obviously are an expert in the area you already know that.
Another thing is, the Cw value is only a factor telling us how slippery the body/shape is, to get the air resistance you have to multply with the Area to puch thrue the air.
So in low speed like the 0-60 sprint the air resistance in of no practical importance, only weight to power ratio ( and gearing, traction...)
but from high speed like 155 mph it is off huge importance and weight are getting a realivly lesser and lessser important factor.
And why dont you stop the stupide Gustav bashing, it does not give you much credit.
Originally Posted by Erik
Cant see what was wrong with that.
![nix](https://mbworld.org/forums/images/smilies/nixweiss.gif)
Although, the E55's Cd is 10% better, so maybe it will offset some of that gearing advantage.
![thumbs](https://mbworld.org/forums/images/smilies/thumbsup.gif)
Originally Posted by Erik
I as I said, I did not use this as an argument to defend or support the contest E55 vs M5. It was only a general observation....
#147
Originally Posted by krispykrme
I don't think is a typo.
Typicall car test are not done based on a single run. It may appear that the test result is inconsistent. But it could be that one run is done when driver shifted a better point, the tire had better grip, may be wind was a factor. Or simply the measuring equipment also made a variation in test result.
No matter how many order of polynomial is used. When you are fitting data to a equation, there are always error. Simple as that. Equation is good as it give us an good insight to an objects behavior in a comman language we can understand. But by nature it is an approximination.
When magazine set up to test, they usually test different objective one at a time. Not all at once. Hence, each result i.e. 0-100, 0-110, 0-120, 0-119 could be results of seperate test. Hence each results are indep and can't be duplicate as I have stated earlier the control in this experiment is not perfect.
In that particular test, I find it's highly likely that the result would be a 119 not 117. Simply because you and i can't 100% certain that 0-119 and 0-120 test results are from the same test.
This is a car driven by human. Even in a machine controlled testing environment you are going to have variation.
In this case, the variation in data points that you have used in your equation is more than the result you are claiming. Hence, in statistical term, your conclusion can't be supported.
Unless you have thousands of data points, in which you have identical control in the experiment, you really can't use data points from different test with different car samples, different control (wind, driver.....) to make arrive to a conclusion. Especially the variation in data point inputed is greater then the result.
I am not saying the concept behind your post is wrong. It's just the way the data are inputed in your model is wrong.
Typicall car test are not done based on a single run. It may appear that the test result is inconsistent. But it could be that one run is done when driver shifted a better point, the tire had better grip, may be wind was a factor. Or simply the measuring equipment also made a variation in test result.
No matter how many order of polynomial is used. When you are fitting data to a equation, there are always error. Simple as that. Equation is good as it give us an good insight to an objects behavior in a comman language we can understand. But by nature it is an approximination.
When magazine set up to test, they usually test different objective one at a time. Not all at once. Hence, each result i.e. 0-100, 0-110, 0-120, 0-119 could be results of seperate test. Hence each results are indep and can't be duplicate as I have stated earlier the control in this experiment is not perfect.
In that particular test, I find it's highly likely that the result would be a 119 not 117. Simply because you and i can't 100% certain that 0-119 and 0-120 test results are from the same test.
This is a car driven by human. Even in a machine controlled testing environment you are going to have variation.
In this case, the variation in data points that you have used in your equation is more than the result you are claiming. Hence, in statistical term, your conclusion can't be supported.
Unless you have thousands of data points, in which you have identical control in the experiment, you really can't use data points from different test with different car samples, different control (wind, driver.....) to make arrive to a conclusion. Especially the variation in data point inputed is greater then the result.
I am not saying the concept behind your post is wrong. It's just the way the data are inputed in your model is wrong.
#148
Originally Posted by krispykrme
I don't think is a typo.
Typicall car test are not done based on a single run. It may appear that the test result is inconsistent. But it could be that one run is done when driver shifted a better point, the tire had better grip, may be wind was a factor. Or simply the measuring equipment also made a variation in test result.
When magazine set up to test, they usually test different objective one at a time. Not all at once. Hence, each result i.e. 0-100, 0-110, 0-120, 0-119 could be results of seperate test. Hence each results are indep and can't be duplicate as I have stated earlier the control in this experiment is not perfect.
In that particular test, I find it's highly likely that the result would be a 119 not 117. Simply because you and i can't 100% certain that 0-119 and 0-120 test results are from the same test.
Typicall car test are not done based on a single run. It may appear that the test result is inconsistent. But it could be that one run is done when driver shifted a better point, the tire had better grip, may be wind was a factor. Or simply the measuring equipment also made a variation in test result.
When magazine set up to test, they usually test different objective one at a time. Not all at once. Hence, each result i.e. 0-100, 0-110, 0-120, 0-119 could be results of seperate test. Hence each results are indep and can't be duplicate as I have stated earlier the control in this experiment is not perfect.
In that particular test, I find it's highly likely that the result would be a 119 not 117. Simply because you and i can't 100% certain that 0-119 and 0-120 test results are from the same test.
Originally Posted by Car & Driver
Our sophisticated VBOX system doesn’t use accelerometers, precisely to avoid this problem. Instead, the VBOX relies on the global-positioning system (GPS) to measure a vehicle’s motion. To work, the VBOX must be in communication with at least four of 24 GPS satellites orbiting the earth. As a vehicle travels down the test venue, there’s a minuscule shift in the arrival time of the radio signals that travel between the satellites and the VBOX’s antenna. By measuring this shift, the VBOX calculates speed, acceleration, and distance. Racelogic, the company that makes the VBOX, asserts that its devices are accurate to within 0.06 mph.
Originally Posted by Racelogic's site
Report Generator
We have just released a powerful stand alone software module for the VBOX called the "Report Generator".
This software is designed to produce custom table based text reports from VBOX data files or live on a laptop. The table contents are fully configurable and can show any channel reorded by the VBOX.
Up to 34 columns can be configured to show the parameters, and also you can pick to show the maximums, minimums and averages. The scale factors and offsets can be configured, and you can also pick a number of standard parameters like stopping distances, MFDD, lateral accel, longitudinal accel etc.
There is a comprehensive range of trigger conditions for starting, stopping and creating a line, using speed, distance, time, trigger inputs or virtual (position based) gates. (A useful bonus feature allows you to convert a vbox file into a file that can be loaded directly into Autoroute for plotting the vehicle's journey)
Registered users can download this software from the Downloads section of this website, so if you haven't bought that second years support yet, here's another very good reason!
We have just released a powerful stand alone software module for the VBOX called the "Report Generator".
This software is designed to produce custom table based text reports from VBOX data files or live on a laptop. The table contents are fully configurable and can show any channel reorded by the VBOX.
Up to 34 columns can be configured to show the parameters, and also you can pick to show the maximums, minimums and averages. The scale factors and offsets can be configured, and you can also pick a number of standard parameters like stopping distances, MFDD, lateral accel, longitudinal accel etc.
There is a comprehensive range of trigger conditions for starting, stopping and creating a line, using speed, distance, time, trigger inputs or virtual (position based) gates. (A useful bonus feature allows you to convert a vbox file into a file that can be loaded directly into Autoroute for plotting the vehicle's journey)
Registered users can download this software from the Downloads section of this website, so if you haven't bought that second years support yet, here's another very good reason!
Can you elaborate?
#149
Impro-man, I'm open to discussion on the subject but fro my experiences I find a car with a wide torque plateau & power drop-off at high rpm doesn't benefit from shorter (or more) gears as much as a car with high rpm power.
Reason being its lways better to stay in a lower gear if your curve is flat. A ower gear multiplies torque by a larger amount than a higher gear. Let's say 2nd gear multiples torque by 17% more than 3rd gear. If your curve is flat its always better to stay in the lower gear as you are getting more torque to the wheels. When you go to a higher gear you are getting less torque to the wheels.
But a car with high rpm power can use shirt gearing to keep the revs hig between shifts so as not to drop off the powerband. A big V8 or roots type SC makes peak power long before redline, so gearing it to stay at high revs will make it slower. If I get the time I will illustrate this with figures.
Reason being its lways better to stay in a lower gear if your curve is flat. A ower gear multiplies torque by a larger amount than a higher gear. Let's say 2nd gear multiples torque by 17% more than 3rd gear. If your curve is flat its always better to stay in the lower gear as you are getting more torque to the wheels. When you go to a higher gear you are getting less torque to the wheels.
But a car with high rpm power can use shirt gearing to keep the revs hig between shifts so as not to drop off the powerband. A big V8 or roots type SC makes peak power long before redline, so gearing it to stay at high revs will make it slower. If I get the time I will illustrate this with figures.
#150
Originally Posted by Compensated BMW advocate M&M
Impro-man, I'm open to discussion on the subject but fro my experiences I find a car with a wide torque plateau & power drop-off at high rpm doesn't benefit from shorter (or more) gears as much as a car with high rpm power.
In the 60's, American cars which redlined at a whopping 5,000 rpm routinely performed better at drag strips by the simple substitution of a lower rearend.
Originally Posted by Compensated BMW advocate M&M
Reason being its lways better to stay in a lower gear if your curve is flat. A ower gear multiplies torque by a larger amount than a higher gear. Let's say 2nd gear multiples torque by 17% more than 3rd gear. If your curve is flat its always better to stay in the lower gear as you are getting more torque to the wheels. When you go to a higher gear you are getting less torque to the wheels.
Originally Posted by Compensated BMW advocate M&M
But a car with high rpm power can use shirt gearing to keep the revs hig between shifts so as not to drop off the powerband. A big V8 or roots type SC makes peak power long before redline, so gearing it to stay at high revs will make it slower.
Originally Posted by Compensated BMW advocate M&M
If I get the time I will illustrate this with figures.
![nix](https://mbworld.org/forums/images/smilies/nixweiss.gif)
What makes for optimum acceleration is to have gearing which keeps the engine in its peak power band. This holds true for cars which redline at 4000 rpm, 6500 rpm or 8200 rpm, because the integral of the power curve is what counts. The M makes less peak torque but over a wider range, whereas the E55 makes a higher peak over a lesser rpm range.
The difference is that the torque multiplication of the M's extra gears enjoys a 10% advantage in fourth and fifth gears. This is substantial, but the upcoming seven speed auto in the AMGs will alleviate this. See the current Car & Driver for a discussion of how the seven speed keeps the engine in its peak power range better than the previous five speed and dramatically improves acceleration, or compare the acceleration times for the five speed S500 with the seven speed S500. The car picked up nearly a half second in 0-60 and 1/4 mile times compared to the five speed.