Trail Braking Revisited
08-05-2004, 07:05 PM
#31
Still plays with cars.
Lifetime Rennlist
Member
"Bob? What has contradicted my instincts? This article fully supports what I have been saying, at least I think so (nicer attitude offered for you)."
CC - thanks for the "nice". I haven't found the magazine yet so I can't comment on the article. My reference applies to the original post which began the debate - the one that espoused "trail brake with full ABS to the apex". I think we're past that now and please rememberI was careful to refernce everything to a 911. There are some cars which want to be trail braked to the apex in many turns (not just slow tight ones) but any 911 I've driven does not fall into that category. One think that is apparent for these discussions is what trail braking really is. I was careful to define it an NOT being the act of turning in as you fade off the brakes - that's the right way to initiate a turn anytime. Some people call *that* trail braking. I don't.
In sum it's nice to see you back Color, you always stimulate thought and lively discussions.
Regards,
CC - thanks for the "nice". I haven't found the magazine yet so I can't comment on the article. My reference applies to the original post which began the debate - the one that espoused "trail brake with full ABS to the apex". I think we're past that now and please rememberI was careful to refernce everything to a 911. There are some cars which want to be trail braked to the apex in many turns (not just slow tight ones) but any 911 I've driven does not fall into that category. One think that is apparent for these discussions is what trail braking really is. I was careful to define it an NOT being the act of turning in as you fade off the brakes - that's the right way to initiate a turn anytime. Some people call *that* trail braking. I don't.
In sum it's nice to see you back Color, you always stimulate thought and lively discussions.
Regards,
08-05-2004, 09:39 PM
#32
Addict
Rennlist Member
Rennlist Member
Thread Theft;
Huh? Turns are just crooked straights to me. Can't take 'em as fast, but I just add 'em to the boring stuff that comes after them!
Huh? OK... you do that. Assuming we have similar HP, If I'm behind you and I time it right so I don't have to come to a complete stop to keep from hitting you in the Inner Loop, I'll pass you going into 6... EASILY! 
Good drivers know that fast laps come from maximizing speed on the straights not speed through the turns.
At the Glen I trail brake hard all the way through the first part of the bus stop which allows me to brake at the 300 mark rather than the 400 mark. Exit speed is meaningless here.
08-05-2004, 11:16 PM
#33
Three Wheelin'
Thread Starter
Join Date: Aug 2003
Posts: 1,686
Likes: 0
Received 0 Likes
on
0 Posts
Bob, I was much more careful than you on your definition of trail braking with my theoreitcal argument of ABS jam trail braking and trail braking/fc/g-max/gsum argument that all went to hell and it looks like it partly still is.
I will provide data this weekend to see how the car does with jammed abs turning (PSM off of course - but it may re-engage) and we can compare segment times to see if P abs is up to the task. I am also selective on who I reply to so this appears to create much less friction, ie. it takes two to tangle.
Bob, when you trail brake on your GT-2 (correct?), do you have to do it much less than mid or front engine cars because the rear stepps out quite easily (in comparison), and then do you modulate the throttle around the apex to control attitude? This is what I have been experimenting with and I don't know if it is good/fast yet. Comments?
I will provide data this weekend to see how the car does with jammed abs turning (PSM off of course - but it may re-engage) and we can compare segment times to see if P abs is up to the task. I am also selective on who I reply to so this appears to create much less friction, ie. it takes two to tangle.
Bob, when you trail brake on your GT-2 (correct?), do you have to do it much less than mid or front engine cars because the rear stepps out quite easily (in comparison), and then do you modulate the throttle around the apex to control attitude? This is what I have been experimenting with and I don't know if it is good/fast yet. Comments?
08-05-2004, 11:29 PM
#34
Three Wheelin'
Thread Starter
Join Date: Aug 2003
Posts: 1,686
Likes: 0
Received 0 Likes
on
0 Posts
Sunday, I did have you on my ignore list but I will try this once and we'll see.
I can drive a line at neutral throttle and the max lateral g's.
OK
I can then go back and drive the exact same line, but a bit slower while acceleratiing.
No. That additional acceleration will change the line later through the turn. Think about a vector. If you are on the fc, the vector is exactly defined by the line. Any change in vector (lat or long g) will change the line of the car (again, when on the fc). Does it make sense, any difference in vector from the original line pushes the car in a slightly different direction and therefore a different line. If you do a drawing it might help.
I am on the friction circle in both cases and on the same line in both cases.
No. Not the same line through the whole turn.
There are multiple solutions as described above because I have an intinite set of possibilities between lateral and longitudinal g's.
Yes, but not when you are on 1 line and on the fc.
I think the other posters to this thread understand that and are also questioning you about it.
Nope. If they question it they are most likely similarly confused.
Just for the record, I have created the integral of g sum formula in my das and it does works but you have to throw in the caveot of no unnecessary g creation. I really blew a turn on one lap and even though it made me slightly slower than the faster lap, my g-sum was barely higher and this is wrong. When I don’t make major mistakes, it works perfectly (so far).
I can drive a line at neutral throttle and the max lateral g's.
OK
I can then go back and drive the exact same line, but a bit slower while acceleratiing.
No. That additional acceleration will change the line later through the turn. Think about a vector. If you are on the fc, the vector is exactly defined by the line. Any change in vector (lat or long g) will change the line of the car (again, when on the fc). Does it make sense, any difference in vector from the original line pushes the car in a slightly different direction and therefore a different line. If you do a drawing it might help.
I am on the friction circle in both cases and on the same line in both cases.
No. Not the same line through the whole turn.
There are multiple solutions as described above because I have an intinite set of possibilities between lateral and longitudinal g's.
Yes, but not when you are on 1 line and on the fc.
I think the other posters to this thread understand that and are also questioning you about it.
Nope. If they question it they are most likely similarly confused.
Just for the record, I have created the integral of g sum formula in my das and it does works but you have to throw in the caveot of no unnecessary g creation. I really blew a turn on one lap and even though it made me slightly slower than the faster lap, my g-sum was barely higher and this is wrong. When I don’t make major mistakes, it works perfectly (so far).
08-05-2004, 11:39 PM
#35
Rennlist Member
Join Date: Dec 2001
Location: Houston, TX
Posts: 537
Likes: 0
Received 0 Likes
on
0 Posts
I shouldn't have to post this but the mathematical ignorance here is killing me.
[QUOTE]
I can drive a line at neutral throttle and the max lateral g's.
I can then go back and drive the exact same line, but a bit slower while acceleratiing.
I am on the friction circle in both cases and on the same line in both cases.
There are multiple solutions as described above because I have an intinite set of possibilities between lateral and longitudinal g's.
I think the other posters to this thread understand that and are also questioning you about it.
[QUOTE]
If you have a given line, and are maximizing G forces on the friction circle and start the line at the same speed each time, there is only one way to do it. Ever taken a basic calculus course and seen how if X and Y axis acceleration are defined in terms of t (time) that you can determine where an object is along a curve at a given time? Same conecept...if you don't get it now then there is no hope. If you change one of those acceleration parameters, your curve changes! Simple as that.
In your second case where you drive the same line but slower, you already blew the concept of equal entry speeds. Equal entry speed is essential for this to work. If you look back through the posts, Color specifically that the corner entry speed must be the same to work.
[QUOTE]
I can drive a line at neutral throttle and the max lateral g's.
I can then go back and drive the exact same line, but a bit slower while acceleratiing.
I am on the friction circle in both cases and on the same line in both cases.
There are multiple solutions as described above because I have an intinite set of possibilities between lateral and longitudinal g's.
I think the other posters to this thread understand that and are also questioning you about it.
[QUOTE]
If you have a given line, and are maximizing G forces on the friction circle and start the line at the same speed each time, there is only one way to do it. Ever taken a basic calculus course and seen how if X and Y axis acceleration are defined in terms of t (time) that you can determine where an object is along a curve at a given time? Same conecept...if you don't get it now then there is no hope. If you change one of those acceleration parameters, your curve changes! Simple as that.
In your second case where you drive the same line but slower, you already blew the concept of equal entry speeds. Equal entry speed is essential for this to work. If you look back through the posts, Color specifically that the corner entry speed must be the same to work.
08-06-2004, 12:29 AM
#36
Lifetime Rennlist Member
Originally Posted by FormulaOne10
I shouldn't have to post this but the mathematical ignorance here is killing me.
If you have a given line, and are maximizing G forces on the friction circle and start the line at the same speed each time, there is only one way to do it. Ever taken a basic calculus course and seen how if X and Y axis acceleration are defined in terms of t (time) that you can determine where an object is along a curve at a given time? Same conecept...if you don't get it now then there is no hope. If you change one of those acceleration parameters, your curve changes! Simple as that.
In your second case where you drive the same line but slower, you already blew the concept of equal entry speeds. Equal entry speed is essential for this to work. If you look back through the posts, Color specifically that the corner entry speed must be the same to work.
If you have a given line, and are maximizing G forces on the friction circle and start the line at the same speed each time, there is only one way to do it. Ever taken a basic calculus course and seen how if X and Y axis acceleration are defined in terms of t (time) that you can determine where an object is along a curve at a given time? Same conecept...if you don't get it now then there is no hope. If you change one of those acceleration parameters, your curve changes! Simple as that.
In your second case where you drive the same line but slower, you already blew the concept of equal entry speeds. Equal entry speed is essential for this to work. If you look back through the posts, Color specifically that the corner entry speed must be the same to work.
Maybe some clarification is in order too. I thought this was about finding the fastest way around the track, not drivng on the fc on a given line. Of course there are going to be small changes in location/line if you get on the throttle, but msot people would take 'same line' to mean within a few inches or fractions of an inch, not within 0.00001 inch.
Seems to me if you constrain the problem to "Max entry speed, set turn in point and line and ride the FC" that yes, you have defined a unique set of parameters and there is only one solution. Buy WTF does that have to do with the fastest way around a track.
And if you want to demonstrate the mathematical correctness of your argument, how about using math instead of stating that I am math ignorant. Us mechanical engineers are pretty dumb but I just might be able to follow along some fancy calculus and dynamics if you type it out real slow.
08-06-2004, 12:32 AM
#37
Lifetime Rennlist Member
Originally Posted by ColorChange
Sunday, I did have you on my ignore list but I will try this once and we'll see.
I can drive a line at neutral throttle and the max lateral g's.
OK
I can then go back and drive the exact same line, but a bit slower while acceleratiing.
No. That additional acceleration will change the line later through the turn. Think about a vector. If you are on the fc, the vector is exactly defined by the line. Any change in vector (lat or long g) will change the line of the car (again, when on the fc). Does it make sense, any difference in vector from the original line pushes the car in a slightly different direction and therefore a different line. If you do a drawing it might help.
I am on the friction circle in both cases and on the same line in both cases.
No. Not the same line through the whole turn.
There are multiple solutions as described above because I have an intinite set of possibilities between lateral and longitudinal g's.
Yes, but not when you are on 1 line and on the fc.
I think the other posters to this thread understand that and are also questioning you about it.
Nope. If they question it they are most likely similarly confused.
Just for the record, I have created the integral of g sum formula in my das and it does works but you have to throw in the caveot of no unnecessary g creation. I really blew a turn on one lap and even though it made me slightly slower than the faster lap, my g-sum was barely higher and this is wrong. When I don’t make major mistakes, it works perfectly (so far).
I can drive a line at neutral throttle and the max lateral g's.
OK
I can then go back and drive the exact same line, but a bit slower while acceleratiing.
No. That additional acceleration will change the line later through the turn. Think about a vector. If you are on the fc, the vector is exactly defined by the line. Any change in vector (lat or long g) will change the line of the car (again, when on the fc). Does it make sense, any difference in vector from the original line pushes the car in a slightly different direction and therefore a different line. If you do a drawing it might help.
I am on the friction circle in both cases and on the same line in both cases.
No. Not the same line through the whole turn.
There are multiple solutions as described above because I have an intinite set of possibilities between lateral and longitudinal g's.
Yes, but not when you are on 1 line and on the fc.
I think the other posters to this thread understand that and are also questioning you about it.
Nope. If they question it they are most likely similarly confused.
Just for the record, I have created the integral of g sum formula in my das and it does works but you have to throw in the caveot of no unnecessary g creation. I really blew a turn on one lap and even though it made me slightly slower than the faster lap, my g-sum was barely higher and this is wrong. When I don’t make major mistakes, it works perfectly (so far).
The real point that I was making, as were others, is that you have a choice to apply throttle earlier or later on essentially the same line. One way will be faster then another. Is that what all this analysis is trying to accomplish or is something other than fastest lap or winning races the measure of success?
08-06-2004, 01:00 AM
#38
CC,
I'm not sure what it means for your theory, but let's realize the fact that the "friction circle" looks nothing like a circle. I'm willing to guess that the max amount of lat G's that you can pull is more than the amount of long G's that you can pull (hard to tell with some of those monster cars you own).
So, if we were to believe that we wanted to maximize G's, we would end up concluding that we wanted to corner under neutral throttle for as long as possible. I think even the most basic DE participant realizes that you want to get on the gas as soon as he can.
I'm not sure what it means for your theory, but let's realize the fact that the "friction circle" looks nothing like a circle. I'm willing to guess that the max amount of lat G's that you can pull is more than the amount of long G's that you can pull (hard to tell with some of those monster cars you own).
So, if we were to believe that we wanted to maximize G's, we would end up concluding that we wanted to corner under neutral throttle for as long as possible. I think even the most basic DE participant realizes that you want to get on the gas as soon as he can.
08-06-2004, 01:08 AM
#39
Rennlist Member
Join Date: Dec 2001
Location: Houston, TX
Posts: 537
Likes: 0
Received 0 Likes
on
0 Posts
Honestly, I have read the article in question and am not interesting in explaining it all. If you are interested in reading it, its in the latest issue of Racecar Engineering. I only wish to clarify the point about a single discrete way of negotiating a line while on the friction circle.
That's terrific that you are a mechanical engineer because I am too. I never singled you out as specifically ignorant and didn't mean to offend you, and you probably do know what I am saying...but many here are not technically oriented. I thought this was somewhat basic but...here goes some "fancy" calculus.
Say you have a point at rest at the origin of an x-y axis. It is undergoing an acceleration of 5t in the X direction and 3t^2 in the Y direction (totally random values). After a certain length of time you can determine exactly where that point is (by integrating over your time interval). More importantly, there is more than one way to get to that point, but not along that same line with different accelerations. The same concept applies to cars.
And if you want to demonstrate the mathematical correctness of your argument, how about using math instead of stating that I am math ignorant. Us mechanical engineers are pretty dumb but I just might be able to follow along some fancy calculus and dynamics if you type it out real slow.
Say you have a point at rest at the origin of an x-y axis. It is undergoing an acceleration of 5t in the X direction and 3t^2 in the Y direction (totally random values). After a certain length of time you can determine exactly where that point is (by integrating over your time interval). More importantly, there is more than one way to get to that point, but not along that same line with different accelerations. The same concept applies to cars.
08-06-2004, 09:31 AM
#40
Three Wheelin'
Thread Starter
Join Date: Aug 2003
Posts: 1,686
Likes: 0
Received 0 Likes
on
0 Posts
Brian, It doesn't matter that the fc is not a circle (more of a heart shape). What matters, is that being ON the fc is a very specific point for a given set of conditions on the car, specifically, a vector, (direction and magnitude) at a certain attitude.
Yes, that lat g's are lower than the max long g's, especially in the tt because of the rear weight bias helps in braking, but again, the shape doesn't matter because the shape is very similar for a specific car and specific tires. The fc for that car changes very little as most of us don't have downforce cars, but even in a df car, once you specify the line and you say you are going to drive the line on the fc, there is one and only one way to do that for any car.
So, if we were to believe that we wanted to maximize G's, we would end up concluding that we wanted to corner under neutral throttle for as long as possible.
No, this would be maximizing lateral g's. We want to maximize g sum (the combination of lat and long g's in the approriate direction.
Finally, if you read the article, they specifically make the case that the fastest line does not have the maximium corner speed. This is made up for by the speed gain through trail braking, and then the speed gain from earlier, larger acceleration.
Hope this helps.
Yes, that lat g's are lower than the max long g's, especially in the tt because of the rear weight bias helps in braking, but again, the shape doesn't matter because the shape is very similar for a specific car and specific tires. The fc for that car changes very little as most of us don't have downforce cars, but even in a df car, once you specify the line and you say you are going to drive the line on the fc, there is one and only one way to do that for any car.
So, if we were to believe that we wanted to maximize G's, we would end up concluding that we wanted to corner under neutral throttle for as long as possible.
No, this would be maximizing lateral g's. We want to maximize g sum (the combination of lat and long g's in the approriate direction.
Finally, if you read the article, they specifically make the case that the fastest line does not have the maximium corner speed. This is made up for by the speed gain through trail braking, and then the speed gain from earlier, larger acceleration.
Hope this helps.
08-06-2004, 10:10 AM
#41
Three Wheelin'
Thread Starter
Join Date: Aug 2003
Posts: 1,686
Likes: 0
Received 0 Likes
on
0 Posts
See some of what I wrote above. I will give you that throttle application will make a very slight change in line, but I can steer to counteract those forces and get very close to the same line. If we are talking about the real world of how to go around a track the fastest, then fractions of an inch difference in line do not matter.
Again I disagree. If you are on the fc and have slightly higher acceleration, you can't turn and maintain the same line because you would skid. If you have a significantly different acceleration, and you are on the fc, you will take a significantly different line. What I think you are saying in the real world is that most drivers are not on the fc so that they can change their g sum and still stay on the same line. This is true, but if your goal is driving the fastest, you have to be on or near the fc. The closer you are to the fc, the less chance you have to vary your g’s along your chosen line.
The real point that I was making, as were others, is that you have a choice to apply throttle earlier or later on essentially the same line.
Only when you are not on the fc, and most drivers aren’t.
One way will be faster then another. Is that what all this analysis is trying to accomplish or is something other than fastest lap or winning races the measure of success?
Let’s just stick with fastest laps for simplification. The goal of the fastest lap is to maximize g’s in the appropriate direction. This is exactly what the article says.
Again I disagree. If you are on the fc and have slightly higher acceleration, you can't turn and maintain the same line because you would skid. If you have a significantly different acceleration, and you are on the fc, you will take a significantly different line. What I think you are saying in the real world is that most drivers are not on the fc so that they can change their g sum and still stay on the same line. This is true, but if your goal is driving the fastest, you have to be on or near the fc. The closer you are to the fc, the less chance you have to vary your g’s along your chosen line.
The real point that I was making, as were others, is that you have a choice to apply throttle earlier or later on essentially the same line.
Only when you are not on the fc, and most drivers aren’t.
One way will be faster then another. Is that what all this analysis is trying to accomplish or is something other than fastest lap or winning races the measure of success?
Let’s just stick with fastest laps for simplification. The goal of the fastest lap is to maximize g’s in the appropriate direction. This is exactly what the article says.
08-06-2004, 10:38 AM
#42
Rennlist Member
Join Date: Feb 2003
Posts: 2,819
Likes: 0
Received 0 Likes
on
0 Posts
All this fighting is making my head hurt! I am on my way to Homestead tomorrow, and will again see that even the most experienced of us DE folks and even the racers don't enter turns at the correct velocity. So forget the theory, get out and drive better. Only the guys we watch on TV really get close to realizing the full potential of their cars.
Now back to this discussion, I don't remember Tim stating that entry speed has to be set. If you control all the variables then yes, I agree that there is only one solution. Same entry speed, correct direction of the force vector at all times, on the fc, correct line, no unecessary g-force creation etc.... This is the perfect lap, something we all strive for. (although I am sure I left out some other conditions). Theory is interesting, but the reality for most of us is that we can't drive that well. How many of us have the consistancy to get the entry speed right most of the time (that means less than 1 MPH variation)? We are not machines! Well, maybe MS is but none of us are. This theory is interesting and I really enjoy all of Tim's posts, but let's get out and drive and have fun.
Now back to this discussion, I don't remember Tim stating that entry speed has to be set. If you control all the variables then yes, I agree that there is only one solution. Same entry speed, correct direction of the force vector at all times, on the fc, correct line, no unecessary g-force creation etc.... This is the perfect lap, something we all strive for. (although I am sure I left out some other conditions). Theory is interesting, but the reality for most of us is that we can't drive that well. How many of us have the consistancy to get the entry speed right most of the time (that means less than 1 MPH variation)? We are not machines! Well, maybe MS is but none of us are. This theory is interesting and I really enjoy all of Tim's posts, but let's get out and drive and have fun.
08-06-2004, 11:16 AM
#43
Three Wheelin'
This thread is bench racing/driving with eng. degrees. Me, I am going out to drive my car on the track this weekend, fast. Sometimes I will get the lap right, other times I wont. I will have fun!! FC and brake pad wear be dammed!!!!
08-06-2004, 11:31 AM
#44
Three Wheelin'
Thread Starter
Join Date: Aug 2003
Posts: 1,686
Likes: 0
Received 0 Likes
on
0 Posts
Mitch: see yesterdays post where I say "The key here is on the same line. Once you define the line, there is only one way to reach the fc and stay on that line. You can have different throttle or braking points and be on the fc, but these would lead to different lines. I am making the assumption that conrer entry speed and postition are the same."
08-06-2004, 11:49 AM
#45
Redlineman... My point is that if the goal were to maximize speed during the turn everybody would take the "optimal" geometric line... we all know that drivers frequently take a suboptimal line to trade entry speed for exit speed based on where the straights are located.... I didn;t mean to imply that you don;t take the car to the limit in the turn or that turns aren;t fun.
Secondly, divide the Bus Stop into three sections. Entry, which ends at the second (inside) curb. Setup, which ends halfway between the second and third curb, and the Exit. In the Setup section, I must feather the throttle to setup for the exit, then go to full throttle.
The Setup section is basically a throwaway, so I can trail brake hard in the Entry, and then gather the car up in the Setup section wihout losing any exit speed.
BTW you very well may pass me in turn 6 but that is because I;m a pansy, not becasue of how I took the Bus Stop.
Secondly, divide the Bus Stop into three sections. Entry, which ends at the second (inside) curb. Setup, which ends halfway between the second and third curb, and the Exit. In the Setup section, I must feather the throttle to setup for the exit, then go to full throttle.
The Setup section is basically a throwaway, so I can trail brake hard in the Entry, and then gather the car up in the Setup section wihout losing any exit speed.
BTW you very well may pass me in turn 6 but that is because I;m a pansy, not becasue of how I took the Bus Stop.