Notices
Racing & Drivers Education Forum
Sponsored by:
Sponsored by:

Racing Techie question regarding braking performance

Thread Tools
 
Search this Thread
 
Old 12-15-2015 | 05:53 PM
  #61  
morsini's Avatar
morsini
Three Wheelin'
 
Joined: Sep 2009
Posts: 1,347
Likes: 0
From: Bawston
Default

Originally Posted by winders
First off, again, don't lump me in with Kibort.

Second, how is this discussion, in the off season, sucking fun out of the sport???

Thirdly, I'll take my life over yours any day.....
Lol, yup, you've got it made. We're not worthy. Blah, blah, blah.

Yours truly,

Dr. O.
Old 12-15-2015 | 06:11 PM
  #62  
winders's Avatar
winders
Race Car
 
Joined: Oct 2010
Posts: 4,602
Likes: 925
From: San Martin, CA
Default

Originally Posted by morsini
Lol, yup, you've got it made. We're not worthy. Blah, blah, blah.

Yours truly,

Dr. O.
Look, this has nothing to do with being worthy or not. You tried to tell me that I don't know what life is about. Well, you and I have very different ideas about what life is about and I am VERY happy in my life and would not trade it for your life. That is no way suggests that you are not happy in your life. Maybe you are and maybe you are not. All I know is I would not be. So don't try to tell me what my life should be like. You live life the way you want and I will live life the way I want. Pretty simple, isn't it?

Not a Dr. and just fine with that.....
Old 12-15-2015 | 06:34 PM
  #63  
ProCoach's Avatar
ProCoach
Rennlist
Basic Site Sponsor
 
Joined: Jun 2008
Posts: 19,229
Likes: 3,382
From: Durham, NC and Virginia International Raceway
Default

Originally Posted by PorscheG96
Unless you're a materials engineer working in the tire industry, it would be virtually impossible to know the variables necessary for solving Pacejka's magic formula. For those who don't know - and apologies to all who already do know - it's basically a sin of arctan function with about 30 different coefficients and scaling variables. Each one of these variables is unique to each and every brand, make, and size of tire. I think that Pacejka solutions are influenced by internal construction and tire dimensions, as well. Furthermore, the forces generated laterally are almost always different than those generated longitudinally, so it doesn't help to compare cornering force of different masses to braking, etc.

So, without knowing ANY of these magic numbers that tire manufacturers spend hundreds of millions to discover [mostly by persistence and luck], hence working vigorously protect, any calculation would equate to guessing and almost certainly be incorrect. As a proxy, one can collect acceleration data for analysis. Pro Coach is probably the user in best position to some answers, assuming he has a dataset for same pro driver, same day, same car with xxx lbs more ballast.
Bingo.

http://www.edy.es/dev/docs/pacejka-9...hensive-guide/

I studied Pacejka's "Magic Formula" in an effort to tweak behavior on home and higher level simulators, modeling similar to that used on Sports Car (Multimatic) and Formula Car (Toyota, first, then everyone else in F1). This is one of the things that makes the behavior of the sim car more "real."

The National Tire Research Center is right across the street from my office at VIR. Frank, the head guy, is amazing to talk to... http://www.sovamotion.com

I'll sniff around, I am fairly sure I have something, but it will be a few weeks.
__________________
-Peter Krause
www.peterkrause.net
www.gofasternow.com
"Combining the Art and Science of Driving Fast!"
Specializing in Professional, Private Driver Performance Evaluation and Optimization
Consultation Available Remotely and at VIRginia International Raceway






















Old 12-15-2015 | 06:51 PM
  #64  
mark kibort's Avatar
mark kibort
Thread Starter
Rennlist Member
 
Joined: Feb 2003
Posts: 29,956
Likes: 180
From: saratoga, ca
Default

Originally Posted by ProCoach
Bingo.

http://www.edy.es/dev/docs/pacejka-9...hensive-guide/

I studied Pacejka's "Magic Formula" in an effort to tweak behavior on home and higher level simulators, modeling similar to that used on Sports Car (Multimatic) and Formula Car (Toyota, first, then everyone else in F1). This is one of the things that makes the behavior of the sim car more "real."

The National Tire Research Center is right across the street from my office at VIR. Frank, the head guy, is amazing to talk to... http://www.sovamotion.com

I'll sniff around, I am fairly sure I have something, but it will be a few weeks.
Putting all the other crap that is flung back and forth there, this is interesting stuff to me. I dont have the answers and the factors ive listed obviously are loose ends and are major influencers for the answer i was looking for.

anything you can come up with would be very interesting to me, and I'm sure others as well. I'm sure the newer simulators might have algorithms that have the tire load sensitivity factor included. Be interesting to see what that is.

but, for me, i think the answers lie in any data you speak of. can you show two cars with near the same set up and driver, in a decl condition with and without substantial ballast? Be even more interested in the cornering forces. (my own personal interest and I'm sure others as well)

the Pacejka program looks interesting.. there is a factor for longitudinal load variance though!
Old 12-15-2015 | 06:58 PM
  #65  
winders's Avatar
winders
Race Car
 
Joined: Oct 2010
Posts: 4,602
Likes: 925
From: San Martin, CA
Default

Originally Posted by PorscheG96
Unless you're a materials engineer working in the tire industry, it would be virtually impossible to know the variables necessary for solving Pacejka's magic formula. For those who don't know - and apologies to all who already do know - it's basically a sin of arctan function with about 30 different coefficients and scaling variables. Each one of these variables is unique to each and every brand, make, and size of tire. I think that Pacejka solutions are influenced by internal construction and tire dimensions, as well. Furthermore, the forces generated laterally are almost always different than those generated longitudinally, so it doesn't help to compare cornering force of different masses to braking, etc.

So, without knowing ANY of these magic numbers that tire manufacturers spend hundreds of millions to discover [mostly by persistence and luck], hence working vigorously protect, any calculation would equate to guessing and almost certainly be incorrect. As a proxy, one can collect acceleration data for analysis. Pro Coach is probably the user in best position to some answers, assuming he has a dataset for same pro driver, same day, same car with xxx lbs more ballast.
Originally Posted by ProCoach
Bingo.

http://www.edy.es/dev/docs/pacejka-9...hensive-guide/

I studied Pacejka's "Magic Formula" in an effort to tweak behavior on home and higher level simulators, modeling similar to that used on Sports Car (Multimatic) and Formula Car (Toyota, first, then everyone else in F1). This is one of the things that makes the behavior of the sim car more "real."

I'll sniff around, I am fairly sure I have something, but it will be a few weeks.
All of this is why I did not attempt to put numbers to the difference when weight alone is changed.

It is clear that the "tire load sensitivity" phenomenon is present in lateral and longitudinal acceleration of a vehicle. In other words, when cornering and braking. What's not clear is how much weight makes a measurable difference and in what situations would that be apparent.

The data that Peter has from all of his data analysis gigs is probably not useful in determining the "tire load sensitivity" for any specific tires or tires in general. Why? Because the vehicles are dissimilar enough and the drivers are dissimilar enough to make a valid comparison impossible. Unless, of course, he did some very specific testing removing or adding weight to a car with a driver able to get 10/10ths out of the car. Even then, the weight change may have had an effect on the cars handling by changing the weight distribution or cg.

Without extensive testing I think all we can say for sure is that, all other things being equal, making a car lighter will improve its cornering and braking. The more weight that is removed the bigger the improvement.
Old 12-15-2015 | 07:26 PM
  #66  
onefastviking's Avatar
onefastviking
Race Car
 
Joined: Apr 2008
Posts: 3,549
Likes: 2
From: Texas
Default

Originally Posted by winders

Without extensive testing I think all we can say for sure is that, all other things being equal, making a car lighter will improve its cornering and braking. The more weight that is removed the bigger the improvement.
But that condradicts all the calculations and what MK says, braking should be the same regardless of weight.
I think extensive testing should be done.
Not by you though Winders, you might be biased. Not by me or PK either as we both too seem biased.
Hmmmm, what to do.......
Old 12-15-2015 | 07:28 PM
  #67  
ir_fuel's Avatar
ir_fuel
Drifting
 
Joined: May 2010
Posts: 2,168
Likes: 5
From: Belgium
Default

Originally Posted by ProCoach
I look at a variety of data for similar cars at different weight and I see little difference in decel rates, as measured in g's...

Originally Posted by ProCoach
Agreed.

So first you tell us that you don't see any difference in braking performance, and next you agree with the guy who writes that weight impacts braking performance?

I'm confused here
Old 12-15-2015 | 07:58 PM
  #68  
mglobe's Avatar
mglobe
The Penguin King
Rennlist Member
 
Joined: Apr 2006
Posts: 9,838
Likes: 119
From: Houston
Default

Originally Posted by winders
And Mike, that penguin slap is a faked video too.....
Do not disrespect The Penguin.
Old 12-15-2015 | 08:17 PM
  #69  
mark kibort's Avatar
mark kibort
Thread Starter
Rennlist Member
 
Joined: Feb 2003
Posts: 29,956
Likes: 180
From: saratoga, ca
Default

Originally Posted by winders
All of this is why I did not attempt to put numbers to the difference when weight alone is changed.

It is clear that the "tire load sensitivity" phenomenon is present in lateral and longitudinal acceleration of a vehicle. In other words, when cornering and braking. What's not clear is how much weight makes a measurable difference and in what situations would that be apparent.

The data that Peter has from all of his data analysis gigs is probably not useful in determining the "tire load sensitivity" for any specific tires or tires in general. Why? Because the vehicles are dissimilar enough and the drivers are dissimilar enough to make a valid comparison impossible. Unless, of course, he did some very specific testing removing or adding weight to a car with a driver able to get 10/10ths out of the car. Even then, the weight change may have had an effect on the cars handling by changing the weight distribution or cg.

Without extensive testing I think all we can say for sure is that, all other things being equal, making a car lighter will improve its cornering and braking. The more weight that is removed the bigger the improvement.
took the words right out of my mouth. I do say, for what my point is (and below, viking has NO clue) i think that the load senstivity factor, is going to subrtract something. how much? hard to say

Originally Posted by onefastviking
But that condradicts all the calculations and what MK says, braking should be the same regardless of weight.
I think extensive testing should be done.
Not by you though Winders, you might be biased. Not by me or PK either as we both too seem biased.
Hmmmm, what to do.......
GEEZ, you have to take your pill now Viking. I NEVER said braking should be the same regardless of weight. I said coefficient of friction would go down with weigiht. same exact thing as Winders is debating with Peter. thats the point of all this. you don't get anything , other than how to be an ***.
so, in the end, what winders just said is ANYTHING but contradicting what i said. you need to read with a little better comprehension....
Therefor, if the mu changes with tire load sensitivity, the now lower friction coefficient (sometimes called mu) from THE ADDED WEIGHT , will cause the KE to be dissipated at the higher rate with the lighter car and will be harder, not easier on the brakes

viking, here are my "calculations" please show us smart guy, anyplace where i have remotely indicated that braking is the same "regardless " of weight.

Two cars , CAR A (light car at 3000lbs) vs Car B (heavy car 500lbs heavier at 3500lbs). They get the point that of the exact same initial braking point, the two cars with the same 400HP are running:

car A 120mph (53.64 m/s)
car B 114mph (50.9 m/s)

the simple part of this example is that we can just plug in the numbers if we agree that the heavy car will slow at 8% less deceleration rate. we will call this 1g vs .92g respectively. this is because of the changing of the mu for the tires as weight is added to them. (weight on the tire, is the total weight/4 plus weight transfer weight/2 based on deceleration rate)

This means in 4 seconds the light car slows to 33mph (15m/s)
this also means in 4 seconds, the HEAVY car slows to the same speed

keeping it really simple. the KE at the start is:
1,991956 J Light
2,085,914 J Heavy

the KE at the end at the final same exact speed is:
153,405 J light
178,977 J heavy

this ends up with the lighter car dissipating a total of 1,838,551 for the light car
and 1,906,937 J for the Heavy car. heavy car dissipates more energy by 3.7%
not anywhere near the increase that the 15% increase of weight would first indicate

So,

Because the lighter car slows to the same speed in 4 seconds (3.94secs) and the heavy car slows to this same speed in 4 seconds too the RATE OF KE DISSIPATION is HIGHER for the heavy car by about 2%. no surprise here

625HP/sec for the light car ( example : 1,838,551 J /4 sec (3.94actual) /746watt =hp/sec)
639HP/sec for the heavy car ( example : 1,906,937 J /4 sec /746watt =hp/sec)

(only about 10hp/sec greater dissipation rate)


BUT, since we know the turn in point will happen at a slightly lower speed for the heavy car (13.5ms or 30mph), the result is: . Because now we are slowing at the same rate of .92g but for 4.15 seconds. NET NET 626hp ave dissipation rate....... the heavier car has the same of KE dissipation as the light car

The heavy car:
2,085,914 J at 50.64/s (114mph)
144,971 J at 13.5m/s (30mph)
Total KE dissipated of 1,940,942. / 4.15 seconds = 626hp ave dissipation


The rate of heat dissipation for the HEAVIER CAR is the same vs the light car for its decel rate and its lower top speed at the moment of braking (at the same spot on the track) and slightly lower target speed upon brake release for turn in.

the interesting thing here is also, with both cars activating their brakes at the same point, they end up at the same spot before turn in, 320f vs 316ft, if you use the decel rates of 1g vs .92g .

Therefore, this shows that adding weight, under these normal conditions, actually doesn't change the burden on the braking system.
Old 12-15-2015 | 08:20 PM
  #70  
mark kibort's Avatar
mark kibort
Thread Starter
Rennlist Member
 
Joined: Feb 2003
Posts: 29,956
Likes: 180
From: saratoga, ca
Default

Originally Posted by ir_fuel
So first you tell us that you don't see any difference in braking performance, and next you agree with the guy who writes that weight impacts braking performance?

I'm confused here
I think he is a little confused....
Old 12-15-2015 | 08:20 PM
  #71  
ProCoach's Avatar
ProCoach
Rennlist
Basic Site Sponsor
 
Joined: Jun 2008
Posts: 19,229
Likes: 3,382
From: Durham, NC and Virginia International Raceway
Default

Originally Posted by ir_fuel
So first you tell us that you don't see any difference in braking performance, and next you agree with the guy who writes that weight impacts braking performance?

I'm confused here...
The logged data (in essence, the end result) does not, on the surface, suggest that there is a significant (or even more than the standard variation of driver execution) difference in decel rates, as measured in g's, between similar cars with the same driver that differ in weight.

I agreed with Scott when he said, no matter what the numerical minutiae, extra weight hurt braking and cornering performance.

There is much more to braking performance and cornering performance than steady-state measures, and a lighter car can change direction, in multiple axis, quicker than a heavier one.

Straddled that fence well, huh?
Old 12-15-2015 | 08:21 PM
  #72  
winders's Avatar
winders
Race Car
 
Joined: Oct 2010
Posts: 4,602
Likes: 925
From: San Martin, CA
Default

I am not debating with Peter. Peter and I are having a discussion and I think we agree or mostly agree. onefastviking and I are getting on fine too. That leaves you and Orsini as the odd men out!

Old 12-15-2015 | 08:38 PM
  #73  
mark kibort's Avatar
mark kibort
Thread Starter
Rennlist Member
 
Joined: Feb 2003
Posts: 29,956
Likes: 180
From: saratoga, ca
Default

Originally Posted by ProCoach
The logged data (in essence, the end result) does not, on the surface, suggest that there is a significant (or even more than the standard variation of driver execution) difference in decel rates, as measured in g's, between similar cars with the same driver that differ in weight.

I agreed with Scott when he said, no matter what the numerical minutiae, extra weight hurt braking and cornering performance.

There is much more to braking performance and cornering performance than steady-state measures, and a lighter car can change direction, in multiple axis, quicker than a heavier one.

Straddled that fence well, huh?
Yes, that is a straddle of the fence, and is that your final answer?

Peter, does "hurting braking perfomance" = less deceleration rates?
I think you now say that the heavy car will slow at a lesser rate than the lighter car? as Scott and i both said. (even though your observations have shown otherwise, and i have SEEN otherwise, so, now yes, then you agree? forget the reasons, if that is true, and all i am asking for is the .92g vs 1g differences, then my calculations point to the lighter car being harder , not easier on brakes vs adding weight.

Im partially joking as well with you , but the difference might be the 8% loss to braking force based on a loss of mu of 8%, depending on the tire load sensitivity factor. base on the calculations i did in my example, it would be very hard to see a difference in that little change, however, the ramifications in power dissipation, (where all this stems from) is significant, in that the extra weight doesn't add much of a heat load to the system, on the track.

Now, we all know, also based on my simulation and common sense, that the lighter car will accelerate at a faster rate. the speeds it will reach for a given distance will be higher but yet, do to the physics dictating that the energy change due to the speed out pacing the added weight, the trade offs equal a very slight change if any, in the rate of KE dissipation. this is true, even if the the heavy car could slow as fast as the light car (3.7%). But, in actuality, i think what we want to see is the actual difference from the data in some reasonably controlled environment. heck, i would settle for the data from a cup car and a P2 car in the 140 to 50mph decel range on the same tire. i think it might be quite telling. better would be a couple of cup cars , in WCGT , one with 250bs of rewards weight and the other not.
Old 12-15-2015 | 08:44 PM
  #74  
mark kibort's Avatar
mark kibort
Thread Starter
Rennlist Member
 
Joined: Feb 2003
Posts: 29,956
Likes: 180
From: saratoga, ca
Default

Originally Posted by winders
I am not debating with Peter. Peter and I are having a discussion and I think we agree or mostly agree. onefastviking and I are getting on fine too. That leaves you and Orsini as the odd men out!

why do you even have to take sides.. such school yard behavior. I dont care who side im on. actually , i do.. im on the side of the truth or the thinking man. so, in this case, i am on the side of those that think that weight hurts braking slightly, to an extent of the tire sensitivity chart. looks like to me, you add 15% more weight, and you lose deceleration rates from 1g to .92 g. doesnt have to be accurate, just a rough estimate. and what you are saying is just that. "deceleration rates are less with more weight"
And you think you "agree with peter".. he went out of his way to say that there is no loss in mu because his observations didnt detect any loss in deceleration rates for adding weight. you and i both sited the tire sensitivity load factor, and Viking, thinks i said something I completely didnt say (quite the opposite).

this is a strange place!

I will say, if there is a loss of decel, that is only "one standard deviation" away, that is significant enough to cause a reduced deceleration go equal KE dissipation rates , and that was my only point. My original question to the list, was " by how much does the mu change with weight??"

EDIT: and Peter, in this case, a standard deviation might be a pretty big spread... dont you agree?
Old 12-15-2015 | 09:34 PM
  #75  
outline's Avatar
outline
Racer
 
Joined: Apr 2008
Posts: 368
Likes: 0
Default

A quick look at the most basic of equations needed to answer the question from a first principles perspective is enough to get a sense of some of the hurdles.

Tire friction coefficient is a function of vertical loading
Vertical loading is a function of weight transfer
Weight transfer is a function of deceleration
Deceleration is a function of tire friction coefficient

Nonetheless, it is perfectly possible to simplify it down to get a sense of the magnitude of the effect of the different weights as per the original question.

Using a platform with a known weight distribution, and a plot of tyre friction as a function of vertical loading for a known tire allows the max deceleration possible on first application of the brake to be calculated (assuming car is neutral when pedal applied).
From there, weight transfer starts to work against deceleration performance slightly. Using height of CG, wheel base and the initial deceleration it is possible to calculate weight transfer and the new vertical loading.
From there, the new tire friction coefficient can be calculated and from that the new rate of deceleration.

It is highly simplistic as a model but sufficient enough to gauge magnitude.

Using a 911 platform, since that is what I have data for with CG heights etc. I found that for a 500lb weight increase, the difference between the cars on deceleration was 0.5% initially, rising to 1% after weight transfer. This corresponds to less than a meter in terms of distance, or 2 hundredths of a second for the conditions given in the original question.

I would expect the natural variation in driver performance, and that of real world conditions to be sufficient to leave such a deviation well within the error bars for any empirical measurements.

All that aside, I wouldn't want the extra 500 lbs in my car for many of the reasons Peter listed a few posts earlier, amongst others.

Last edited by outline; 12-15-2015 at 10:10 PM.


Quick Reply: Racing Techie question regarding braking performance



All times are GMT -3. The time now is 05:28 AM.