* CORNER balanced - numbers from this...and the weight

[Cdnone1]

sorry what I meant to say was the tire carrying the extra weight needs more force to slow thereby since we do not have AGS that extra force is being sent to the less weighted tire and will cause it to lock up first
Steve

[Wil Ferch]

Cdine1:

We're open on discussion and no one will "jump" down your throat.

When situations come up like this..do like I did and assume "big" numbers to help us all "visualize" what's happening.

If one front corner carries more weight ( by design...say there is a lead-shot bag on the LF), and the RF doesn't...and further assuming the braking ("clamping") forces are equal for the LF and RF brakes....when the brakes are applied....what do you think would happen under a threshold braking condition?

I offer the opinion that the "lightly loaded" wheel will lock up first.

The question we're wrestling with is the two methods of corner balance under discussion. If we "set" the weight to be equal on both front tires ( but in reality the shot bag is really on the LF ...and the LF *should* be carrying more weight...but is not..in other words the car is "weight-jacked")....I think this set up would induce one wheel locking. I may be wrong.

However, there is a thought that says if the LF actually carries its apportioned weight....and the resulting corner numbers are ( naturally) different. ( LF being higher to reflect the "Actual" heavier weight there)...wouldn't this promote equal braking?

That's what's under discussion.....

Wil

[Cdnone1]

Thanks
I think I got it. I still think it will come down to basic physics. Equal clamping forces on uequal weighted tires should cause the less weighted tire to lock up first
Steve

[Randy Blaylock]

I don't have a problem at all discussing this stuff, and other things. I don't think anybody is getting their personal feelings hurt either.

I am intrigued by the concept of apportioned weight. My analysis is that there may indeed be a natural balance or distribution of weight in a car as it sits, I guess in a relaxed state. This is how a chassis is loaded with "dead weight".

Changing these natural percentages is by definition jacking weight. I readily acknowledge this. I do this intentionally, seeking certain objectives. This is, in and of itself, not incorrect nor lacking in sound application of basic chassis tuning principles. There are many things that can be acheived by weight jacking. That's the whole point of what I was doing.

A tire under a given vertical load will have a finite coefficient of friction, maximum slip angle, or "grip" available. The higher the load, the higher the grip. This is why downforce is valuable. Now, if a pair of tires are clamped equally under braking, unless the load is identical, then the more lightly loaded tire will lock first.

In my racing experiences, I have had the opportunity to try many different theories, and have found the strategy I have outlined of seeking to equalize the front weights and the diagonals to be very effective. It's just my experience, yours may be different.

There are many ways to cut a pie......

[Craig 930 RS]

When I take my car in for the alignment next week, I'll post my subjective impressions of the results.

Hats off to Randy's hard work and knowledge here. He worked his arse off for 2 solid days making everything correct F&R. More attention was paid to detail than other times where I took my car to some spendy race shops - and wrote some big checks at the rate of almost $100 per hour.

[Jeff Alton]

Like I said earlier, I am confused? My question now is"

Can the fronts left vs right, and the rears left vs right both be made equal? Or is this just a pipe dream? If not, I understand Randy's theory.


Jeff

[Randy Blaylock]

Not likely in a production chassis, without physically moving weight inside the car. You can sometimes improve things by relocating the battery, or situating ballast strategically if the extra weight is required to meet minimum weight rules. Typically you will almost always end up with a bias on the side where the driver sits.

I have a pal that had a BMW M roadster that had corner weights that were within 5 pounds in all four corners, with him in it, very unusual.

[Jeff Alton]

Ok, great. I hope I am not asking stupid questions. I really did pay attention in class, honest!! I am getting mine done shortly, so I am just curious.

Jeff

[patkeefe]

As I have to get my SC aligned and balanced next week, I find this very relevant.

We have a line drawn from front to rear centrally through the car such that there is geometric symmetry from L to R. Randy, are you saying that the CG should be on this line, and that if not, you transfer weight to make it on the line? And Wil, is your argument that the CG does not necessarily have to line on this line? I sort of simplified this for my benefit.

And, when you balance, you perform this operation with the sway bars detached from their mounting points?

Thanks!
Pat

[Randy Blaylock]

It's not so much about CG as it is about individual wheel loading.

A car with one diagonal ( LF + RR or RF + LR) loaded more than the other will turn better in that direction, in extreme examples to the point of being undrivably prone to oversteer.

In my opinion, an even diagonal bias, 50% for either pair, is an objective that is weighted (pardon the pun) equally with trying to acheive equal front corner weights.

Most production car chassis, because of the physical placement of things like battery, driver etc.... as Wil said, will necessarily be left side biased, so playing around with jacking weight usually ends up being a compromise to meet both objectives of equal front corner weight and even diagonals.

I find that one usually has to concentrate more on evening out the front corner weights, and the diagonals end up being within a few percentage points of even. This typically isn't an issue because production unibody chassis aren't particularly stiff, and radial tires don't grow much with heat to affect the tire diameter and the attendant increase in corner load and it's affect on the diagonal bias.

[randywebb]

Bottom line: your girlfriend should weigh as much as you do... I hate to break that news folks...

[Zeke]

Wil, could you give us a complete set of numbers off your car including your weight so as to have a target of an ideal setup?

And, are there any particular handling idiosyncrasies in any turns at any tracks you run given that you have an ideal setup that you could share?

[lateapex911]

Random thoughts and a question.

Randy, just a clarification...you said: "A car with one diagonal ( LF + RR or RF + LR) loaded more than the other will turn better in that direction"
So, to be specific.... Lets say that the LF-RR diagonal is at 52%. Do you define that diagonal as being the biased diagonal? And therefor are you saying that the car will turn better to the left?

Second, swaybar links. Disconnecting them during the setup is important, but it's important to remember that soe of the "good work" being done during a good corner set up will be undone if the links are not adjustable. Again, Randy's point about the chassis flex is an important caveat, but adjustable links are cheap, so why not get some anyway?

Finally, could chasing "perfect weights" around on a street chassis result in a car with some "interesting" ride height settings?

Ultimately, actual weight placement within the chassis is the way to go. THe McLaren F1 road car was designed with the driver in the center for a number of reasons, not the least is weight distribution. In the P car, moving everthing and anything to the right helps!

[911pcars]

Randy has set up the car the way he wants it, and if it works for him that's fine. But IMO, in achieving his objective of equal LF and RF weight distribution, the RR will be light under braking and will tend to lock up before the LR. Furthermore, the chassis will tend to oversteer more turning left than right. There's an additional compromise with the LR + RF diagonal 100 lbs. heavier than the other diagonal.

Since others have offered their chassis balancing weights, let's look at them.

The following are Chad's numbers:
LF 398
RF 359
LR 680
RR 641

LF/LR weight proportion is 58%
RF/RR weight proportion is 57%

Front-to-rear weight proportion (total) is 57%.
LR+RF diagonal = 1039 lbs.
RR+LF diagonal = 1039 lbs.

The following are cvfncrew's numbers:
LF 515
RF 469
LR 778
RR 730

LF/LR weight proportion is 66%
RF/RR weight proportion is 64%

Front-to-rear proportion (total) is 65%
LR+RF diagonal = 1247 lbs.
RR+LF diagonal = 1245 lbs.

I'd say both chassis are pretty well balanced. The total vehicle weights are irrelevant for this discussion.

MHO,
Sherwood

[Wil Ferch]

LOL !....Zeke..my numbers are far from "ideal" but they'te pretty good..

I've posted them before in my discussions with Chuck on this subject. I'll try to repost the link.

Randy...I understand that purposeful "weight jacking" can get you characterisitics that you might want to get....but this requires a high level of knowledge and racing expertise to achieve.

I still think my original question is on the table.....which scenario will promote one front wheel locking?. The weight-jacked ( equal weight RF and LF)...or the car that apportions the load to where the weight actually is in the car?

Interesting sidebar: maybe the conclusion we can draw ( if...for example...we say that both methods work)...might be that the actual *numbers* used for both methods might not be very far "off" from one another....and therefore even the "wrong" way ( in one person's view)...might still work. Meaning? The "correct" or "incorrect" way might not make that much of a difference if the individual corner weights are within ( say) 20 lbs ...comparing one method to another. You might not be able to tell the difference.

Wil

[Randy Blaylock]

Great discussion.

Diagonal biasing and it's use as a tuning tool is usually used in an oval track scenario. The theory applies to any chassis configuration from champ car to front wheel drive hobbyist racer.

Using an example we are all probably familiar with watching the 800 pound gorilla on Speed channel, you'll hear stock car racers talking about making a "wedge" adjustment. They're actually talking about adding or subtracting corner weight, which adds to or subtracts from diagonal bias. They usually do this by inserting a "L" handled wrench with a long extension through a tube that comes up to the back window. This tube guides the wrench so that it engages a long coarse threaded bolt with a plate mounted vertically through the frame, that is actually the upper spring perch. When they give it "a round of wedge", they're turning the wrench one revolution, either adding weight or subtracting it from either the RR or LR depending upon how critical the ride height at that corner is. Incidentally, they do this at the rear because of rules constraints, not from any sound engineering application. This changes the diagonal percentage a small increment to affect the overall balance. Asphalt cars usually run a diagonal bias on the LR/RF pair in a range from 52% to 58%.

FWIW, running the car at a left side weight that's similar to the diagonal bias yields even front corner weights for braking stability.

Just FYI, when they talk about making a track bar adjustment, they're raising or lowering the panhard bar at the chassis end typically, in the same manner as the adjustable spring perch, making a rear roll center/roll couple adjustment to affect the rear end grip character.

It's very, very important to take into account that true race car chassis are extremely stiff, and making small adjustments will make a distinguishable change in balance. An analogy for coil-overs is that a 1/4 turn of the adjusting ring on fine threads is also a distinguishable difference.

This is simply not the case in most production unibody chassis, even the types more oriented to sporting use, these chassis are simply very flexible. This is why trying to draw a correlation between race cars and street cars is difficult, because things are not relative.

So to answer one question, a 2% diagonal bias in a production chassis is most likely not going to be noticable at all.

Spring rates are another good example. On a local track here, there are PCA/SCCA etc... races and occasionally stock cars. A typical RSA race car will use spring rates up to 800/1100, in a car that weighs 2700 lbs +/-, on 8 and 10 inch wheels. A stock car that weighs 2900 lbs, on 10 inch wheels might use 500 lb fronts and 200 lb rears. The chassis stiffness of the stock car allows the use of much lighter spring rates because the chassis isn't absorbing half the spring travel as the car pitches and rolls.

Taking this production chassis challenge into account, I personally think chasing track specific setups to be an ineffective use of time. There is probably more time to be gained from simply investigating different lines using the setup in the car that you're already familiar with. I may play around with sway bar or rebound settings a little if a particular track has a critical corner with a distinguishing characteristic that can be improved upon, but usually a good solid baseline setup is very effective on many tracks. Of course a legitimate race car is an entirely different
prospect.

Regarding the actual scaling procedure, it's critically important to either use the exact same location for the scale pads each time you use them for repeatability, and even then you can't say for sure what the actual weights are relative to true level ground. Shops that use four post lifts for scaling are giving up a margin of accuracy because the position of the scale pads relative to each other in a vertical plane cannot be determined to be dead even because of the movement of the machine. This may not be relevant in a way, because they're usually scaling production based cars, and the chassis flexibilty sort of minimizes the margin of error range.

True accurate scaling requires the use of some sort of individual scale pad levelling devices, or a platen, so that all the pads are levelled to each other side to side, front to back and diagonally. I do this until the pads are all within one tenth of one percent of each other.

You do also need to disconnect the sway bar links, and adjustable links are the only way to attach the bars without introducing preload into the bar. You also need to have the car sitting as it will be run, with whatever fuel load you want, the ballast placement and quantity done if applicable, and tire pressures set very accurately, especially with bias ply tires.

Regarding the ride heights, they are only a starting point. I like to keep a car within a range, raising or lowering individual corners to change individual corner weights while staying in a certain ride height and rake range. An example is if I want a 911 to have approximately 24 inches at the rear fenders and 24-1/2 at the front, if the car starts out there, then I will take a little off the heavy diagonal on one corner and add a little on the other at the same end of the car. It takes a lot of litttle adjustments to get the results you want, kind of like truing a bicycle wheel. You will see measurable changes in corner weight without a lot of change in ride height in the same corner.

It's been my experience in production chassis that a 2% bias diagonally is a very acceptable load to accept when trying to acheive even front corner weights. I have not been able to discern any handling differential left vs right. I think the bias would have to be upwards of 5% before it might be noticable. This assumes a car without a welded 6 point cage minimum.

A car with weight jacked diagonally will turn more effectively in the direction of the diagonal if you drew an arrow from rear to front on the diagonal. Visualize a car turning right. The left side tires are more heavily loaded than the rights, on 911s the inside front tire sometimes even in the air (far from ideal but that's another story). Now imagine if the LR had more load, making it prone to exceed it's maximum grip, or be further into it's slip angle sooner. Imagine also that the inside front tire is now more heavily loaded, contributing more to pulling the front end to the inside. You may have noticed when making a long turn on a track, if you just put the front inside wheel up on the curb a little bit, the car will rotate better if you otherwise had a little understeer there. What you did was jack a little weight into that diagonal for the time the front tire was ridind the curbing.

Just to throw a monkey wrench into things, asphalt stock cars run the diagonal bias in the LR/RF diagonal, to stabilize the car. The corner weights end up being even in the front, with the LR carrying approximately twice as much as the RR. They sometimes have upwards of a 60% left side bias. They also run positive camber on the LF, caster split and stagger. They turn left like crazy.

It's all about the physical loading a tire sees in a vertical plane. Two identical tires loaded equally will develop identical grip assuming identical surfaces. This is specifically why I seek a balance on the front corners, the balance being defined as even physical load as meassured in pounds on a scale pad. The G sum developed on the brakes in a 911 transfers weight from the rear to the front so that the F/R bias changes. The original bias is quite rear heavy, so that with weight transfer under braking, the rear wheels are still carrying a significant load. This is why the RR doesn't lock. FWIW, I have never, and I mean never, had individual rear wheel lockup on heavy braking, even entering turns while trail braking. No ABS. This despite the fact that I like to run more rear brake bias than typical.

It's true that we all may be using different techniques and philosophies to get to the same point. I have to admit though that if a different philosophy seeks an end result where the front corner weights are not close to even, then I'm not in agreement, based upon my practical experiences.

Happy Easter

[Zeke]

Originally posted by Wil Ferch
LOL !....Zeke..my numbers are far from "ideal" but they'te pretty good..



Interesting sidebar: maybe the conclusion we can draw ( if...for example...we say that both methods work)...might be that the actual *numbers* used for both methods might not be very far "off" from one another....and therefore even the "wrong" way ( in one person's view)...might still work. Meaning? The "correct" or "incorrect" way might not make that much of a difference if the individual corner weights are within ( say) 20 lbs ...comparing one to another. You might not be able to tell the difference.

Wil

My request might be seen as a bit hostile (and I can't be totally absolved from that accusation). I see this subject as both theoretical and practical.

I learned of Randy's prowness the hard way on a thread about a broken steering part earlier this year. I have the utmost respect for the practical analysis of corner balancing from someone who uses this to drive. Personally, I wouldn't know the difference between the nuances of one method over the other. I drive what I've got to the best of my ability which is not anywhere near that of many other's experience.

Wil, in the past, you have contributed many well thought out axioms that are being somewhat challenged here. I'm just a bystander. I'm pretty much buying into Randy's thoughts that you drive what you have and find a way around the track.

I was asking you to present the practical side of your opinion. Can this equal front tire weight and potential wheel lock up be that critical? Or is it only theorectical? You do seem to say that in your last statemant.

So many different conditions and circumstances seem to intercede.

BTW, I expect to undergo an alignment and CB soon and I want to interpet the numbers when I get them. As with any car, my car may not cme out on the optimum side due to many issues. If I have to compromise, I want to know which way to lean. (I guess the pun was intended. )

[Wil Ferch]

Randy , thank you for continuing this interesting and (hopefully) informative dialogue....

Let me try to get us closer to the original question, and make some simplifying points so we can all view these points.

First off...we need to start with non-controversial *facts*.

Fact #1.
....no matter how you change corner balance....the starting and ending result of any adjustments will STILL have ANY TWO ADJACENT wheels carry the same total pound load. You need to noodle with the numbers to convince yourself this is true....but it is. Whether your corner balance is "spot-on" or way off base....any two adjacent wheels will carry the same TOTAL ( for these two) loads. The only way to change this is to change where the masses of the car are located ( move battery or engine, etc).

Fact #2:
- if you increase the loading of a LF.....you will increase the loading of the diagonal opposite corner ( in this case...RR). Same if you decrease the weight on LF..the RR will show less.

Now...just to throw big, ridiculous numbers on this...lets's say the total (front) weight of a car is 1500 lbs. BUT, the way the car is built...you have 1000 lbs on the LF and 500 lbs on the RF.

Method 1--> you corner balance such that the LF actually carries 1000 lbs at rest , and the RF carries 500 lbs. My opinion is that during threshold braking...the heavy loaded LF is "properly" loaded..and the RF ( although lighter) is also properly loaded....and one brake will not be prone to lockup before the other, all else being equal. The opposing view will say that regardless of this.....the RF wheel will lock up first because it's normal ( downward force vector) force is less and therefore will lock up first. Friction is a function of normal ( downward) load.

Method 2--> you corner balance such that each front wheel carries 750 lbs ( remember...the TOTAL front weight....1500 lbs.... can't change !). So...the LF is now set up to carry less than its "apportioned" ( 1000 lb)load......and the RF is set up to carry more of its "apportioned" ( 500 lb) load. Each wheel is "off" ( or weight-jacked) by 250 lbs....in different directions from "ideal". The normally "heavy" LF corner is now (relatively) lightly loaded....and the RF ...normally "light" corner...is carrying more than it should ( it's now relatively "heavily" loaded). The heavy loaded tire will not tend to lock up and will "rotate" longer under threshold braking...just as if aero aids were in-play that load up a car ( downward/ verticaly) to help likewise..

Now....here's where the "size" of the deviation comes into play. So far, one view says the LF will skid first. The opposing view says opposite. BUT...as Randy correctly points out....during threshold braking there is an "apparent" weight shift to the front. A 911 that "sits" at 40/60 front-rear may actually experience 60/40, (or more) during heavy braking. Might these "weight shift" numbers completely over-shadow the small 250 lb corner balance difference we looking at ?? Definite possibility !!!

If you assume the 40/60 *can* become 60/40 during heavy braking...then about 500 lbs is transferred forward on a 2500 lb car. More so if the weight transfer is more, like 70/30. This 500+ lb affect may completely overshadow the "only" 250 L-R discrepancy we were talking about...in comparing the two corner balanvce methods.

I offer the opinion, however, that unless you want or need to "weight-jack" you car in a particular direction, for a particular track....you'd be best off to assume "random number of right and left turns"....and corner balance without weight jacking....so the car can corner and brake at its best for whatever track or street you encounter.

BTW.. a little factoid to consider:
Some people like to put a "fine edge" on their corner balance, by placing themselves or equivalent weight on the driver's seat. This is OK if your "sweet spot" is with one driver. This is counter productive if the normal state is with two people on-board.

Anyway....if you do this...the effect of a 150 lb weight ( in the driver's seat) of a 911 results in THIS effect on the original corner weights:
LF= Adds 61 lbs
RF= Ads 11 lbs
LR= Adds 50 lbs
RR= adds 28 lbs

Different weights ( other than 150) can be scaled up or down accordingly.

Wil

[Randy Blaylock]

If we were able to have this discusssion over beers, a common activity in the circle I travel in, then we could all see each other's facial expressions, hear voice inflections, and gesture wildly to help add emphasis to our points.

Unfortunately the format of the net doesn't allow for this, and as such sometimes things are assumed.

I am not posting here to prove my intellect to myself, I am sure that there are others that have forgotten more than I will ever know. I would like to think that each of us has theoretical ideas and practical experiences that can be shared that can be discussed and learned from.

I'll apologize if the tone of my posts seem to exclude other theories, that's not my intention. The discussion adds to the concept for everyone.

I propose that we all imagine that we're sitting in a group in a place that you personally enjoy, partaking of an adult beverage if that's your thing (or not), as we discuss this concept.

By the way, there is some good racing on Speed right now.

In fraternity....

[Wil Ferch]

Randy, I'm enjoying this.....and I think you are too. I've already come clean to apologize if my earlier posts appeared confrontational...

From all sides.. I think they're not... and I share Randy's views that this is the E-version of the bar-stool round-house.

Zeke..indeed this can get to theory more than practical application. At some point, when the numbers get BIG and out of whack...then even theoretical nuances become "meaningful". I guess I'd like to share dialogue , facts , experience, and ( yes) opinions....so that the "theoretically-correct" method emerges. I'm open to correction, fer sure !

Wil