When the Results Don't Match The Principals or Formulas

[Ron Sutton]

This is a thread I'm starting for Jason Rhoades, myself & others who are interested in figuring out what is the cause ... when the track or test results don't match established principles or formulas.

Post #2 through #9 ... are moved from another thread ... so we could continue discussing this, without completely hijacking that thread.

Posts #10 & later, are the continued discussion.

[j-rho]

Here's my first post from the old anti rollbars thread-


Ron you've written a lot of good stuff, and pretty much all these folks have the same basic 2 problems:
1. Poor front:rear tire size ratios considering the cars' front:rear weight biases
2. Not enough resulting negative camber on the outside front tire at terminal roll.

The solutions for everybody are pretty much the same:
1. More front wheel/tire
2. More negative camber - if you aren't replacing enough components to get great geometry (whether for budget or ruleset), that means it'll need more static caster and negative camber

Another common element is how the motion ratio of the stock geometry cars is awful, most are running very low ride frequencies with far-too-low damping rates for performance applications. The Corvette geometry front subframes like Matt's DSE stuff allow for reasonable spring and damper rates to achieve a performing setup, but those still on the stock 60's geometries require 4-digit spring rates and absurd damping forces (which basically nobody is doing) before they actually start to work.

I think some of your statements are a little misleading - like sway bars "engaging the inside tires more" - adding, or stiffening a swaybar at one end of car, will increase that end's share of the lateral load transfer distribution in a corner, decreasing dynamic loading on that end's inside tire, potentially to the point of lift - which sounds like the opposite of what you said. Sway bars do not decrease total lateral load transfer (ignoring minor second-order dynamic CG height changes) so they can't add net weight to the inside tires at a given lateral acceleration ("get them more engaged") - only a wider track or lower CG can do that.

I think you probably already know this and are trying to simplify it for the folks here, so I offer this only for those that are maybe "starting to get it" but are having a hard time with those particular elements of your posts.

[Ron Sutton]

Hi Jason, thanks for chiming in.

Ron you've written a lot of good stuff, and pretty much all these folks have the same basic 2 problems:
1. Poor front:rear tire size ratios considering the cars' front:rear weight biases
2. Not enough resulting negative camber on the outside front tire at terminal roll.

The solutions for everybody are pretty much the same:
1. More front wheel/tire
2. More negative camber - if you aren't replacing enough components to get great geometry (whether for budget or ruleset), that means it'll need more static caster and negative camber.
You’re correct. But, I hate to add any more Camber than needed, because as much as it helps the outside front tire achieve an optimum contact patch, it hurts the inside tire angle, preventing it from achieving an optimum contact patch. All of this is caused by the KPI/Caster Split favoring the KPI too much.

Of course some cars can’t achieve enough caster for an optimum KPI/Caster Split favoring the caster … for a variety of reasons ranging from rules & owner limitations to parts availability & packaging issues. But it makes sense to start with the best KPI/Caster Split we can … then achieve the rest with camber gain & static camber.

Another common element is how the motion ratio of the stock geometry cars is awful, most are running very low ride frequencies with far-too-low damping rates for performance applications.
Agreed. If people … including the shock engineers at most pro touring shock companies ... saw the shock dyno sheets & saw how much dampening rate we’re running on the nose (0 -1/2 IPS) & total rebound at 1-3 IPS … they would faint or be in total disbelief.

The Corvette geometry front subframes like Matt's DSE stuff allow for reasonable spring and damper rates to achieve a performing setup, but those still on the stock 60's geometries require 4-digit spring rates and absurd damping forces (which basically nobody is doing) before they actually start to work.
I agree & even then … on the stock geometry cars … everything they do with springs, bars, shocks, etc … are crutches for the horrible geometry.

I think some of your statements are a little misleading - like sway bars "engaging the inside tires more" - adding, or stiffening a swaybar at one end of car, will increase that end's share of the lateral load transfer distribution in a corner, decreasing dynamic loading on that end's inside tire, potentially to the point of lift - which sounds like the opposite of what you said. Sway bars do not decrease total lateral load transfer (ignoring minor second-order dynamic CG height changes) so they can't add net weight to the inside tires at a given lateral acceleration ("get them more engaged") - only a wider track or lower CG can do that.

I think you probably already know this and are trying to simplify it for the folks here, so I offer this only for those that are maybe "starting to get it" but are having a hard time with those particular elements of your posts.

I respectfully disagree, but Total Roll Stiffness, Roll Couple & the net Roll Angle all play a role in this.

“Adding, or stiffening a swaybar at one end of car, will increase that end's share of the lateral load transfer distribution in a corner, decreasing dynamic loading on that end's inside tire, potentially to the point of lift” … happens only when the car has excessive Roll Angle.

In that case, you are correct, because the car is rolling over with so much force, that any attempts to reduce Roll Angle with ARB stiffness, simply unloads the inside tires. It’s funny, because the stiffer ARB DOES make the suspension run flatter, but the car is rolling with such force it just brings the inside tire, on that end, with it. If it’s big enough, you can see the tire(s) lift. But long before it gets that bad, you can really measure it on a skid pad, because that inside tire starts running cooler.

In your AutoX racing you may have seen, or experienced, stiffer ARB’s leading to lifting the inside tire to a degree, reducing the inside tire’s grip, but the cause is excessive Roll Angle, either evenly or unevenly. This excessive Roll Angle can be caused by three things …. excessive chassis flex, too soft of spring rate and/or too long of a CG lever arm over the roll center.* (Addition: Actually 4 things, counting track width, but most folks on here are not narrowing their track width as a tuning tool.)

*For those following along, that means the car’s CG (Center of Gravity) is too high and/or the Roll Center is too low. The CG acts as a lever on the Roll Center. The further they are apart, the more the car rolls.

I have seen chassis flex cause both even … and uneven … excessive Roll Angle … depending on where & how the chassis was flexing. The spring package can cause “even” excessive Roll Angle when all the springs are equally too soft … and uneven excessive Roll Angle when just one end of the car’s spring are too soft … or they’re all too soft and one end is more so. The excessive distance (leverage) between the CG & roll center can cause “even” excessive Roll Angle if the condition exists at both ends of the car … and uneven excessive Roll Angle when the condition exists only at one end of the car.

Here is a USAC Midget hiking the LF wheel on corner ENTRY … because the crew chief put too soft of a rear spring on the car in an attempt to gain more traction on exit.

You can see the LF tire is 3”+ off the ground, while the LR tire is on the ground.



Here is an entertaining photo of a Mustang Autocrossing with excessive Roll Angle.
You can see the LF tire is 4”+ off the ground, while the LR tire is on the ground.
(Hope I am not embarrassing the car’s owner. Just needed an example.)



But, as long as the car we’re tuning doesn’t have excessive roll angle, stiffening the ARB's do keep the car flatter, spreading the load more evenly over all 4 tires, giving the car more grip. In my experience with all forms of race cars, the race team engineers & myself found increasing ARB rate always made the suspension run flatter … and always worked the inside tires more & the outside tires less … but we were not working with excessive Roll Angle situations.

When testing on 7-post rigs, and even more simple pull-down rigs, you can see the increased loading of the inside tires with stiffer ARB rates. It shows up right on the screen with increased wheel load cell readings. Of course, the track is the best test, and more fun.

In professional oval track & road racing, when we're in the "happy window," a slightly stiffer ARB, or a little pre-load, helps us to works the inside front tire … making the car turn better. In oval track racing this is most critical, because the spring packages in front are very soft.

If we stiffen the ARB too much, it will actually start engaging the LR tire more ... which is normal ... tightening the car up, especially in the middle & exit of the corner. It’s obvious to see with our on-board data acquisition systems, but less equipped racers simply look at the tire temps. When a temp goes up, you’re working that tire more. When a temp goes down, you’re working it less. The only thing that tricks some racers here is when they are sliding, pushing or spinning a tire … will also get hotter.

We do run into excessive Roll Angle situations … which is easy for us to see … because the car doesn’t gain load on any of the inside tires with stiffer ARB rates. How we fix this, depends on the class of racing. If we’re on a coil-bind set-up, we have to install a spring with a taller bind height. If we’re on a bump stop set-up, we simply add a shim. If we’re on a “soft spring/big bar” set-up, we increase the spring rate 1 step. We could adjust on the Roll Center, but only if we run out of options or find it was lower than optimum.

Occasionally, with less experienced Racers, we find they have the CG too high, or the Roll Center way too low (under ground) … and these require work I consider above & beyond normal track tuning … more like their baseline chassis design or set-up wasn’t optimum.

And of course you really see excessive Roll Angles in a lot of street cars with big power, better suspensions, grippy tires & no roll cage to add chassis torsion rigidity. Kinda like that Mustang above.

Best wishes …

[j-rho]

Next post from that thread

Hi Jason, thanks for chiming in.


I think some of your statements are a little misleading - like sway bars "engaging the inside tires more" - adding, or stiffening a swaybar at one end of car, will increase that end's share of the lateral load transfer distribution in a corner, decreasing dynamic loading on that end's inside tire, potentially to the point of lift - which sounds like the opposite of what you said. Sway bars do not decrease total lateral load transfer (ignoring minor second-order dynamic CG height changes) so they can't add net weight to the inside tires at a given lateral acceleration ("get them more engaged") - only a wider track or lower CG can do that.

I think you probably already know this and are trying to simplify it for the folks here, so I offer this only for those that are maybe "starting to get it" but are having a hard time with those particular elements of your posts.

I respectfully disagree, but Total Roll Stiffness, Roll Couple & the net Roll Angle all play a role in this.

Actually, they don't at all. Whether a car rolls dix degrees at one g or three degrees, the amount of weight transferred from the inside tires to the outside remains the same. The only way reducing roll angle could have an effect on total weight transfer, would be if the car gained CG height as it rolled - some cars might but that is not universal and dependent on the geometry. CG might actually sink as the car rolls. In any case the contribution of any CG height change as the result in a change of roll stiffness at one axle is likely orders of magnitude smaller than the handling balance change brought by the same change.

“Adding, or stiffening a swaybar at one end of car, will increase that end's share of the lateral load transfer distribution in a corner, decreasing dynamic loading on that end's inside tire, potentially to the point of lift” … happens only when the car has excessive Roll Angle.

I'm not sure where you draw the line at "excessive" but this happens all the time - especially in the rear of FWD cars, and in the front of mid or rear-engine RWD cars. For 2WD platforms there are plenty of places where you want to saturate load transfer at the non-driven end to maximize potential tractive force at the driven end.

But, as long as the car we’re tuning doesn’t have excessive roll angle, stiffening the ARB does load the inside tire more, giving it more grip. In my experience with all forms of race cars, the race team engineers & myself found increasing ARB rate always made the suspension run flatter … and always worked the inside tires more & the outside tires less … but we were not working with excessive Roll Angle situations.

This is the part where I get a bit worried...the most basic tenet of tuning automotive handling is tire load sensitivity. A pair of tires (Front pair or Rear pair) will generate the most grip when they are most evenly loaded (all else being equal). When a car achieves a given lateral acceleration, CG height and track width alone determine how much load is going to go from inside to outside - the suspension tuner gets to choose how much of this weight is borne by the front and rear axles via their relative roll stiffnesses (contributed both by springs/bars but also link geometries). For a car that started neutral, increasing front swaybar stiffness relative to the rear, will increase the differential in load between the two front tires (UNloading the inside front tire MORE), which decreases the amount of potential grip that axle has to offer at that lateral acceleration - which means shift towards understeer, all else equal. While this happens, the rear axle is enjoying a reduction in lateral load transfer, with the inside rear now more heavily loaded at that lateral acceleration than it had been with the softer front bar.

Total weight on the pair of inside tires (front + rear) does not change directly from a stiffening of swaybars; it leads to decreased load on the inside of the stiffened axle, and increased load on the inside at the opposite axle, for a given lateral acceleration. Same thing could be achieved with stiffer springs.

When testing on 7-post rigs, and even more simple pull-down rigs, you can see the increased loading of the inside tires with stiffer ARB rates. It shows up right on the screen with increased wheel load cell readings. Of course, the track is the best test, and more fun.

The only part of this I agree with is the last part - real world testing trumps just about anything you can do on paper.

In professional oval track & road racing, the stiffer we make the front ARB, through size or pre-load, the more it works the inside front tire …

This should be counterintuititve to you - you could put together a front bar so stiff, the inside front tire is almost always in the air - how can you work a tire that never touches the ground?

If we stiffen the ARB too much, it will actually start engaging the LR tire more, tightening the car up, especially in the middle & exit of the corner.

This makes perfect sense - increase in load transfer across the front axle means reduced load transfer across the rear axle, which means more load/"engagement" from the inside (always Left in your case?) tire.

[j-rho]

My next-

Hi Ron,
Appreciate the perspective here. I wonder if maybe our apparent disagreement isn't really a difference over the underlying principles, but maybe just the way we talk about them.

Using your scales approach from above, let's take a hypothetical "totally symmetrical car" - everything is the same about it any way you cut it in half. 50/50 weight distribution with its CG at vehicle centerline along the longitudinal or lateral axes.

Let's say at rest this car looks like this on the scales:
{At rest baseline}
LF 500# RF 500#
LR 500# RR 500#
Left side total: 1000#
Right side total: 1000#
Front total: 1000#
Rear total: 1000#


Now let's say we did something to simulate a cornering situation - maybe we attach a cord to the CG, and pull directly horizontal with some force F. Our tires stay stuck to the scales, but that force rolls the car just as it would in a corner, and based on this car's setup, now our scales read:
{Cornering force F baseline}
LF 750# RF 250#
LR 750# RR 250#
Left side total: 1500#
Right side total: 500#
Front total: 1000#
Rear total: 1000#

Now, we do something to stiffen the front swaybar some amount. But we did so without adding weight or changing anything else about the car. At rest, it still looks the same:
{At rest stiffer front bar}
LF 500# RF 500#
LR 500# RR 500#
Left side total: 1000#
Right side total: 1000#
Front total: 1000#
Rear total: 1000#

The difference, is now what happens when we apply the exact same lateral force F, now that we have increased the front roll stiffness?

I contend we will see something like this:
{Cornering force F stiffer front bar}
LF 750+X# RF 250-X#
LR 750-X# RR 250+X#
Left side total: 1500#
Right side total: 500#
Front total: 1000#
Rear total: 1000#

I contend we haven't changed the total resulting total load on either the left side or right side tires from baseline, even though the car might visually be leaned less now in response to the same lateral force F (e.g. we haven't reduced the total amount of weight transfer). Whatever additional weight transfer X occurred on the front axle, will be the amount of reduced weight transfer we see on the rear axle.

When I read "Stiffer bars reduce roll angle, engaging the inside tires more." What I hear is one of the classic misconceptions in vehicle handling (which seems out of place amongst your otherwise correct information) - that quantity of roll is directly related to quantity of weight transfer, and the more roll, the more weight transfer independent of lateral acceleration. A corollary misconception is that which says the car that squats more, has more rearward weight transfer. I think the basis of this misconception is that when you see two cars with identical setups, and one is rolling more, then there is more weight transfer - but only because that car is cornering harder (more torque about the CG).

If stiffening the front bar resulted in more weight on the inside wheels at the same lateral force F, where is that load coming from? Since our car can't magically gain weight, it has to come from reduced load on the outside tires. This all means a reduction in front-view torque about the CG, and there is no other force available to counteract the torque presented by the cord tugging at the CG. It doesn't add up.

What does make sense is "Stiffer bars reduce roll angle, engaging the inside tire on the other axle more (but not the "inside tires" as a whole)" - if you can live with that interpretation, then maybe we can find some agreement.

[Ron Sutton]

I don’t really like to discuss the new tuning concepts I've learned in the last decade or two, where it conflicts with established engineering principals, because I find it hard to properly explain. But I wouldn’t want anyone to dismiss what we’ve been talking about here, because they think I’ve lost my mind.
Just so you know … I’m not saying I haven’t lost my mind. It happens often … but fortunately for me … I keep finding it again.

On my regional 9-car race team we had 3 Engineers … and an impressive data acquisition program in-house, including our own satellite based DAQ system … when combined, were critical to our race team’s progress & winning successes (we won 45 races the last 3 years … 2009-2012). The engineers did everything from analyze data on race cars, measure chassis flex, g-forces, tire temps, track friction & slip angles to engineer cutting edge suspensions, aerodynamic aids & torsionally stronger, lighter chassis.

In my profession, we have to be careful not to take 100.00% of engineering training as gospel. It was, after all, developed by human beings based on their understanding of how things work … at the time they developed the principal … with the technology of the day. We occasionally … VERY RARELY … but occasionally run get results that conflicts with what was written by engineers.

This discussion of ARB loading tires is one of those areas.

-------------------------------------------------------------------------------------------------------------------------------

Another is Track Bar/Panhard Bar angle. A few years back, a team was testing a Dodge NASCAR Craftsman Race Truck on a 7-post rig … and the Crew Chief was testing the Track Bar at different angles for an upcoming road race, but with the center always at the same height, so the roll Center stayed the same. He started with a baseline with the left side at 10” & the right side at 10” and ran the test, producing a baseline set of load numbers that were even on the left rear tire & right rear tire.

When he put the left side at 8” & the right side at 12” & tested it … the 7-post rig computer showed more load on the left rear tire & less load on the right rear tire … dynamically … and the Truck had a decreased Roll Angle turning left & increased Roll Angle turning right. When he put the left side at 12” & the right side at 8” … the 7-post rig computer showed less load on the left rear tire & more load on the right rear tire … dynamically … and the Truck had an increased Roll Angle turning left & decreased Roll Angle turning right.

We have known this for years, before doing 7-post testing. This NASCAR Crew Chief wasn’t testing to see IF it worked … he was just testing to quantify the tire load numbers. The Dodge Engineers said there must be something wrong with the 7-post rig, because that is not possible, quoting various engineering experts & engineering principals. All of their engineering training says the CG Height, Track Width & Roll Center location alone affect tire loads … and the angle of the track bar doesn’t matter.

I do not know why the results conflict with the engineering principal. On the track … as on the 7-post rig … Track Bar angle affects tire load. Maybe it's time to update the engineering formula to take that into account.

-------------------------------------------------------------------------------------------------------------------------------

That example doesn’t mean we don’t value engineering. Just the opposite. Engineering has led to almost all of the advances we see today in modern racing suspensions. But there are a few, VERY RARE situations where the test results we see today … conflicts with what was written years back. I surely don't know why they conflict. When we see something like that, we talk about it a bit & move on.

We are all clear you can not outrun the laws of phsyics & fundamental engineering principals will always apply.

Sometimes the formulas are not 100% accurate. I contribute that to we are all human, even the guys that worked out formulas years ago. The professional engineering community continues to learn new stuff & advance as well as the professional racing community. That is what makes progress in engines, suspensions, cars & just about everything in our lives. Otherwise our cars would perform at the same levels they always did.

There are a few rare exceptions where the formula is not accurate or complete, and having experienced them firsthand, I have learned to keep an open mind.

-------------------------------------------------------------------------------------------------------------------------------

An example of a formula that was considered gospel for years, was the quarter mile drag racing power/weight/speed calculation formula.
It is (.00426 x mph)cubed** x weight = HP
**(I couldn't figure out to put the small 3 where it goes.)

If you enter a car with 1000# & 100HP & compare it to car weighing 4000# & 400Hp … both having a 10-1 W/P ratio … the formula provides the same quarter mile top speed mph (106.76 MPH) for both cars, which is then turned into an estimated ET. In the real world, we know that is not accurate. Even though both cars have the exact same weight to power ratio, the lighter car will produce a substantially quicker ET & slower MPH … and conversely the heavier car will produce a substantially slower ET & faster MPH. Said another way … to achieve 106.76 mph … the lighter car will need more than a 100 HP & the heavier car will need less than 400hp.

The formula was created ions ago … and the creators are dead … so we can’t have a beer with them & discuss this. But all indications are they created the formula based on “averages” … and a simple weight-to-power ratio. They did not build into their formula a correct method … or any method … for adjusting for lighter & heavier cars with the same W/P ratio. So the formula is not an absolute. That’s ok. It is still helpful. It’s just not gospel.

Years later, a formula was created as a correction factor to be used with the original formula. It goes …

a. For cars weighing 2300# use the original formula.

b. For cars lighter than 2300# … Car weight/2300 = X … 1.00-X = Y … Y times .3875 = Z … Z+1.00 = B … B times the result of the original HP in the formula … provides the corrected, real HP to achieve that mph.

c. For cars heavier than 2300# … … 2300/Car weight = X … 1.00-X = Y … Y times .3500 = Z … 1.00-Z = B … B times the result of the original HP in the formula … provides the corrected, real HP to achieve that mph.

Now let’s re-run the calc’s …
The 1000# car actually needs 121.9HP to go 106.76 MPH in the ¼ mile.
The 4000# car actually needs 340.5HP to go 106.76 MPH in the ¼ mile.

This new formula is correct, because some one learned & advanced it. But I still see the original formula being used & relied upon as “fact.”

-------------------------------------------------------------------------------------------------------------------------------

Anti-Roll Bars

In an oval track race car, if we preload the front ARB … it adds “wedge” or “cross weight” (pick your term) to the car statically & dynamically. Meaning it loads two opposing tires. Statically, we can see it on a set of scales. In fact, as a part of most race team’s set-up, they adjust their “Wedge” or “Cross Weight” to a number lower than their final desired % … then add in bar preload to achieve the final target “Wedge” or “Cross Weight” % they are targeting.

Follow me close here … when we do this on the scales … it looks like this:
This car weighs a total of 2570# ... with driver & 175# Lead
Left side weight: 58.2%
Front weight: 48.77%
Rear weight: 51.23 %
*These #’s & %’s do not change with cross weight.

LF 636# RF 618#
LR 860# RR 457#
Cross Weight: 57.49% before preloading the front ARB

LF 616# RF 638#
LR 880# RR 437#
Cross Weight: 59.03% with the front ARB preloaded

The scale under the LF tire gets lighter. This would indicate the LF tire gets worked LESS with a preloaded or bigger sway bar. But dynamically, that is not what we find.

Some people think of testing race cars on pull down rigs and/or on 7-post rigs as “dynamical” testing. I do not. I think of it as “semi-dynamic” testing … because the tires are not rolling, nor are they twisting their contact patches into the slip angle they experience on track. It is helpful … VERY helpful ... to see the loads on the on each corner. In the case of the 7-post “shaker rig” as some call them … the forces of braking, cornering & accelerating are recreated exactly from the data pulled from track testing. It is so accurate the track surface is recreated down to the pebble in turn 3. Teams can test any suspension part to see how well the tire & wheel follow the track surface (contact %) & how much it is loaded in pounds or weight %.

On the 7-post … when the crew installs a bigger, stiffer sway bar, or preloads the existing bar … and the 4 load cells under the tires are monitored on corner entry & middle … the load goes up on the inside front tire & down on the outside front tire … as compared to a smaller, softer bar or a bar with less or no preload.

I think the 7-post teaches us HOW the inside front tires gets loaded in the first place and what keeps it there. When the car is braking hard setting up for the corner, most of the Force is transferring load from the rear tires to the front tires … compressing both front suspension corners & loading both front tires. Because the car is in a slight roll condition ... not driving perfectly straight ... the inside front wheel compresses about 3” & the outside wheel 3-½”. Obviously the outside tire is loaded more than the inside tire. In hard threshold braking, there is little load on the rear tires & a lot of load on the front tires.

Testing shows the bigger the sway bar, the more the inside front tire gets loaded on this braking transition, which is good, because the more we load the inside front tire now, while we have braking force, the better grip it will have on corner turn in.

The question was, if the inside front tire gets loaded on, more or less, straight line braking, what keeps the inside front tire loaded through the corner entry & the mid-corner roll through zone?
The answer is the stiff rebound valving & small bleeds in the shocks … keep the whole front end “tied down.” Otherwise, when the driver lifts off the brakes … the inside front corner would “pop up” pretty quick 7 roll over in the process. In oval track racing the LF is “meaner” on rebound valving. In Road Racing we run both shocks the same, which is why you see a tick more Roll Angle in a Trans Am car than A NASCAR Car. (Look at the splitters).

How do you REALLY KNOW when you’re DYNAMICALLY working a tire more or less?
On the track, where tires are rolling, and they’re twisting their contact patches into the slip angle, the best method of measuring tire load … is tire heat … with on-board infrared tire temp sensors pointed at each wheel. Some set-ups use 3 per tire to measure how well the set-up is working the whole contact patch. Other set-ups just use 1 per tire, pointed at the center. Regardless, the results are the same here.

When the crew installs a stiffer ARB, or preloads the existing bar … and the 4 tires are monitored on corner entry & middle … we see the temps go up on the inside front tire & down on the outside front tire … as compared to a smaller, softer bar or a bar with less or no preload. So fully dynamic ... you can see it in tire temps.

In my experience most all factors in the race car’s geometry, suspension & chassis tend to have an effect on other factors. Some more than others. Some statically … some dynamically … some both. What I shared with you is what we see on highly developed race cars utilizing big soft springs, big bars & stiff rebound shocks … ending up with Roll Angles of 0.75-1.5 degrees. I believe Roll Angle to be a factor. If we let the race car roll too much, the ARB is not capable of keeping load on the inside tires as well, or at all if it rolls too much.

I didn’t want to take the time to get into this conversation, because I knew it would be complex, lengthy ... and is always controversial. I am not trying to be a know it all, because I do not know it all by a long shot. I’m learning every time we test or race. I encourage you to think for yourself & not rely on 100.00% of what you read … my stuff included.

Best wishes !!!

[j-rho]

One more-

Hi Jason & Cobra,

...
In my profession, we have to be careful not to take 100.00% of engineering training as gospel. It was, after all, developed by human beings based on their understanding of how things work … at the time they developed the principal … with the technology of the day. We occasionally … VERY RARELY … but occasionally run across something that conflicts with what was written by Engineers … even top Engineers.

This discussion of ARB loading tires is one of those areas.

-------------------------------------------------------------------------------------------------------------------------------

Another is Track Bar/Panhard Bar angle. A few years back, a team was testing a Dodge NASCAR Craftsman Race Truck on a 7-post rig … and the Crew Chief was testing the Track Bar at different angles for an upcoming road race, but with the center always at the same height, so the roll Center stayed the same. He started with a baseline with the left side at 10” & the right side at 10” and ran the test, producing a baseline set of load numbers that were even on the left rear tire & right rear tire.

When he put the left side at 8” & the right side at 12” & tested it … the 7-post rig computer showed more load on the left rear tire & less load on the right rear tire … dynamically … and the Truck had a decreased Roll Angle turning left & increased Roll Angle turning right. When he put the left side at 12” & the right side at 8” … the 7-post rig computer showed less load on the left rear tire & more load on the right rear tire … dynamically … and the Truck had an increased Roll Angle turning left & decreased Roll Angle turning right.

We have known this for years, before doing 7-post testing. This NASCAR Crew Chief wasn’t testing to see IF it worked … he was just testing to quantify the tire load numbers. The Dodge Engineers said there must be something wrong with the 7-post rig, because that is not possible, quoting various engineering experts & engineering principals. All of their engineering training says the CG Height, Track Width & Roll Center location alone affect tire loads … and the angle of the track bar doesn’t matter.

On the track … as on the 7-post rig … Track Bar angle affects tire load. It’s time to re-write the engineering formula to take that into account.

-------------------------------------------------------------------------------------------------------------------------------

Hi Ron,
When the test was done, what happened to loads on the front tires? Wasn't there an equal and opposite change in load transfer across the front axle?

When the crew installs a bigger, stiffer sway bar, or preloads the existing bar … and the 4 tires are monitored on corner entry & middle … the temp goes up on the inside front tire & down on the outside front tire … as compared to a smaller, softer bar or a bar with less or no preload.

So fully dynamic ... we see it in tire temps.

Now … in my experience I have learned that most all factors in the race car’s geometry, suspension & chassis tend to have an effect on other factors. Some more than others. Some statically … some dynamically … some both. What I shared with you is what we see on full blown, highly developed race cars utilizing cutting edge suspensions … ending up with Roll Angles of 0.75-1.5 degrees. I find Roll Angle to be a factor.

In cars with much higher Roll Angles above 2.5+ degrees … I think the original engineering principal may apply. But I don’t know … because I‘m not involved with high tech testing of those types of cars.

I bolded that last bit just to call it out - those are just the sorts of cars Matt is asking about, and that the people here build and drive.

What I suspect is happening, are a number of other contributing factors come into play in your situation (could be aero, or a change in the bump stop engagement) that combine to trump the fundamental effects of the original change, producing an end result (hotter inside front tires) opposite of what a pure and simple "textbook scenario" would say. It's not that either is necessarily wrong, or even in contradiction; it's that in real life we never really get to just change one variable at a time in an experiment, and the other things that change with one variable, have an effect on the outcome. However, your observed cause-and-effect relationships are likely somewhat specific to your situation, not necessarily the same as would be seen in all (or most of the time in typical street) situations, and perhaps not the best representation of the basic mechanics of these sorts of changes, to a beginning audience like the Pro-Touring forum members. I bet some Pro-Touring cars roll 2.5 degrees just when the driver sits down!

[j-rho]

Last one from the rollbar thread-

Hi Jason,


Hi Ron,
When the test was done, what happened to loads on the front tires? Wasn't there an equal and opposite change in load transfer across the front axle?
Yes. For example, when the track Bar was 8" on the left & 12" on the right, and the RR was loaded less, the LF was loaded more.

So if there was an equal and opposite reaction on the other axle, wasn't the total amount of weight transferred to the outside (weight transferred to outside front + weight transferred to outside rear) the same between the different scenarios? Or at least, close enough that a discrepancy could be explained by a second-order effect like a change in resulting CG height?

[Ron Sutton]

Hi Jason,


Hi Ron,
When the test was done, what happened to loads on the front tires? Wasn't there an equal and opposite change in load transfer across the front axle?
Yes. For example, when the track Bar was 8" on the left & 12" on the right, and the RR was loaded less, the LF was loaded more.


I bolded that last bit just to call it out - those are just the sorts of cars Matt is asking about, and that the people here build and drive.
True.

What I suspect is happening, are a number of other contributing factors come into play in your situation (could be aero, or a change in the bump stop engagement) that combine to trump the fundamental effects of the original change, producing an end result (hotter inside front tires) opposite of what a pure and simple "textbook scenario" would say.
It's not aero or bump stop engagement, because we see it without bumps stops & on the 7-post, where there is no aero. It is connected to low roll angles, because when we get too much roll angle, it stops working.

It's not that either is necessarily wrong, or even in contradiction; it's that in real life we never really get to just change one variable at a time in an experiment, and the other things that change with one variable, have an effect on the outcome.
Too true. Everything affects everything else, which is what makes this challenging & fun.

However, your observed cause-and-effect relationships are likely somewhat specific to your situation, not necessarily the same as would be seen in all (or most of the time in typical street) situations, and perhaps not the best representation of the basic mechanics of these sorts of changes, to a beginning audience like the Pro-Touring forum members.
I agree for almost all passenger cars & most PT cars.

But Matt's situation with his car was he made his suspension too stiff and started fighting a pushing problem. It's a Roll Angle vs Pitch angle equation. A performance or race car meant to handle ... can’t run flat … it needs the suspension to travel. So it’s either got to Roll more and lessen the grip on the inside tires ... or Pitch more and lessen the grip on the rear tires … to turn well. Matt's was neither rolling much or pitching much.

That’s the primary difference between Conventional vs High Travel set-ups. Conventional set-ups pitch less & roll more. High travel set-ups roll less & pitch more. Matt's spring & sway bar combination was simply not traveling enough either way. It is less a problem on big tracks with fast, sweeping corners & more amplified on small AutoX tracks.

I bet some Pro-Touring cars roll 2.5 degrees just when the driver sits down!
That's funny ... because it's true.

[Ron Sutton]

Jason,

We had two topics going on in our previous discussion. One about track bar angle effects & the other about front sway size affecting loads on the inside front & inside rear tires. Which conversation do you want to start with?

[j-rho]

Lets start with #1: I contend total lateral load transfer at a given lateral acceleration is governed by two things: CG height and track width. You believe that roll angle plays a part - that a car that rolls less, transfers less weight, even if CG height and track width are kept constant.

#2 will be the contribution of sway bars. I contend stiffening a swaybar at one axle, increases the amount of weight transferred across that axle at a given lateral acceleration, while decreasing the amount of weight transfer across the opposite axle. But as above, total weight transferred to the outside does not change. My post #5 in this thread is an example of this question.

[j-rho]

If we're addressing common fallacies and misconceptions, another one I see:
When one stiffens their shocks, there is no question that the visible transient motions of the car are slowed down. But does that shock stiffening decrease the actual rate of load transfer between tires, or increase it?

The reality is that stiffening the shocks increases the rate (the speed) of weight transfer. It doesn't reduce quantity of weight transfer. However it is very commonly believed (even by a few of the shock vendors selling stuff on this site) that stiffening shocks slows weight transfer, and some appear to believe it reduces the total amount. I believe this misconception goes hand-in-hand with the ones we're talking about above -that reduced roll (and its side view buddies squat/pitch) are indicators of reduced weight transfer. It's based upon a flawed but fundamental belief that the chassis' displacement of each relative corner must be tied, at that moment in time, to the load present upon that corner's tire. Once people get over that, their depth of understanding can rapidly advance.

[Ron Sutton]

Hi Jason,

First, for us to converse well, I need us to speak more "car guy" and less engineer. I'm not an engineer. I'm a car guy/racer that happens to design, build & tune winning race cars. In my racing world, the Engineers & I respect each other backgrounds because we both contribute to this overall goal, so they tolerate my simple language & at times, I find myself trying to understand what they're saying. I use some engineering terms ... correctly ... and sometimes I use them wrong. For Pete's sake, let's not discuss polar moments. I understand it, use it, live & breathe it ... but don't explain it correctly. This drove my lead engineer nuts.

One of my other engineering friends said, Ron is a race car designer that did not go to engineering school, so he speaks "car guy". I think it was a compliment, but I'm not 100% sure. Engineering speak is like French to me. I understand some of it, but if I have to go look it up, I'm not having fun.

Lets start with #1: I contend total lateral load transfer at a given lateral acceleration is governed by two things: CG height and track width. You believe that roll angle plays a part - that a car that rolls less, transfers less weight, even if CG height and track width are kept constant.

Let's use terms like force, load, weight, weight transfer (not literally), working the tire, suspension travel, roll, pitch, dive, lift, rise, rate, preload, etc.

Roll & pitch play a role in loading & working the tires ... and at the end of the day ... all I really care about is working the tires in an optimum fashion, for optimum performance.

When we brake hard, in a soft sprung front end race car, and the front end suspension travels far & fast ... and the rear end rises similarly ... creating a higher degree of pitch angle ... the front tires are loaded & working more ... while the rear tires are loaded less & working less ... than if we have a stiffly sprung front end race car, and the front end suspension travels little & slowly ... and the rear end rises similarly ... creating a smaller degree of pitch. The same concept works with roll angle.

What part of that do we not see eye to eye?

[djfxall]

Awsome thread getting into shocks is like opening a can of worms along with sprung weight I know on a bike you can load the suspension into a corner and use that for exiting the corner. I am new to the 4 wheel race world but valving controls the speed of the compression allong with decompression. Is there dual speed shocks for cars which have different rates for high speed compression?

[Ron Sutton]

Awsome thread getting into shocks is like opening a can of worms along with sprung weight I know on a bike you can load the suspension into a corner and use that for exiting the corner. I am new to the 4 wheel race world but valving controls the speed of the compression allong with decompression. Is there dual speed shocks for cars which have different rates for high speed compression?

Yes. We can tune shocks internally with parts ... or externally with knobs ... in ways that control the low speed valving (where handling is) and high speed valving (where bumps are controlled). Of course 2, 3 & 4 way shocks are spendy.

If you'd like to discuss it, let's start another thread. We're created this one to work on reasons "why" track & test results conflict with proven principles occasionally.

[j-rho]

Let's use terms like force, load, weight, weight transfer (not literally), working the tire, suspension travel, roll, pitch, dive, lift, rise, rate, preload, etc.

Roll & pitch play a role in loading & working the tires ... and at the end of the day ... all I really care about is working the tires in an optimum fashion, for optimum performance.

When we brake hard, in a soft sprung front end race car, and the front end suspension travels far & fast ... and the rear end rises similarly ... creating a higher degree of pitch angle ... the front tires are loaded & working more ... while the rear tires are loaded less & working less ... than if we have a stiffly sprung front end race car, and the front end suspension travels little & slowly ... and the rear end rises similarly ... creating a smaller degree of pitch. The same concept works with roll angle.

What part of that do we not see eye to eye?

Let's say we have two cars almost exactly the same - both have the same CG height. One is softly sprung, but when it brakes, the sinking front and rising rear combine to keep the CG height unchanged. The second car is different only in that it has a totally solid suspension that can't move at all.

Both cars brake at 1g. The soft car's nose dives and its rear end rises. The stiff car doesn't move at all. Which car transferred more weight to the front tires?

You'd say of the soft car, "the front tires are loaded & working more". I say, the front tires of both cars are equally loaded, because they have the same CG height, wheelbase, and are both braking at 1g. This is just basic physics and nothing discovered in the last 10 or 20 years makes this basic principle any less valid under any circumstance.

This isn't to say both cars would work equally well on a racetrack. A simple example like this is only used to explain the fundamentals of what's happening with weight transfer. In reality there are many variables, and sometimes when you make a change it may have a net effect (after all other affected variables are considered) opposite of what a simple textbook says happens when just that one variable is changed in isolation.

A classic example is stiffening a front swaybar on a crummy car - the textbook says doing so will increase relative weight transfer at the front tires, loading the outside front more and inside front less, which will reduce total grip of the front axle, which will shift the car's balance towards understeer. But in real life, your crummy car might have really awful front geometry, where a stiffer front swaybar does the opposite - shifts balance towards oversteer - because the additional roll stiffness of that bar, kept the resulting alignment of the outside front tire at a "happier" place, resulting in more grip from the front axle than before. It's not that the swaybar somehow magically wasn't having its fundamental textbook-defined load-transfer-altering effect; it's that a second-order effect (resulting alignment change) altered the equation in a way that trumped the change's primary effect (load transfer change).

In my experience there is no such thing as track tests conflicting with proven principles. There is only failure to observe and properly account for all the variables that go into the equation. Technology has improved our ability to measure, record, and model the contribution of these variables to the end result, and as development progresses under ever-changing rulesets, we often see the importance of certain variables overtaking others in the process.

Ron, you aren't the first person I've come across to have a successful career in something, without having a very thorough understanding of what was going on under the covers. While that probably sounds disrespectful it isn't meant to be, and speaks to the importance of a combination of other great qualities like your kind personality, hard work, dedication, and wisdom to choose the right employees to compliment your own abilities. I thought it might be interesting to hear about more cases like the swaybar example I gave above, but it is clear now we aren't even starting with a compass of solid fundamentals, which will make progress in exploration of apparent real world vs. theoretical contradicions, a wild goose chase.

Last tangential note not to Ron but to those reading - I don't normally post much on message boards these days, but one thing that gets me fired up is people selling you junk, or the wrong parts. It's part of why I put so much effort into my blog - I want to help everyone learn, so they can get better, and be better competitors, and maybe turn into people I can learn something from someday. The Pro-Touring movement as a whole I see as a big community, hungry to learn about vehicle dynamics and handling and driving well. Not all vendors, but many, don't really know what they're talking about, but won't let that stop them from taking your money. Educate yourself on the basic principles of tuning your car's handling. Make your own choices, think for yourself. And practice driving! You can't figure out what car changes to make if you don't know how to drive and evaluate present state.
http://www.rhoadescamaro.com/build/?p=338

[Ron Sutton]

Zing! Duck! Ouch !

Help! I’ve been insulted & can’t get up.

Jason I know the formulas for calculating g-forces & weight transfers too.

Let’s take the formula for Acceleration Weight Transfer, which is:
(G x W x CGH)/WB = T

Let’s take a real car, like the one in the photo below, that has a 109” wheelbase, weighs 3000#, with 54% on the front (1620#) & a CG height at 16”, launching with 2.0 G’s.

The formula says (2.0g x 3000# x 16”cgh)/109”wb = 881# of weight transferred from the front end … to the rear tires. But the front end weighs 1620# … and it’s in the air … and all of the weight of the 3000# car is on the rear tires. Will you be my huckleberry and educate us all on how that computes?

Feel free to change the numbers around, but keep them realistic.





---------------------------------------------------------------------------------------------


While we’re having fun, let’s take this cool PT car at the AutoX track. (LOVE this car.)





The formula for Cornering Weight Transfer is:
(G x W x CGH)/TW = T

Let’s say it weighs 3200#, has a track width of 58”, a CG of 16” & pulls a g-force of 1.2.

The formula says (1.2g x 3200# x 16”cgh)/58”tw = 1059# of weight transferred from the left side … to the right side tires. But the left side of the car … with the driver, weighs at least 1600# … and it’s in the air … and all of the weight of the 3200# car is on the right side tires.

How do we use the formula to make that compute?

[CA B4C]

Gentlemen, thank you very much for starting this thread and discussing vehicle dynamics in a professional manner. I truly appreciate each of your points of view and it kind of reminds me of the ol Guldstrand vs Adams debates from the 80's.

Ron, in the photo of the SpeedTech Nova, I think you would have to use a launch equation and not a lateral weight transfer equation. It appears from the photo there is an elevation change in the course. Granted photos are only 2D and freeze a moment in time.

Again, thank you very much for the highly intelligent discussion.

[Ron Sutton]

Gentlemen, thank you very much for starting this thread and discussing vehicle dynamics in a professional manner. I truly appreciate each of your points of view and it kind of reminds me of the ol Guldstrand vs Adams debates from the 80's.

Ron, in the photo of the SpeedTech Nova, I think you would have to use a launch equation and not a lateral weight transfer equation. It appears from the photo there is an elevation change in the course. Granted photos are only 2D and freeze a moment in time.

Again, thank you very much for the highly intelligent discussion.

No worries. And no on the elevation change ... it's flat at Irvine.

[CA B4C]

This was at El Toro also, pretty good elevation change at speed