What's your procedure for leveling scale platforms?

[dereklola]

OK - now that we've decided how to level our platforms to .001" etc etc - let me throw in my usual rejoinder -

We never turn turn left and right at the same time so what's all this crap about must have this aerospace accuracy left to right?

Assume we do achieve it what's the advantage? The track is never level or otherwise equal under both left and right sides of the car? And all tracks have different importance of left and righthand corners.

Not suggesting we can be way out (weigh out?) but extreme accuracy is of no significant benefit.

First taught to me by a well known F1 designer.

Fire away!

Derek

[formulasuper]

I have to admit I have been thinking the same thing while reading all this about scales and accuracy, but was afraid there must be more to it than I understand. Scott.

[Purple Frog]

It is for consistency in setups. Period.

If you start with the same platform base for each setup, then you will know which setup you are putting in the car. Even if the car worked best with 55% weight on the right side last year at a certain track, then this year you want to dial in 55% again when you go to that track, not 56% nor 54%.

If the car is working sweet. That's when to put it on the pad and measure EVERYTHING. Next time at that track, you can't put all those sets back into the car, unless the pad is set up the same way as before.

[R. Pare]

Derek;

I'd have to ask just how long ago that guy was in F1, as accuracy and repeatability is of paramount importance. A world class driver can feel the difference of only a couple of pounds of cross weight and has to adjust his driving accordingly. While we at this level of driving skills may have a hard time discerning a change of 20 pounds, the effects are still there.

To illustrate, a story I sometimes tell in our seminars:

A friend here in town has Mario's '79 Lotus in his shop. A publicity photo shoot with the car and Mario was done at Watkins Glenn 2 years ago. The car was set up exactly to the original Glen setup sheet. Mario took a few laps, came in, and said that he thought that there was more cross weight in it than he might like, that it felt like about 15 pounds heavy on the right front. Sure enough, that's what the setup sheet called for!

Mario wanted to dial it out to see if it would get the effect he wanted. Greg (the mechanic) said that he hadn't brought a setup pad, and wasn't sure that he could change it with any accuracy. Mario just said "No problem. Do this, do that, and I'll take a lap to see where we are".

After two tries, Mario was satisfied that he had it within "about 2 pounds", and proceeded to have some fun with the old beast.

When Greg got the car back on the scales at his shop a the next day, sure enough, the cross weight was 2 pounds!!!!!!

Gads... I'm gonna be sick...........

[RussMcB]

I would think that the biggest reason for getting the cross weights very close to 50% would be so that the car reacted similarly in left and right hand turns. For instance, if you have 20 lb. more weight on the left front, then the car might turn in better initially for right hand corners, but understeer ultimately because porportionately more weight has transferred on the front compared to the rear. Conversely, the same car in a left hand corner might not turn in as quickly, but might have a tendency to oversteer in the corner.

Admittedly, I may be full of it, but if that's true, it's got to be a little tougher for a driver to comfortably get to the edge of performance.

I don't know what .001" equates to lbs. in cross weight. My hunch is that super accuracy isn't needed when leveling platforms. If I weigh the car in both directions and the difference is less than 10 lbs., I try to be satisfied (but it is hard to stop trying to get it exactly the same).

Something I've been thinking about, but probably isn't realistic, is to create a rig that could be set on the scales to check the cross weights in both directions. If I could create a jig, something like a four legged stool with known and repeatable weight jacking, that would be a better measuring tool than the race car itself, because it would not have springs, friction, etc.

What do you think of that idea? Silly? Or worth thinking about more? It appeals to me because my driveway is cramped and my race car's turning radius is about a half mile, taking about 20 "k" turns to get it turned around. I can use a floor jack under the rear wing to move it around quicker, but then I've introduced another possible variable when the car gets loaded back down onto its suspension.

Food for thought ...

[R. Pare]

Sensetivity to every .001" of off-level depends on the contact patch wheel rates ( wheel rate due to the spring in series with the tire rate). For every 100#/in of rate, every .001" of off-level will produce .1# of cross weight (100/1000). For a decently stiff FC with 450#/" contact patch rate at the front, that's 4.5 x .1 or .45 pounds. For every .010", that's 4.5 pounds, etc.

We have found that cross weight changes have a very measurable affect on reletive tire growth (bias ply tires, of course), which in turn affects how the car balance changes thruout the course of the race. At certain tracks, it may actually be desirable to bias weight one way or the other so that the car is continually "coming to you" during the race, rather than start off perfect and have it go away. A change of only 10 - 15 pounds can make all of the difference, and 18 pounds can be too much, causing the growth to be opposite of what you want.

So, if 15 pounds is the number you need, and 18 is too much, how important is it then to be accurate and consistant?

Very!

[Dane Whaley]

I can't help but to wonder if all this focus on leveling isn't the answer to Racer Russ's problem. You seem to have the pads leveled OK. Are you backing off all the damping settings on the shocks, and bouncing and rolling the chassis to 'settle' the car upon initial placement and after every tiny adjustment? it's not unusual for 20 lbs crossweight to vanish once the car is settled.

[Jim Garry]

I solved my problem when I discovered that one of my scales was a different height than the others. I shimmed it and now scaling works out fine.

[formulasuper]

Is it just me or have any of you guys noticed that racing slicks are not always perfectly round, even when new? Would this not have an effect on leveling? Scott.

[RussMcB]

Originally posted by formulasuper:
Is it just me?

Scott, you know when you go screaming into Road Atlanta T10a and stomp on your brake pedal with all your might and the chicane is enveloped in smoke for the next few laps? Don't use those tires when setting up your car on the scales. .

All seriousness aside, that's a good point. I think it's a good idea to at least check the circumference of tires to see if any are out of line.

[carnut169]

it's a good idea to at least check the circumference of tires to see if any are out of line.

Yes, everytime. You write the measurement on the sidewall so you know that if you are changing tires and want to keep the car balanced you know exactly how much to raise or lower it... and you might not even need your platform if you have previously figured what a flat does for you in terms of height...
Can someone post the formula to figure height from circumference?

[formulasuper]

Ya, ya, ya, I know about circumference difference between different tires. I was at the seminar. I'm talking about out of round, not perfectly round, meaning taller on one side of the wheel than the other. Why set up to the .001" leveling when the other side of the wheel is 1/16" or more different height?!?

[RussMcB]

[quote]Originally posted by carnut169:

Can someone post the formula to figure height from circumference?

I'll take a stab.

Circumference divided by Pi gives diameter, then divide that by two to get the radius.

I think you may still need to do some calulating after determining radius, depending on what you're checking. For instance, if you put on a new tire on who's radius is 1/4" taller than the one coming off, I'd guess you'd raise that end of the car by half that much (1/8").

Did I get it right? It's been a few decades since a teacher tried to learn me that stuff. Hyuk.

(PS. Scott. I knew you knew. And I was pretty sure that you knew I knew you knew, you know?)

[formulasuper]

Russ, Ya,I knew you knew I knew, but you didn't answer my question. What about the not so round tire issue? scott

[RussMcB]

Well, I don't know if I know the answer for sure, but my gut feeling is that if the tire is out of round enough to throw off corner weights, then you'd feel it when you were driving. That hasn't happened to me in quite a while, so my guess is that the tire companies must be doing an okay job at getting them pretty round.

[formulasuper]

RacerRuss, Have you ever had your car off the ground and gave one of the tires a spin? They are not perfectly true. Like not .001" nor even close. Can be as much as .125" variation from round, (and Goodyear says this is normal) so why try to level to .001" when you don't know if your tire sitting on the scale is the tall side or the short side?

[RussMcB]

Good point. I guess it's not hard to imagine that being true.

I'll venture a WAG.

Maybe even with that large amount of runout, once the car's weight is on the tire, since the tire is not a rigid object, the tire's sidewalls flex, allowing the tire to transmit all of the weight it "feels" from the suspension. The scale pad sees that transfered weight, plus the unsprung weight of the tire/wheel.

[formulasuper]

You should be in politics, but I'm not buying your WAG. Any engineers out there have an explaination for this? Scott.

[Tom Sprecher]

I'm not an engineer but I play at work, wait, I quess that makes me one. Anyhow, I think the whole idea of leveling you scales to whatever precision, as Purple Frog mentioned, is to ensure repeatability. If you have a level and can get a beam that can go to .001 then why wouldn't you. If it can only do .005 then so be it.

The thing is since this measurement is the basis that all other chassis set-up will be measured from it's a good idea to get it as accurate and precise as possible. A couple of thou off the precision may or may not make a difference, it depends on if you looking at it relatively or as an absolute.

Also, I think I read somewhere that to do it right a nice new shiny set of unused tires should be used to do your measuring with. They should be saved and used every time as well, like we all have a tire budget for that.

In operation the tire will flex, roll, deflect, etc. due to track conditions so again it's that relative or absolute question again. I think it's easier and more relevent to work with the relative over the absolute.

[Scott Hanba]

Another option would be to not use tires, but bolt a support to each hub that represents a nominal tire size. This was shown in a recent Race Car Engineering in an article about how Jordan sets up their F1 car. Then you would just have to accept any error due to tire size variation. Their set-up was slick in that they don't use scale pads, but build load cells into the support that bolts to the hub.

[Purple Frog]

Tom,

I'd rather drink the absolute in the presence of relatives. [img]graemlins/skull.gif[/img]

[Tom Sprecher]

That's a good point that I think everyone overlooked. Also, we should not forget about presents from relatives of absolute. [img]smile.gif[/img]

[formulasuper]

Mr. Hanba, Thank you. Now I remember seeing the pictures of metal disk used to replace the wheels and tires during set up on F1 cars and it all starts to make sense. Scott.

[R. Pare]

The problem with using either "control tires" or setup disks is that, at least with bias ply tires, your corner weights and ride heights can/will be off, and maybe way off, due to differing diameters and subsequent growths.

You CAN do that with radials.

[RussMcB]

Originally posted by formulasuper:
Is it just me or have any of you guys noticed that racing slicks are not always perfectly round, even when new? Would this not have an effect on leveling? Scott.

Scott, check out Wooly's post about the Sherline alignment system in the FF section. Pretty interesting. It eliminates tire and wheel variables.

http://www.apexspeed.com/ubb/ultimat...099;p=1#000001