Heh, true but either value works out to be the same change since A *cos(alpha) = C and the ratio cancels out the trig function. I will say though, the same measurements need to be used, if the hole to hole distance is measured then it is more accurate to measure Leg Length ("C" as Iowa999 pointed out).

BTW, if anyone wants to double check my work I'll snap a picture of it. My Linear Algebra is rusty and there certainly could be an error.

 

The two parts of the denominator of Puhn's formula relate to different things. The ".4244 A^2 B" part concerns the leverage for the twisting of the bar (which is where the A^2 part comes in) and the amount of bar being twisted (which is B). The ".2264 C^3" part is actually the bending of the arms. A gets squared simply because you have arms on both ends ... thus, two of them. C gets cubed because the ease of bending a bar depends on cube of its length.

For completeness sake: the reason that D is raised to the fourth power (in the numerator of Puhn's formula) is that the resistance to twisting of a round bar is proportional to the fourth power of its diameter.

edit: I should note that the above is actually beyond my knowledge of physics and could be one of those rules of thumb that is technically wrong ... please take with grain of salt and/or feel freer than usual to tell me that I'm wrong

 

You are correct on what Puhn's equation is doing, I basically just took his equation and set it equal to itself times a percentage gain. Then assuming D is constant I canceled out what I could, pulled out like values, solved for P and made some linearization assumptions. So I took his complete equation which factors in everything he does to simplify it.

Also, I added A * cos(alpha) = C so I could get rid of one of the values.

The assumption I made was :
(0.4244*B + 0.2264*A1/cos(alpha))
-------------------------------------------- ~ = 1
(0.4244*B + 0.2264*A2/cos(alpha))

The reason this is true is because the 0.4244*B is much larger than 0.2264*A. Try plugging in some numbers to validate this and you'll see.

So the change is left with the factored values of A1^2/A2^2 or C1^2/C2^2.

 

Yep. Totally agree. And because what most people do is either change bar thickness and/or change the lever-arm length, your two simple formulae in your previous post for seeing what the changes will do, in percent, are probably what 99% of all people need.

But I, at least, need more. I need to know the OE bar's contribution to both single-wheel rate and roll resistance, because I know that I need a slightly lower single-wheel rates when the bars are disconnected (for gravel) and much more total roll stiffness when the bar are connected (for tarmac). That's why I've been working through the details. If this isn't the thread to do this in, I'll run back to the Evo X sub-forum. But, I have to say, you folks with older Evos are way ahead of those of us with Xs when it comes to have the raw data required to do the math. And I really, really don't want to go through the hassle of measuring the car ... it took two weekends to measure the DSM and then all that happened is that Dennis said my numbers were wrong but refused to say which and/or share his.

edit: in case you don't know, I'm also jtmcinder

 

Is the OP still around ?

Would have been helpful if he noted why he needs this info rather than asking some off the wall question that few would know

 

FWIW, most recent recommendation from Ralliart Italy for Evo IX recommends 27mm FSB for gravel and 29mm for tarmac using softer spring rates than previous - what the latter are I do not know. Rear bar is also beefy in both instances, forget how thick exactly.

 

Weird. Everything that I've read says that a car set up for gravel should be getting less than half its roll resistance from the bars. As a point of reference, a bone-stock Evo X gets about two thirds of its roll resistance from bars which is why most people disconnect at least the front for rallycross, if they've done nothing else to the car.

 

Yeah I'm still here,the thread came alive and ive been busy trying to make money lol
The reason im asking is there are aluminum swaybars and hollow steel bars that have obvious advantages over the regular steel stock shaped bar,id rather not get into specifics of exactly why i need this info but these bars are measured in llbs and i need that rate as a starting point for my project
i should also mention that this is for a full blown track car

 

what im personally after, is to have a complete understanding of my new setup. Ill be switching to ohlins soon (fingers crossed) with 8k/10k springs, which given the approximate rear motion ratio gives equivalent spring rates front and rear. I run whiteline front and rear bars, and like to keep the car as neutral as possible for now due to my rookie track experience. thus, id like to understand how much "spring rate" the bar contributes to the overall rear number.

much more basic than what everyone else seems to be talking about doing! lol

 

Pounds per what? Because if it's just pounds, they are merely jumping up and down about how light the actual bars are which is just about irrelevant. Something that low on the car can be as heavy as it wants to be. Hollow sway-bars are, IMO, a gimmick.

 

I think the Ralliart is the only 29mm FSB I've ever seen. I'm really not sure why they prefer more bar and less spring for rally setups.

 

this is the article that gave me some ideas and a way to calculate although a very long way of doing it,mitsubishi must have a starting rate for the stock bar
http://www.circletrack.com/chassiste...s/viewall.html
the llb rate has nothing to do with the weight of the bar,the bars are measure at 5 degrees of twist and how many llbs it takes to twist it to 5 degrees,much like rating a spring and how many llbs it takes to compress it 1"

 

Im pretty sure what he was eluding to was that you needed proper units. How would we know you mean lb/5deg. I know we all say 700lb springs but its commonly accepted that its lb/in and we're all just too lazy to type it out. lb/5deg is an uncommon unit.

 

i mentioned it in a previous post,i guess ill keep searching for the answer

 

Got it. But the standard is actually pounds per degree, not pounds per 5 degrees.

The reason that most people just report pounds per inch (from the Fred Puhn formula) is that pounds per degree requires that you know the track of the car, which depends, of course, on wheel offset. Pounds per inch only requires that you know the motion ratio of the swaybar, which is rarely if ever changed.