Ralphy: 
   Have you ever driven a 289 Cobra, Ace or a Zagato?   I have .  samples of all three.   Many times.  They actually handle and drive unbelievably good.  Even with the skinny bias ply tires that were on the ones I drove ,  they gave the driver a sense of confidence and were quite predictable.  You knew just how far you could push them before the tires would break loose.  Then if you knew how to drive a corner with the throttle you could take it to the next level and push a drift with confidence.  The cars were ridiculously light and responded in a predictable manor.  They felt down right nimble and were so incredibly neutral that the only way I can describe the way I felt driving them hard was confident.  This type of suspension was crude and the leaf spring playing a dual role as a weight support and UCA. would seem to flex too much.  However the less than 2100 lb weight of the Zagato and similar weight of the Cobra, they did not exceed the limits of this design.  For those who do not recognize the Zagato,This car was built by AC cars of England.  The Zagato was a hard top and the Ace was a roadster but aside from that they were the same car. The Ace looked identical to the 289 Cobra but had a different nose.   Underneath the skin the frames and suspensions were identical. The engines in both the Zagato and Ace  were  all aluminum 6 cylinder Bristol aircraft engines.     The roof line of the Zagato was like a fast back with lumps in the roof to accommodate the drivers and passengers heads similar to the late model Corvette's roof line.   
       Now when the 427 was to be installed in the Cobra Because the Corvette was to be equipped with the ZL1 aluminum 427 the following year.   Planning ahead they knew they had to re-designed the chassis and suspension before  combine the two.  The leaf spring design was abandoned entirely.  Heavier tubing was used in the ladder frames and a heavy duty suspension with coil overs were used at all four corners among many other changes.  In short, the 427 was never installed in a 289 type car. The rear suspension that you posted with out a leaf spring was a common design for the time period.   It was not invented by Klaus.   It was used in many different F1 cars since the late 50's and everyone that used it fought the inherent quirky handling it provided.  Lotus and Ferrari both struggled with this same design and were able to solve some of it's flaws.
      BTW  its roll OVERSTEER that was the hated part of that type suspension.   Nearly every factory IRS sold to the public had some roll under steer designed in for a safety factor to novice drivers.  The eary 427 Cobras had serious issues.  They took a light little 289 sports car that the men who raced them loved the neutral feel and even though they had less horse power  they could do circles around the competition.   Then they added a engine designed as a truck power plant, added 500lb steel to the frame and suspension, and a ill handling IRS.     What they made was an over weight pig of a sports car that was fast due to shear horse power and little else.   The427 Cobra wasn't undrivable, rather when compaired to the 289 it was a tank.  It had none of the predictable neutral steer characteristics the 289 car was known for.  It was no longer nimble.    It required an entirely different style of driving technique from the 289.   The drivers  relied on the throttle to create the neutral drift effect and due to this was totally unforgiving to driver error.   It  no longer had a designed in neutral steer feel what so ever.   None of the light .nimble neutral feeling little sports car remained.  It became a hairy chested brutus   like monster in weight and feel.     The races they won with the 427 were the result of guts and lots of hp. not superior handling!

Last edited by tyrellracing (1/21/2012 4:13 am)

TR,
That's some really good feedback and writting in those last three posts to me. As far as your forward lower link. What I am trying to say is that if that link given the same angle but shortened to half length say. Attaching to the chassis one foot further back or so,  would tend to give a very less favorable cornering effect. I like the fact it goes that far forward.

I think that another point not mentioned or the name does not imply with anti squat. That when braking the same angles that create anti squat, help hold the rear of the chassis down. Aiding the anti dive of the front end in a braking condition. Although the mass of an IRS sprung weight is not a lot. However, it's more.
Ralphy

Last edited by Ralphy (1/21/2012 7:35 am)

That's what a kinematic study is.  You incrementally cycle the suspension when drawn or construction model in 3d.     The LCA controls the toe setting on the free body diagrams. The way to think of these interrelationships is like the bmw or datson 510 trailing arm where the arms are angled away from parallel with the center line of the car and will have a large camber and toe change with wheel travel.  Most people who race these cars make their suspensions very stiff to hold them in as close to ideal as possible.  The change is so dramatic that very little wheel travel is required before the tire grip begins to deteriorate from incorrect camber and or toe.   Your roll steer would have all the same change in toe with wheel travel as the trailing arm  and I am failing to see a benefit.
     A year ago I downloaded some images of a front full race suspension.  I then used the images to scale drawings copied from Griggs racing.    They are rather proud of their suspension packages that can make any  year mustang produce 1+ lateral G's.  They show a Griggs stang cornering at 1.5 G's on DOT tires.  The GR350's are available for 65-70 .  This kit goes for over$ 9000.00 for the front.    Google GriggsRacing and click on his photo gallery. Look at the  65 fast back  and its  LCA mounting angle  I made a copy of my own.  I have a spare 67 mustang nose section with this copy installed, level, and aligned .  I have been doing all this testing off the car because I have to cut off all the original motor and lower suspension mounts to install..    The primary concerns are measuring bump steer , Camber gain,and caster change.    This project was put on a back burner at the end of last summer

As to your $64000  question.    I have the car on a center lift with the wheels at full droop.  This gets them out of the way for painting.   My best guess on the leading link would be between 25 and 30 degrees from the bottom of the hub carrier up to the frame at normal ride height.    The trailing link goes down  from the top of the hub carrier to the frame aprox. 10 degrees The live axle links were at a similar angle but to compair the two would be pointless.  The live axle uses the rotational force equal and opposite as the force required to turn the tires to act on lever arms welded to the bottom of the axle tubes vertically 6 in. long.  On an IRS all these loads are kept in control at the frame.  none of it is applied to the hub carrier.     With the live axle under acceleration the rotation  force not retained by the leaf springs is directed to the frame just ahead of the spring eye. through the lever arm and link.   The hub carrier has no rotational force to do this .  The forward force is divided between the leading and trailing links.  Since the trailing link attachment point is much closer to the axle center line, it gets more load then the leading link.  The leading links upward force gets smaller every time I look at just what it does.

Last edited by tyrellracing (1/22/2012 5:46 am)

Its nowhere near as tough as Daze thinks!  (when you have a big lathe)
      The inboard brakes were a problem.  Having rotors that were wider and 1in. larger in diameter only made that problem more severe.  The solution was straight forward.  Since the shape of the control arm makes little difference as long as it is ridged and holds all pivots in their assigned places.  It can look like a pretzel with out problem.    My solution was two fold,  First My rotors may have been larger but they incorporated almost  no off set.  The  rotor is directly over the companion flange and due to this further from the differential.  This was also required for ducted cold air routing .       I used Kelsey hays 4 piston calipers at the rear so new brackets were required. With the caliper further offset away from the diff.the mounting bolts are much easier to get to than the Jag config.  The inner pivot  requires a single heim.  Since this CA supports the weight I used a 3/4 heim with a 5/8 hole.  With the rotor relocated, I had an additional  couple of inches clearance to work with but still wasn't enough to fit on center where I wanted it. That is why I offset the CA 2 in. forward of center.  The CA has a kind of U to it just at the inner pivot.  This is for rotor clearance.  It works just fine and I will down load pic's as soon as I can.
       Daze: 
   At the time I down loaded the images of my cross member I had all ready made the rotors, rotor hats, caliper mounts, and three different sets of LCA. mounts.  I had also put a few hundred miles on the IRS trying to solve the issues I had described in  post's at the time.  I drove the car to work every day last summer with the K.A  linkage in place.  Satisfied with the mechanical aspects    I  decided to do the body work when the monsoon rain came at the end of Sept.   The body work has been done since before last christmass.  The car is totally masked off and currently free of oily finger prints.  I plan on keeping it contamination free till the paint is done.   It is still dismantled waiting  for warmer weather to be painted.

Last edited by tyrellracing (1/24/2012 1:56 am)

Here is the page on bump steer. Titled,IRS Bumpsteer

"At 2" of bump travel, the factory roll-understeer and bumpsteer were considerable - roughly 0.35 degrees of toe change at each wheel"
http://mdmustangs.com/David_Simmons_Cobra/irsbump.html

David does get a little crossed up in his explanation, here he contradicts what he had said previously. "(i.e., toe-in at the inside wheel and toe-out at the outside rear wheel)"
Then here he talks, adding over-steer when in fact his new numbers show a more toe neutral IRS. Which is what he does want. "The new toe-curve provides 'roll-oversteer' at the rear of the car."

Last edited by Ralphy (1/23/2012 6:06 am)

Ralphy
      I have always used a surface ground  1/2 plate 24in. dia.   I made this years ago on a Blanchard grinder with the 5 on 4.5 bolt circle.  With my spring  removed and sway bar disconnected I cycle my suspension with a floor jack while I use a hinge plate with a single indicator like the one shown.     I took multiple readings of the toe on the rear after installing the K.A. links  before I ever put the tires on.  This is just part of my alignment regiment .   I have a pair of old Hunter alignment heads with lazers installed for setting toe.  I also use a longacre magnetic caster/camber gauge.   At normal ride height I set the toe at .125 12 in. forward of axle center.  .0625 per side.   When the wheel travel went below ride height , the change in toe was less than .020.  at -3in.   Above ride height I observed  readings that  were additional toe in.    Not out.  The change in toe per side at + 3  was from .0625 to.095 toe in .  The small amount of roll steer measured is not likely to play much of a roll in the handling as a whole.  Especially when you stop to consider that the roll bars combined with the stiff springs together  create a large amount of roll stiffness.  If you were to try to bounce the front of my car with your whole body weight , you would not get more than 1/2 inch of movement.  Some of that is spring  but most is shock dampening.   
  The amount of body roll is so small that the 1.375 front sway bar pulls the nose down instead of allowing lean  when cornering hard.
     These are my observed readings.  I dont think the theoretical method of determining roll steer is as universal as you have been led to believe.  It may work with twin parallel leading links like the Vette has or with a more traditional linkage but it did not predict  the results of the KA set up
  I have no idea why,    I just know my IRS never produced toe out as you predicted.  In a nut shell it makes a very small amount of roll understeer!
I cannot feel .030 change in toe in the front,  Can you?    That would be less than the tread distortion caused by simply rolling.   As far as the toe readings in the charts you provided.   So those numbers mean nothing here because they came from an entirely different suspension.   Wouldnt that be like saying motor homes are top heavy. your vehicle has four wheels there fore your vehicle is top heavy.  The chart has as much in common with my suspension as a motor home!        In other words you are trying to say that the readings I got cause effects I am not aware of?   Just rolling down the road your tires distort more than .020 when they meet the road surface.

Ralphy wrote:

phantomjock,
You got me going on this roll center height. I'm sitting here graphing a C3 half-shaft/camber rod geometry with a UCA. What I see is if you were to add a UCA the roll center lowers visibly. So when Pier built his design he also lowered the roll center. The IC moves way out!

Hmmm.......

Ralphy

Ralphy - I'll have to take some spare time and do a plot (BonesApart) of the C3 and then Pier's approach to it.  Pier's is a bit more C4-like - dual TAs etc.  Maybe even do a C4 for baseline comparison.

I do like the idea of being able to design the possibility of :

The new toe-curve provides 'roll-oversteer' at the rear of the car. In roll-oversteer, body roll induced by cornering results in toe-out at the outside rear wheel and toe-in at the inside rear wheel - effectively steering the rear of the car around the turn.     [from; http://mdmustangs.com/David_Simmons_Cobra/irsbump.html ]

BUT, I'm afraid  there would have to be some dramatic changes to really get the effects of all-wheel steering!

As a point in that direction - I am planning on adding tabs to the leading edge of my rear spindle uprights -   strategically placed if I want to add the TAs later. 
Cheers - Jim

Deleted

Last edited by tyrellracing (1/31/2012 5:24 pm)

phantomjock,
When you mentioned the extended ball joints for front suspension. My thinking was this would not lower my CG, but now I see it does. Nice! The only negative effect I see maybe is this would lower my inner point of the UCA. Then maybe alter my bump steer. There are bump steer kits for the MII's.

Would you want to raise the inner point of the UCA? To bring the camber curve back a little?

Ralphy

TR,
For some reason it seemed you might have tried to estimate your LCA angle. And that you never had the opportunity to do an actual measurement. It seemed tome or I thought maybe you did not. That's why I asked or eluded to, to see if you did measure the roll steer. From the numbers you posted. I see you had app. 1/16 to 3/32 roll toe @ 3" bump. So if you started at 1/32 static toe plus 1/16 roll. Real roll toe would be .448 degree, almost 30 minutes. If I were to start at your high number 3/32 plus 1/32 static, total at 3" bump would be 1/8" or .596 degree total. Seems to be so small I know, but your falling right in line with that Mustang guy above. I find the numbers interesting because there are so few people who have been able to give any. There are only two I have seen now, you and that Mustang guy above.
.448 to .596 total degree at 3" bump? Or 3/32 to 1/8 total inches at a 12" radius app.?

Seriously, thanks for posting your findings (settings).

Ralphy

Just read your post again, you started at .o625 static toe per side. That's even higher, going from a minimum of 1/8" at 3" travel to 5/32" at 3".

What I have seen on the OEM cars. Such as the C4 IRS, bump or droop, you will have toe in. Verses your KA, which rotates the toe in the same direction, in-out, in regard to body roll. Going through a turn loaded side, toed in, unloaded, toed out. True roll steer! I know the loaded side becomes the dominant side. However I can't see how this doesn't have some negative effect on the C4, giving the KA design a superior edge. Maybe these OEM designs are the ones with the really noticeable, is it, snap steering issues?

T5, KA toe curve


So now let's go to another point you made. Roll steer was not started by KA? Interesting, but did all the other designs work like the C4 always toeing in, bump or droop? Did KA design a superior racing idea? Hmmm...........

Last edited by Ralphy (1/25/2012 7:06 am)

Ralphy:
        All your free body diagrams clearly started with three attachments to the vehicle frame that could never produce roll steer or  for that matter even move with out breaking.  Then the revised free body diagrams now clearly have two  hinge points on each LCA that were out of parallel with the center line of vehicle .  One of these designs you had labeled as ARNING.     The free body diagram you drew had nothing in common with the T5 LCA.     Then you went on to state that this diagram was how my suspension was to create roll steer even though the diagram was not anything like the T5.   What was the deal with that?
      I have a working copy of my suspension.  I have cycled the thing many times during the alignment process.  I wrote down all the spec.'s as I cycled the suspension in my log book.  Both incorrect alignment and  the alignment I approved.  You do not have the benefit of this data to draw from,  I do!
     The leading and trailing link's do not respond in the manor you described and I have the data to back this up  nor do they share any characteristics with the Goulstrand twin trailing arm you always want to fall back on as your example.   Using angles that would be impossible to obtain on my car is a waste of time.  In reality it still proves nothing.
     Separation of my spring s from the LCA was to enable me to cycle the suspension with a jack.  How would you measure all the data with them in place as easy any other way? 
      My example being :     From a side view it would be impossible to make a Vettes wheel travel in a perfectly linear line.    It is entirely possible with a T5.  All I would have to do is replace my hub carrier. with no other changes.  This is the fundamental difference you choose to ignore.  No matter what direction you move your links , The twin trailing arms will always follow the arc path around the anchor bolt by the   link length no matter what angle you move them to.
     The T5 for comparison,   can be made to curve forward, rearward, or perfectly linear.  with out moving any link points what so ever.  All I would have to do is move the position of the axle on the upright.   This cannot be done with the Goulstrand .  This is a quantum leap away from twin trailing arms and why all your spastic moving of links to non align able positions has no real meaning. Yes the designs can be changed to the point it no longer can track down a road.  Where is the point in that.
   The T5 does not have a Camber rod or a toe rod.  It has a lower control arm.  A toe rod usually connects a steering arm on the rear of the upright to the frame near the diff approximately the same length as the LCA so bump steer is not produced like the designers of the Goulstrand did on the Fiero   Another poorly designed IRS from GM.   There is no such thing as a Camber rod!    Camber is the product of the upper and lower control arms on the upright. Neither one can take be responsible for camber with out the other.  That's common sense! 
     There is little to be gained through discussing relocating the links to positions that would not allow the suspension to be aligned.  That is a waste of time.  Once you have moved the pivot points on a T5 or a Goulstrand you no longer have either.  The fundamental part of each design is its geometry.  Once you significantly alter it you no longer have a Goulstrand,.
      I have my log book with my notes from the last 25 years. This includes my real IRS with all my drawings, alignment spec.'s.  Not something I down loaded off line that some one else owns.   
My notes are from my  real car that I own and can walk out to my shop, put a key in it and fire it up.  You don't appear to have any running examples of your roll steer design.  I have yet to see anything  you downloaded that you own.   Even when you needed to provide a image of your car you posted a image of some one else s  that was made by the same company as yours.  In theory.   I am sure you probably have a kit car.in pieces somewhere.   .  How many years has this cobra been a topic of discussion but never actually worked on.  After finding this site I had a running Jag IRS  in my Shelby in two months.    It is easy to have a car that can do anything when you have not finished it .  Who knows,    at the current rate of completion it may  never be a driver.   Well I don't know that but it sure seems that way.   
    Going through my log book,  I can see  that even when the alignment wasn't right because the wheel wasn't centered in the fender it did not create toe out.  When the camber was excessive the toe didn't change.  The toe was measured with Hunter lazer alignment heads.  These have a resolution of .005 inch.   Each time I made a significant change I would cycle the suspension and take toe and camber measurements each inch. traveled.  they were entered in my log book right then and there. 
        I guess the most important point I have to say is: I am writing about a running driving vehicle that I have had finished!       The Jag Irs has been done since 2010,the T5 conversion has been installed for over 9 month.  I  drove it to work all last summer. with out any problems  If the paint were done.  I would be  driving it daily still.               
        You have yet to explain how a vehicle that has such high roll stiffness that body roll in is nearly non existent could provide roll steer of any significance.

I explained that I have Hunter alignment heads that use lazers to measure toe.  They measure in thousandths not in degrees.  However, converting to degrees would present no serious problem since I use trig every day at work.   I am a machinist remember?  Try laying out a 17 bolt circle on on an oval if you feel cocky.  Dividers  and trammels not allowed.  Figure out the angles between each bolt hole when all the bolt holes are equidistant.

I too have the toe curve graph.  So what.   Did you check that info with a working example?    I did and It isnt the curve for Klaus Arnings IRS.  I don't  know where they came up with that graph but it doesn't describe the suspension I built nor does it describe Arnings.    Therefore I don't care what they described it as.   I have known this for 9 months now  Again So what.     Camber rod?   LOL

Typical machinist attitude. We are some really messed up folk, including me! It's a product of the environment.
Some of the brightest blue collar people you can meet! LOL!

Seriously!

Damn, that was fun.



Give Mr. Ego A Chance
http://www.youtube.com/watch?v=4B5EfetAa0w

Last edited by Ralphy (1/25/2012 9:23 am)