Tires

[robfrey]

IO,
Could you run the "N" model for us? This is great stuff.

[John Burk]

An independent diagonal (N linkage) and a separate anti-roll bar would be safe and simple .

[robfrey]

I believe that an "N" type bottom bar setup puts all the lateral load on one rod end. Maybe I'm wrong? It is still loading the rod end in a way that it was not intended to take load therefore not really any better than what I have. Maybe IO can answer this for us with his FEA.

[Stan Back]

I don't understand any of this.  I bet I could after a week or two.  But I do know that most landspeed cars get in trouble when they are not going straight.  So counting on them to be straight all the time may not be wise.  Drag cars get "straightened out"  and contained by close guardrails.  We don't have them.  Is this too simplistic (from a roadster driver)?

Stan

[jl222]

  Rob.... Shouldn't the front heim-joints be solid joints? I know you have the bars Xed but will the heim-joints allow the  rear end to move when pushed to the side?

                    JL222

[johnneilson]

Rob,

wouldn't the "N" design give you different roll rate per side? R&L

John

[robfrey]

  Rob.... Shouldn't the front heim-joints be solid joints? I know you have the bars Xed but will the heim-joints allow the  rear end to move when pushed to the side?

                    JL222

I'm not sure what you mean but there is absolutely no side to side movement except maybe the slightest amount of slop in the rod ends but I can't feel it. The rectangle that is created here would have to go out of square for the rear to move from side to side.

[robfrey]

Rob,

wouldn't the "N" design give you different roll rate per side? R&L

John

The "N" designs that I have seen are not welded solid but have diagonal link with rod ends so everything can float a little.

[johnneilson]

Rob,

yeah, that makes sense to me.

In all the work I have done in suspension design, it seems counter to create a suspended plane.
The "N" design would work allowing the suspension to roll without "uncontrolled rising rate of spring" component.
It would introduce less bump into the car when one wheel hits one.

Back to my hole, John

[John Burk]

Looking at the N layout I realize it has a fault . If one wheel tries to raise more than the other one bar travels in a vertival arc and the other three travel in a cone shaped path and everything sees high tension or compression .

A layout that would work is to have the lower bars like \ / with the rear rod ends close together under the rear and a separate anti-roll bar .

[Interested Observer]

With the model changed to an “N” configuration, the maximum stress results are tabulated below, with plots of element numbers and the end-to-end stresses of the top surface following.

The N diagonal was connected to the same point (node) as the endpoints of the bottom trailing arms as a matter of convenience.  In real life they would be located separately but nearby.  The results would not be much different.  Since the N configuration does not supply anti-roll capability by itself, a rudimentary anti-roll bar and links were added.  The arms and crossbar are (arbitrarily) of the same 1.25” dia, .125 wall tubing.  (The unshown “axle” connects the aft ends of the upper links to the lower links.)

As can be seen, the links and shanks now have much lower stress levels and the higher stresses, due to bending in the arms, are in the anti-roll assembly, where they can be more easily managed.  The diagonal, element 24, in this case is in compression (1384 lb/3782 psi)  The diagonal shanks, 13 and 14, at 7598 psi  are carrying the same load, and are much improved over the shank situation of the previous X solutions.

Regarding John Burk’s last comment--I agree that the four points (pivots) at the axle, for instance, will, in general, travel out-of-plane with uneven suspension travel side to side.  However, given the limited travel likely to be allowed and the even more limited roll travel likely to be allowed, the mismatch may well be taken up by the slop in the system, elasticity in the system, or some designed-in elasticity.

ELEM    VMX
       1   2456.0     
       2   2456.0     
       3   666.86     
       4   666.86     
       5   661.48     
       6   661.48     
       7   661.31   
       8   661.31   
       9   1705.6     
      10   463.09     
      11   459.35     
      12   459.20   
      13   7598.5   
      14   7598.5     
      15   14353.
      16   12423.
      17   14351.
      22   801.44   
      23   799.68   
      24   3782.2   

 MAXIMUM VALUES
 ELEM         15           2
 VALUE    14353.      2456.0   

[robfrey]

Well the computer doesn't lie. I still think that we are missing a variable or two out of the equation but I will need to think on it some more. Looks like I need to get rid of the bar across the front and add a separate sway bar or re-manufacture the bottom bar assy to use a different bearing. maybe I don't even need a sway bar with my stiff coils. I really like the original concept though as it completely eliminates torque steer like a ladder bar car but puts the instant center out as far as I want. A ladder car's instant center is at the front pivot points. Eliminating the torque steer caused by chassis roll is so important on a high hp car with a narrow wheel track like a lakester because it will really change the on / off power steering bias.
Since I have such little real estate to work with I will probably re-manufacture the bottom assy.
It is sort of neat to know that the majority of the load is coming from the anti sway function. Thank IO.
I use computer sim for everything I can. Wonder how we ever got along without it.

[robfrey]

Looking at the N layout I realize it has a fault . If one wheel tries to raise more than the other one bar travels in a vertival arc and the other three travel in a cone shaped path and everything sees high tension or compression .

A layout that would work is to have the lower bars like \ / with the rear rod ends close together under the rear and a separate anti-roll bar .

I've seen this work well before on some road race applications.