Register now to get rid of these ads!

4-link debate

Discussion in 'The Hokey Ass Message Board' started by sideweeler, Apr 4, 2013.

  1. Dan Timberlake
    Joined: Apr 28, 2010
    Posts: 1,283

    Dan Timberlake

  2. sideweeler
    Joined: Sep 28, 2010
    Posts: 127

    from searcy AR

    im not talking about the link bars, I'm talking about the rod ends, and mainly the "pivot point". to me it seems like the pivot point should be parallel to the rear end. the guy that e-mailed me back from thorbecke bros said the same thing. and thats how I have mine. I don't claim to be an expert by any stretch of the term. so I'm relying on feedback from others that have built them, and it seems like everyone says the same thing, their way is right, and the other way is wrong. I think Hackerbilt understood what i was asking, and gave the best answer so far.
  3. mob53
    Joined: Sep 6, 2010
    Posts: 129


    Is your post (#51) from thorbecke? If so, I wonder why they don't offer those bars on their website, or at least I don't remember seeing them. Did you have them make them custom for you? I think they make a good point and seems hard to argue against.
  4. CutawayAl
    Joined: Aug 3, 2009
    Posts: 2,144

    from MI

    The same is done of the rear beam axles of some front wheel drive cars. The suspension is two simple arms and the hat shaped axle twists to accommodate independent wheel movement. That's essentially what's happening with split wishbones on an old Ford.
    Last edited: Apr 5, 2013
  5. n.z.rodder
    Joined: Nov 18, 2008
    Posts: 1,016


    I take it you're still in the build stage of your car. If so, set the frame on some stands at ride height, place a jack under the diff head and remove the wheels and shocks, bags etc. Next, remove the front bolts that hold the four bars to the chassis and replace them with something to hold the bars loosely in place (welding wire will do) now grab one side of the diff and move it up and down, take a close look at what the front of the four bars are doing now bolt the bars back up and try to do the same up and down action with the diff, you will notice a LOT less movement, this is called BIND. The other way with the mounts not parallel to the diff is a lot more forgiving.

  6. earlyv8
    Joined: Jan 13, 2007
    Posts: 194

    from oklahoma

    All the above designs above are with significant compromises.
    The primary design using tubular links in themselves create binding as the tube is very unlikely to twist so the rotation stress is either from the flexing of the bushings (rubber or otherwise) or the brackets themselves that are taking a boatload of stress.
    The only way to eleminate the rotational twist (stress) is to use balls like ball joints, heim ends (which I do not like for this purpose ), or Ford's proven unsplit wishbone design.
    Personally, I dislike the triangular set up because of additional stress in the form of pushing the frame rails out under braking or pulling toward the center under acceleration. Minimal but there.
    I have used parallell links ( initially using Chev rear panard rods for links) for approximately 40 years rather than split type trailing arms after observing a u shapped piece of wire on the garage floor and noticing it would twist the axle during movement.
    Hope I have not offended anyone but MHO.
  7. CutawayAl
    Joined: Aug 3, 2009
    Posts: 2,144

    from MI

    Spherical ends will eliminate torsional binding, but even with those there is still bind in the system. For the suspension to articulate freely the links need to get longer and shorter. Without compliant bushings to allow that, even if the links are parallel and equal length, the suspension binds.
  8. earlyv8
    Joined: Jan 13, 2007
    Posts: 194

    from oklahoma


    With all due respect.

    Please explain the binding without the bushings allowing length change.

    Is that because of the back and forward motion of the attached axle as it travels up and down moving back and forward at the same time?

    The longer the links they are the less binding, because the back and forth motion is reduced because of that extended length.

    As far as binding with the ball type ends, the manufacturers compensated for that when using channel type links (that twisted) during movement.

    Olds used trailing arms in 49/50 similar to Chev pickups with coil springs. The olds used massive rubber disc insulators 4 per side similar to Ford biscuit motor mounts at the rear axle.


  9. CutawayAl
    Joined: Aug 3, 2009
    Posts: 2,144

    from MI

    It's not the easiest thing to visualize. It's most obvious when the links are short, the travel is high, and the difference in height between one side and the other is high. Consider how the links on one end rotate the axle housing as the suspension moves up and down. When the opposite end is at a different height those links are trying to impose a different rotation to the axle. That results in binding that can only be eliminated if at least one of the links can change length. The situation is at its minimum when the links are parallel to each other and to the ground, the links are an equal distance above and below the axle's vertical centerline and the links are as long as possible. When thinking about this you can also see there is roll steer, but that's a different discussion.
  10. impala631
    Joined: Oct 26, 2012
    Posts: 78

    from INDIANA

    Here is a better pic of what he is after, notice how the bushings are mounted parallel so that they match the lowers, it's just a slight variation to the normal triangulated 4link, both will work fine, this way will increase your bushing life a little, that is all

  11. El Polacko straightened you guys out on this on page 2, but no one was listening...

    Either way you mount the ends, you will get bind. The angled ends bind in straight bump, the parallel ends bind in roll (...YES, they do).

    Think about it, for the rear end to be able to roll, flex, articulate, whatever you want to call it; the bolts on the housing will no longer be parallel to the bolts on the frame. That is bind, pure and simple. If you can't see that, put the welder away and have someone build it who understands suspensions. With any tri-4 link, the ONLY way to eliminate bind in both directions is to use spherical rod ends or "johnny joints".

    There is no "right" answer to this question. The suspension should be built to suit the most likely motion. For a bagger, that would be straight bump; for a road racer it would more likely be roll.
  12. blowby
    Joined: Dec 27, 2012
    Posts: 7,047

    from Nicasio Ca

    Geez all this makes the original wishbone design sound pretty good.
  13. speedyb
    Joined: May 12, 2010
    Posts: 484

    from socal

  14. BootleggerMatt
    Joined: Aug 17, 2011
    Posts: 258


    It doesn't make that big of a difference. Calling a guy out on his hard work, just to point out that your way is better, isn't helpful. If he asked you what you thought was better before he started, thats different. Telling him he did it wrong after the fact, given thousands of cars on the road with the same setup and no problems, only strokes your own ego.
  15. LM14
    Joined: Dec 18, 2009
    Posts: 1,936

    from Iowa

    The best way is to float the mounting points on the axle housing in birdcages. We ran parallel 4 bar on race cars for years. You have to add a panhard bar and a torque control device but there is no bind anywhere with that setup. Too many pieces for most street cars, but very smooth.


  16. chopt top kid
    Joined: Oct 13, 2009
    Posts: 959

    chopt top kid

    FWIW most four bar suspensions are normally built so that the link in either tension or compression thus the reasoning behind the rod ends being mounted parallel to the normal axis of the link. When you mount the rod end on an angle to the normal axis of the link, you introduce a torsional bending moment into the link. The more severe the angle between the rod end and the normal axis of the link, the more severe the torsional bending moment. This can usually be be controlled by the addition of a panard bar or similar device to control the lateral loading. Otherwise, you should calculate the maximum torsional bending moment of your links and size them accordingly. The rod ends will also be subject to the severe lateral loading and could be subject to pre-mature wear...
    Last edited: Apr 8, 2013
  17. You do NOT want to use a panhard, etc. with a triangulated 4 link. You now have 2 different lateral locating devices; and unless they are perfectly aligned, neither one is free to react correctly. How the link ends are attached (angled or straight) has NO bearing on the amount of side load or torsional bending moment on the link. The ends of each link are still trying to move in exactly the same directions, depending on chassis loading. The directional loads on the BOLTS do change substantially.

    (If you want to mount the bolts parallel to the housing, go right ahead. What you DON'T want to do is bend the links themselves to square the ends with the mounting sleeves on a high power, high traction car. See below)
  18. blowby
    Joined: Dec 27, 2012
    Posts: 7,047

    from Nicasio Ca

    Just looking at the picture above, it looks to my uneducated eye like under hard cornering the links would have a hard time keeping the rear end from moving side to side, at least a little bit. I'm running my rear tires just a bit inside the fenders, any side movement and it's rub time.

    Some drag racing 4 links run a diagonal link from the forward lower link mount at one side to the rear lower link mount on the other side. Would this work on the street under hard cornering in place of a standard panhard?
  19. banginona40
    Joined: Mar 5, 2007
    Posts: 758


    What is IG please?
  20. Drivinford
    Joined: Aug 24, 2012
    Posts: 730


    What are Hemi joints.;)
  21. jcapps
    Joined: Dec 30, 2008
    Posts: 473

    from SoCal

    This is the answer to your question
  22. I have been thinking about this since this thread started, trying to decide which if either was better. I finally decided to run it by a friend who is an engineer at one of the Big Three in the Michigan area;
    "Hey Don, good to hear from you. I looked at both of the pictures you sent and can tell you that my first thought was the parallel mounting of the bushings to the axis would not work well. I decided to run some simulations on the computer in 3D.
    Now before we get into the problems we have to understand that all triangulated 4 links are a compromise to start with. When properly designed they work very well for their intended purpose on a street car. If I were building a track car I would go with a 3 Link and Watts Linkage (or at minimum an adjustable Panhard)
    The version with the link bushings parallel to each other on the frame and axle exhibited zero bind in a straight bump situation with no articulation of the suspension. As we started to articulate the suspension, the bind increased at an exponential rate putting unbelievable amounts of force on the bushings initially, then the bolts and eventually the mounts. Since bumps in the road are rarely equal on both wheels, this is not a good design and will create large amounts of rear roll stiffness while inducing undue stress. We believe the mounts will fail in this design if used in real world driving conditions as they are most likely the weakest link.
    The version with the bushing mounted perpendicular to the links themselves, like all of the OE designs I have ever seen, is a much better compromise. Although there is some inherent binding in all types of travel, it is much less than the other design as you are starting in a more neutral position, and since most bumps and therefore travel have some articulation, this design works very well. In most OE applications, the bushing themselves are designed to absorb some of this binding action as to not induce roll stiffness. In many cases they are overkill as we also have to deal with NVH.
    Although either systems will work, for a car that is being built to be driven on all forms of roads, and in particular if it has a performance slant to it for some track work, I would never consider the version with the parallel mounting of the links. It is just not good suspension design or theory.

    So there we have another opinion ......
  23. Thats pretty cool to be able to test theory like that.
  24. Yes it is very cool! And it's not a theory if we can prove it works.

    A huge thank you to your friend there Don, Saved me from doing the same. I was waiting to (and still might) take the time to draw this up.
  25. And there you have it...AGAIN.
    Joined: Aug 7, 2009
    Posts: 2,069


    You should "Cut It All Up" & Do a "Suicide" 4-Link
  27. You should have seen the simulation he and his engineering buddies did on the Cray concerning Mustang II front suspension and early Mustangs and Falcons back about 5-6 years ago. They were curious about the relocated forces and in particular why convertibles were breaking windshields after a conversion to Mustang II. It was some really cool stuff and they came up with some solutions to the problem in the way of bracing to redistribute the loads.
  28. Yes, very much I would love to see that data.

    I have my own ideas of what they would have found, and would be interesting to see their results.
  29. ^^^Agreed.^^^ If you can get that info, Don, I'd be very interested. We teach a lot about the MII, and I agree that there are some definite issues with that setup.
  30. Pops1532
    Joined: Jun 19, 2011
    Posts: 544

    from Illinois

    The OP seemed unconvinced after several others posted basically the same thing. Maybe hearing it from engineers from a car company will help. Thanks to you friend for running the simulations and for the explanation. Maybe that will save someone from building something that won't work well.

Share This Page

Register now to get rid of these ads!


Copyright © 1995-2020 The Jalopy Journal: Steal our stuff, we'll kick your teeth in. Terms of Service. Privacy Policy.

Atomic Industry
Forum software by XenForo™ ©2010-2014 XenForo Ltd.