Register now to get rid of these ads!

Technical How much side flex is normal in a triangulated 4 Link?

Discussion in 'The Hokey Ass Message Board' started by tlmartin84, Feb 14, 2022.

  1. gimpyshotrods
    Joined: May 20, 2009
    Posts: 23,264

    gimpyshotrods
    ALLIANCE MEMBER

    Having read your previous posts, and intending no offense to you, you are quite deep into overanalyzing this.

    The danger here is that you will build a vehicle that could pull 1.2G on a 100' skid pad, take a trophy at Laguna Seca, and have completely unacceptable manners on a public road, leaving you with a vehicle that you will not want to drive, and will regret building.

    And yes, as @X38 says, a parallel 4-link, with a Watts link would be better.
     
    Tman likes this.
  2. FrozenMerc
    Joined: Sep 4, 2009
    Posts: 3,093

    FrozenMerc
    Member

    This statement has me a bit concerned. There should be no motion between the inner bushing sleeve and the bolt. The ID of the inner bushing sleeve should not apply any appreciable load to the OD or shaft of the bolt. If there is, you have a clevis style joint that will lead to rapid bushing failure, and/or bolt failure due to bending of the bolt. Bolts are tension members and have a very short fatigue life when put into bending load.

    A proper clamped joint relies on the bolt and nut to clamp the mount flanges to the ends of the inner bushing sleeve. This clamp load is produced by the bolt as it is stretched during torquing. The load is transferred directly from the suspension arms into the bushing and then directly into the chassis mount flanges through the friction generated by the clamping force produced by the bolt. The bolt never actually carries any suspension loads.

    [​IMG]
     
    gimpyshotrods likes this.
  3. gimpyshotrods
    Joined: May 20, 2009
    Posts: 23,264

    gimpyshotrods
    ALLIANCE MEMBER

    I suspect that folks might have overlooked the presence of the sleeve.
     
  4. tlmartin84
    Joined: Jul 28, 2011
    Posts: 1,030

    tlmartin84
    Member
    from WV

    That's why I am here asking the questions. How are you supposed to know what you are building, and what the outcomes will be if you have never installed and driven one?

    I am specifically NOT trying to build a drag car. The problem is a lot of these "bolt on" parts, are designed for a race application, and when you talk to the manufacturers techs, they can't tell you what to expect....

    I've seen a ton of these installed online, and yet I can only find a few people who have actually driven one, and even fewer of those understand how they work.
     
  5. tlmartin84
    Joined: Jul 28, 2011
    Posts: 1,030

    tlmartin84
    Member
    from WV

    The threaded portion of the bolt was slipping in the flange. Granted it was minimal movement but once the slop in the rubber bushing was taken up, the threaded end of the bolt would take up the remaining slop in the hole (maybe 1/32"). I still think a bolt with a full shoulder all the way through both flanges and the inner bushing would be better.
     
    SlamIam likes this.
  6. gimpyshotrods
    Joined: May 20, 2009
    Posts: 23,264

    gimpyshotrods
    ALLIANCE MEMBER

    None of what I mentioned has anything to do with drag racing.

    A 100-foot skid pad is a circle, designed to be driven around, to test maximum lateral adhesion, measured in fractions of the force of gravity.
    [​IMG]
    https://fastestlaps.com/lists/top-grip-kings

    Laguna Seca is a world-famous road race track on in Monterey California, and my home track.
    [​IMG]
    While it might look somewhat flat in the picture, the backside is a hill, with 180-feet of elevation change, descending into four corners, known as The Corkscrew.

    What wins races there is uncomfortable to drive across the paddock for tech inspection.
     
    Ned Ludd, Oldiesmiles and rockable like this.
  7. AccurateMike
    Joined: Sep 14, 2020
    Posts: 631

    AccurateMike
    Member

    I think when they say "no movement compared to a panhard", they mean none of the side to side deflection that a panhard bar creates. As the suspension moves through it's travel, the panhard moves through an arc. On a hot rod, with only a little travel, the deflection is small. By selecting the mounting points carefully you can minimize this deflection. On my OT Early Bronco with a foot+ of travel, the deflection is inches out to one side at ride height and inches out the other side at full compression. Landing a jump with panhards front and rear is disconcerting :) .This is why there are watts linkages and triangulated 4-links. What you are looking at is just bushing deflection, a necessary evil of a triangulated 4-link. Depending on your travel and geometry, it is almost certainly less deflection than a panhard will impart moving through full travel. Your fender/tire clearance moves from one side to the other (or more to just one side) with a panhard too. BTW, my OT Bronco is parallel 4-link front and rear with panhards. There was no way to do a triangulated with the front punkin going up into the pan, Bronco style. Similarly hard to fit a watts in there. I did it more than 40 years ago (Bigfoot was still on leaf springs). I had to look at hotrods for inspiration, I was way ahead of the curve as far as off road suspensions go. Once you got over the weird panhard deflection, you could land Fall Guy jumps :) , Mike
     
    Oldiesmiles and bchctybob like this.
  8. gimpyshotrods
    Joined: May 20, 2009
    Posts: 23,264

    gimpyshotrods
    ALLIANCE MEMBER

    As for a Watts link, there is still side-to-side movement, but a whole lot less than a Panhard bar.

    With a Watts link the axle moves up and down in a very slight S-curve. This is way better at protecting your fenders and tires from collision than a Panhard bar.
     
  9. SlamIam
    Joined: Oct 8, 2007
    Posts: 468

    SlamIam
    Member

    FrozenMerc and others are right - sleeves inside the bushings must be clamped tight between the mounting ears so movement at joints is all in the rubber. The holes in the ears should also be a tight fit for the bolts. I have this setup in my pickup, frame mounts boxed against flex, bolt holes tight, bolts torqued 80 foot pounds, DOM seamless arms, upper arm angle 90 degrees, coil-overs 22-1/2 degrees leaning in at the top side-to-side, which also resists side movement. Bushings take a beating but cheap and changed annually. The truck tracks straight under hard acceleration and tires don't hit fenders under hard cornering. I find it an easy design to live with for a street-driven vehicle.
     
    Oldiesmiles and bchctybob like this.
  10. tlmartin84
    Joined: Jul 28, 2011
    Posts: 1,030

    tlmartin84
    Member
    from WV

    Any idea how much side to side movement you have?

    Any pictures of your setup?
     
  11. tlmartin84
    Joined: Jul 28, 2011
    Posts: 1,030

    tlmartin84
    Member
    from WV

    I have a 71 Bronco!

    I am not too concerned with movement, just curious what is normal and the fact that every thread on here states mount your bars at a minimum 60 degree angle, and 45 preferable. Mine are around 26 degress ( the kit was designed for 27).

    I have a pretty tight fender clearance, which is part of the reason I steered away from a panhard... in hindsight with as little movement as there is that probably would have been better. It comes with it's own set of challenges though.
     
    AccurateMike likes this.
  12. bchctybob
    Joined: Sep 18, 2011
    Posts: 5,190

    bchctybob
    Member

    Out of curiosity I took the old Panhard bar from my roadster (40 3/4" long) put it on my workbench and traced the amount of displacement through 6" of travel. In 6" of up and down travel (3" up - 3" down) it moves the axle less than 3/16 of an inch side-to-side. There may be another 1/16" or so in bushing compliance but it's still less than you say you have with the triangulated bars just pushing it sideways by hand.
    On Panhard bar installations, I usually use a Heim joint at one end which adds slightly more stability and rubber/urethane at the other just for vibration/noise suppression.
    If I were you, I would try not to overthink the issue. Put it together with good bolts and bushings and try it. It will probably be just fine unless you plan to be the fastest at open track day at your local road course. I would, however, put off powder coating the rear axle housing just in case you decide to make changes later.
     
    Last edited: Feb 15, 2022
    loudbang and gimpyshotrods like this.
  13. afaulk
    Joined: Jul 20, 2011
    Posts: 1,194

    afaulk
    Member

    I'm a long time drag racer so I ALWAYS run a Wishbone with the wide end to the front and the heim joints located horizontally, parallel to the ground and with the single point mounted to the rearend housing also mounted parallel to the ground. This eliminates virtually all side to side movement under any load conditions. With the proper chassis preload the car will always launch straight. Shocks, springs and Instant Center are another whole conversation. Cheers and good luck.
     
    Oldiesmiles and bchctybob like this.
  14. twenty8
    Joined: Apr 8, 2021
    Posts: 2,318

    twenty8
    Member

    A Panhard bar will induce some lateral movement. A triangulated 4 link or 3 link will not. It is the removal of the Panhard bar's inherent lateral movemnt of the rear end that is the marketing claim. Either system will still have, and need, some lateral compliance, which is usually provided by the bushings.



    With regards to the angle of the triangulated bars, 45* is optimum for the bar to do it's job as a trailing link and to minimize lateral movement. The more you move away from 45*, the less the bar is effective at controlling the lateral movement.
     
    Last edited: Feb 20, 2022
  15. goldmountain
    Joined: Jun 12, 2016
    Posts: 4,443

    goldmountain

    If you want to do an easy real world check on how much play is OK, just find an old stock GM rear wheel drive car with the triangulated 4 bar and give it a wiggle.
     
    Tman likes this.
  16. rockable
    Joined: Dec 21, 2009
    Posts: 4,430

    rockable
    ALLIANCE MEMBER

    :):D
     
    Oldiesmiles likes this.
  17. twenty8
    Joined: Apr 8, 2021
    Posts: 2,318

    twenty8
    Member

    I think your bars are actually mounted at a 64 degrees (90* - the 26* you say you have). More angle will definitely help limit lateral movement. Proper bushing bolts and jamb nuts on the bars will make a it all tighter. In my opinion, the brackets on the rear end tubes are a bit of a weak spot, and may be twisting and allowing some lateral movement. It is easy and inexpensive to redo these using the 'overkill' method of design and construction.

    I think it will be ok if you address the issues above, and if you can get the bars on a bit more angle, even better.....
     
  18. krylon32
    Joined: Jan 29, 2006
    Posts: 9,397

    krylon32
    ALLIANCE MEMBER
    from Nebraska
    1. Central Nebraska H.A.M.B.

    I have built several 100 chassis over the years with triangulated 4 bars. I always weld the top rear brackets toward the center of the rear end with the front of the bars to a bracket welded to the frame. The rear of the bottom bars attaches to a bracket welded toward the outer end of the rear end with the bars going forward to a bracket welded to the frame. never had any problems with side sway. Have I been doing this wrong all these years?
     
    Nominal, Tman and twenty8 like this.
  19. twenty8
    Joined: Apr 8, 2021
    Posts: 2,318

    twenty8
    Member

    No sir, you have not been doing it wrong, and your experience should not be questioned. The method you use is by far the one that is used most. Triangle forward or triangle backwards will both work equally well to locate the rear end laterally. The more common way is like you do it, but essentially either will work, as long as the geometry is correct and their is minimum bind. I think there would be other differences that you would get, such as roll center, but lateral location doesn't care either way.
     
  20. lostone
    Joined: Oct 13, 2013
    Posts: 2,820

    lostone
    Member
    from kansas

    I prefer the upper bars be mounted as high as possible and towards the middle of the housing myself. With the lower bars straight with the chassis and parallel with the ground and mounted low under the housing.

    I like as much distance between the upper bar mount on the top of the housing and lower bar mount on the bottom of the housing as I can get. I like as much leverage between the 2 as possible. I also like the upper arm mounts as close as I can to the frame rails on the front of the upper bars to help prevent as much flex as possible.

    Here's all the room I had but took full advantage of what I had.

    Also note I built all my own brackets and reinforced the upper mounts on the housing.
     

    Attached Files:

    Oldiesmiles and twenty8 like this.
  21. Mimilan
    Joined: Jun 13, 2019
    Posts: 1,230

    Mimilan
    Member


    NO it doesn't.

    If the height difference at the frame pivots is the same as the height of the pivots on the bell-crank the arc of the 2 link arms counteract each other.
    The centre pivot moves vertically on a Watts

    You can also mount a watts link bell crank horizontally [with 1 link in front of the axle and the other link behind]
    and it achieves the very same result.
    I've seen this on Clubman racers where they need a really Roll Centre with a watts.
    Early Clubman Racers preferred the Low Rear R/C to soften the rear roll stiffness [pre-Limited slip diffs]

    You can also achieve the same low R/C results with a "Woblink" variation of the Watts Link
    upload_2022-2-23_15-48-46.png

    All this does is "Re-invent the wheel", and ending up with the same end result as Colin Chapman's triangle 3-link.
    Which give a low R/C height and also controls forward thrust, axle torque reaction, brake torque, and lateral acceleration with only 3 moving pieces
    AND
    The R/C height never changes with suspension movement.
    upload_2022-2-23_16-0-11.png
     
    Last edited: Feb 22, 2022
    gnichols and twenty8 like this.
  22. gimpyshotrods
    Joined: May 20, 2009
    Posts: 23,264

    gimpyshotrods
    ALLIANCE MEMBER

    I have observed and measured it on several track cars. You can disagree, but lasers do not lie.

    I am not on this board to argue. I know what I have seen.

    If you want to argue this, contact the maker of this: https://www.baumer.com/us/en/produc...es-up-to-1700-mm/om70-l1000-hv0700-vi/p/38588
     
    Last edited: Feb 23, 2022
  23. AccurateMike
    Joined: Sep 14, 2020
    Posts: 631

    AccurateMike
    Member

    watts-linkage.gif
    Mike
    (Edit, another version is an angle, see ~42 seconds in)
     
    Oldiesmiles and Ned Ludd like this.
  24. Ned Ludd
    Joined: May 15, 2009
    Posts: 5,026

    Ned Ludd
    Member

    The angle of the line of movement seems to be perpendicular to the angle of the arms at that moment when they are parallel to one another. Of course that would have to be verified: there are a thousand ways to cheat an animation. But it could account for the lateral movement @gimpyshotrods mentions: the bars might be out of parallel at rest, or there might have been compromises in locating the outer pivots. It this works it's a solid basis for a heuristic to set up a Watt's linkage.
     
    gimpyshotrods likes this.
  25. twenty8
    Joined: Apr 8, 2021
    Posts: 2,318

    twenty8
    Member

    Set up with the arms the same length and at equal angles, and therefore the offsets up and down equal, the center pivot will move in a perfectly vertical line. If there is too much travel, you will start to get some deviation at the extremes, but we here don't need to worry about rock crawling stuff. In short, a perfect Watt's linkage will work perfectly. The more you move away from perfectly equal arm lengths and angles, the less well it will work. With space restraints, sometimes compromises are made, and that will lead to the deviations that may be observed.

    So, I guess you are both right, depending on the way the Watt's link has been set up.:)
     
    Last edited: Feb 23, 2022
    Mimilan likes this.
  26. For many years Falcons here had a Watts link rear suspension. (pics are from bushing manufacturers)

    [​IMG]

    [​IMG]
     
  27. Ned Ludd
    Joined: May 15, 2009
    Posts: 5,026

    Ned Ludd
    Member

    The rotation at the ends of the links happens about three axes: assuming about 3" of travel in each direction, you might see 7-12° of rotation in each direction about the bushing's axis in both bump and roll, about 3-4° of rotation in each direction about the link's own axis mainly in roll, and some fraction of a degree of rotation in each direction about the axis perpendicular to the other two. So, to prevent your linkage from binding, i.e. ceasing to be a linkage at all and instead becoming a structure, the links need either to be spherically jointed, or to have compliant bushings capable of accommodating at least 4° of rotation about an axis perpendicular to their own, at both ends.

    Due to their angle, the compliance at the angled links' bushings is amplified geometrically under side loads. This is by no means a bad thing. The linkage's job is to prevent longitudinal and lateral movement while allowing vertical movement and roll movement, and to locate the roll centre. For various reasons it is conventional to use a triangulated 4-link to establish a fairly high rear roll centre — so much so that many believe that this is the only possible correct application of a triangulated 4-link. The principle can in reality be applied to put your roll centre wherever you want it. It's a lot of design work, though, so it's more suited to a street situation where you try to establish an overall optimal balance and then lock it in for the rest of the car's life, rather than a road-race scenario where you'd want to be able to tune the car for each track etc., preferably by changing as little as possible. Still, the triangulated 4-link has the facility that it is possible to project the roll centre well outside the bounds of the physical hardware of the car.

    Solid axles tolerate high roll centres because they are relatively immune from the lateral jacking effects which affect every form of independent suspension I've heard of — at least as long as the RC height is fairly constant and there is a positive roll moment arm. Notwithstanding all the good reasons to keep roll centres low, exploiting a solid axle's tolerance to a high RC might be a perfectly valid choice. For instance, a high roll axis might be a way to enable very soft spring rates to be used while keeping roll within reasonable bounds. The downside to this approach would be that vertical wheel movements over bumps result in small lateral movements of the car's sprung mass, proportional in amplitude to the RC's height above the road, which might be detrimental to ride quality.

    That brings us back to those compliant bushes. And that is how that geometrically amplified lateral compliance could work in our favour. It could eat up that lateral jitter over bumps without affecting much else. Because we're dealing with a linkage, where all the stresses in the links are either compression or tension, we know that deformations due to compliance will only happen along the links' axes. We are therefore able to plot where the roll centre is going to move as a result of bushing compliance, and unless you've got some seriously weird geometry, it isn't going to move very far. So, a lot of compliance in the angled links isn't materially going to compromise the lateral location of the axle. We'd have to look at what it does to axle rotation under power and braking, but that's another exercise. What I see coming out of this is that it might be worthwhile looking at very stiff bushings or even Heims or other spherical joints on a triangulated 4-link, but with a very soft compliant bushing at one end of each angled link only.
     
    Last edited: Feb 23, 2022
    twenty8 likes this.
  28. tlmartin84
    Joined: Jul 28, 2011
    Posts: 1,030

    tlmartin84
    Member
    from WV

    Both bars are mounted 26 degrees from centerline or 52 degrees of separation.

    I planned to box the back side of the tabs on the axle. HOWEVER one issue with boxing, is it takes out some of the ability for the tabs to "clamp" to the inner bushing as discussed earlier in the thread.

    I've been looking at several of these kits, the kits supplied by one of the Hamb vendors (Ron Horton Welder series) comes with urethane bushings. As a matter of fact, a lot of them come with urethane bushings... I was told this kit was a welder series when I bought it, however I do not think it is based on the bushings and some other things I have noticed. I think it is a Chinese knock off, and they used Ron's pictures in their Ad.

    Anyhow, I believe if urethane bushings are installed and everything is tight the movement will be drastically less.
     
  29. tlmartin84
    Joined: Jul 28, 2011
    Posts: 1,030

    tlmartin84
    Member
    from WV

    Finding one is easier said than done!
     
  30. tlmartin84
    Joined: Jul 28, 2011
    Posts: 1,030

    tlmartin84
    Member
    from WV

    The center section is a cast housing, if it was a 9" we wouldn't be having this conversation. I would have done them as you suggest.
     

Share This Page

Register now to get rid of these ads!

Archive

Copyright © 1995-2021 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.