+ Reply to Thread
Page 1 of 28 1 2 3 11 ... LastLast
Results 1 to 10 of 277

Thread: Beam Axles - Front, Rear or both.

  1. #1
    Hey Guys,

    I am very keen to discuss our new beam axle design. I will be posting pics on photo bucket with the user name oz_olly. I will make another post a little later with links to the other significant beam axle threads however, there isn't really a thread for consolidated discussion.

    For 2009 ADFA presented a car with front and rear dependant suspension. The front was a peg and slot beam axle (inspired by Model T Ford) and the rear was a de Dion Twist Axle. So far the performance hasn't really been investigated as we clocked up more km's at the comp than in testing.

    I am keen to develop the design looking for simplification opportunities, weight reduction and stiffness increase. I will be designing an experiment over the christmas break to do some installed compliance testing.

    The beams themselves are fabrictated from 4130 tube and sheet. Thr front unsprung, disconnected completely from the car weighs 25kg and the rear 24kg. We are also using the Hoosier 18x7-10 R25B which is about 4kg. I would like to look at the 6.0x18.0-10 in the LC0 compound.

    Cheers

    Olly

    ACME Racing
    UNSW@ADFA
    Olly

    Academy Racing 04-07, 09-11
    UNSW@ADFA

  2. #2
    Senior Member
    Join Date
    Sep 2004
    Location
    Rochester NY
    Posts
    1,061
    I've been waiting to see good pics of this thing. While I cannot comment on the overall effectiveness of beam axles compared to an 4 corner irs car the one that i designed at UB had a 4 link solid axle rear end (spool) and a irs front and was rediculously out of tune on the engine,cvt and suspension but it managed to finish 18th so the potential is there to finish high with something odd like our cars. Kinda cool to hear a spool car won Australia. I have plenty of input on linked suspension systems if you would like to listen. I even have a sweet idea for running a beam axle front with the four link pictured here that will allow you to tie all link and shock forces at the main roll hoop and turn the chassis into a person holder. Run a satchell link in the front as well as the rear. Again how effective it would be is all up to doing homework but mechanically it would work since most sprint/dirt cars do a variation of it now.

    http://www.eng.buffalo.edu/Stu.../formula_web/car.php

  3. #3
    Hey Rob,

    Thanks for the interest. It's a shame Z has moved on from the forums these days because I am sure he would love to be in on the discussion. I am planning to give him a call over the coming weeks to give him an update on how it all went.

    The decision to go with beam axles was based around a few different ideas. The first was that we had a very inexperienced team this year (75% first timers) so we needed something easy to understand and easier to manufacture with less reliance on getting the chassis really pick up points really accurate. The other idea was from a higher level, looking at FSAE as a cool project management problem. If we could design a car that maybe has C level performance but we have an A level team that develops and extracts the performance out of the C level car then we may be on to something. We are a small university with few people willing to volunteer their time for something like FSAE, so our goals have to be set appropriately. Unfortunately we didn't get much development time this year as we didn't finish the build until very late. I believe if we had of gone for a bells and whistles design we wouldn't have made it and our program probably would have been cancelled. So enough of the back ground more about the why...

    Compared to the conventional double wishbone car, our car requires far fewer accurate pick up points. A double wishbone car with direct actuating dampers requires 24 pickup points, our car requires 10 (four on the front and 6 on the rear). This significantly increased the ease of manufacture on the chassis.

    The beam axle suspension is also relatively easy to understand. We don't have to worry about the scrub vs camber change compromise. We can use castor in the front for increase cornering camber. Camber at the rear is shim adjustable as our drive train is sprung. You can set a static camber and then what you lose is due to compliance. It was quite easy to determine the roll centres for roll stiffness calculations and you can physically touch the roll centre if you are so inclined. Bump/roll steer on the front has not been measured yet but by placing the steering rack on the unsprung it was significantly reduced from what it may have been.

    The big downside seems to be that disturbing one wheel sympathetically disturbs the other dependant wheel. It will change the vertical load and camber of the opposite wheel thereby changing either lateral or longitudinal force. Although if you have a fairly stiff antiroll bar on an independent suspension car, hitting a single wheel bump will also affect the vertical load on the opposite wheel.

    I would like to take this concept further and see just how much we could simplify it. One of the things we are looking to test is a spool, we ran one in 2007 when our overcomplicated gear drive failed. The car didn’t accumulate many test miles so I am not in a position to make a very informed decision at this point. If we did run a spool and it looked to be successful then I would most certainly take a look at a solid axle like the one Buffalo has. I am a strong believer that a really simple car, well executed, using nicely machined parts in the right places has a real chance of being very successful. Sure in absolute performance terms it may be hard to beat a team like Stuttgart or Delft or UWA etc without the high levels of performance those designs may be capable of. The one thing I am sure of is that we could never pull off a car of that technical magnitude in one year let alone two. Our best chance of being really successful is to build the best C or B level performance car we can with a solid understanding of vehicle dynamics and test it, test it, test it, make it fast and then test it some more.
    We are currently running a rack and pinion steering on the front unsprung but we are going to look to change that to a go-kart style for reduced unsprung mass and further simplicity/lower cost. One other benefit I forgot to mention before was the reduction in the number of spherical bearings required. Our design requires 27 in total (not including ARB) where as previous designs have required 40. So that in itself is a pretty decent saving. Looking at Rob’s design, I think he would need fewer again.

    My biggest concern with the beam axles at this stage is compliance. I don’t think it is a show stopper though as it can be fixed through revised mechanical and structural design. I have also had time to think about how I would constrain the suspension in FEA modelling differently to how I did it. Constraining the model more realistically should help to target the causes of compliance.

    Cheers

    Olly

    ACME Racing
    UNSW@ADFA
    Olly

    Academy Racing 04-07, 09-11
    UNSW@ADFA

  4. #4
    Senior Member
    Join Date
    Sep 2004
    Location
    Rochester NY
    Posts
    1,061
    Olly,

    Just a side note did you see Eric Bana in the movie "Love the Beast". Such a good movie. Nothing will ever be as good as "Dust to Glory" but it was up there

    Do you have Erik Zapletals contact info. I have and old phone number but I was hoping to latch onto an email I could contact him at since I am thinking about building a z bar suspension for my 1982 Toyota Starlet.

  5. #5
    Olly,

    I'm an ex Team bath Racing Member from 2006. Funny you should mention beam axles - myself and an extreamly well respected special builder are currently developing the latest mallock racing car to use a de dion style rear end. However instead of using a a de dion tube we are using a space frame. The reasons.... well the gentleman that is helping me has in the past built 6 racing cars all around the beam axle principle. Two cars of which competed in F3.

    Email me on jonty_hair@yahoo.co.uk, send me some photos of your car and we'll discuss further!

    Remember the minute you put a roll bar on an independent suspension car it becomes dependant - also how big are the bumps on a race track???

  6. #6
    Senior Member
    Join Date
    Sep 2004
    Location
    Rochester NY
    Posts
    1,061
    Just thought I should mention that Satchel links work on dedion rear ends. It is in my opinion one of the best all around link configurations for beam axles/dedions.

  7. #7
    The section below comes from the March 2010 issue of 'Chassis Newsletter' by Mark Ortiz:

    DE DION TUBE DESIGN AND LOCATION

    Question:I am trying to decide on the arrangement to locate a DeDion axle in an autocross car that has a transverse engine/transmission directly in front of the axle. I plan to use a watts link for lateral location. Single trailing arms on each side with a third central link does not seem appropriate because the transverse engine would dictate a very short center link.

    I am considering parallel trailing arms on each side. However, I have seen applications that converge the arms on each side to a single front mount. It seems to me that in this situation there might be bending forces applied to the arms when one wheel rises and/or the other falls. Perhaps these are not significant because the axle will have quite limited vertical movement.

    Which do you consider more appropriate, parallel arms or triangular arms converging to one front mount on each side?


    Answer: The simple answer is to go with parallel arms. Converging arms or hairpin-style ones with a single pivot will bind in roll, unless the DeDion tube has a swivel in the middle.

    A swivel in the middle complicates the DeDion tube, but this was actually a feature of many designs when DeDion suspension was popular in F1 cars – for example the Mercedes W154. That car used a tube assembly that was rigid in bending and tension/compression, but not in torsion, and single trailing arms, with outboard brakes. Lateral location was provided by a roller in a slot machined into the back of the differential housing.

    The Rover 2000/3500TC used a tube that both swiveled and telescoped: it was rigid only in bending. Lateral location was then provided by fixed-length halfshafts, eliminating the need for any plunge accommodation at the shafts, and the need for any additional lateral locating mechanism. Single trailing arms were used, and brakes were inboard.

    Both of these designs offer somewhat more than 100% camber recovery in roll (ignoring tire deflection), and a roll center a bit higher than the halfshafts or the roller.

    The Mercedes design afforded very ample anti-lift in braking, due to the combination of outboard brakes and single trailing arms. The Rover design does not have comparable anti-lift, despite similar side-view geometry, because the brakes are inboard.

    Either of these designs could also use four trailing links, and they wouldn't have to be parallel. With outboard brakes, that would permit having any desired anti-lift in braking, without significant bump steer, which is not possible with parallel trailing links.
    ---------------------------------------------

    This seems fairly relevant to the discussion above and the type of De Dion we did on our car (bit of a bump too). The way we got around the kinematic binding which occurs with single pivot trailing arms was to use a torsionally flexible beam that was still fairly stiff in bending (a tube with a slit cut down the bending neutral axis.

    Cheers

    Olly
    ACME Racing
    UNSW@ADFA
    Olly

    Academy Racing 04-07, 09-11
    UNSW@ADFA

  8. #8
    One of the perceived weaknesses of beam axles is lack of camber gain in roll. I would say the solution to that is fairly simple:

    Use a front solid beam axle zero steering axis inclination and enough castor until you have the camber gain you want.

    In the rear use a twist axle style beam. We did it using a de Dion style design. Depending where you place the trailing arm pivots and torsion section of the beam, you can design in a reasonable about of camber gain.

    I would say the difficulty with beam axles is designing in enough installed stiffness in both camber, toe and hub to hub stiffness. Bump/roll steer can be an issue but there are definitely solutions available to reduce those effects. The unsprung mass can be quite high at 14.5kg for each corner (29kg total)on the front and 14kg in the rear. Our initial design was fairly conservative and I definitely didn't nail the load paths and material distribution first time around.

    Hopefully the beam axle discussions can continue here and continue to grow. Many of the reasons that beam axles fell out of favour are no longer relevant for our application for example we haven't exactly been crippled by steering shimmy.
    Olly

    Academy Racing 04-07, 09-11
    UNSW@ADFA

  9. #9
    I just spent the last week manufacturing a suspension compliance test rig to try and put some numbers to the compliance of our beam axle design. Photos of the rig are available at:

    http: // s928. photobucket. com/albums/ad127/oz_olly/Suspension%20Compliance%20Test%20Rig/

    (just remove the spaces)

    The first picture shows the rig set up with the car to measure camber stiffness and the other pictures (taken during fabrication) show the set up for toe compliance. Notice the leaning tower of G clamps for good measure. I did some quick tests this afternoon just to make sure everything worked and it all seemed to go quite well. The angle results were only accurate to +-0.05 deg as I was using an electronic camber gauge. When I do the full set of tests I will use either linear pots or LVDTs to get better resolution.

    My initial results are 285Nm/deg in camber with 1.6g*80kg*0.2413m giving a moment of 303Nm. So we lose 1.07 deg camber at 1.6g with 80kg static corner weight. That's much more compliance than I would like and I have a feeling the toe compliance is going to be worse but now that we have a way to measure it we can quantify improvements.

    Cheers
    Olly

    Academy Racing 04-07, 09-11
    UNSW@ADFA

  10. #10
    bump! My post above was delayed even though I broke the link.
    Olly

    Academy Racing 04-07, 09-11
    UNSW@ADFA

+ Reply to Thread
Page 1 of 28 1 2 3 11 ... LastLast

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts