+ Reply to Thread
Page 13 of 29 FirstFirst ... 3 11 12 13 14 15 23 ... LastLast
Results 121 to 130 of 284

Thread: Beam Axles - Front, Rear or both.

  1. #121
    As any other design, there are up and downsides on a beam axle setup. The hard task is to identify the possibilities and flaws of each design, and then choose and go with the one that suits you better ACCORDING TO YOUR TARGETS. There is no thing such as "optimum solution". I am pretty convinced that a beam axle car can have similar (if not better) dynamic performance than a "conventional" FSAE car, and there are many examples out there that favor a simple-as-hell reliable car. A recent example I can remember of...

    http://fsae.com/eve/forums/a/t...20065151#11120065151

    Keep in mind that we are talking about GFR's 2011 car in comparison, which is one of the fastest FSAE cars out there. Also take a look at Rob's old school (Uni at Buffalo) and teams like ADFA, CalPolySlo, Mini Baja cars etc... At least I was surprised! We might surprise quite a few in the not-so-distant future too...

  2. #122
    Senior Member
    Join Date
    Sep 2004
    Location
    Rochester NY
    Posts
    1,061
    BTW I looked up the results in 1996 from the Akron Bajamula and they finished 11th overall even after their fuel pump quit in the enduro. They ran a pure rotation swingarm for a rear suspension and just swapped the 18hp Briggs V-Twin for the spec Briggs single. Their baja car that year if memory serves me correctly was 320lbs and there is a 30lbs difference between the motors so they were probably at 350lbs for FSAE comp.

  3. #123
    Senior Member
    Join Date
    Sep 2004
    Location
    Rochester NY
    Posts
    1,061
    Dr. Paasch's post also really got me thinking that it would make a hell of a novel idea to do a Bajamula again but with Z's interconnected suspension utilizing front leading arms and rear trailing arms (2CV style) and just replace the "black box" or trailing/leading arm linkage configuration along with engine (45 degree leaned single and 90 v twin) in between events. It would make a much cheaper vehicle that would would be very competitive and be able to run dual competitions.

  4. #124
    Senior Member
    Join Date
    Mar 2005
    Location
    Australia
    Posts
    1,688
    Charles,

    "Wouldn't a car that's soft in twist be less responsive to changes in ERMD/LLTD? I remember hearing about some older cars that couldn't have o/s or u/s tuned out of them because the chassis was too soft in twist - would you have to make changes to the twist stiffness to get some changes out of this?"

    Big subject here, but briefly.

    The issue of "not enough chassis torsional stiffness" started, in circuit racing, when wings caused the springs to become much stiffer to take the aero loads. So, for example, the race engineer might have wanted the front tyres to carry most of LLTD so he stiffened even more the front springs/ARB. But the engine, mounted at the back, kept "leaning" on the already stiff rear springs, transfering all its load (ie. most of the car's mass) to the rear wheels. The pop-rivetted aluminium chassis didn't have the stiffness to pass any of the engine's LLT to the front tyres. So oversteer always, even with completely rigid front springs.

    However (!), way back in the early 1900s the chassis was deliberately made soft. The production car would typically have its large and rigid engine mounted at four points (2F-2R) to the twin-rail chassis to torsionally stiffen it, mainly to stop the doors rattling, windows cracking, etc. For racing the engine was mounted at three points, one at the nose and two either side of the rear bellhousing. This allowed the chassis to twist, which was useful for grip on the rough roads (which was not helped by the stiff leaf spring suspension).

    This arrangement was quite similar to the top-left of the Z-Bar Sketch, with the sketch's longitudinal leafsprings being the chassis rails, and the centre structure being the engine. During cornering the engine "leant" on the centre of the chassis rails, thus distributing its LLT about equally front to rear. If the engine had its mounts with 2F and 1R, then all its LLT would go to the front wheels and there would be massive understeer (all else equal).

    I believe a similar thing happens with go karts when the seat mounting points are changed? The driver is the largest mass, and if during cornering his weight is transfered, via the seat mounts, closer to the front of the chassis, then US. If more to the rear, then OS.
    ~~~o0o~~~

    Regarding complexity of De Dion versus rear wishbones.

    In both cases the engine/diff/halfshafts/stub-axles are the same.

    I would NOT do a "sliding-tube" DD just to avoid plunging CVs. I would do a beam as in the sketch, but without the two inner bearings.

    So,
    1x largish, but simple, triangular tube structure for the beam,
    1x heavy duty apex BJ,
    1x P&S, as in sketch,
    2x (optional) bolted connections at "beam-ends/uprights", for camber/toe adjustments.

    Again, the chassis now only needs two accurately placed hard points on its centreline at floor level (using stringline to align these and front points), and two less accurate points for the spring-dampers. The problem with wishbones is that more points have to be located accurately in 3-D, and their preferred locations don't always match the chassis nodes (eg. upper side-impact tube has to be at Z = 350mm (memory?), which might not be the best kinematic location for the upper wishbone mounts).

    (Edit: Oops! I forgot that a DD needs different side-view kinematics. The above list should be changed slightly, but overall the "complexity" (time to build) is about the same. I might do a sketch of this option one day. It can be done with only 3 chassis hard points for the rear beam (the P&S is dropped).)
    ~~~o0o~~~

    "Farm tractors..."

    Ahhh, I love my tractors... I better not bore you.
    ~~~o0o~~~

    "As drawn, you have moving front suspension arms [the front beam] inside the cockpit and a heavily loaded front suspension pivot - at the back of a leading arm no less - that is ...
    ... in very close proximity to a rather delicate part of the driver's anatomy."


    I would have both front and rear beams mounted UNDER the chassis's floor. The BJs would be attached to chassis with vertical bolts in single shear, to act as "fuses". In a big impact the bolt fails, chassis remains undamaged, and the beam slides under the floor (or car slides over beam). This also makes mounting of underwings easier. (I have grown fond of the "family jewels", and would not risk losing them.)
    ~~~o0o~~~

    "A passenger car's suspension doesn't actually have to work over bumps!
    Almost every consumer willingly accepts a car that rides and handles like crap."


    Agreed. Sad, but true.

    In the past I have suggested getting rid of the mandatory "+/-1 inch suspension" rule, and just scattering a lot of bumps around the track (say, molded rubber "cowpats", 2-3m (10ft) diameter, 0.1m (4") high in the middle, and tapering to very little at the edges).

    Teams can then build a cheap no-suspension go-kart and pay the penalty of driving around the cowpats (or launching over them!), or else build a car with good suspension that allows them to take the fastest racing line, regardless of bumps.

    This could be a long term benefit to consumers of production cars.

    Z

  5. #125
    Originally posted by Z:
    Important De-Dion note. The side-view kinematics, or anti-pitch, changes when the diff is chassis mounted. A layout as in the sketch, but with De-Dion, will have lots of pro-squat under power. This can be ignored with a low power engine, or a beam-centre bump rubber (aka "third-spring", or "lateral Z-bar", as mentioned by Harry above) can be used to limit squat (axle bounce) but allow roll. Alternatively, a different kinematic location can give anti-squat (higher pitch axis) while maintaining a low roll axis. This is a quite simple structural change, but will require a sketch to explain it properly...

    Z
    Z,
    I might have completely misunderstood, my kinetmatics knowledge isn't great, but the pro-squat under power you talk about, is that due to the reation moment transmitted into the diff causing a rotation about it's axis. In which case displacing the diff decreases the effect this has on the chassis? For a moment there I thought I'd got it visualised in my head, but then I lost it, I need more experience.

    I would love it if you could enlighten me. (If it sounds like I'm gunning for that sketch you mentionned, that probably because I am.)
    Dunk
    --------------------------------------------------------
    Brunel Racing
    2010-11 - Drivetrain Development Engineer
    2011-12 - Consultant and Long Distance Dogsbody
    2012-13 - Chassis, Bodywork & Aerodynamics manager

    2014-present - Engineer at Jaguar Land Rover

  6. #126
    Senior Member
    Join Date
    Mar 2005
    Location
    Australia
    Posts
    1,688
    Originally posted by Dunk Mckay:
    ... the pro-squat under power you talk about, is that due to ...
    If it sounds like I'm gunning for that sketch you mentionned, that probably because I am.
    Dunk,

    That sketch might take some time, so here's some that might help until then.

    Figure 7 below illustrates "side-view antis" (eg. anti-squat under power, or anti-lift when braking) in terms of a suspension's "n-lines". N-lines are simply the lines "normal" to the direction of travel of any point in a linkage. Alternatively, the n-lines can be thought of as the directions of "no" motion.

    Importantly, a suspension's n-lines are all you need to know, in order to understand its "antis".

    In the figure the point being considered is the wheelprint (simplified to a point). Since the suspensions shown have only one degree-of-freedom (this needs more explaining, but...) the wheelprint can only follow a single, roughly up-and-down path. So the wheelprint has a "planar pencil of n-lines" that is perpendicular to its path of motion. (A "PP" is a bunch of lines, lying in a flat plane, and all passing through the same point, like the spokes of a wagon-wheel). Here we are interested in the particular n-line that also lies in the side-view (vertical-longitudinal) plane.

    Importantly, in any linkage, forces can ONLY travel along n-lines. So, in the figure, any component of force acting on the wheelprint that lies outside of the plane of n-lines must necessarily be reacted by some other structure, which here is the spring-damper (not shown). So if the same arbitrary force acts on the wheelprint in all of the examples in the figure, then in each case the "component of force oustide of the n-line plane" will be different, so the spring-damper will be squashed to a different length, and the "anti" behaviour will be different.



    Regarding an FSAE rear beam-axle, consider Figures 7a (similar to a "live" axle), and 7b (similar to a "De Dion" axle). In both cases consider the swing-arm pivot (towards top-right of each little sketch) as being much closer to ground, so well below the axle height. This gives similar side-view kinematics to my earlier "Twin-Beam" sketch, with the "apex BJ" = "swing-arm pivot".

    Figure 7a now has a side-view n-line that is close to horizontal with only a slight rise to the front (right of pic). This gives a small amount of anti-squat.

    Figure 7b (= De Dion) now has its n-line sloping quite steeply down to front. This gives quite a large amount of pro-squat. The question is how to push this n-line up-at-front, for a little anti-squat, while keeping the lateral n-lines close to horizontal, for good cornering behaviour. (Answer later...)
    ~~~o0o~~~

    Meanwhile here is another figure that might help understanding of the distribution of wheelprint forces between the control-arms (ie. along their n-lines) and the spring-damper. Shown is a car accelerating longitudinally (eg. forward to right), but same principles apply for lateral acceleration (ie. cornering). Black arrows are the forces acting on the wheelprint or car. White arrows are some of the possible components of the black arrows. Fca = Control Arm force. Fsd = Spring-Damper force. H = Horizontal. V = Vertical. G = Gravitational. I = Inertial. F1,2V = Vertical forces on wheels 1 & 2. Etc...



    (BTW, I uploaded the above pics somewhere (Picassa?) sometime ago, and I can access them easily via my old posts on this forum, hence this quick reply. New pics mean that I have to go shopping for a new cable for the old (perfectly good) scanner, or buy new scanner, or???, so will take more time.)

    Z

  7. #127
    That's awesome, really helpful, I can visualize it properly now.

    Ok, so I've been pondering this twin beam-wing concept of yours, Z, since a friend suggested it a while ago. Starting at the rear wheels (front can come later, much later) I've been trying to figure out the best way be able to adjust camber and toe.

    The best ideas so far are these:

    -big solid uprights that mount onto the beam via interchangeable plates/spacers which have the camber and toe incorporated. The issue with this is that for stiffness these will have to be quite bulky. And even though you'll generally only need one setup for running the car you'll need to fine tune it to find that first, so you'll have to machine a large variety of parts to that kind of angular tolerance, which is gonna be costly (time or money or both).

    -and upright mounted with a single B.J. at it's base to the beam end, then two adjustable tie rods between the upright and either arm of the beam. These of course will impede air flow (if only a little and above the wing), and load the beams such as they will be bent in the middle unless they are made fairly bulky as well. One also ends up thinking that perhaps double wishbones might not be that much more complicated.

    Much as I'd like to say "stuff it" and come up with a design with fixed camber and toe, both design judges and the academic staff at my school would rip it out of us if we made that. I also think being able to find the ideal setup is not really something worth sacrificing for the sake of oversimplifying things.
    Dunk
    --------------------------------------------------------
    Brunel Racing
    2010-11 - Drivetrain Development Engineer
    2011-12 - Consultant and Long Distance Dogsbody
    2012-13 - Chassis, Bodywork & Aerodynamics manager

    2014-present - Engineer at Jaguar Land Rover

  8. #128
    Aha! I've just had another idea, and feel really stupid for not thinking of it in the first places, my normal work is obviously draining my brain.

    Essentially, similar to the from beam ends on your original sketch, the beam forks with the top fork extending upwards. An additional arm would branch off forwards or backwards, and then a very light simple upright would mount to these three point via some Shim mounted BJ's. Still fairly bulky but...less aero disturbance and less complicated (calling it more elegant would be a push, no?)

    EDIT- Three roughly evenly space branches (2 mottom one top) might end up more optimal, depending on loads and packaging issues for caliper mounts, provided you are mounting brakes outboard that is.
    Dunk
    --------------------------------------------------------
    Brunel Racing
    2010-11 - Drivetrain Development Engineer
    2011-12 - Consultant and Long Distance Dogsbody
    2012-13 - Chassis, Bodywork & Aerodynamics manager

    2014-present - Engineer at Jaguar Land Rover

  9. #129
    Senior Member
    Join Date
    Mar 2005
    Location
    Australia
    Posts
    1,688
    Dunk,

    The De Dion is my preferred option (the sketch shows a spool because of earlier discussions on "Suspension for Spool" thread). With a De Dion I would definitely have some means of camber and toe adjustment, both to correct for manufacturing tolerances, but more importantly to tune handling.

    This is a good opportunity for some original detail design (at Geoff's "Reasoning..." Level 1), mainly because there aren't many examples to copy. I can think of several ways to do it that would add less than 1kg to the total weight, and I might post some sketches later.

    Meanwhile, how about asking all your team members (even the newbies) to solve this "Engineering Design 1.01" problem:

    Design a joint in the middle of a length of horizontal (70mm diameter? x wall thickness?) tube, so that an accurately adjustable "bend", in both planes (ie. camber and toe), can be put in the tube, and the joint has the same strength and stiffness as the original tube (in bending, torsion, etc.). Low weight, ease of manufacture, and ease of "bend" adjustment being priorites. Camber adjustment +/- 5 degrees in 0.5 degree steps, toe +/- 1 degree in 0.1 degree steps, or better.

    Who knows? Maybe a newbie will come up with a cracker!

    Your last edit seems close to my preferred idea, but details are important. Otherwise, think about a wheel bearing housing that can be adjusted for camber/toe inside a larger, concentric, clamping sleeve???

    Oh yes, KISS!

    Z

  10. #130
    Any team members still around are busy finishing off the car at the moment. I'm actually on an work placement, although I'll be taking a long weekend in a weeks time to go help them finish build before FSUK. I'll be managing next year so I'm planning ahead, getting some ideas together for when I go back in August.

    Sketches of ideas are good, sketches are always good. I can't sketch very well however although my technical drawing is excellent, seems a little contradictory I know.

    However I can CAD reasonably well. I've had a crack about your problem but I think I got carried away, I have a soft spot for U-joints for some reason (worked on driveshaft and CV joint stuff last year, no actual u-joints though).



    There's not real scale or dimensioning as such, just the concept. Additional turnbuckle could make it stronger but would also fight each other if not adjusted properly.

    EDIT--my grammar is terrible.
    Dunk
    --------------------------------------------------------
    Brunel Racing
    2010-11 - Drivetrain Development Engineer
    2011-12 - Consultant and Long Distance Dogsbody
    2012-13 - Chassis, Bodywork & Aerodynamics manager

    2014-present - Engineer at Jaguar Land Rover

+ Reply to Thread
Page 13 of 29 FirstFirst ... 3 11 12 13 14 15 23 ... 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