Consider a double-trailing-link suspension mounted to a subframe, an assembly superficially similar to a VW Beetle front suspension. Consider now that one trailing link is connected to its counterpart on the opposite side by a relatively large-diameter tube, free to rotate within a tubular member of the subframe, so that both trailing links act together as a virtually solid unit. The action of the trailing links is sprung by any of various arrangements of spring(s) and damper(s), perhaps a coil in torsion, or torsion bar(s), or hydropneumatics, or whatever.
The subframe is mounted to the vehicle by a system of lateral links that allows it to rotate as a unit about a centre defined by the geometry of the links, which centre would be the roll centre of the system. The rotation is resisted by a second arrangement of springs and dampers, which again may be of various sorts. The space required for the rotation of the subframe is limited as no vertical movement and only a small angular movement needs to be accommodated. The first spring (or springs) deals only with bump movement, and the motion is as for a double-trailing-link system. The second springs handle only roll movement.
The advantages of this system are that there are (nominally) no camber changes with suspension travel; that bump and roll spring rates may be optimized separately and damped properly; that the roll centre may be placed accurately wherever convenient, thus limiting roll couples to a range where bump and roll spring rates are of similar orders of magnitude, and the demand for torsional rigidity of the vehicle structure is reduced; and that high roll centres may be used without geometries that are prone to jacking. The system is equally suitable for front and rear suspension, for driven and non-driven wheels, and for installation with links trailing or leading. A suitable steering gear might be tricky to arrange, but by no means impossible.
I've added a few sketches. The colour isometric shows the general principle in notional form. I've shown a coil in torsion and rotary damper for the bump spring, mainly for clarity, but other spring types may be used.
The four end-view diagrams compare the system here proposed (right) with a conventional double-trailing-link system (left). What applies to the latter applies equally to a torsion beam or any kind of pure leading or trailing arm system. Note the roll-induced positive camber (alpha) and the long roll moment arm H with the conventional system. The proposed system has no roll-induced positive camber and a very short roll moment arm H (and consequently very little CG displacement D - though lateral CG displacement is seldom great enough to be much of a factor with any system. I mention it only to pre-empt objections on that point.)
I have two progressions on the above:
Firstly, the roll springing may be interconnected between front and rear, as on a Citroën 2CV, BMC 1100, or 1955 Packard. The advantages regarding torsional compensation etc. are retained without the concomitant instability in pitch; i.e. the vehicle won’t sit on its bump stops at one end if, for instance, a load alters the weight distribution beyond a narrow range. Coupled with a high roll axis, the need for a torsionally rigid vehicle structure is thus virtually eliminated.
Secondly, though the camber remains nominally constant, tyre compression will result in a small positive camber in the wheels on the outside of a turn. An arrangement may be incorporated to induce a compensatory negative camber. If this is referenced to roll motion it is a relatively simple matter of additional lateral links that locate the upper trailing-link pivots (assuming the lower trailing links to be coupled). Ideally (or anally, depending on how you look at it) it would be referenced to lateral acceleration, and require geometrically-determined compliant joints to transfer lateral forces to an instant centre somewhere below the outside tyre’s contact patch. That is another exercise, however.-- Ned_Ludd, Mar 08 2007 Here's a drawing http://i88.photobuc...eCoetzee/RTLS01.jpgBlue components act in bump only. Red components act in roll only. The rocking subframe is the green thing. [Ned_Ludd, Mar 09 2007] Here are some diagrams http://i88.photobuc...eCoetzee/RTLS02.jpgA conventional double-trailing-link on the left; the system here proposed on the right. Exaggerated bump movement at top; exaggerated roll movement at bottom. [Ned_Ludd, Mar 09 2007] Gee This looks awfully familiar http://www.railway-...sweden/sweden6.html [jhomrighaus, Mar 09 2007] Bose suspension system http://www.bose.com...n_solution.jsp&ck=0via bose.com [acurafan07, Mar 10 2007] Dax Cars http://www.daxcars....l-suspension-01.htmSimilar camber compensation? [Anchovy, Mar 10 2007] Sorry, [Ned], I need diagrams.-- angel, Mar 08 2007 I have considered the info but am left wondering about what it is all for. Making a beetle handle just a little better? or can you provide a better overview of the concept?-- the dog's breakfast, Mar 08 2007 As I said, there is a superficial resemblance to a Beetle front suspension. Note the word "superficial".
I'll see if I can find time for some explanatory sketches.-- Ned_Ludd, Mar 08 2007 no. read it all again. it would drive like a boat. no diagrams necessary.-- the dog's breakfast, Mar 08 2007 [TDB], would you care to explain?-- Ned_Ludd, Mar 08 2007 Torsion Beam suspension would serve the same purpose but is significantly lighter and less complicated. Torsion beams are essentially trailing arm systems with a flexible beam in between which allows for differential movement without changes in the camber of the wheels as due to the trailing arm aspect of the system. Springs and dampers connect to the trailing portion of the arm and sway bars are readily mounted. This is a common rear suspension design in use on most GM small and midsize cars and I believe on many minivans as well. I do not see the advantage of your system over this design.-- jhomrighaus, Mar 08 2007 With a torsion beam, as with a conventional trailing-link system, camber changes with roll. Because the camber orientation of the wheels remains constant in relation to the vehicle structure with these systems, the camber orientation of the wheels must change in relation to the road when the angle of the vehicle structure changes during roll.
Camber is a useful measurement if it is taken in relation to the road. It doesn't mean much when it is taken in relation to the vehicle structure. It is the way a wheel relates to the road that induces camber thrust, for instance. It doesn't matter how the wheel relates to the car's body. Thus, when we criticize a swing-axle system, for instance, for undesirable camber changes, we mean changes in camber relative to the road, not relative to the car's floor.
So, a torsion beam won't do the trick. With a torsion beam, 1deg of roll means (nominally) 1deg of positive camber over the static spec. That is why they are used on the back of front-drive cars that would understeer woefully if not counteracted by "deliberately bad" rear suspension. Moreover, the torsion in the torsion beam is undamped, and one has an intrinsic problem determining a damping setting that satisfies both bump and roll requirements. Then, you're stuck with a roll centre on the ground and a considerable roll couple. In short, a torsion beam provides none of the advantages I have listed.-- Ned_Ludd, Mar 08 2007 Would it be more understandable if I added gratuitous electronic controls?-- Ned_Ludd, Mar 09 2007 yes. more linkages-system mounted to subframe by series of linkages etc. Going through any more links from frame to wheels creates more movement, even when one bit is on a floating subframe. So the body would roll around like a pig. My main problem though is with the title/description. If you had it down as "a tricky way to make a trip to the shops seem like a pleasure cruse" then I would bun it. anyhow on the gratuitous electronic controls have you looked at just using dampers like on new ferrari that change their internal viscosity when you zap them?-- the dog's breakfast, Mar 09 2007 [TDB], you're worried about lost motion through cumulative compliance? i.e. a little motion being lost at every connection and the whole lot adding up? Please answer, as I would like the opportunity to address your concerns.-- Ned_Ludd, Mar 09 2007 Okay only once more. Yes the suspension system works. But. Only if you have no more connections to the body than you have drawn in the diagrams to the right, then massive body roll will occur even with any godawful number of links. If you also connect to the body where you have drawn the springs then the assembly WILL behave as in the diagrams on the left. HOWEVER. PLEASE NOTE. describe WHAT it is that it is supposed to do- or do poorly (not how) so that i can decide to bun it or bone it, which has been the guts of my scratchings all along.-- the dog's breakfast, Mar 09 2007 Your first question last: the purpose is firstly to allow a better ride quality without sacrificing handling, by separating bump and roll spring functions; and secondly to allow vehicles that have limited torsional stiffness, like hot rods, for instance, to ride and handle well.
Note that the tube connecting the lower trailing links is not a torsion bar. It is not intended to twist during roll. For that reason it would probably be a hollow tube of relatively large diameter.
Note also that the red springs in the isometric are positioned notionally. The idea is merely that there are springs that resist rotation about the roll centre, wherever one chooses to locate the roll centre and whatever spring medium one chooses. Those springs resist only roll and are not affected by bump motion at all. To understand it, forget the lateral links and consider that the subframe has only a single pivot to the body, at the roll centre, like a bushing with a bolt through it.-- Ned_Ludd, Mar 09 2007 Wouldn't your system suffer all the issues inherent to solid axle designs when you introduce a varying road surface? While you are attempting to address body roll, the applications you site are those which suffer least from body roll due to lightweight and low center of gravity.
Your system is in effect identical to the active tilt systems used in high speed trains accept it is passive rather than active.-- jhomrighaus, Mar 09 2007 This is an interesting idea and I think it would probably work as you've described it. I don't think the extra cost, complexity, and weight would be worth it, but you might find some sort of application. I've enjoyed your somewhat wacky suspension designs so far, so keep 'em coming!
One question: how exactly would you separate roll and bump? It seems to me that both of the suspension systems you've put together would allow for bump and roll, unless I missed some linkages in the picture.-- discontinuuity, Mar 09 2007 Ned after reviewing your other idea I am confused about your application of these linkages from the perspective of their use in vehicles. In both designs I see your concern about camber and roll center conditions and about the location of the roll center of the vehicle, but both designs seem to have the same flaw in that the only way that one wheel can rise above the other involves a tilt of the entire vehicle about its roll center rather than allowing the wheel to ride up and over a bump independently of the other wheel. In a simple scenario it would appear that if one wheel drove over a 3 inch bump that the opposite side of the vehicle would have to be lifted by some corresponding amount to allow the linkage to actuate. This is more apparent in your previous linkage but as you describe this one I see a similar fault. The wheels cannot follow the road without the entire body of the car shifting in response.(hit bump with right wheel vehicle tilts to right)
While your applications to roll center and tread orientation certainly seem reasonable I think your missing one of the key features of an automobile suspension that is the ability to deal with potholes.-- jhomrighaus, Mar 09 2007 Now I've gone over the idea and I think I understand it better. Like I said, this would probably work in theory, I'm just not sure how well or whether it would be worth it.-- discontinuuity, Mar 09 2007 Thanks for the railway link: I wasn't aware of that application of an instant centre.
I hear what you say about single-wheel bump behaviour. Yes, single-wheel bumps would be handled effectively as roll and, yes, that would mean a momentary camber change in the opposite wheel (the entire vehicle would tilt to some extent less than the suspension travel, but so would it with any other system). I submit that the importance of that effect has been exaggerated in our minds. I think we've all read up on suspension design and have thus all been exposed to GM's Knee-Action ads from the '30's. That gives opposite-wheel effects as the most important, if not the only, reason for the change to independent suspension. The history of suspension design would seem to indicate otherwise.
There are a number of problems with solid axles. Unsprung mass is a factor with live axles; with dead axles there really is very little difference. The most important problem is that a conventionally located and sprung solid axle has a spring base that can be as little as 40% of the track, which means that it affords relatively little roll stiffness at any given spring rate in bump. In other words, decent roll stiffness can only be achieved with a very harsh ride. This is a perfectly valid problem, to which independent suspension is one of many perfectly valid solutions.
Another problem with solid front axles, which belies some slight cynicism on the part of the motor industry, is that designs of the '40's were going nose-heavy and highly polar in order to achieve the 1.3-1.5Hz resonant frequencies we've since come to expect. This was done by shoving the heavy engine forward between the front wheels, and a moving solid axle was simply physically in the way.
An unfortunate side-effect of ifs was dropping the roll centre from about a foot/300mm above the road to about 1 1/2"/40mm, and similarly increasing the roll moment arm length. For that reason early non-unitary ifs-equipped cars had heavier frames than their solid-axle predecessors. Compare for instance the simple, delicate frame under a MG TC with the gross excrescence supporting a TD. This was necessary to ensure that any attempt at suspension tuning wasn't simply lost in chassis flex.
I can't explain the Chrysler Airflow and Lincoln-Zephyr at all. The use of unitary construction with solid axles can arise only from some combination of marketing and insanity. The mainstream motor industry had reasons to favour unitary construction, not all of them honourable.
What I'm trying to do is to go to the historical juncture where the problems of solid axles were first addressed, and investigating other approaches with the benefit of hindsight. I am particularly interested in simple, 2-D, flexible frames, for various reasons.
[raspberry re-tart], of course suspension movements have simultaneous bump and roll components. My intention is not to separate the movements so only one can happen at a time, but to spring and damp the components separately.-- Ned_Ludd, Mar 10 2007 I have been struggling through these gearhead ideas of yours, [Ned], looking for one that I understood well enough to give you a bun, or a bone - no luck. I appreciate that these are real ideas, meant to work: a valuable commodity on the HB.
I am not sure why [acurafan] has not visited any of your ideas. His (her?) ideas seem cut from the same cloth as yours.-- bungston, Mar 10 2007 I've visited them- not exactly "cut from the same cloth" (although still kind of similar). Aside from the fact that the idea is hard to understand for me even with the illustration (which was actually done quite well), it's hard for me to be interested in any ideas about suspension systems after seeing the Bose suspension. No offence to [Ned] or anyone with a half-baked suspension idea, I just don't think that there is currently a way to do it better than the Bose system. But this idea certainly is creative, so bun for that.-- acurafan07, Mar 10 2007 Thanks for clarifying, Ned. I can see that this would work, but I also see that you could have problems with everything binding up or rocking the body too much if you hit a pothole. I suppose the only way to know for sure is to build it or make a computer model. I'm sure that you could make a prototype out of a VW fairly easily. Until then, I think that independent suspension, a solid axle, or a DeDion axle would be best in this situation.-- discontinuuity, Mar 10 2007 Thanks for the Dax link. It wants some close study, with my Morris Minor in mind.
As for the Bose system, I was very excited about that sort of thing myself, when I was young and foolish and hadn't yet fallen victim to perfidious electronics as many times as I have. It wouldn't work for my way of living now.
[rasp], to some extent this idea is an attempt at a more space-efficient version of my previous solid-axle locating idea. The separation of bump springing was almost an unforeseen bonus.
Thanks, people, for trying to get your heads around this stuff in earnest.-- Ned_Ludd, Mar 12 2007 random, halfbakery