Interconnected Suspension: How?

I'm nowhere near as well-versed in the mechanics as you are, so let me know if I'm way off base. I have given this some thought as well.

If we're trying to keep cost down, perhaps hydraulic bumpstops could be explored? They're reasonably attainable at less than $200 per unit, very simple in design, and built to take a lot of abuse. I've seen some sites list them around 5-8lbs, depending on the model. Could one theoretically machine some way to permanently attach them to the suspension/chassis, plumb some hydraulic lines to them along with an accumulator, and get some semblance of interconnected suspension? Maybe even have them act on levers attached to individual torsion bars at each corner, similar to a sway bar?

*EDIT* Just watched the kyle.engineers video and it looks like there's quite a few more parts needed for this to really get the results we're looking for.

Or perhaps remote reservoir shocks? They're already plumbed and ready to hook up, just remove the reservoir and fab up a hydraulic line that hooks to your accumulator system.

So the goal would be the analog version of the BOSE suspension?

I think you should consider off-the-shelf parts - specifically, using Citroen parts. They're usually cheaper than decent steel springs and hydraulic dampers.

They worked well enough to get permanently banned in WRC - twice!

You could totally do a reasonably priced DIY implementation of this with a set of cheap used rebuildable (some are adjustable too) motorcycle rear shocks run in parallel, a pair of ports drilled and tapped into their sides for AN, a bunch of AN line, and a couple remote reservoirs as accumulators. Your best bet from a system efficacy, reliability, weight, tunability, and quite possibly cost standpoint is hydraulic. a la McLaren 

In reply to WillHoonForFood :

My understanding of hydraulic bump stops is that they're basically just a small fat shafted monotube damper, with a gas chamber to provided a spring force and a valved piston on the shaft to provide damping. So to use these would still require eliminating the fluid transfer across the piston.

In reply to WillHoonForFood and bumpsteer:

Yes, I've thought about using shocks, but I'm skeptical their design would hold up well. First it would have to either be ordered with a custom zero-port piston, or would have to be rebuilt and modify the piston to fully seal all ports. Beyond that I'm concerned about the much higher internal pressures it would be required hold a perfect seal against due to the relatively small piston diameter. As a damper, a small amount of imperfect long term sealing reliability in the name of reduced stiction might be an acceptable compromise. As a hydraulic cylinder it would be considerably less so. They also tend to have considerable thinner walls than true hydraulic cylinders, which under such high pressure could also potentially move enough to affect sealing.

That being said, it's still something to consider.  While I agree on the efficacy and tunability advantages, I guess I'm just not sold on the reliability of sliding seals in a closed circuit application where it's the only thing carrying the weight of the car, as I was hoping to avoid the complexity of pumps and controllers and whatnot. In my limited understanding, beyond addition to the extra maintenance of fluid changes to keep it working right, leak s are still a common problem with the purpose built hydraulic cylinders in lowrider suspensions. Remember the cost and complexity of the Mclaren system is also still only functionally replacing the sway bars.

If Hayabusa shocks can be readily DIY rebuilt, I do have a couple of those sitting around doing nothing.

In reply to NOHOME :

More or less, yeah, the goal would be kind of like a passive analog (purely mechanical) version of that.

In reply to chaparral :

I seemed to recall the Moulton systems parts being in short supply and not serviceable even in Europe, let alone State side... So I guess I just assumed the same of the Citroen system. That's definitely something to look into. If nothing else, these should also be usable as accumulators... And I guess by extension, other accumulators could be used similar to this. Hmmm, something else to think about.

Despite my concerns that the compressible flow through the connetecting tube might not only be helpful for damping the system, but conversely also lead to responsiveness issues, my mind actually keeps coming back to air bags as possibly the easiest to implement system demonstrator with the fewest moving parts. I'm also not sold on it either way though. Thinking more along the lines of bicycle 'air shocks', with the 'positive' and negative' air chambers, it could even function similar to the double acting hydraulic cylinders typically shown for interconnected hydraulic systems. The cost/benefit of that is debatable though, since air/hydraulic connections are really more helpful for the front/rear connection than they are the left/right connection.

I've even thought about using oil filled air bags the 'cylinders' in a much lower pressure hydraulic system, but the added complexity of this still requiring an accumulator would really only be necessary if the compressibility of the air connection did have responsiveness issues.

I'm loving the idea generation and critical thinking continuing to poke at different ways of doing things. Thanks!

Are you dead set on having the hydraulic struts support the vehicle's weight? I feel like the lower budget constraint would make that incredibly difficult, whereas a system that works along side conventional suspension. That could also give room and budget for developing better hydraulics.

WillHoonForFood said:

Are you dead set on having the hydraulic struts support the vehicle's weight? I feel like the lower budget constraint would make that incredibly difficult, whereas a system that works along side conventional suspension. That could also give room and budget for developing better hydraulics.

In a FRICS setup, coil springs support the vehicle's weight, with the rates chosen primarily for comfort, and the hydraulics control body motion.

Is this the simplest answer? At least for one end of the car...

Still using springs to support the weight of the car is a bit fo a 'belt and suspenders' approach. It reduces the benefits of the interconnections, but obviously can work well though too. Of course, you still can't do away with the damping either, so off hand I see two main ways of doing it:

The elegant one is to replace the standard dampers with ones modified into hydraulic units. This would package quite nicely. The challenge there, beyond modifying the dampers into double acting hydraulic cylinders, is that there now needs to be some trick custom machined bits to incorporate functional (locked in place) damper valves into the hydraulic circuit, as the high-dollar professionals have done.

The brute force one is to keep the standard dampers and add a second set dampers modified into hydraulic cylinders. This would be the more DIY-able option, although at that point could also just use off-the-shelf hydraulic cylinders as well. The big drawback there is the additional bulk, complexity, and redundancy of having all the components required for two functionally independent (conventional and interconnected) systems, when not even getting all of the benefits of one.

In reply to Asphalt_Gundam :

Unless I'm missing something, that looks to be a conventional formula car style inboard suspension with conventional (U-bar effect) functioning T-bar style sway bar. Unfortunately, the only piece that would be relevant to this discussion, is the one that it doesn't have... The '3rd element', which acts as a lateral Z-bar.

Yeah that's an inboard version of a conventional coil spring + shock + sway bar arrangement.

The last few days I've been coming back to this torsion bar arrangement for the longitudinal linking. It's more complex than an air line between two air bags, but really only about as much so as an inboard suspension, while also still providing a positive mechanical connection. With the long distance between the firewall and front axle on a 7-esque design giving the bar a place to reside, the partial frame in my garage should actually be pretty well suited to it. Not to mention that the 'U' bar (still functions as a 'Z' here) gives good clearance for steering lock.

The above diagram shows 12 pivot points added by this "Z" bar arrangement.  In my years of racing we found "stickson" in the suspension was the enemy to it working as designed.  We worked hard to make sure each pivot moved as freely as possible without being so loose as to loose proper calibration.  We were serious about rod-end suppliers and the such.   Somehow adding 12 more pivots on one side,and 12 more pivots on the other side, while in theory a solution, really adds a huge amount of work to build it so it moves perfectly smooth.  Probably why engineers that attempt it turn to hydraulics.

Just an opinion.  Back to the regular scheduled programming.

In reply to Purple Frog (Forum Supporter) :

Counting it out, I see:

(4) Rods, 2 spherical joints each, 8 total spherical joints.

(2) Bellcranks, 1 revolute joint, 2 total revolute joints.

(1) Torsion bar, 2 revolute joints.

(2) Dampers - not shown, 2 spherical joints each, 4 total spherical joints.

Performing the same analysis on the entirely conventional and uncontroversial inboard suspension conveniently posted by Asphalt_Gundam above, I count the exact same BOM. Now I still need a pitch spring, which I have an idea for that would only require:

(1) 3rd element - not shown, 2 spherical joints.

So now I still have the exact same BOM as an entirely conventional and uncontroversial inboard suspension with 3rd element, and all of it even necessarily occurs twice on the car too. It's only 'adding' joints relative to traditional outboard suspensions, just as a traditional double wishbone adds joints compared to a strut, but is really exactly the same as other more 'advanced' suspension configurations.

Is it a lot of joints to keep in good working order? Sure is, but that's far from unique to this design. I'm sure you share similar concerns about stiction in nocones' LMP360, with the upcoming addition of sway bar and 3rd element to the inboard configuration. And yet, while I'm sure it has been tried, I've never previously seen or heard of anybody (engineer or otherwise) promoting hydraulic actuation over jointed rigid linkages for inboard suspensions.

In reply to Driven5 :

Yes. That'll be a mess to keep in good working order.

Citroen parts are readily available from European companies with about a 2-week lead time. They're cheap - ordinary passenger-car parts cheap. You can only use LHM in the system.

In reply to chaparral :

From what I'm finding, it looks like the Citroen system also requires the extra cost, complexity, and weight of the entire engine driven pump system and control/return circuits. Apparently they got around my concern with cylinder leakage by simply designing them to intentionally leak under pressure, and incorporating a recirculation circuit. Unfortunately, it seems that the only component actually useful in a simplified Moulton style system would be the accumulators... Which aren't much different than what is available from the domestic lowrider suppliers.

The number of joints in the above sketch could be reduced by switching from a lateral a-arm in the rear to a semi-trailing arm direct acting on the pull rod, but would that actually make it any better overall? Regardless of whether it's racecars or OEM's, time and time again the choice gravitate towards fewer joints when cost is the driving factor, and more joints when function is the driving factor. And considering spherical joints are also regularly used on OEM's and not just racecars, it's not like there is no such thing as a reasonable life spherical joint either. It should also help that none in the sketched system would ever see reverse loading either, so even if they did loosen up there still wouldn't necessarily be any 'slop' to speak of.

Hmm... If I turned the torsion bar vertical I might also be able to trade out some forward joints in exchange for some unsprung weight. Maybe I'll sketch that out too. But again, which is the better trade-off.

In reply to Driven5 :

What about this?   Seems like it functions as a Z bar with only 1 pivot and 4 rod ends.   Though the bar would have to be stiffer due to its length you may be able to make it larger diameter and lighter.

Yes I'm aware that care would have to be taken to keep the effective actuation rates at what you actually want them to be front to rear because of the different angles of the pushrods.  Though you may actually want a difference for weight distribution and you could probably angle the Torsion bar to make the pushrod the same angle.  

In reply to nocones :

You've got the right idea on tuning the front vs rear ratio to change the LLTD. But every time I've tried to think about that style of rod actuated, heavy (albeit carried mostly sprung), full-length torsion bars for a 2-seater, I get stuck on how to get it far enough inboard to clear the front steering lock yet still not need to pass through the squishy organic bits or drivetrain. I'm open to suggestions there though.

Usually this just leads me back down a path of the similar Packard style setup from the first page, which would probably require a more bespoke chassis design, but is certainly still one to consider.

How about something like this? The non-twisting tube would be in the same plane as the lower control arm. The push rods would also control the suspension springs somehow.

In reply to VolvoHeretic :

That's a creative concept! I really like that it includes damping directly into the springing, rather than relying on corner damping... Although some corner damping may still be helpful with minimal twist/warp mode resistance. 

It does have 2 main hurdles I can see though. One is that it requires enough additional clearance under the car to pass whatever diameter tube  is used... Which leads into the other issue being that on a ~8' tube it's very hard to reasonably get it to the point of being effectively rigid. When directly using the tube as the spring, like nocones drew, I recall an appropriate thick-wall 1.5" OD (for packaging) tube weighing on the order of 30-35 pounds.