Suspension Design - "Trailing Arm"

[Brandon T]

Hey all, Brandon here, I'll be the new suspension design lead for my school this coming year.
Right now our rear suspension has a swept back "trailing arm" geometry. We are discussing where to go for next year and are stuck on whether to redesign the system or refine the current one. The judges didn't seem to like the setup because of forces acting on the A-arms and putting so much stress on one of the bars, but on the other hand there are teams like Cal Poly Pomona that have a very similar setup and did very well in design at competition. Maybe it's just a matter of validating the test data to the judges.If anyone has suggestions it would be really helpful, thanks!

[Jonny Rochester]

Would you be able to post pictures of your current car with this design, or diagrams?

And can you identify which bar was in theory subject to high stress? I can't comment till I see the design.

[SomeOldGuy]

Looking at pictures of the cal poly car I do not see trailing arms. Are you referring to swept a arms?

[Marshall Grice]

I suspect the OP might be trying to say that the outer ball joint is rearward of both inner ball joints.

[Brandon T]

Swept A-arms is what I mean, apologies for the confusion. Here is a photo of the current setup, with the high stress members in red that the judge pointed out.

[Tim.Wright]

If you can manage any compliance issues (and more importantly be able to demonstrate that you have!) then there is no real reason to change it. The main challenge will be to minimise the amount of toe out you see on the outside wheel in a cornering condition which will be adding a destabilising effect right when you don't want it.

Yes swept back arms will need to be heavier than in a traditional setup to maintain the same stiffness and mitigate this toe compliance, but a traditional (unswept) setup would require your chassis to be extended back another 300mm or so to clear the chain and this will be much heavier than a set of beefed up arms.

The judges can't (or at least shouldn't) burn this design without understanding the reasons that you did it and I guess this was missing last time your presented it. If you can show them your toe compliance curves and prove that this solution is lighter than extending the chassis back then the design is sound.

Don't fall into the trap of burning a design because a judge doesn't like it. Depending on your targets, anything could be possible. Oversized rod ends in bending can be justified if your target is ease of adjustability. Single shear upright mounts can be justified if your target is reduced part count and reduced assembly time. etc etc etc...

T

[Brandon T]

Thank you for the strong points Tim!
On your point of the toe-out issue, is there a direction we can head in to mitigate the problem? Right now the toe link is ungrounded, on the rear of the upright. Would a repositioning of the link and mount help, maybe having it forward of the upright to cause toe-in during bump and counter the toe-out tendencies? We haven't been sure how to go about creating camber curves and toe travel curves to show.

[Yessamgerg]

In regard to you question about the toe-out tendency:

Due to the overturning moment during cornering, the upper control arm will be in tension and the lower in compression (draw a FBD if you can't picture it). With this the upper ball joint will deflect outwards and the lower ball joint inward. As the you appear to have low amounts of KPI & Castor offset, and an OK toe base (coulc be bigger) lets say the track-rod is under relatively little load (just the self aligning moment). Because of this when the UBJ moves out and the LBJ in, the wheels gains toe out. In RCVD (don't have my copy on hand, I can look up the page later if you want) there is a picture showing "ideal" track-rod locations. As you want toe out in the front (understeering trend) and toe-in on the rear (again understeering) this leads to the following "ideal" locations

Front axle:
Lower Leading or Upper Trailing

Rear axle:
Lower Trailing or Upper Leading

[Sky Squillace]

Did they say anything about RIEB (rod end in bending)? If they catch you having RIEB's then the rest of your design presentation will be rough.

[Z]

Brandon,

I can see a lot of things that can be improved on your suspension, but ditching your "swept back" design is NOT one of them. IMO the idea of having minimal structure behind the rear-axle line is a worthy goal.

Tim and Greg have given you some good advice. From a quick look at your picture, here are some more suggestions (in no particular order).

1. The lower wishbone carries greater LATERAL loads than the upper one (approximately 2 x). Therefore, IMO, it makes structural sense to have two lateral members down low, and only one up high. This also lowers the CG. So put the toe-link down low, and as rearward as possible, as suggested by Greg. Also move the bottom-wishbone-outer-BJ further forward for a wider "toe-base". Make this "toe-base" as wide as possible, at least 15+ cm (6+").

2. The above is probably easiest done by turning your current uprights upside-down (and swap left-to-right if you want calipers in same place). It looks like the uprights have bolt-on "horizontal-bracket+spacers" for camber adjustment. If so, then these are now at the bottom. And you do not need the "REIB" for camber adjustment, so better would be a "spherical" at the end of the wishbone. The toe-adjustment can stay as is (but low down).

3. Depending on your various chassis node positions, you might be able to move your lower-wishbone-to-chassis-front-BJ further forward. This makes the wishbone's forward member more longitudinal, and the wishbone "base" wider, for lower stresses. A good place for this BJ is the MRH, which must be there by the Rules, and is quite strong (~2.5 mm wall)

4. You have a Push-Rod-&-Rocker taking the vertical loads to a heavy Spring-Damper ... that is mounted as HIGH as is physically possible! Aaaaarrghhhh!!! :( Yes, I have ranted at length about this, but WHY??? A Direct-Acting-S-D, approximately following the line of your pushrod, would be so much better (ie. stiffer, lighter, lower CG, less-friction, cheaper, quicker-build, +++). The bottom-end of this DASD can act directly on the upright (rather than on the lower wishbone), maybe via a BJ on that bolt-on horizontal-bracket.
~o0o~

If you really feel that you have to impress the Design Judges, then:

1. Draw up lots of "toe and camber kinematic curves".

These are easy to do. Support chassis on some blocks of wood, remove SDs, and then progressively lift the wheel through its suspension-travel range about 1 cm at a time (maybe with 1cm thick pieces of plywood/steel under the wheel) while measuring toe and camber. Regardless of impressing DJs, you really should have a system in place for quickly measuring these. I suggest a horizontal string-line down the side of the car for toe (accuracy ~+/-0.5 mm at the rim), and a vertical bubble-level for camber (here accuracy can be much less than with toe, but both easily done with a steel-rule and eyeball).

2. Draw up lots of "toe and camber compliance curves".

Only a little harder. Chassis again up on blocks, and "dummy-SDs" fitted (ie. short tube replacing SD) that fix the wheels at "ride-height". Buy some "ratcheting tie-down straps" (ie. they look like seat-belt webbing...) and a heavy-duty "spring-scale" (ie. like those for weighing fish, but say 0-~300kg). Wrap the straps around the bottom of L and R wheels, with the straps acting at the "wheelprints/contact-patches", and pull the two wheelprints together, with the spring-scale somewhere in the middle. Measure the inward Fy loads at the wheelprints (ie. via the spring-scale) and the resulting toe and camber deflections. Repeat for every ~500 N from 0-~3 kN. Or until something breaks (oops!).

You can also measure compliances due to longitudinal Fx forces at the wheelprints. For this you have to lock the wheels to prevent them rotating, and pull the front and rear wheelprints on one side of the car together.

Once you have all the (quite cheap) equipment required for above, the actual measurements should only take a few hours to do. I suggest that all "newbie" team members be given the task of drawing up such K&C curves for all the past year's cars you have in the shop. This would be a good start to the "knowledge transfer" process.

Z

[Jay Lawrence]

I second what Tim says about not changing designs just because some judge doesn't like it. There's a good chance another one will, and an even better chance that they all will if you can justify whatever you've done (and unfortunately a chance that they are not interested in any kind of logic, but you can't do much about that).

On another note, it looks like you are going to place some bending loads on the upper outer rod end. If you have shim (or similar) adjustment already on the upper end of the upright then you shouldn't need a rod end there (and would be better off moving the connection point of the upper outer swept arm outwards as close as possible to the upright and replacing the rod end with a spherical). If that makes sense...

[Brandon T]

1. The lower wishbone carries greater LATERAL loads than the upper one (approximately 2 x). Therefore, IMO, it makes structural sense to have two lateral members down low, and only one up high. This also lowers the CG. So put the toe-link down low, and as rearward as possible, as suggested by Greg. Also move the bottom-wishbone-outer-BJ further forward for a wider "toe-base". Make this "toe-base" as wide as possible, at least 15+ cm (6+").



Z, thank you so much for the extended advice. A note on your suggestion about the "toe-base", I'm not sure which BJ you mean. Are you talking about the bottom-wishbone-to-frame-forward mount? Also, we were thinking of shortening the wheelbase by moving the rear wheels forward, and possibly making the rearward members of the A-arms perpendicular to the wheels, i.e. creating right triangle shapes.

[Brandon T]

Jay, thank you for the suggestions. I'm not clear on your description though, do you mean where the "point" of the A-arm triangle connects to the top of the upright? I.E. to the black aluminum mount bolted to the gold upright?

[Tim.Wright]

Thank you for the strong points Tim!
On your point of the toe-out issue, is there a direction we can head in to mitigate the problem? Right now the toe link is ungrounded, on the rear of the upright. Would a repositioning of the link and mount help, maybe having it forward of the upright to cause toe-in during bump and counter the toe-out tendencies? We haven't been sure how to go about creating camber curves and toe travel curves to show.

Yes one possible approach would be to design in some bumpsteer to counteract the compliance steer. Its not a perfect solution (because there is a phase delay between bumpsteer and lateral compliance steer) but its an easy one to do.

Be careful though: I'd guess that putting the toe link as a seperate member in front of the wheel centre might result in it be in very lightly loaded in comparison to the a-arms. Why would this be a problem? Well if you consider the outside wheel, the A-arms will deflect under load towards the centre of the car while the toe link MAY remain more of less its same length due to the low forces. In this case its easy to imagine the result will be EVEN MORE compliance toe out than what you have now and will need to be further compensated for by the bumpsteer.

Its a tricky one to address and it highlights the importance of knowing what your link forces are because it gives you a good idea on what the compliance might be doing without having to do any FEA.

If the relocation of the toe link doesnt do what you want, you can just make the suspension as stiff as you can and then rebalance the car using the roll distribution. Another trick could be to lower the rear roll centre to reduce the forces seen by the links.

[Yessamgerg]

As for what Jay is means is that your trailing upper tube is cantilevered by ~1.5" from where the leading arm connects to it. If you are going to run REIB (which you shouldn't), you should at least connect the leading arm directly next to were the rod end attaches.

The "Toe-Base" the distance that the track-rod is from the kingpin (in your setup effectively the distance from the UBJ to the TR Outer BJ. As the tire generates lateral force it also creates a "self-aligning moment" which tries to straighten the wheel (toe out on the outside wheel, toe in on the outside wheel) By increasing the toe base, for a given moment the force in the track rod is reduced. Additionally if you have some play in any of your linkages (from ball joints, poorly toleranced holes etc.) a larger toe base will make a given slop distance a smaller angular effect.

It looks like you are on 13" wheels (correct me if i'm wrong, its hard to tell from the pic) For 13" wheels a reasonable (Z's 6+" would be great but a bit generous) toe base should be ~4.5+ inches. On 10" wheels that drops to probably 3.5+ inches.

I agree with Z that the low mount TR is probably your best bet, another thing to try would be to move you LBJ forward. This allows a larger toe base, would eliminate the some of the "swept" arm inefficiency (for a given set of inboard pickups) and as its the rear your castor doesn't really matter.

A good example of the entire setup we are describing would be the picture below. I can post pictures of our new rear suspension in a few days if people want, it basically emulates everything talked about here.

nitronracing . files . wordpress . com / 2010 / 06 / img_4762sm1 . jpg

[Yessamgerg]

Also what is your current weight dist. and wheel base? As you said, a good option would be to reduce the wheelbase by bringing the rear wheels forward assuming the weight distribution allows. Some others may want to chime in here but most good teams are <64" wheelbase and unless you have reasons to do otherwise (Terp's massive undertray & 93" wheelbase) the short wheelbase is generally good for performance.

[Brandon T]

I now see what was meant about the leading and trailing members meeting each other at the top.
Also the toe base description is now clear, thanks! Both should be relatively easy adjustments to design in. And yes, 13" wheels.
Right now we're at roughly 45F/55R, coming in at 509 pounds. Wheelbase is 66", so it should be favorable to shorten this a bit, along with our frame to help with some weight.

[Tim.Wright]

Here's an interesting design:

http://www.tobymoody.co.uk/EMPIREWEBSITE/empireracingcars/Gallery.html#73

On first inspection it appears to be breaking all the sacred design rules. But I can see some thought has gone into the loadpaths. There is a significant "caster trail" in this design which allows the toe link to be loaded in compression for the outside wheel. This will give a toe in component to the overall toe compliance. Of couse the total toe compliance will depend also on the relative stiffness of the control arms too but loading the toe link in this way will at least give you a fighting chance of avoiding a toe-out response.

This is exactly what I was talking about in my previous post.

[Yessamgerg]

As promised a few days ago, here is the new rear suspension I am working on for our 2015 car. Its actually very similar to what Tim just posted (other than that REIB). Like he sad lots of castor trail (which isn't really a bad thing on the rear) and some tricks with the control arms to ensure toe-in under cornering loads. The uprights also happen to be symmetric which will be great for spares and machining time.

279

[Claude Rouelle]

Yessamgerg,

"....lots of castor trail (which isn't really a bad thing on the rear"

Why do you assume this is "not really a bad thing on the rear"? Because there is no rear steering? What about compliance? And what about rear KPI trail (also called scrub radius)?

[Tim.Wright]

As promised a few days ago, here is the new rear suspension I am working on for our 2015 car. Its actually very similar to what Tim just posted (other than that REIB). Like he sad lots of castor trail (which isn't really a bad thing on the rear) and some tricks with the control arms to ensure toe-in under cornering loads. The uprights also happen to be symmetric which will be great for spares and machining time.

279

Which way is the front of the car? To the left? Its always a good practice to present any design images with an axis system showing left/fwd/up. If the first thing your boss has to do when you present a design is figure out which way is up, your aren't off to a good start.

Have you done anything to validate your toe-in compliance? Because it looks to me that you have more stiffness in front of the wheel centre rather than behind it which will give you toe-out (assuming the driving direction is to the left).

Generally you want a negative "caster trail" on the rear axle because it implies that your "compliance centre" is behind your contact patch so you "should" get a toe in response from a lateral force.

Why all the "quotes"? Because the geometry is only half the story. The relative stiffness of your toe link compared to your control arms is the other half. You can get a toe in response with a positive, negative or zero caster trail so these parameters don't tell the full story.

I'm interested to hear your explanation.