quote Originally posted by Bloozberry: Somewhere back around page 12 or so I had originally considered using the lower hole in the knuckle for the upper control links, but it was impossible to make the roll center behave in that configuration. I played around a LOT trying to make it work exhausting a dozen or so ways to salvage the idea back then, but it just falls flat. I'm nearly certain that using the lower hole would mean major concessions in the kinematics.

Interesting. WAAAYYYY back in the day I developed a design for an SLA conversion for the '88 rear suspension.
I lowered the cradle 1" from stock, so the lateral links pivots were 1" higher at the knuckle than the body.
I put the outer pivot for the A-arm or lateral link between the knuckle and the brake rotor so that the setup would fit inside the stock 15" wheels.
I positioned the inner pivot just outside the frame rail and MORE than 1" below the outer pivot.
The design was extremely sensitive to the elevation of the inner pivot.
However, by playing with that one variable, I was able to develop a system with PERFECT camber behavior, which also kept the roll center within a 4x6" rectangle through 5 degrees of body roll.

However, the roll center was in the range of 4-6" off the ground (don't remember right off) static. If you're looking for a roll center at the ground surface, you won't find a combo that gives you good camber performance.

However, the higher roll center works well in a Fiero, because it carries its rear end weight higher in the body than it carries its front end weight, especially with a DOHC engine. As the body experiences lateral g's, having the roll moment arms more equal front to rear helps ensure that the rate at which contact pressure rises is closer to the same front and rear. This goes a long way toward making the car neutral and the suspension easier to tune.

Reproduced to aid visualization.

quote Originally posted by Bloozberry:

Also, the closer you can get your upper and lower links to the same length, the easier it will be to make the roll center location stable. It's not surprising that you have a hard time getting roll center stability, because your links are very different lengths. The upper and lower links in my design were much closer to the same length.

It also looks like your lower link inner pivots are higher than your lower link outer pivots... so you're "static ride height" as defined by suspension pickup locations is about 2" higher than mine was.
Of course, it's relatively easy for you to lower your inner pivots without lowering your car.

[This message has been edited by Will (edited 09-07-2013).]

quote Originally posted by kennn: Does your latest proposal shorten the wheel base, or are you relocating the forward trailing link pivot to the rear to compensate for switching the uprights?

I kept the center of the wheel in the same location (no wheel base change) and moved all of the inboard upper and lower link mounts to compensate for the asymmetry of the knuckle.

quote Originally posted by kennn: If I understand correctly, you wanted to maintain the stock length of the trailing links.

Not exactly... I bought HT Motorsports entire 6" rear track increase package (3" per side) which included the stretched lower lateral links (not using them), strut bottom adapters (not using them either), and trailing links that were longer and designed to compensate for the knuckle that was moved outboard, yet still pick up the stock inboard cradle mounting point. I'm planning to use these trailing links, not the stock ones. What I meant a few posts ago is that by swapping knuckles side to side, even though it would pull the trailing link further backwards, I wasn't planning to change the length of the HT trailing arms but rather compensate for the relocation by sliding the inboard mounting points back to meet up with the trailing arms.

quote Originally posted by Will: WAAAYYYY back in the day I developed a design for an SLA conversion for the '88 rear suspension...the design was extremely sensitive to the elevation of the inner pivot.

I found that the roll center movement is extremely sensitive to the elevation of the inner upper pivots.

quote Originally posted by Will: If you're looking for a roll center at the ground surface, you won't find a combo that gives you good camber performance...the higher roll center works well in a Fiero, because it carries its rear end weight higher in the body than it carries its front end weight, especially with a DOHC engine. As the body experiences lateral g's, having the roll moment arms more equal front to rear helps ensure that the rate at which contact pressure rises is closer to the same front and rear... I developed a design for an SLA conversion for the '88 rear suspension... the roll center was in the range of 4-6" off the ground (don't remember right off) static.

I wanted my roll axis sloped downwards towards the front for those very reasons, and achieved a front static roll center located 2.8" above the ground and the rear at 7.4".

quote Originally posted by Will: I was able to develop a system with PERFECT camber behavior, which also kept the roll center within a 4x6" rectangle through 5 degrees of body roll.

My current design gives a 2" suspension drop over stock, and gives a 0.7:1 degree camber angle to body roll ratio through 6 degrees of body roll, which is what I aimed for. My rear roll center moves within a 1" tall X 15" wide box through 6 degrees of body roll... a far cry better than the stock '88 configuration at the taller stock ride height which allows the rear roll center to migrate through an 18.5" tall X 82.5" wide box through the same range. That box became significantly larger with the stock configuration on 2" lowering springs.

quote Originally posted by Will: Also, the closer you can get your upper and lower links to the same length, the easier it will be to make the roll center location stable. It's not surprising that you have a hard time getting roll center stability, because your links are very different lengths.

Finding the sweet spot may be more difficult with larger differences in the link lengths, but now that I've found it, I believe my camber gain and roll center stability are pretty darn good. Did you ever make a thread showing your design? I can understand not wanting to divulge the details, but even a schematic would be interesting to look at.

[This message has been edited by Bloozberry (edited 09-07-2013).]

Well the results are in... thanks Zac!

For those that don't care to look at more graphs, but are interested in the impact of flipping the knuckles around just the same, the short and sweet of it is that the only parameter that changes is anti-squat. The static value at ride height doesn't change since I kept the same trailing link angle, but rather than anti-squat increasing as the suspension compresses, it decreases, probably as a result of the trailing link attachment point behind the centerline of the axle at the knuckle. It would be better if the anti-squat increased instead. The magnitude of the decrease is small enough that I am going to go ahead with swapping the knuckles side to side and carry on building up the suspension accordingly. For what it's worth, the stock '88 suspension also experiences a decrease in anti-squat in jounce as well.

On to the graphs for those that are more inclined to see the proof of the pudding. You'll note that the blue lines are stock Fiero performance, rust lines are my previous SLA design with the knuckles on the correct sides of the car, and the pink lines are the same SLA design with the knuckles swapped left to right (and minor modifications to account for the asymmetry of the knuckle). In most cases you can't even see the rust colored line because the pink line is directly on top of it, meaning there was no change between the original SLA design and the flipped knuckle design. Remember that these graphs represent the suspension's behaviour over a huge 6" total travel from full jounce to full rebound, and a crazy 12 degrees of body roll from full tilt left to full tilt right. I don't expect to allow that much movement but the large range shows the effectiveness of the design nevertheless.

I'll start with anti-squat since I've already summarized the results above.



You'll note in the graph above that both SLA designs appear to have significantly less than what the stock configuration has at ride height, but then I'm not entirely convinced that the Lotus software calculates anti-squat properly for a McPherson strut arrangement. Everything I've read says that in such a configuration, anti-squat is technically limited to a max of about 28% or so. Again, my understanding of how anti-squat is derived for different configurations is limited so I'll let others argue the point if they see fit. The important characteristic of this graph is that it shows by flipping the knuckle around, that anti-squat decreases with jounce. For best performance, it should increase or at least stay relatively constant, like the original SLA design did.

For all the remaining graphs there is virtually no difference between the flipped and non-flipped designs... but each offers better performance over stock.











And because the above graph isn't completely intuitive (at least for me) here's what the camber numbers mean at the full 6 degrees of body roll for the flipped knuckle design (the outside tire camber being much more important than the inside tire because of the weight transfer):



This gives me the green light to forge ahead with the flipped knuckle design, so once again, many thanks go to ccfiero350 for the brilliant idea, to Zac88GT for the use of his software, to Will for his sanity checks, and to everyone else for helping me brainstorm through this unexpected problem.

(Edit: reformat roll center to body roll graph)

[This message has been edited by Bloozberry (edited 09-07-2013).]

quote Originally posted by Bloozberry:

What's going on here? This doesn't make any sense. If your lateral and toe links are parallel and the same length, there shouldn't be any toe change with suspension travel.
However, the graph shows way more toe change with suspension travel than stock.

[This message has been edited by Will (edited 09-07-2013).]

I'm glad it worked out so well.



I had the same issue with with the 17x11 rims and the strut being very snuggly.



One thing to consider, give yourself some options and make the anti squat adjustable with multiple front tie in points. Sometimes little or no anti-squat works best for a given track.

------------------
yellow 88 GT, not stock
white 88 notchie, 4 banger