PDA

View Full Version : Separating ride from roll...?



exwestracer
05-29-2010, 08:23 PM
I've been keeping this idea under my hat for a couple of years now, but things change; and most of my fellow competitors don't follow this site...so here's something to ponder, and give feedback on. We've all (well, a lot of us anyway) spent a lot of money on sway bars and trick digressive shocks which try to control roll forces differently than ride (bump) forces. I think there may be a better way, at least for solid axle systems.

I've been intrigued by the Mumford lateral locating linkage ever since I learned about it; and having followed F1 suspension design through the 80s, I noticed the similarity to a "pull-rod" inboard suspension design.
https://static1.pt-content.com/images/pt/2010/05/ME_MUMFORD_3COLJPG-1.jpg
One of the benefits of the Mumford design is the free rotation of the axle about the roll center (the linkage only operates in bump and droop). So what if...the mumford linkage was used to operate a monoshock coilover to hold the weight of the vehicle? The spring and shock would operate in bump, but offer NO resistance to roll forces.

So what to do about roll force? If we consider the physics of roll, what we have is a torque force applied by the vehicle center of gravity about the roll center. Torque is expressed as lb-ft., or pounds of force with a 1ft lever arm. So, if a spring and shock are placed 12" above the roll center to resist roll torque, the units cancel out and what is left is straight pounds of side force. Mount the coilover horizontally, and it has very little, if any, effect on bump movement. So what comes of this is a way to control roll force and damping totally separately form bump force. No more sway bars and 4 way adjustable shocks unless we want them.

51201

ace_xp2
05-30-2010, 11:38 AM
Cool idea, but I can't imagine what you'd use it for without serious downforce. After all, we typically need more roll force resistance in the form of ARBs, which points to our spring rate being compliantly low as is, just high enough to constrain pitching forces.
So I can't see dropping center spring rate unless we add a lot of anti geometry to control pitch and heave, which isn't really the best addition to be making (though not always a deal breaker.).

But you could do that with a normal system, and the lower the spring rate, the less impact it has on roll anyways.

The mumford link arrangement is a cool idea, though you could just mount the spring on the back of the axle at center and get the same effect, the packaging of both axle locator and center spring arrangement is a cool idea. more compact than trying to package a watts or a panhard even further back from the pumpkin.

And I can't really see the purpose of lateral springs versus two weaker than the center spring in the typical location. Springs can't work until there is motion and since the axle is located through the linkages you'll still operate those springs in single wheel bump( the vehicle still experiences "roll in relation to the axle"), which is a good thing as otherwise single wheel motion would be undamped.

exwestracer
06-01-2010, 09:45 AM
While it's true the Mumford design offers a VERY low roll center (if desired), the point isn't to allow the chassis more freedom to roll. The lower roll center causes less inside wheel jacking effect, no matter what the level of roll resistance. With a typical spring/shock system however, the rates needed to control roll with a very low R/C are much too stiff in bump, so we run softer springs and add a bar to keep the chassis upright under roll forces. The bar works the same way no matter what type of suspension it's used on; it controls chassis roll by forcing both sides of the suspension to try to do the same thing, in essence converting any setup into a swing arm. The lower the roll center, the stiffer the bar needed to keep the chassis upright, so the more inside jacking we get and the more compliance we give up. The standard solution is to keep raising the rear roll center to keep the car upright without stiff springs or a giant bar. With a sold axle suspension, we always end up in roughly the same place, no matter what route we take to get there...

The purpose of the modifications is to separate control of roll forces and bump forces as completely as possible. Any system with (typical) outboard vertical springs/shocks requires that the spring and shock act as both bump resistance and roll resistance, therefore we are always going to have a compromise in spring and shock selection between what is needed for bump forces and what is needed for roll forces.

Now, there is obviously shock technology out there which allows pretty good results, and as mentioned we can artificially move the roll center with bars; but we are sacrificing traction and bump compliance in doing so.

The idea behind the horizontal coilover for roll control is that it has little or no effect in true bump. The ride coilover can therefore be tuned to resist bump and pitch forces, without affecting roll resistance. In true roll however(lateral g on smooth surface), ONLY the horizontal roll coilover is active, so it can be tuned to deal strictly with the moment arm of the suspension. A single wheel bump (i.e. running over a kerb on the track) is really a roll force imparted by the axle instead of the chassis. With typical stiff springs/shocks and big bars, this type of force is usually over-damped, as evidenced by the way most race cars are thrown into the air when they run across the kerb. By separating the springing and damping of roll forces, we can now eliminate the bar and use the available shock technology to add a whole new level of control never before possible.

Jaybob
09-25-2011, 12:27 PM
Hi Ray,
I'm new to this site and actually found it because I was searching for info as I design the chassis and suspension of my project vehicle. I am intrigued by the concept you have posted here. I know that this is an old post and I am wondering if you finished and tested your design. I am currently modeling my design in 3D, (Solidworks to be more specific). My project is not a Pro-Touring car, but more along the lines of Pro-Street. I'm intrigued because the packaging of the suspension could possibly work for me. I have a very narrow rear frame set 25" outside to outside on the rear frame and I currently have a parallel 4 link with a watts setup and 2 coil-overs in a vertical arrangement. Feel free to contact me directly if you like. Thanks! -Jay

Bjkadron
09-26-2011, 02:58 PM
Wow, How did I miss this thread the first time around?

Norm Peterson
09-27-2011, 03:25 AM
You weren't the only one that missed it, Ben.


Norm

Jaybob
09-27-2011, 05:02 PM
Here are a couple images of the rear suspension I'm designing... The diff housing is also designed by me... Work in progress... - Jay
49568
49569

Norm Peterson
09-28-2011, 06:13 AM
I think I know what's going on with it, and I'm going to guess that you wouldn't be placing the ride springs and their dampers in the usual places. Maybe closer to what jaybob's pictures show . . .


Norm

exwestracer
10-04-2011, 07:36 AM
Hi Ray,
I'm new to this site and actually found it because I was searching for info as I design the chassis and suspension of my project vehicle. I am intrigued by the concept you have posted here. I know that this is an old post and I am wondering if you finished and tested your design. I am currently modeling my design in 3D, (Solidworks to be more specific). My project is not a Pro-Touring car, but more along the lines of Pro-Street. I'm intrigued because the packaging of the suspension could possibly work for me. I have a very narrow rear frame set 25" outside to outside on the rear frame and I currently have a parallel 4 link with a watts setup and 2 coil-overs in a vertical arrangement. Feel free to contact me directly if you like. Thanks! -Jay

Jay,
Thanks for the interest. This project is stalled for a couple of reasons... Due to packaging constraints in this chassis, I am using a through rod type spring/shock assembly (not exactly a coil over). I had been working with the lead development engineer at Penske, but a lot of the info is proprietary to MotoGP and F1 teams. I think I found out some stuff i shouldn't have because he isn't returning my emails anymore...:ssst:

I also purchased another supermodified which was supposed to be a bolt it together and run it deal. Turned out to be a junk pile, but I'm in it now and need to get it finished.

https://static1.pt-content.com/images/noimg.gif
We will get the new car (on the bench at rear of this photo) done next year, and I'm confident that the suspension will work the way it should.

As to your question, there really isn't a lot of benefit in your application. Unless you are really running 10/10ths through the corners, your pro-link with watts setup should be just fine. You would have to reverse the Mumford link (pivots attached to housing) with that narrow a chassis. The unsprung weight goes up with no real performance advantage.

exwestracer
10-04-2011, 07:46 AM
I think I know what's going on with it, and I'm going to guess that you wouldn't be placing the ride springs and their dampers in the usual places. Maybe closer to what jaybob's pictures show . . .


Norm

Norm,
There's only one ride spring and damper... Think further outside the box...

Cobra 498
10-05-2011, 11:58 AM
Norm,
There's only one ride spring and damper... Think further outside the box...

1992 VanDieman Formula Ford?

exwestracer
10-06-2011, 01:59 AM
1992 VanDieman Formula Ford?

No. LOL. There are a number of pull/push rod monoshock IFS systems out there. None that I know of has any sort of damping associated with the roll motion (very limited travel). Most use some sort of compression washer stack to allow a slight amount of roll movement.

This system I've outlined above allows full freedom of motion in roll, and the opportunity to separately dampen roll spring motions. And it's used with a solid axle.

Norm Peterson
10-06-2011, 07:24 AM
What I'm thinking is . . . well, I wonder how many circle track racers would sort of recognize the components but wonder why they were turned sideways in the car.


Norm

exwestracer
10-06-2011, 08:06 AM
What I'm thinking is . . . well, I wonder how many circle track racers would sort of recognize the components but wonder why they were turned sideways in the car.


Norm

We're gonna find out next spring...:fingersx:

Jaybob
10-06-2011, 01:28 PM
Jay,
Thanks for the interest. This project is stalled for a couple of reasons...

Hi Ray,
Thanks for the reply... I tend to agree with your comments regarding my current project... However, now I'm already thinking forward to my next project which is planned as a pro-touring car. I'm thinking a 1967/68 Ford Falcon slung low with a coyote small block maybe... Anyway, I digress...

Here's my thoughts... As I model these various live axle rear suspension centering devices, (watts, mumford, satchell), in 3D and apply side loads, (as in cornering loads), I find that all three of these have a similar problem with respect to "jacking", (maybe not the correct term). What I see when I allow the 3D models to articulate is that it wants to "push" the body/chassis to roll to the outside, (in a counter productive direction). So then I add in some loading to represent "anti-roll" and I see that the side, (cornering), loading is, (in effect), trying to "lift" the outside tire from the ground. Am I missing something here?
Thanks! -Jay

exwestracer
10-06-2011, 05:03 PM
Jay,
I don't think your model is accurately representing the loads placed on the system... As long as the C/G is above the roll center, you are going to have increased loading on the outside tire. The anti-roll bar does not change the load going onto the outside corner. What it does do is to keep the car flatter by unloading the inside tire (picking it up off the ground in extreme cases). Normally the weight on the inside corner exceeds the lift generated by the bar, so the car "falls" back toward a level state. This isn't the greatest thing for traction, obviously...but that's a topic for another thread (coming soon).

You are correct that a lower roll center (as offered by the Mumford linkage I am using) does cause more side shift of the chassis over the tires, BUT the low roll center also reduces the amount of lift on the inside, which improves total traction. It's common knowledge that a larger front bar will increase understeer, so the theory is backed up in practical application. We don't like to feel like we're going to roll the car in hard cornering; so the bar does tend to improve driver confidence, which can easily reduce laptimes. I'm still employing plenty of anti-roll force...just in a way that is much more controllable and doesn't reduce inside tire load nearly as much as a fat sway bar...

Jaybob
10-07-2011, 01:27 PM
Hi Ray,
I went back through my loading and I cannot believe that I completely forgot to put the respective vehicle mass on the tires... Makes a hell of a difference... Okay, for this project I'll stick with the watts setup... I've got some ideas on a variation of a double A-arm front suspension that allows for VERY quick & easy camber & caster adjustment... Give me a few days and I'll post some images...
-Jay

exwestracer
11-10-2011, 07:13 AM
Just a quick update on this... We have the actual components fabricated and installed on the car. As mentioned above, the rest of the car isn't finshed, so it hasn't been track tested yet. What I did find from cycling the suspension on the bench is that there is very limited motion of the roll shock in the expected range of travel. Like less than 1/2"... Keep in mind that I purposely set up the roll shock 12" above static roll center so lb/ft of roll torque becomes straight pounds of side force. I think damping and spring rate are going to be pretty ridiculous, so I'm considering using some sort of simple bellcrank to increase the motion ratio between the axle and the shock/springs. This presents some packaging issues in the supermodified chassis, and I really haven't got much further than scratching my head over that.

Still no response from Penske, but I think we are just going to have to soldier on....

Jaybob
11-10-2011, 06:06 PM
Hi Ray,

Thanks for the update... I have been laxxed in working on the front end design... I just keep thinking that maybe I can eliminate the Watts setup in lieu of a Satchell, (triangulated), 4 link setup. I just can't make myself comfortable with angled links and such a narrow frame spacing... Even with short upper links, (which I'm not fond of)... Anyway, can I talk you into posting a couple more pics of your Mumford setup? Maybe I can help with your Bell-Crank/dampening concepts...

Cheers,

exwestracer
11-11-2011, 06:48 AM
Jaybob,
You said the front end design??? I'm guessing rear, and you are correct, that frame width is pretty marginal for a triangulated setup. Personally, I use a minimum of 60 deg included angle between the upper links for a street car. Yes, you will see kits out there that run less...

If you stick with 60deg, I think you will find the effective length of the upper links will be shorter than optimum. On a big tire, narrow chassis car, I would avoid the true Satchell design that triangulates the lower links instead.

exwestracer
11-12-2011, 04:59 PM
Jaybob (and everbody else:)...) here are some pictures to mull over. This is where the car stands now.

51199
Above shot from the front of the car showing the basic Mumford setup with cross link and 2:1 crank arm for the ride shock. At upper left is the "top" mount for the coilover.
51198
LF corner of car showing pull rod and bell crank for ride shock. Behind the aluminum cap on the pivot is a needle and thrust bearing arrangement. In the background is the engine mount plate. Notice how the Chevy bolt pattern is rolled 30deg to the left.
51197
RF corner. Back side of pull rod spud on axle is threaded for the roll link. End of roll shock shaft is just visible under upper frame rail
51200
From right side of car. Roll shock and adjustable mount at top of picture.
51196
Closer view of double end roll shock and mount. The shock will be held in place by a spring perch adjuster nut on either side of mount. Moving the shock side to side will change cross weight or "wedge" on front axle. The small spring on the far end of shaft represents how the roll springs will be mounted on both ends. The spring force will balance against the shock body to hold the car upright at static ride height. Keep in mind the car will have about 66% of the total weight on the left side tires.

arue333
06-27-2012, 09:23 AM
So any news? Updates?

exwestracer
07-05-2012, 04:37 PM
So any news? Updates?

Bought another car that was supposed to be "bolt together and go". It turned out to be a parts pile, but I'd spent the money, so now I'm finishing it first.

Short answer, no.

BrianP
10-23-2012, 04:45 PM
Why wouldn't adding a coil over setup just shove the axle to the side until the other linkage is at the limit of it's travel?

cornfedbill
10-23-2012, 05:54 PM
Ray,

I did some sketching of the Mumford link suspension. If I am not mistaken, mounting the pivots to the frame, as in your photos, causes the roll center to move towards the outside tire under roll. Isn't it more desirable for the roll center to move inboard during roll?

If the pivots were located on the axle (not desirable for unsprung weight) the roll center could be made to move inboard on a corner and thus cause the body of the car to begin rolling downward, reducing the center of gravity (slightly) and making the car more stable.

We did this on a student design project years ago. It was for the SAE Mini-Baja. The front and rear roll centers were very low and moved inboard significantly during roll (which was much more than a street car). The body would actually begin to squat under hard cornering. It was nearly impossible to roll even with the relatively high roll center and modest track width.

What do you think?

exwestracer
10-24-2012, 05:07 AM
Ray,

I did some sketching of the Mumford link suspension. If I am not mistaken, mounting the pivots to the frame, as in your photos, causes the roll center to move towards the outside tire under roll. Isn't it more desirable for the roll center to move inboard during roll?

If the pivots were located on the axle (not desirable for unsprung weight) the roll center could be made to move inboard on a corner and thus cause the body of the car to begin rolling downward, reducing the center of gravity (slightly) and making the car more stable.

We did this on a student design project years ago. It was for the SAE Mini-Baja. The front and rear roll centers were very low and moved inboard significantly during roll (which was much more than a street car). The body would actually begin to squat under hard cornering. It was nearly impossible to roll even with the relatively high roll center and modest track width.

What do you think?

Bill, your theory is correct, however there isn't enough roll motion in these cars for it to be a concern.

I did Mini-Baja for CSU Fresno for 3 years back in the late 80s. Formula SAE as well, although I never made it to a competition. Great training!

cornfedbill
10-24-2012, 07:26 AM
Bill, your theory is correct, however there isn't enough roll motion in these cars for it to be a concern.

I did Mini-Baja for CSU Fresno for 3 years back in the late 80s. Formula SAE as well, although I never made it to a competition. Great training!

Yes, you are probably correct. It is fascinating to see how effective moving the roll center towards the inside tire can be at stabilizing the car in hard cornering.

It is amazing to see how many fast cars break all of the theoretical "rules".

I do wonder if idealized roll center movement would be effective in a multi-purpose PT car. Maybe roll stiffness could be reduced to give a more compliant ride without a measurable sacrifice in cornering. It may even be effective when cornering on an uneven surface to prevent the tires from bouncing, thus keeping more rubber on the road.

Just thinking out loud.

exwestracer
10-24-2012, 10:28 AM
Typically, we'd like the roll center migration to be as small as possible. Again, it's one of those things that would be different for each corner we went around, so predicting where the R/C would be and how it would affect cornering becomes mind-boggling.

I've had thoughts about the effectivness of the "modified" Mumford on the rear of a P.T. type car. It would need a LOT of bar, as the R/C is MUCH lower than any other type of lateral locator (except a WOB link) would give us.

I guess in my mind it's not really "necessary"...

mitch_04
01-30-2013, 10:10 AM
Awesome thread, thanks for sending me to it. I'll reply with more in a couple weeks when I understand what's going on!

AMC Racer
01-31-2013, 10:11 AM
Sorry if this is a bit off topic, but with the Mumford link, if you put the chassis mounts and axle mounts higher, can you get a roll centre height more typical for a PT car? Should give better ground clearance and maybe require less substantial bracing than a typical watts?

Thanks.

exwestracer
02-03-2013, 06:53 AM
Sorry if this is a bit off topic, but with the Mumford link, if you put the chassis mounts and axle mounts higher, can you get a roll centre height more typical for a PT car? Should give better ground clearance and maybe require less substantial bracing than a typical watts?

Thanks.

You can change the R/C height by raising (or lowering) the chassis mounts. This involves taking the included angle out of the drop links. I haven't mapped this out, but I believe the Mumford wants a certain minimum amount of angle in the links. If the links become too close to horizontal, they have to "stretch" to allow any body roll. As I said, I'm not sure where that definite point is.

The mounting can also be reversed (as can a Watts link) to put the pivots on the axle housing. With a Watts, one mounting gives us a R/C height fixed to the chassis, the other gives us a R/C fixed to the ground. With the Mumford, the R/C is still moving with the suspension, but the direction of motion is reversed. With the pivots mounted to the chassis, the R/C drops with suspension compression. With the pivots on the AXLE, the R/C raises with compression.