"Link-X" suspension 1

[Dave K]

I just heard about, and started looking into the "Link-X" suspension system. The concept is interesting, but I haven't had a chance to lay one out in the comptuter yet, to see how things work.

The setup requires crossing the A-arms, ie attaching the "upper" A-arm's inboard pickup points below the inboard pickup points of the "lower" A-arm. The concept is to work the cars natural roll against itself, to resist the roll, eliminating the need for an anti-roll bar.

I'm curious if anyone has looked into this type of setup, and any pro's, con's, or thoughts regarding it.

Some links:

The Link-X inventors webpage:

http://www.wageng.com/

Articles on it:

http://www.autofieldguide.com/articles/040203.html

http://www.autofieldguide.com/driven/0602dri05.html

http://www.caranddriver.com/article.asp?section_id=29&article_id=4816

-Dave

http://www.moslerauto.com

 

[WagEng]

I presented a SAE 'small engine' paper on our ATV work, but haven't presented anything to the big SAE Congress. I didn't remove the original thread. I guess it just went unused for too long. Mostly the same questions came up, though.

About a year ago we looked at Graviman's idea of a floating subframe to give us more rebound travel. It seemed too complicated and the need for further droop wasn't compelling enough for us to try it.

We're working with the Military on the next-gen Hummer type machine (all indep susp). The Hummer rolls over a lot more than people know. Also, the 50mm machine gun is hard to control because of body roll...even when the vehicle is sitting still! Our moment-inversion principle can keep the platform stable.

The ADAMS model predicts a higher rollover threshold as well because of lower cg displacement & lower rolling momentum (less stored energy because the sway bars are gone).

 

[ProEpro]

The humvee will be a chalenge.

good luck,



ProEpro

http://www.whitelightdesign.com

Pro/E FAQ

http://www.whitelightdesign.com/servicestips.htm

 

[GraviMan]

Been thinking about this a while now. More I think about it the more it makes sense. Basically it allows the vehicle mass centre to flow along the same axis over rough terrain. The only down side is the need to allow lateral movement of the wheels. The only ideal method is to have the steering linked to a mass constrained to the axle by a torsional compliance (along x in car line) - has to be the axle, not the vehicle.

Presumably this is what you mean by a load based toe changes? The system ends up looking like McPhearson strut, with an axle between, using the axle rotational inertia to control the wheel toe angle. The struts take the vehicle weight. Not sure how the steering input goes in yet, but I'll get there! Packaging wise this looks good too, since the wheel tops stay about the same position (in plane of mass center).

How big can these axles be made? I take it 20 tonne axle weight is out of the question! ;-)

Mart

 

[mloew]

Here are links to two SAE Technical Papers on the "link-X":

2004-01-1545 Link-X Suspension for Roll and Pitch Elimination
2003-32-0071 Link-X Stability System


Best regards,

Matthew Ian Loew
"Luck is the residue of design."
Branch Rickey

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[GraviMan]

Thanks for this - these are now ordered.

After further thought I am having doubts as to how well such a system might respond to the step input of say a curb strike. Steering would not allow the wheels to move sideways quickly enough. OK this is fairly extreme, but I know that active suspension and ground height roll centre will cope with this, and still eliminate roll. We are talkng about an off-road suspension system...

BTW is there any merit in designing a conventional beam axle to have bump steer? I'm thinking steering towards bump, to reduce centre of mass lateral movement. The trucks I'm involved with have very high cabs that would benefit greatly from reduced lateral movement (Articulated 6x6 tractor trailor combination), not to mention higher ground speed from reduced mass center disturbance.

Mart

 

[WagEng]

GraviMan,

The lateral motion that you're talking about is not near as large as you're envisioning because the instant center of rotation moves up with the tire. Load-based toe change in the case of my suspension means that upon an impact on the front of the tire, the bushings are designed to generate a tiny bit of toe-out so the tire steers its way over the face of the bump. It works beautifully!

I would think that a lot of the lateral movement you're getting in the tall cab is from the roll of the vehicle coupled with the large distance between the roll axis and the position of the driver.

Link-X will reduce that distance and generate less overall roll, so lateral movement of the driver himself should be lower...I think.

 

[GraviMan]

"The lateral motion that you're talking about is not near as large as you're envisioning because the instant center of rotation moves up with the tire..."

But roll centre is still above ground height!?!
I certainly agree with the idea of load based toe change. I'm trying something similar (albeit less sophisticated) with the truck system. This helps the truck to rotate about it's mass centre, rather than roll centre.

Trouble with bump steer avoidance is that it really needs to be linked to rate of change of wheel height. Load based toe change would approximate this due to the damper load, and I can certainly see how this would work with smaller bumps. When the bump contacts nearer tyre front, I can't see how steering would help. An example of this is curb strike - steering just ain't gonna help reduce the contact patch scrub.

"I would think that a lot of the lateral movement ... large distance between the roll axis and the position of the driver."

Yup!

"Link-X will reduce that distance and generate less overall roll, so lateral movement of the driver himself should be lower...I think."

Well, yes and no. Truck will need lower roll stiffness, true. Unless bump steer works perfectly, large lateral movement with high roll centre will cause high lateral accelerations. I can see how this sytem would cope very well with less rough ground, but in the extremes that out trucks have to deal with...

The final concern is cost. Don't forget the hidden costs of reliability, something I have to think about every day. There are hydraulic systems for active spring adjustment, which become more practical every day. I have read the paper sent to me, but it dealt more with the theory than application. Engineers love pictures of oily bits...

Mart

 

[GraviMan]

One last thought I forgot to mention: Jacking. Examples of this include: Triumph Herald; Volkswagen Beetle (earlier design, i think); Chevy Impala; and indeed any swing axle type of system. Basically car continues to roll, after a critical angle is reached (dynamically unstable). The net result is the the vehicle eventually stops rolling, once the roof has intervened. ;-)

I can understand the need to keep the driveline length constant, but having tyre roll centre outboard of centreline is asking for trouble. As a fan of the "handbrake technique", I would be worried about sudden and unforeseen roll above a certain cornering g. This is why I prefer, at least, the idea of a floating subframe (keeps contact patch distances across car the same).

It's good to think outside the box, as long as you see the big picture! Believe me I'm the first to get overenthusiastic. ;-)

Mart

 

[patprimmer]

GraviMan

I think the Chevy you mean is the Corvair.

You forgot early Porsche and some old Mercedes Benz and Auto Union

Regards
pat pprimmer@acay.com.au
eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[WagEng]

Ahhh the dreaded swing arm! Link-X has a high roll center like a swing arm, but the similarities end there.

We have less camber gain, ability to build-in a

The most important difference is that the Link-X roll center doesn't migrate. A swing arm's roll center moves up as the wheels droop - that is unstable.

 

[GraviMan]

OK. Good points well made. I have then, only two concerns:

1. Ride comfort (lateral acceleration) in extreme off-road conditions - bearing in mind my point about bumps contacting near tyre front.

2. Structural stability. The need to have overlapping links reduces the ability to reinforce against lateral loads - bearing in mind we see at least 20 tonnes per axle.
This usually ends up requiring, at least, expensive (in small volume) ADI castings.

Don't get me wrong, I really like your concept. I just remain to be convinced that it can handle extreme loads and terrain without becoming prohibitively expensive. Admitedly the size of the tyres we use will help to keep contact patch at centre. We have, at this stage, even had to reject hydraulic suspension on cost grounds - although the view is still that this offers the best dynamics...

Mart

 

[Dave K]

Excuse me if this is obvious, but...

I'm curious on locating the rollcenter of the Link-X setup. In a standard SLA setup the instant center is on the opposite side of the car centerline, so the 'instant center' to 'tire contact patch center' line crosses the car centerline. Thus diagraming the instant rollcenter is easy.

http://www.miracerros.com/mustang/t_roll_center.jpg

On the Link-X, the instant center looks to be between the wheel and the frame, where the link bars cross. Putting it on the same side of the car centerline as the wheel, thus the 'instant center' to 'tire contact patch center' line never crosses the car centerline...so how is the rollcenter height determined?

-Dave

http://www.moslerauto.com

"Everything should be designed as simple as possible, but not simplier"

 

[mloew]

Dave,

The geometric roll center determined by the intersection of lines through the each suspension's contact patch center and IC. This suspension, just as a swing arm suspension, indeed has the IC on the same side of the centerline as the CP as you noted, but this does not change the method to determine the roll center. Lines are infinite, line segments have a finite length.

Best regards,

Matthew Ian Loew
"Luck is the residue of design."
Branch Rickey

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[Dave K]

Ah, gotcha...it should have clicked to extend the line back the other way.

Thanks!

-Dave

http://www.moslerauto.com

"Everything should be designed as simple as possible, but not simplier"

 

[WagEng]

Our third patent issued a month ago. It applies to the concern GraviMan has about arm length. Note that there are some 'trade secrets' that work in conjunction with the new method of Link-X.

We use arms that are similar in length to a traditional SLA, but still exist on the same planes as would be the case with a 'long arm' Link-X setup. The IC is still in between the wheel and vehicle centerline so the rollcenter is still very high. The knuckles are very short, so unsprung weight is reduced.

The roll center can be designed to be absolutely motionless and as an auxilliary benefit total camber gain is reduced and droop camber gain is considerably less. High-droop, off-road setups are more reasonable in this configuration.

We have found that you still get about 85% of the roll reduction just from the high roll center without extending the arms all the way to get 'moment inversion' benefits. Crossing the arms is still better, but considering the cost/weight/footprint savings...

This is the setup we have on the supercar, and it works spectacularly well. The turn-in is almost magical.

t

 

[shanba]

As you raise the front RCH, aren't you impacting the effectiveness of shock/spring/ARB tuning to the overall roll moment distribution. Or are you pretty much accepting that change and using the rear only to balance the desired understeer. How effective is body roll being damped? Seems to me the front is now a go-kart (except over bumps) which is why the turn in is more responsive.

 

[shanba]

After thinking a bit more, raising the front RC only reduces the sprung mass roll moment, which reduces total roll angle. The turn in response may result from the front high roll axis causing the body to yaw into the corner when rolling, vs. conventionally it would yaw out of the turn (which I believe has been mentioned by someone else).

 

[WagEng]

The high front RC does contribute to the immediate turn-in. The tires leaning into the corner slightly also helps (camber thrust into the corner vs out of the corner on other setups).

A high front RC will typically make the car feel like it's rolling onto it's rear. We've compensated for that and achieve slight forward pitch with roll. This has been proven to inspire the most confidence in drivers.

 

[HendrikJan]

maybe I am completely wrong, but isn't it also true that a swaybar controls wheelloads in corners?
As one wheel travels in more than the other, the wheelload is compensated by the anti swaybar.
I can imagine the x-link setup has no compensation for wheelloads when cornering.
Offcourse the vehicle stays horizontal with the x-link layout, but what happens to the wheelloads?
Don't you need a swaybar for this?