22 replies to this topic

[Greg Locock]

One reason I like a Watt's link on a Hotchkiss is that at least you know where something that you might like to think of as the roll centre is.

That's about the end of the good news. Here's some runs I did messing about with the height of a Watt's link on the back of a truck. Some of the results are obvious... some less so.

http://members.optus.../rollcentre.zip

or failing that in the gallery at http://www.geocities.com/greglocock

You'll need a ps viewer eg ghostview to look at them.

[McGuire]

Originally posted by Greg Locock
One reason I like a Watt's link on a Hotchkiss is that at least you know where something that you might like to think of as the roll centre is.

LOL well put.

[LS 1]

Originally posted by Greg Locock
One reason I like a Watt's link on a Hotchkiss is that at least you know where something that you might like to think of as the roll centre is.

Without the link, isn't the roll centre of a Hotchkiss the intesection between the axle line and the horizontal centerline of the car? So being able to easily locate the roll center isn't an advantage of the link? It's just trading one kind of simplicity for another, highly related kind?

If I got this right, this will be the only time I ever have and ever will "catch" you on a suspension issue. Well, sort of. I'm so proud.

[McGuire]

Sorry, we don't believe in geometric roll centers around here.

[Greg Locock]

I didn't believe in them, and regularly get confused by them, but would have to say that the GRC gave me a very quick handle on things when we were developing a suspension recently. (That is, the handling changes we saw seemed to follow changes in GRC).

But at the moment my favourite definition of a roll centre height is : track*change in the vertical force at a contact patch/change in lateral force at that patch.

It is simple to measure, robust and fundamental. It bears a very strong relationship to instantaneous centres of rotation, so in some ways it is connected back to the GRC.

But the results you get are spectacularly different to the GRC, particularly as the suspension jounces and rolls.

LS 1 - I'm not ignoring you. You raise an interesting point. I think we'd agree the RCH of a pure Hotchkiss is somewhere between the ground and the highest leaf spring mounting point. You can make arguments for anywhere in between, and as a rule of thumb certainly the wheel centre height is not a bad one. The SAE FBRC concept works well with Hotchkiss, because it is a practical test, but it is an absolute bear to work out analytically - the FBRC height changes if you stiffen the chassis up, for example, which is ridiculous. The ADAMS model above probably assumes that the roll centre is the instantaneous centre of motion of a one-wheel lift - ie the other CP for a Hotchkiss (roughly). Other suggestions include - the middle of the spring in line with the axle, halfway between that point and the front eye, or even the rear eye.

[McGuire]

Roll center is exceedingly useful. It's just named wrong is all.

[LS 1]

Originally posted by McGuire
Sorry, we don't believe in geometric roll centers around here.

Sigh. I guess this just proves that if you think you've got something figured out and it seems simple, you probably don't have it figured out. Reminds me of the time my number theory prof told me he'd never heard a really convincing definition of the number "one." Say what? 30 years later and I'm still trying to figure that one out.

[LS 1]

Originally posted by Greg Locock


LS 1 - I'm not ignoring you. You raise an interesting point. I think we'd agree the RCH of a pure Hotchkiss is somewhere between the ground and the highest leaf spring mounting point. You can make arguments for anywhere in between

Argghh! So there is an argument for placing the RCH of a Hotchkiss at just above the surface of the ground? I don't even want to know what the argument is! But at least now I understand your original reference to a Watts link.

Geez. Here I am walking smuggly along thinking I'm finally getting a handle on this suspention thing and you guys show me I'm all wrong. I feel like a creationist who wandered into a biology symposium.

[Ray Bell]

Does anyone know where I can get a photo of a Watts Linkage and Panhard Rod rather urgently? I mean a hi-res pic...

e.mail me if you can, r@ybell.net

[Greg Locock]

http://www.teamvenom...attsLink-TA.JPG

http://home.att.net/...a/rear_lift.jpg

Why high definition?

[Ray Bell]

To use in our magazine...

Can we get hi-res of that Watts Link?

[Greg Locock]

No, that's off the web.

what I can do is take a picture of my car's watts link tomorrow - I couldn't see to do it last night!

Might not be as clear as that as the fuel tank is in the way.

[McGuire]

Originally posted by Greg Locock
http://www.teamvenom...attsLink-TA.JPG

http://home.att.net/...a/rear_lift.jpg

The Buick GNX Panhard bar linked above is sort of an interesting deal... it was a one-off conversion of the standard GM G-body setup, which was a converging four-link geometry with no lateral link. To improve forward bite on the GNX version (around 500 produced) of the Buick Grand National they discarded the outward-pointing upper links, replacing them with a single "torque strut" or central third link, which required the addition of the Panhard bar to provide the lateral location.

[Greg Locock]

Fuel tank is a problem

It's a very dirty panhard rod in fairly high resolution

http://members.optus...ock/panhard.jpg

[Ray Bell]

Thanks guys, Pat has sent me diagrams that are sufficient for the article...

[Fat Boy]

Are any of you familiar with the watt's link arrangement so that the diamond shaped pivot is attatched to the underside of the axle 'pumpkin' (rotating about a vertical axis) and the links attach from there to the chassis? I saw it on several Trans Am cars in the late 90's-early 00's. It seemed like a fairly clever way to run a lower rear roll center.

[phantom II]

McGuire posted a picture of a Mustang Eleanor with a Watt's as you describe with a Reese bar.

Originally posted by Fat Boy
Are any of you familiar with the watt's link arrangement so that the diamond shaped pivot is attatched to the underside of the axle 'pumpkin' (rotating about a vertical axis) and the links attach from there to the chassis? I saw it on several Trans Am cars in the late 90's-early 00's. It seemed like a fairly clever way to run a lower rear roll center.

[Greg Locock]

That's pretty funny. Very ugly. But as they say, if it is stupid and it works, it isn't stupid.

So one really weird question is why didn't we drop the height of the Watts link on production cars?

I guess there is a ground clearance issue, but more from laziness than it being a real issue.

[Fat Boy]

Originally posted by phantom II
McGuire posted a picture of a Mustang Eleanor with a Watt's as you describe with a Reese bar.

Not in this thead, right?

There is the picture that Greg put up that had the pivot rotation about a longitudinal axis and is mounted to the rear cover. That would have the same basic effect as on the bottom, except for the obvious height differences.

There are a couple different ideas about how to mount the pivot. What are people's ideas on the matter?

If you mount the pivot to the sprung mass, then the roll moment stays constant. If you mount the pivot to the axle, then the roll moment varies with ride height (including bumps and track undulations), but the jacking forces stay constant.

It's an interesting problem. I haven't ever really had to make this call, but I always liked the pivot under the axle. I think that was a Riley & Scott thing. It just seemed like that's about where I'd put the roll center if it were an independent suspension, so why not stick it there on a live. Anyway, I'm curious what others thing or have found.

[phantom II]

Check the first page on the thread below. Ill take a picture of my first roadster live axle setup with Watts and 3 trailing arms maybe later today just for you. I gotta move 3 cars to do it. Some things you cant get around are high roll center, unsprung weight and poor damping. The 3 links can be arranged to provide roll under steer. I have changed it to IRS and there is just no comparison not even in straight line stability in a 1/4 mile. The Reese bar shown on the Mustang only adds to the above problems.



Koenigsegg CC8S

http://forums.autosp...p?s=&forumid=18

http://forums.autosp...anor watts link

Originally posted by Fat Boy


Not in this thead, right?

There is the picture that Greg put up that had the pivot rotation about a longitudinal axis and is mounted to the rear cover. That would have the same basic effect as on the bottom, except for the obvious height differences.

There are a couple different ideas about how to mount the pivot. What are people's ideas on the matter?

If you mount the pivot to the sprung mass, then the roll moment stays constant. If you mount the pivot to the axle, then the roll moment varies with ride height (including bumps and track undulations), but the jacking forces stay constant.

It's an interesting problem. I haven't ever really had to make this call, but I always liked the pivot under the axle. I think that was a Riley & Scott thing. It just seemed like that's about where I'd put the roll center if it were an independent suspension, so why not stick it there on a live. Anyway, I'm curious what others thing or have found.

[imaginesix]

Originally posted by Fat Boy
Are any of you familiar with the watt's link arrangement so that the diamond shaped pivot is attatched to the underside of the axle 'pumpkin' (rotating about a vertical axis) and the links attach from there to the chassis? I saw it on several Trans Am cars in the late 90's-early 00's. It seemed like a fairly clever way to run a lower rear roll center.

Wow, I must have the wrong picture in my head because that layout sounds like the holy grail of lateral links for rear solid axles. Zero lateral movement, RC anywhere you want it (above ground)... Why isn't it used everywhere?

[Fat Boy]

Originally posted by phantom II
Some things you cant get around are high roll center, unsprung weight and poor damping.

The 3 links can be arranged to provide roll under steer. I have changed it to IRS and there is just no comparison not even in straight line stability in a 1/4 mile. The Reese bar shown on the Mustang only adds to the above problems.

I'll go with you on the high amount of unsprung weight. There's just not a lot of way to get around that. Beyond that point, though, a live rear can be made to be a pretty damn good suspension.

The roll center issue can be addressed how I described. What you lose with that arrangement is ground clearance. It isn't a big deal on most racecars or street driven high performance stuff. If a raccoon runs out in the middle of the road, though, you don't want to try and straddle him right down the middle...get him off to one side.

I don't know where you get the contention that live rears have poor damping. Check out one of the better (now dead) Trans-Am cars or a late model that someone has really sorted. They can run on pretty rough pavement and still do an impressive job of getting the power to the ground. Cup cars on a road course look reasonable, and they are super heavy awful things with truck arm suspension and a panhard bar.

If you were having straight line stability issues, then you had something else going on. A common error is to try to stick a bunch of anti-squat in a live rear because it's pretty easy to do and people think it will cure corner exit understeer. All it does is encourage wheelspin.

I've written this before, but a well sorted live axle can beat a poorly executed independent. If they are on equal ground in terms of development, then the independent is the way to go. It's a superior concept. It isn't a clear winner just because it's independent, though.

[McGuire]

Speaking of beam axles and Watt's linkage, here is another application. The old Indy roadsters such as the Watson used beam axles front and rear of course, with Watt's linkages to locate the axles longitudinally -- the front trailing links and rear leading links doubled as lever arms for transverse torsion bars. Shown here is Calhoun, the 1963 Indy 500 winner.