Warning! Roll Centers & Front-End Lowering

[FJCamper]

I see a lot of posts about lowering the front end of a VW for cosmetic reasons ("in-the-weeds" look) and need to post a warning about how this can really affect handling.

The "roll center" is the center point from which the chassis and body leans.

VW's (and older Porsches up through the 911's before they lost the torsion-bar trailing arms) have the rear roll center considerably higher than the front. This is one of the things that makes a VW want to swing its tail out in a turn.

The tail-swing happens not because the rear loses adhesion first, but when the front has lost adhesion and throws more load to the rear, and bingo, the tail comes out.

The VW was designed to make the front do as much work as possible. The rear tries, but it is at the mercy of the front.

Prime example. When we had our '70 Ghia level, it used both front and rear 19mm sway bars, and chassis tuning was predictable. When we put a Puma beam on the front to permit lowering (in anticipation of also lowering the rear, (surprise) we discovered the adjustment wouldn't let us start at standard ride height. We got lowered two inches like it or not.

After some wild handling experiences, we found that removing the rear sway bar was necessary for survival, not just chassis tuning. Remember, we are doing extreme maneuvers at high speeds at Road Atlanta, Roebling Road, etc.

You may not be road racing, but a sharp maneuver on the street can cause trouble. Roll centers don't care where you are.

FJC

[Marc]

While I'll not dispute the conclusion - most radically-lowered or raked Beetles/'Ghias handle and ride worse than stock - I must point out that it isn't changing roll center height that's the culprit.
The roll center height of the front trailing-arm suspension is always going to be at ground level; by its very nature, it cannot be anywhere else.

[FJCamper]

Hi Marc,

Yes, the front roll center is at ground level, like my drawing.

Lowering the nose is not greatly changing the roll center itself, but has the effect of leaving the rear center of mass higher (in relation to the front), which affects the forces on the rear suspension.

The front end lowering actually aides in jacking some weight forward, and up to a point it makes the front suspension work harder and handle better until that bad snap-oversteer moment occurs.

We found that removing the rear sway bar totally was essential, because it throws cornering loads forward (diagonally wheel to wheel) and caused the front end to wash out earlier, and when the front goes, the rear is a heartbeat behind it.

I can't lower the rear until I get the engine dry-sumped, so until then we run no sway bar and as little front drop as the Puma will allow us.

The people who lower their front ends tend to also use very narrow tires, which in an emergency dodge or weave will lose grip quickly.

I like cosmetics as well as the next guy, but not when it costs so much in handling and safety. I hope we never get a fad of painting the windows black -- people would try to drive like that!

FJC

[fortyeye]

As you may notice ... suspension geometry means little to the vast majority of those lowering their vdubs. Overly narrowed beams, removal of shock towers, bump stops, sway bars (and sometimes brakes), total disregard to toe and caster settings, and that's just the front end. Moving to the swing-axle rears with with sagging negative camber (think airkewld) and again, no concern for toe settings. It's good to see most of the mistakes haven't migrated to the Ghia community. Road race on!

[Flow]

Is this applicable to superbug? Otherwise, what attention should be paid on suspension tuning for superbug?

BTW, FJC's articles are always interesting and constructive.

[FJCamper]

Hi Flow,

The Superbug suspension is very much like the 911's before they got fancy. We built an IMSA Superbug and raced it in the mid-1970's. We were able to construct what today would be considered a tube-frame chassis, but we kept the floorpan. The front tub was cut away to allow for our fuel cell, the chassis braces for it, and the fact we built in our own camber plates and used BMW front struts so we could have disk brakes.

We did use the Superbug swaybar and suspension linkage bar to the chassis. That was some fancy welding, as we needed to look sort of like a VW front end. It was a good thing the IMSA tech people didn't know what VW parts looked like, and we told them the disks came from the European Supers!

Like with a 911, you can lower it but keep the suspension level.

The strut front and diagonal arm rear suspension combo is very good and has entirely different dynamics than the torsion bar front. VW built in massive low-speed understeer to the Super by allowing the front tires to tilt way left and right as you steer off center, breaking tire adhesion.

At high speed, the understeer is actually a good thing, and is lessened as the rear suspension begins to lose adhesion.

The "nose-down" look is a visual hold-over from the 1950's hot rod cars that needed front-end rake to raise the tail because when they took off, the tail was going to squat.

The rake became associated with the appearance of "performance."

We don't want "the appearance of" on our cars, we want the actual performance.

FJC

[helowrench]

We don't want "the appearance of" on our cars, we want the actual performance.

FJC

Ahh, truer words have never been spoken right there.



Rob

[Flow]

I am modifying my Superbug (1303) for road racing.

The front suspension has Golf Mk1 (Rabbit) struts (coilover modified) with Bilstein HD inserts, Porsche 944 spindles and 5/8" swaybar. It is lowered by about 4in from stock . The spring rate is 100lb/in right now but mayl be increased to 120 or 150lb when seeing the race track.

I am doing the rear right now. It will have 944 torsion bars (23.5mm) and I will install some adjustable spring plates to level it with the front. I will leave the rear without swaybar at the beginning but if necessary I will add the SAW 3/4" bar or Whiteline adjustable bar. The shocks are Koni red which I will set them to medium stiff.

I will set the camber front 0 to -0.5* and rear -1 to -1.5*.

Brake system is Porsche 944 NA all corners.

The wheels/tires combination is front 6x15"/195-55-15 and rear 7x15"/225-50-15

Is there any obvious "mistake" to start with this setup?

[Piledriver]

Don't forget that at speed. aerodynamics matters.

Most (all) beetles and possibly Ghias should ride with the nose a bit higher than the tail.

I'm sure if I Google long enough I can find the reference, pretty sure it's true.

[FJCamper]

Hi Piledriver,

Running an inch lower in the rear is a good setup for hillclimbs. It is subtle but effective. Aircooled.Net once had an Aussie hillclimb account and suspension setup in general where this was recommended. The idea is to get the rear roll center down even more in relation to the front.

For long high-speed straights, if you have no airdam, you want as low a nose as you can get (but it is still limited to about an inch) to fight front end lift. The air gets under our cars and creates more front lift than rear.

Look at the B&W image of our IMSA Herbie. See those air deflectors under the front end? We were not allowed an airdam, so we improvised and called the sheet-metal bits part of an air-guide system to our brakes, which was allowable!

Also note we sit fairly level, as low as the rules allowed.

With this car, we set and still hold the official US speed record for a stock body VW on a closed course, run at the Talladega Alabama super speedway.

No kidding. Keeping it on the ground was tough.

FJC

[FJCamper]

Hi Flow,

Your specs look good to me. Make sure all this is tied to a good stiff rollcage. And I suspect you'll need the front swaybar.

Do you have a sponsoring body such as NASA or SCCA in mind? What class in which you might compete? I'd recommend NASA, and spend a season running the car in HPDE 1, 2, 3, and finally 4 to get it sorted out.

A car set up as you describe should have at least a 7000 rpm 2-litre in it, and have two gearboxes, high speed and medium track.

FJC

[Flow]

Hi FJ,

I will just take part in some local club races in Asia which don't have strict rules. Sure it's got a standard rollcage. It has a Bugpack 7/8" front swaybar.

I've got a used 1835 motor (stock besides the big bore) from a friend. I plan to stroke it to 74mm and drop in some hotter heads and cam. One of the racing clubs limits the intake so I probably have to be conservative on the cam, may be top at 6500rpm or even lower.

Right now I have the stock gearbox (4.375 final drive) on the car. I do have a spare one with 3.875 which I can swap if need.

[raygreenwood]

Excellent....you have discovered what I have been ranting about on the 411/412 forums for several years...because too many people are excessively lowering the front end of the 411 412.

Maybe I can add some to your explaination...and I see (I think) where Marc may have some reservations. I am not sure if "roll center" is quite the right word....but you have seen the detrimental effect.
In a nutshell ...what is causes can "look" like a roll during cornering...but is actually an inertial "pitch" forward. It looks like a roll because during corning that includes deceleration you get a lifting of the rear, a lowering of the inside corner and a raising of the opposite corner diagonally in therear. So on a braking right sweeper you get the right front corner dropping, the left rear rising....and because of lack of adhesion and cenrifugal force ...the rear wants to spin to the left.

If you were in a straight line braking...you get the forward inertial dive. If you were powering through teh corner...that same right sweeper...but no brakes so no inertial shift forward....you get the opposit...if you get anything....which is the right hand front slightly lifts...and since the body pivots across its length...the left rear corner drops.

A bit of a case study:

In the 411/412 the nose from the factory is artifically high for load purposes (better with a full trunk). Lowering it flat (parallel to the grround at the rocker panels)...greatly helps. Lowering the front excessivly already hikes the rear up...pivoting about the axle. This helps castor a little in the front ...but there is a limit.
You already have a rear weight bias...meaning a serious enertial shift forward and more lifting of the rear. What it does is very quickly reach the limit of the rear shocks...taking pressure from the rear tires. Also..at this point your sway bar starts to work against you.

It works better to keep the ratio of front lowering to rear...sane. No more than a couple degrees. As you go even lower in the front...drop the back to keep the angle the same.

One method I have seen to counteract the jacking of the body and excessive unloading of the suspension...is one Saab and a few other manufactures had. use "leashes" for the rear trailing arms. These are opposite pivot outere links attached to the body at the trailing edge of the fender. usually steel with a bushing in each end. as the bosy lifts either through enertia or sway/roll in cornering the body can only lift so high before it is limited by the arc of teh rod. these are not springs or sway bars. they are arc limiting links. Look at say...an 82 Saab 900 suspension diagram to see what I mean.

By the way, this is refeered to as a "watts link". It usually was used to counter the torque and enertia of braking. it could also beused in some systems to get rid of the need for an panhard rod as these links could arrest lateral body forces if not too long and spindly. Its primary function is to decrease roll center.

http://www.google.com/imgres?imgurl=htt ... image&cd=1

Ray

[FJCamper]

Hi Ray,

Yes, before the actual roll centers kick in, we have a lot of pitch and yaw that goes on. This causes weight transfer, and because our VW's have almost no anti-dive or anti-squat in the suspension geometry, handling and stability gets dicey.

On our racing Ghias, I counteract roll, pitch, yaw, and weight transfers as best I can by keeping the suspension firm and the Konis on medium. All I can do is minimize it.

Too hard is asking for trouble and too soft is deadly.

Watts linkages, panhard rods, DeDion rear suspensions, etc., all work but are for different suspensions than our own.

I've mentally designed anti-dive/anti-squat devices, but never built one. I did watch Porsche fight with this problem on our racing team. They kept it simple, only once going to an exotic adjustable rear sway bar controlled by the driver in the 935's.

VW/Porsche handling? If it were easy, anybody could do it.

FJC

[raygreenwood]

Hi Ray,

Yes, before the actual roll centers kick in, we have a lot of pitch and yaw that goes on. This causes weight transfer, and because our VW's have almost no anti-dive or anti-squat in the suspension geometry, handling and stability gets dicey.

On our racing Ghias, I counteract roll, pitch, yaw, and weight transfers as best I can by keeping the suspension firm and the Konis on medium. All I can do is minimize it.

Too hard is asking for trouble and too soft is deadly.

Watts linkages, panhard rods, DeDion rear suspensions, etc., all work but are for different suspensions than our own.

I've mentally designed anti-dive/anti-squat devices, but never built one. I did watch Porsche fight with this problem on our racing team. They kept it simple, only once going to an exotic adjustable rear sway bar controlled by the driver in the 935's.

VW/Porsche handling? If it were easy, anybody could do it.

FJC

Not really. this is where people get too hung up in one system type or another. Suspension systems are themselves collections of methods of tampering with physics. Only a few get named after someone who invented the first one. The components that conrol differnt physical movements get chopped up and grouped together all over the place in differen tsystems.

Are you aware that Watts linkages are used on a great many independent rear ends in cars made as late as this year?

A for instance of an other type: The motion and camber problem of a solid rear beam like the saab had...is nearly identical (though not as extreme) as the swing axle beetle. On the solid beam (like the saab)....when one wheel is depressed...the other lifts pivoting about its center...and increasing camber. They fought this by installing trailing arms at each end as well to make it a solid beam with an extra link at each end. It combats jacking much like going to IRS from swing axle did.

But...because its not perfect....hit a heavy bump that bottoms the trailing arm while going through a curve that has the cross beam pivoting...and you get camber jacking...and a change in body roll center.

What the Watts link does....is allow only 1-2* of change in camber of the wheel (when the wheel/arm reaches its limit of compression)......in relation to camber of the body.

So when for whatever reason, either curves, braking etc...the body has an enertial shift change forward...pivoting or rolling about its center....the watts link acts as a leash...and limits not only the forward motion of the body but how much it can lift off of the rear suspension. It controls how far the suspension can unloaded. It limits the pivoting arc of the trailing arms.

The funny thing is that the Saabs like all vehicles now was a motly collection of solid rear beam with pivoting trailing arms and a panhard rod...which was totally rendered unneeded....by installing the watts links.
The watts links were more oft than not seen on A-arm and transverse A-arm suspensions.

Because this watts link is NOT ball jointed (whcih the cheesy diagram did not show)...but is actually a barrel and bushing assembly at each end ....just like a control arm....it also does not allow the body to sway sideways out of alignment with the wheels through cornering.

This type of link is not a product or characteristic of any particular suspension style. It can be readily used on any suspension that incorporates a trailing arm.
It can also be used on both front and rear suspension and has been on many vehicles. In fact...some of the front four link and multilink suspensions are simply derivatives of a watts link.

On a front suspension...a watts links would generally trail from forward. It could also lead from backward on a struts suspension that does not employ a radius or diagonal arm. The exception is that on a front suspension ...a watts link (probablycalled something else when used up front)....does not counter body roll. It counters random directional changes and bump steering by countering compression of control arm and radius arm bushings that come from the friction of forward motion of the tires, striking ridges and bumps.

Most modern multilink suspensions have details from several classic original suspension systems.

Don't get lost in the thought that "Our cars are not using that sort of suspension". Look instead at what characteristic of physical reaction a given suspension component is countering...and decide if that one sub-component could be useful. See how other people solved the problem you are having...and steal/borrow from their effortsRay