Technical Fixed the Death Wobble

And more importantly understood the "why" for at last my particular case.

I am converting an old Harwood Super-Gas '27 roadster for street use. They use a straight axle with Anglia spindles and in this case a torsion bar spring. Not your average street roadster. However, it still has a front-located tie-rod reversed corvair steering box and tubular drag-link.

It has been around for a long time and was raced successfully. I thought I would be smart while reassembling it and add some caster as I was once told, "you can't have too much caster with a sold axle." Not so smart.

What I learned is this: 1.) Use toe-in with rear tie-rod, use toe-out with front tie-rod. Toe-out with front tie-rod puts the rod in tension where it is strongest; toe-in with front tie-rod pushes the tires and tie-rod inward, causing the tie-rod to deflect and create the death-wobble. This stumped me for a day and now seems so obvious once realized. DOH! 2,) I suspect that adding caster reduces the effect of toe and will require it to be increased. Think of it at it's extreme; if you have 90* caster the toe adjustment would have absolutely no effect. Hope this saves someone else from some head-scratching.

Ended so-far with 8* caster and about 3/16 toe-out. Waiting for belts and helmet before exceeding about 60 MPH but so-far so good! May try to add a little more caster as the steering is pretty light.

 

Lots more people post their problems than their solutions - thanks!

BTW, the cure for death wobble on my car was simple: get rid of the old '33 Ford wires and outdated bias-plies, and install a set of new steelies and Coker Classics. Death wobble gone!

 

This makes sense. The more I learn the more I'm convinced the classic, elusive death wobble is most often caused by bowing oscillations of the tie rod.

 

Theres’s a bit more involved than that.
Generally Toe is used to negate the effects of “camber thrust”. Negative camber needs toe-out , positive requires toe-in. [think of 2 cones driving towards each other]
Early cars had lots of positive camber because of the curvature in the roads therefore needing toe-in.
Death wobble [on cars with good componentry] is cause by forces fighting each other, all it takes is a bump to start a chain reaction.
Too much scrub radius is the biggest culprit.
Designers of geometry try to put the king-pin inclination close to [or slightly inside the centreline of the tyre ]
Now because of king-pin inclination if there was zero caster the wheels would “flop over” when turned either direction.
So to correct this ,the manufacturers add positive caster so the outside wheel is closer to vertical during cornering. [the side effect is the car will straighten up when the steering wheel is released]
Generally sharper turning circles would require more caster, but manufacturers avoid this and try to keep caster to a minimum for normal turning at normal speeds [ scrubbing tyres in a carpark isn’t generally life threatening ]
Theoretically If there was NO tie-rod the caster would try to toe the front wheels inwards [due to vertical forces], so to reduce these loads the common practice is to add a bit of scrub radius.
If there was NO tie-rod the scrub radius would try to toe the front wheels outwards, so the 2 forces counteract each other.
Now with the common practice in hot-rodding is to fit smaller diameter and wider wheels [compared to the 30’s], The moment smaller diameter wheels are fitted to the same centreline the scrub radius increases [ due to king-pin inclination ] Wide offset wheels make it worse.
Excessive scrub radius isn’t an issue in a straight line with no bumps because the drag [toe-out forces] counteract each other, but it does load up the componentry more [causing bent tie-rod on rear steering etc]
When only one wheel hits a bump with excessive scrub radius, that particular side will sudden toe-out snapping the steering wheel [I have yet to find a road with zero bumps so this is the norm]
Now another common practice in hot-rodding is to increase positive caster with the myth that it increases straight line stability [because hot rodding and drag racing go hand in hand so this is common]
Excessive scrub radius combined with Excessive positive caster in a lightweight vehicle is where the oscillations will start [all it takes is a bump for the scrub radius to start the chain reaction]
The lighter the vehicle, the worse it is, because caster will lift the vehicle instead of forcing itself to return to a straightline position.
If the LH wheel hits a bump, the steering will snap to the left, causing the caster on the left to try to steer to the right. These oscillations will go past the straight ahead centreline because there is excessive caster for the weight of the vehicle.
The fix is to reduce caster and reduce scrub radius ,all other methods are patch ups [providing the componentry is in good condition]

 

Good explanation, Kerry.

 

Good one Kerry. Radial tires , (maybe not real traditional) makes a BIG difference. I remember the 1st time I fitted radial tires to a car I was driving it felt like I could drive on the curbs without the steering wheel grabbing.


Ago

 

KerryNZL did a pretty good job summing this up but adding an excessive amount of caster is not the way to solve the issue. Too much caster produces what is best explained as a shopping cart wobble, maximum caster should not exceed 6* with larger tire tread width.

 

I forgot to add 2 more modifications that don't help the situation.
One is "split bones" , the original A frame does help to support lateral wobble , any sideways movement needs to Arc on a radius [from the ball joint]
By splitting the bones, the whole front end can "parallelogram" itself.
The other mod is "reverse eye" springs and/or reset springs. When the eyes are on the topside of the arch it is slightly wider between the eyes [ flattening the springs also does the same thing ]
With the spring eyes wider , the shackles are more vertical so there is less lateral control.
These mods don't actually trigger death wobble, but allow it to happen more easily.

If the spring is re-arched or reversed, the eyes should also be rolled inwards so the shackles end up being closer to horizontal .[unfortunately this is easier said than done]

 

A panhard bar on the front could take care of both of these problems in one swoop.

 

I read the shackles should be closer to vertical. Not even 45 deg. as mentioned above, panhard bar will take care of side motion.


Ago

 

Close to vertical with a Panhard bar; flatter than 45° without. The reason for this is that without a Panhard bar the shackles set up a secondary roll centre that becomes undefined when the shackles are vertical. This is bad if you're looking to the shackles to provide lateral location but good if a Panhard bar is doing that job. You don't want different roll centres trying to define the same motion: angled shackles and a Panhard bar will want to fight one another.

 

I guess that is why my car with vertical shackles and panhard bar works so problem free.


ago

 

There are always examples out there that work. As a road-racer, I personally like "too-much" caster and less static negative camber. [there is less camber thrust in a straight line ]
But what you haven't mentioned is the wheel combo.
I seen more than 14 degrees work OK , but usually on a drag car with skinny wire wheels where the scrub radius is close to zero [or even negative scrub radius where the king-pin centreline intersects at a point outside the tyre centreline]
The ones with negative scrub radius can get a violent death wobble when reversing

 

This is what I have done on just about every dropped axle set-up I have had over the years. It always works. It is a VW part or buy from SoCal. The bracket on the wishbone I fabricated.

 

Those do work well to even out the highs and lows.
Do you have a VW application to avoid paying the SoCal tax?

 

The one shown is SoCal. I straighten the end to 90 degrees that goes to the wishbone.

 

We fabbed this panhard bar for my suicide front end. I had a very violent death wobble. The panhard with adjusting the wheel alignment with trial and error took out 98% of the wobble. We put on a damper for that last 2%.

I have spoken to many hot rod owners with a suicide front end. I've never met one that didn't have the DW problem. I canvassed the Goodguys at Indy show last September. I think I found 3 hot rods that had the suicide with tie rod out front. One guy told me he has the DW, and he just lives with it (his car was fully traditional and was competing in the Top Hot Rod competition). That's not a scientific study, but it makes me think the suicide is more prone to DW.

Panhard bar:







The damper uses welded on tabs. While standing in front of the car, the grille shell masks it.







And yes I've read the old threads on the single leaf spring.

 

I use a 1960 - 78 VW damper, available from NAPA.

 

Thanks!

 

I never considered that 'splitting' the bones would allow for the 'Parallelogram' effect, Thanks for pointing that out!

 

so can anyone else confirm that toe out is good if your tie rod bar is in front of the axle? Mine is in front and i seemed to fix my wobble by having about a 1/16th of an inch toe in but I've never had it up to highway speeds yet

 

Completely unrelated, but my o.t. ram truck is toed in. Made a world of difference compared to the other shop towing it out. 4 bar setup with from steer. So it makes sense to me that you would see improvement. Sorry for the comparison, just went through similar situation....

Posted using the Full Custom H.A.M.B. App!

 

Geometry is geometry.
The only difference between front and rear steer is whether the tie rod is under tension or compression when there is drag on the wheels.
With more positive offset wheels [ eg: FWD types ] the drag would try to toe-in the front wheels so a rear steer tie rod would be under tension
It is all to do with king pin inclination and wheel offset [ where the king pin inclination intersects the tyre footprint ]
Toe-out should only be used with negative camber to counteract camber thrust.
Excessive negative camber can also increase drag on the inside of the tyre, and toe-out can also put more drag on the outside of the tyre because of positive caster
Toe out can be quite unstable during braking.

 

I use much less caster 3 1/2 pos at the very most on my cars which includes the one to the left. They steer very well. The only time I have trouble with DW is on a very bumpy track and then at very slow speed. The VW damper completely eliminates that stopping the repeat action after the initial bump. My theory of what causes it is a bit different but that is unimportant. A damper should be part of any straight axle steering is the conclusion I have come to after many years. I run toe in , 1/16 to 1/18 depending. I believe that when at speed the natural tendency of the wheel to want to walk around the kingpin brings that very close to zero at speed. Toe out is often on front wheel drive vehicles for the same reason as under power the wheels would wrap around forward if the was no tie rods. With toe out the deflection and there is always some brings it close to zero when driving.
don

 

This shock absorber style drag link snubber is a waste of $35.00+, if you have to add this band aid to control your steering resistance you got something else wrong. After going over this with a number of customer issues I have found that although the snubber does mask some wobble issues it seems that the wobble issue is created by the lack of having the suspension installed correctly in the vehicle or loose parts. My position has been to carefully insure that the axle is installed squarely to the chassis and the spindles are square to the axle before setting the toe, carefully setting the caster to not exceed 6* and insure that the steering box and all ends are lacking any play. Since the tires are coupled by the tie rod any action created by incorrect tire alignment is magnified as the scrub angle initiates the return to center action. Thus if you have a wobble you have something wrong as simple as having one tire patch ahead of the other because you welded the radius rod bracket on one side ahead or behind the other side bracket and actually skewed the axle.

 

Too bad you hadn't been there for Volkswagen, Jeep, GMC and the rest who install them at the factory including SOCAL.. just think of the money they could have saved!
don

 

Dolmetsch, I realize that there are a bunch of doubters because they saw a part advertised in a fancy catalog and was to be the next best thing to ice cream but doesn't necessarily make it correct, just a band aid cure. I have a significant amount of time building I beam performance vehicles and have yet to rely on a steering shock dampener. The reason that this is used on other vehicles is to compensate for a design error and it's the cheapest way to CYA without a total redesign. Most of this wobble is due to design error, vehicle over loading , incorrect wheel offset or tire size as well as incorrect alignment so I can stand by my position of not being in favor of the overrated horizontal shock absorber. Once an I beam is installed correctly the resistance in the steering box as well as king pin friction should over ride wobble tendency. This position is backed up in reviews by designer Maurice Olley as he evaluated the development of IFS suspension for GM.

 

Interesting information which makes a lot of sense. One question, a lot of the shock/dampers are factory installed, not add on. So are you saying the manufactures were using them as a band aid in place of correcting whatever the issue was?

 

All 3 cars I built with I beam axles worked well without DW, never needed a steering damper.



Ago