Running out of ideas - death wobble at <30mph

[FridgeFreezer]

I did realise just as Mike set off that neither set of wheels he owns are likely to fit over the brakes on the 109 but we'll give it a go and a 2nd set of eyes is a useful thing anyway.

[miketomcat]

It looks like success. 

We found a slightly loose TRE (got tightened) and adjusted the tracking to toe out (2mm on the rims). The TRE didn't solve it but it settled rather than getting worse. The tracking then solved it.

Mike

[Wytze]

2 minutes ago, miketomcat said:

It looks like success. 

We found a slightly loose TRE (got tightened) and adjusted the tracking to toe out (2mm on the rims). The TRE didn't solve it but it settled rather than getting worse. The tracking then solved it.

Mike

Nice one!  So, it's about 5 mm of toe out now?  (on a meter)    Happy to hear it's sorted

[simonr]

16 hours ago, FridgeFreezer said:

no steering damper on the Volvo or on the Series

I don't know about the volvo - but on a series, it's built in to the steering relay.  Just a friction damper - but a damper none the less.

While it's true that if there's no play in any of the components, if the castor & toe are set correctly, you don't need damping - but when is that ever the case with a Land Rover?

Long ago, back when I had a Series 2, I had the same problem.  I changed & adjusted everything that could be changed or adjusted - but the wobble persisted.

Eventually I dismantled the relay (complete pig to remove I remember).  Inside there was a spring, 2 thrust washers and two tapered friction bushings.  There wasn't much left of the bushings and the thrust washers were paper thin.  I replaced everything but the shaft - and the wobble vanished!

But Hey, what do I know! 😜

 

[FridgeFreezer]

Well as Mike says, it looks like we got it and it was only a couple of small things that seemed to add up to a big problem;

The TRE was the drag link, the Volvo setup being that the axle end actually lands on a double-balled TRE to clear the portal housing - here you see the volvo setup (left) and my newer steering bars (right), the upper one being the drag link.

 

 

Now, the axle end joint had a very small amount of play in it (having Mike laying underneath rocking the wheels helped massively to detect this). This was giving a small but noticeable "click" when rocking the wheels back & forth with the front end jacked up - I'd previously roped Mrs Fridge into sitting in the cab & rocking the steering wheel but I'm guessing that feeding the system from that "end" was masking the very small amount of play in the joint, whereas getting two of us underneath the truck with the wheels off the floor made it easier to feel & hear - even then you needed your finger across the joint to actually feel the movement.

With this dialled out it was about 90% better but could still be provoked into a gentle wiggle on hitting a bump.

The tracking started off with a little bit of toe-in, moving to a little toe-out helped, moving to ~2mm toe-out made the drive a lot less twitchy and settled the response over bumps, it now feels a lot calmer and any wobble you do get is damped away rather than feeling like it might take hold.

With the configuration of the vehicle and the aspect ratio of the tyres you're always going to get a bit of a "boing" over a bump as the sidewalls allow it, but now it's a single boing rather than a boing that becomes a wobble that becomes a change of underwear Likewise the tyres going square in storage doesn't help the matter but have never been bad enough to cause an issue before now.

 

So, ultimately a very small amount of play in an imperfectly tracked system allowed it to start a wobble and escalate from there.

 

Thanks all for the suggestions - now it's on to the next item on the to-do list and maybe even getting out to play some time!

[B reg 90]

Nice bit of detective work.

portals look nicely refurbished. Portal tec disk conversion by the look of it?

[FridgeFreezer]

4 minutes ago, B reg 90 said:

Nice bit of detective work.

portals look nicely refurbished. Portal tec disk conversion by the look of it?

Yep - did a bit of a rebuild on the ends while I was doing the callipers and rims last year - basically trying to bring everything up to snuff and get rid of a few things that were done in a rush when the truck was built.

[elbekko]

Glad you found something. I can definitely see that kind of TRE setup amplifying issues, since you suddenly have two wheels flopping around with just one loose joint.

[Ed Poore]

On 11/25/2021 at 8:47 PM, FridgeFreezer said:

I'm borrowing a Mike over the weekend to see if the two of us together adds up to a whole brain

Surely that's a negative brain in total... Two wrongs don't make a right you know...

It's amazing how much a little bit of play can add up on these systems - I once pretty much refreshed all bushes and steering joints simultaneously on the 110 and it was sublime to drive - very precise to drive. Then one part got slightly worn, didn't replace immediately which subsequently wore out the next one and then it was a continual game of cat and mouse.

[Anderzander]

Great news 😀

You wouldn’t think that such a small amount of play could amplify to that degree.

[miketomcat]

I think if the tracking had been toe out to start or there was a steering damper the problem would not of manifested so badly. The play would of been noticeable but the death wobble would of been contained. I'm just glad we found it and were able to rectify it.

Mike

[deep]

Great that it's fixed!  Keep an eye on that tyre wear though, as you've moved a long way from where you were with toe out.  I'm still skeptical, from my desk half a world away.  If you didn't make a modification and suddenly ended up with a problem that big, I smell more than a slightly worn tie rod end.  Friend DeRanged knows a huge amount about this stuff and I'd still take his advice and check caster, for piece of mind, if nothing else!

That thing about supermarket trolleys is funny.  They "self-centre" by having the axle trail the point of steering rotation.  As the wheel moves from straight ahead, the drag of the sideways track increasingly pushes it back behind that point (going forwards - if you go backwards, the drag actually increases the angle until the axle has turned 180 degrees).  The effect is light at small deflections and more at larger deflections.  Because of the inertia of the swinging wheel, they don't necessarily come to rest in the straight ahead position and easily overshoot.  In a way, they're designed to do the death wobble!  Of course it doesn't matter because they aren't coupled and are tiny but they are good at demonstrating the problem.

[FridgeFreezer]

I think the big (and semi-square) wheels don't help - they tend to do this sort of thing over bumps, especially if there's any side-load;

 

And not being round after being parked for any length of time means you're putting extra vibration into the system.

As Mike says, if the tracking was set differently to start with the wheels would be pre-loading with a bit of scrub and hence tending to stop oscillations as they're putting the track rod in tension.

However, the drag link TRE having play means that the track rod is free to oscillate side-to-side just a little - and with tracking set incorrectly there's less resistance (around 0 degrees you're encouraging this sort of thing). It also removes/reduces the damping effect of the PAS box which is then on the wrong end of things.

I still maintain my philosophy of not running a steering damper - the truck has driven very well since it was built and has seen a huge range of use and abuse with no issues, I'd rather know there's an issue early than rely on a damper to mask it until it gets really bad. This is still early days / low miles for the heim joints in the steering so if I need to step up my game in terms of checking for play and adjusting on the regular so be it.

I understand the concern/interest about caster angle but as I've said a couple of times, short of a deformed leaf spring it's not changing - the fact the anti-wrap bar still lines up tells me the axle is sitting at the same angle it ever was.

[Snagger]

Glad it’s sorted.

My understanding is that a normally RWD car will have toe in and a FWD or permanent 4wd car toe out.  I don’t think it is defined by the spring type.  Checking the LR specs for the Stage 1 and 101 should answer that question, though.  It is amazing how much effect the tracking has - my RR was leaping all over the road in the 1st and 2nd lane ruts on the M1 and M25 until I found it was a little toe in.  

[miketomcat]

4 hours ago, Snagger said:

Glad it’s sorted.

My understanding is that a normally RWD car will have toe in and a FWD or permanent 4wd car toe out.  I don’t think it is defined by the spring type.  Checking the LR specs for the Stage 1 and 101 should answer that question, though.  It is amazing how much effect the tracking has - my RR was leaping all over the road in the 1st and 2nd lane ruts on the M1 and M25 until I found it was a little toe in.  

Your quite right. It's to do with the component slack being taken up so you end up near enough parallel under load. Nothing to do with spring type but this the reason for bumpsteer on some suspension types 

Mike

[Snagger]

Yep, even with no play in the joints, there is always a little flex or spring in the components themselves which will change the tracking under driving conditions.

[geoffbeaumont]

I guess the toe in on series is because the expectation is that if they're traveling at any significant speed they're in two wheel drive?

[miketomcat]

That would be my assumption. If you were to convert to permanent 4wd I would change it to toe out.

Mike

[Snagger]

13 hours ago, geoffbeaumont said:

I guess the toe in on series is because the expectation is that if they're traveling at any significant speed they're in two wheel drive?

Yes.  The front wheels will drag and so will pull parallel.

[FridgeFreezer]

On 12/14/2021 at 10:46 PM, miketomcat said:

That would be my assumption. If you were to convert to permanent 4wd I would change it to toe out.

It is permanent 4WD these days, and it is toed out now

The transfer box with the 2WD kit is currently in the shed awaiting overhaul, so I threw a standard permanent 4WD one in.

[toenden]

I was thinking that maybe toe in/out have more to do with steering setups than 4wd/2wd. A series have the trackrod in front of the axle (As far as I recal) therefore toein will keep the rod under tension when driving forward. A defender on the other hand has it on the back of the axle, so to keep it under tension it needs to be toe out. By "under tension" I am assuming that the roadwheels will try to selfcenter on the axle therefore and apply A small amount of streaching force on the trackrods. In my mind, A trackrod can fastly become a spring if you put preasure on it, leaving a potential wobble. 

The disclaimer is, that I by no means am an engineer, but I have tried to straiten enough trackrods to remember the thought: "god they are tiny and easy to bend/straighten" 🙂 which tels me that they might be able to work as a spring. 

/mads

[Bowie69]

It's not really 2wd/4wd, it more whether the front wheels are driven. If they are driven, then the wheels get pulled forward a touch under power, and toe them in, if not driven you need to account for rolling resistance, which would pull the wheels backwards and cause toe out.

[geoffbeaumont]

6 hours ago, toenden said:

I was thinking that maybe toe in/out have more to do with steering setups than 4wd/2wd. A series have the trackrod in front of the axle (As far as I recal) therefore toein will keep the rod under tension when driving forward. A defender on the other hand has it on the back of the axle, so to keep it under tension it needs to be toe out. By "under tension" I am assuming that the roadwheels will try to selfcenter on the axle therefore and apply A small amount of streaching force on the trackrods. In my mind, A trackrod can fastly become a spring if you put preasure on it, leaving a potential wobble.

Surely that would work the other way round, and compress the track rod in each case?

[miketomcat]

It's definitely down to driven wheels and taking up tolerances. I've had several mini based cars all the front wheel drive were toe out, the rear wheel drive was toe out at the back (front subframe in the back) and toe in at the front. 

Mike

[Carloz]

Interesting...

I have a bit of the same thing (on much higher speed) with a 1986 Range Rover, completely standard setup with RRC '92 alloys with 235/70-16.

In general use it behaves very well but when hitting angled bumps on the highway (while crossing angled viaducts for example) @ arround 110/120km/h it can get very violent! When that happens you know you are alive, if you were in doubt!

 

All is checked a couple of times. A RRC has the steering damper is new and on the best spot to prevent these kind of things: on the track rod. It happens with and without 30mm spacers. And stil after replacing the steering box (because of a leak).

So... after reeding this it seems difficult to detect the amount of play that can already cause this problem. My RRC has done +300.000km so it seems a rebuild of the front axle is the only way to solve this.