I am in the process of fitting a De Dion axle to my chassis. I was planning on making up simple trailing arms with bush tubes welded on at either end and poly bushes and crush tubes the same as on the front wishbones.

Now the stupid question.......

When cornering, the chassis will be trying to roll around a longtidudinal centreline while the axle is staying relatively parallel with the road surface - i.e. - the axle will not be simply travelling directly up and down in relation to the chassis.Won't this be exerting twisting forces on the trailing arm mounting brackets, bushes, bush tubes etc. Surely in time this twisting is going to weaken the brackets/welds, and therefore isn't it better to use rose joints?

On the other hand I'm just a simple chef - am I missing something?

Thanks in advance,

Martin.

The poly bushes should be compliant enough to allow this much twisting movement of the arm.

Hi you can't rosejoint that setup as it will lock up solid you have to use a compliant bush set

cheers matt

Sorry - rosejointed trailing arms and panhard rod are pefectly acceptable - if not a little pricey. The benefit is sharpness/precision of axle positioning, downside is a harsher ride.

quote:

---

Originally posted by procomp
Hi you can't rosejoint that setup as it will lock up solid you have to use a compliant bush set
cheers matt

---

Hisorry kissey but if you think that a book 4 link setup will work with rose joints you should do some more research as it allows so little roll of the axle and any movement in bump or rebound with a slight amount of roll is just trying to rip the brackets of either the chassis first and then the axle. search on the westfield forum wscc where this has been seen hapening for the last 20 years. all of the fast 7 type cars with tis setup use compliant bushing of some sort.

you may be getting confused with 4 link systems such as on mallock's ans so on where it may look simalar but it's totaly different geometry setup all together.

cheers matt 20 years of building live axle cars for competition.

sorry if that sounds harsh but in this area i do no what im'e talking about

hi syd ime confussed are you for or a gainst rose joints in this aplication.

sorry but ime speed reading and typing as my day job seems to be getting in the way

cheers matt

Matt,
I'd like you to explain how rose(or ball) joints would 'lock up'.

Read what I put, carefully. If you undestand tension, compression, and torsion; then you'll understand how wide of the mark your reply is.

As the major amount of movement is vertical, with only a minor( though not insignificant) amount of twisting, when a rear trailing arm is looked at.

A spherical type joint would have no problems, and is preferable as it would have no compliance, giving pinsharp handling. Downside is harshness, so a poly bush is softer and less noisy. But, the polybush imparts extra loads which are unnecessary(and potentially dangerous), into the arms.

Cheers,
Syd.

If Westfields with Rose joints are experiencing failure with brackets, then it is a problem with the bracket or its attachment, and definitely NOT with the joint setup. As example, if the axle is in single wheel bump, and moves say 4", across an Escort axle this is less than 10 degrees, which is what the joint sees as torsion. Most, if not all, joints would have this in their movement. The only exceptions are some of the very cheap joints.

[Edited on 28/3/06 by Syd Bridge]

I don't get that either.

I've seen rod end bearing used in both axes (axes - not axles) on front setups (although I know which I prefer) and they work fine.

Why would these joints not be more suited to twisting than any form of plain / rubber / metalstic / poly / nylon bush ?

The answer to the problem Mr.Procomp posts, is in his own explanation.

The arms are 'ripping the brackets off', either the chassis or axle. If it was a joint problem, then the joints would be failing, even if oversized as most are. Se7en style cars with rose joints and solid axles have been around since the very first, and an overall problem is not known.

It would appear that Westfield have a car specific weakness, that a non compliant suspenson brings out by 'ripping off' the trailing link brackets.

Syd.

Hi effectivly when the car is in roll and bump one side droop the other you are trying to tension one side and compress the other . at some point this will restrict any further movment of the axle if however the axle trys to carry on moving further the loads are then placed in to the bracketry system and the rest is obvious.This remember is talking about equal lenth trailing arms as opossed to long bottom and short top ones as in the case of mallock's gp4 escorts and so on .set up with compliant bushing allows the axle to move further without resriction thus enabling the car to have more available grip.
probably not verry well explaind and would need a working model to fully demonstrate .

cheers matt

hi syd i think you should build yourself a full size rig and play . if you streghen up the brackets so that there will be no failures all you are going to end up with is an axle that has resricted movement and less grip available in cornering ussually lifting the inside wheel clear of the ground.

i am afraid i think we will have to agree to disagree on this.

cheers matt

In theory both rose joints and Polly bushes are wrong but in the real world you can get away with either.

A locost book style a 5 link system has a redundant member (in theory you only need 3 links + a panhard rod) because of this there is a small conflict of arc when the suspension is in bump and roll at the same time -- however as the chassis and axle isn't 100% ridgid it is compliant enough to work.

Hi british trident is correct may be my first reply was not worded correctly . what i am trying to say is that there is more travel available whith polly bushes than with rosejoints.

cheers matt

on the rally cars we used to run equal links (boreham group 4 style) but with bushes one end and rose joints the other, so you get the travel without binding but keep the adjustability of the rose joints. it was just easy to do it that way, others put insitu adjusters in the links and ran bushes both ends.

Procomp is perfectly correct, a system with 4 trailing arms does bind, it's simply a matter of geometry.

quote:

---

Originally posted by britishtrident
In theory both rose joints and Polly bushes are wrong but in the real world you can get away with either.

---

I'm trying to get my head around this one Syd, and not being at home can't make a little model, but it's not simple!

At first I thought exactly the same as you are saying, but that's not actually quite right.

Matt wasn't talking about running out of roll articulation on the rose joints, he was talking about the fact that with 4 trailing links, all parallel and the same length, in roll, the compression and tension in the links will actually cause the whole lot to bind up. This is due to the arcs of movement of the 4 links, which will cause them to try and attain 4 different distances from the common mounting plane of the 4 links (if that makes sense!) Because the axle is solid in torsion and the links are attached to it, this cannot happen. The bind will only be over a tiny distance, hence bushes will enable it to work, as will bracket flex and chassis/axle twist, hence the system will work, but on a light car will cause fluctuation in loads at the 4 wheels.

Using only three links will remove the problem (but not really possible on a 7), and, if my confused brain is working right, using divergent links will alleviate the effects over the normal working range, but not remove them altogether.

I think though that with parallel links attached to a solid axle which is stiff in torsion, bind up in roll using totally non-tension-compression compliant links will happen, which is what Matt was explaining.

JUst being nosey, but some of you are talking about a 3 link type thing, as opposed to 4 trailing links. Where would these go? Would they be 3 seperate trailing arms?

Thanks nat for explaining what i was trying to say i cant stop work to respond at the moment . clearly syd dose not understand the geometry of the 4 link setup whitch is why he cant understand the point im'e trying to make .

cheers matt

After much head scratching I think I understand, but, for it too bind up its got too travel prob further than the shocker etc would allow ?

or don't I understand........?

The way I see it, and I do understand what both sides are saying, is that the rose joints are better to a point, ie their rotation limit, at which point somthing breaks. However, poly bushes will always transmit some twist to the chassis, but have a generally higher rotational limit, and as such may be better at higher levels of bump/droop. BUT, how much bump/droop do any of you have at the limit? To get the +4"/-4" situation stated above you'll need 8" of travel at each side. And I seriously doubt anyone has near that! Neither system is ideal. So, basically what I'm saying is for most it makes bugger all difference.
At this point he grabs a tin hat and dives for cover......
Forgot to say, I think rose jointing both ends will limit things more. Best bet seems to be rose joints at one end and poly bushes at the other. This should overcome some of the 'vagueness' inherent with poly bushes, but still allow the ease of movement.
Might just have confused myself reading that back.

[Edited on 28/3/06 by DIY Si]

quote:

---

Originally posted by DIY Si
JUst being nosey, but some of you are talking about a 3 link type thing, as opposed to 4 trailing links. Where would these go? Would they be 3 seperate trailing arms?

---

quote:

---

Originally posted by DIY Si
The way I see it, and I do understand what both sides are saying, is that the rose joints are better to a point, ie their rotation limit, at which point somthing breaks. However, poly bushes will always transmit some twist to the chassis, but have a generally higher rotational limit, and as such may be better at higher levels of bump/droop.

---

quote:

---

Originally posted by procomp
. clearly syd dose not understand the geometry of the 4 link setup whitch is why he cant understand the point im'e trying to make .
cheers matt

---

Quote
"Might have just confused myself..."
What i was trying to say was the way it has now been put. Syd says the bushes etc limit things. Others say the system itself is the limiting factor. Seems a bit strange, as any and all suspension systems have their limiting factors. This just happens to be the limit of this type. If the rod ends have a big enough rotational movement the limit is the shockers topping/bottoming out surely? Personnaly I agree with Syd, but I think this'll be a let's agree to disagree jobbie.

quote:

---

Originally posted by britishtrident
In theory both rose joints and Polly bushes are wrong but in the real world you can get away with either.
.

---

i think that with equal length parallel trailing arms, and rose joints, you could twist it til the roses themselves run out of play. It would be similar to a wire cube being collapsed, as the 4 verticals lean over, the top will remain perfectly flat. Unequal or unparallel arms will resist in most cases.

To explain my poor example, take two pieces of card each 20cms square, tie a piece of string to each corner to turn it into an odd cube. Pull the pieces of card apart til the string is taut, then as you twist them relative to each other they will be pulled together, but both remain flat. The fact that they stay flat shows that no forces are acting on the axle or chassis.

So im with syd on this one

However, add a panhard rod an my example falls flat on its face

[Edited on 28/3/06 by JoelP]

I used metalastic bushes on mine and it works
As for the above argument, I'm firmly on the fence!
I'm going to model this up tomorrow and see!

Marcus

I have poly bushes at the chassis end of my trailing arms panhard rod and joints at the axle end and it bounces up and down quite nicely (live axle though but de dion will work same won't it?) I think as Jon says the shock absorbers will stop the sphericals getting to their limits. I have been told that my panhard joint is in the wrong orientation as well but as long as it works I don't care. In theory bumble bees can't fly as they are aerodynamically wrong, but it never seems to stop them as nobody has told them yet.

quote:

---

Originally posted by Syd Bridge

quote:

---

Originally posted by britishtrident
In theory both rose joints and Polly bushes are wrong but in the real world you can get away with either.
.

---

Mr. Trident,

Well then, come on. Tell us what the 'ideal' joint is for this application.

A Universal Joint?Or maybe one of those flash jobbies made from unobtainium!

Cheers,
Syd

---

Since you say the watt's linked 4 link system is the only perfect one, and I'm trying to build one in the near future, what would be best for the joints? Iknew something I'm planning would be good in the end! Huurah

quote:

---

Originally posted by DIY Si
Since you say the watt's linked 4 link system is the only perfect one, and I'm trying to build one in the near future, what would be best for the joints? Iknew something I'm planning would be good in the end! Huurah

---

Firstly a big thank-you to Cheffy for sparking this one off.........I'm finding it particularly thought provoking.

The more I think about this the more I think I may be missing something

I just can't imagine how a fully rose jointed system will 'lock up', (assuming the joints don't run out of radial movement).

I'm off to model this in CAD

Mick

So, the centre point would be inline with the diff centre? Or does it have to be speciafically the back plate for some reason? Just so I get the right pic in my head, the ideal system would have trailing links an equal distance above and below the diff with the shafts horizontal? or does the horizontal bit not matter?

Got it...........I think........apart from why the Watts linkage is different

When the axle moves in bump and droop at the same time the trailing links impart a torsional load in the rear axle and corresponding loads in the chassis. My CAD doodlings suggest that the twist with 1.5" of roll and droop would be in the order of 0.2 degrees or 0.2mm.

Or have I missed the point.........again

Mick

quote:

---

Originally posted by Mix
Got it...........I think........apart from why the Watts linkage is different

When the axle moves in bump and droop at the same time the trailing links impart a torsional load in the rear axle and corresponding loads in the chassis. My CAD doodlings suggest that the twist with 1.5" of roll and droop would be in the order of 0.2 degrees or 0.2mm.

Or have I missed the point.........again

Mick

---

quote:

---

Originally posted by Mix
Got it...........I think........apart from why the Watts linkage is different

When the axle moves in bump and droop at the same time the trailing links impart a torsional load in the rear axle and corresponding loads in the chassis. My CAD doodlings suggest that the twist with 1.5" of roll and droop would be in the order of 0.2 degrees or 0.2mm.

Or have I missed the point.........again

Mick

---

quote:

---

Originally posted by JoelP
i think that with equal length parallel trailing arms, and rose joints, you could twist it til the roses themselves run out of play. It would be similar to a wire cube being collapsed, as the 4 verticals lean over, the top will remain perfectly flat. Unequal or unparallel arms will resist in most cases.

To explain my poor example, take two pieces of card each 20cms square, tie a piece of string to each corner to turn it into an odd cube. Pull the pieces of card apart til the string is taut, then as you twist them relative to each other they will be pulled together, but both remain flat. The fact that they stay flat shows that no forces are acting on the axle or chassis.

So im with syd on this one

However, add a panhard rod an my example falls flat on its face

[Edited on 28/3/06 by JoelP]

---

Being following this all day. Just checked it having got back from the pub! Sorry for opening up a whole can of worms, but glad it wasn't the daft question I thought it was going to be! Thanks for all the input guys!

I think from all this I've deduced the following - if I'm going to use it on the track go for rose joints, for road use go with the poly bushes. Would that be a fair assessment?

PS - Thanks Mix - I aim to please!

[Edited on 29/3/06 by Cheffy]

nat, im just pondering one thing now! If the system i described is symetrical in all dimensions, and the twisting is round the axis through the centre of the cards, how could the strings decide which would go slack, seeing as they are all in an identical situation?

possibly making it square has taken it away from the car simulation, as the cars 'cards' would be rectangular.

[Edited on 29/3/06 by JoelP]

ahh, now if it were symmetrical in all directions then I think you are ok, should move fine, but as you say, on the car its far from that layout!

This really is unf**kingbelievable!!

Grade 8 geometry and trig, and all of you professors seem incapable of simple reckoning!

All of what has been said on this previously put aside.....Someone answer me this...First from you Matt, as you want to be seen as the all knowing expert on this subject..


How do the many hundreds and probably thousands of solid axled racecars that race every weekend around the world, using just the very setup that you say doesn't work, manage to get themselves around a circuit for many laps, without the major failures you say must happen??

And how have those same racecars done so for many many years???

Four link with panhard or Watts is the norm in Nascar and Aussie Supercars, AND THEY ALL USE RODENDS!!!!!

If this is the mentality at Procomp, then my high opinion of the company has just dropped to unknown lows.

Cheers,
Syd.

quote:

---

Originally posted by Mix
Got it...........I think........apart from why the Watts linkage is different

When the axle moves in bump and droop at the same time the trailing links impart a torsional load in the rear axle and corresponding loads in the chassis. My CAD doodlings suggest that the twist with 1.5" of roll and droop would be in the order of 0.2 degrees or 0.2mm.

Or have I missed the point.........again

Mick

---

[quote
has Mix drawn it wrong then Syd?

Syd,

I have just discarded a long post explaing that four trailing links will bind very slightly. I was wrong. Four equal length trailing arms will not bind!

Thank you for posting that, Jon. I'm humbled that you took the time. It's just a shame that the rest have neither the ability to reason this out, nor the strength of character to write as you have.

Thanks Again,

Cheers,
Syd.

God I am so geometrically confused now!!!

Still running this one through my head and STILL can't suss it!

I never was very good at maths!

aaarrrrrrrrggggggghhhhhhhh!!!
can't stop myself getting more and more muddled now!

Now convinced myself that it won't bind up, but I'm still totally confused!!!

If the axle mounts move in a common circle in roll then everything stays right, but they won't surely, as (and this is where I am struggling again!) surely the roll centre with panhard rod migrates in roll????

but then the roll centre is not real and the links are...?????

aaarrgghhh

I give up, need to make a model now just to satisfy my own curiosity/confusion!!

also:

"Roll-steer

When a car rolls into a corner, it is very important that no steering component is added to the suspension geometry. Short trailing arms used to locate live axles are the worst offenders (300 mm trailing arms used on the Locost will induce 5 mm of roll steer in 50 mm travel). A car will never feel 'in-the-groove' or 'take-set' in a corner if the suspension is deficient in this area. The Striker and Phoenix use longitudinal Watts-Linkages to locate the rear axle. This arrangement ensures that the rear wheels never steer. There is a downside with this arrangement. When the car rolls, the 2 watts linkages try to twist the axle in opposite directions. This means that this arrangement can never be completely rose jointed, there has to be at least one compliant joint in the system. "

from Sylva autokits website....................what are the differences between the longitudinal watts linkage that they use and the 4 parallel link that most locosts use in terms of binding in roll????? I can't see any difference in terms of binding but they say theirs does??

anybody please???

Hi syd first off i am not going to get in to a slanging match with you . all i will state is that if you use a compliant bush setup you will achive more free axle movment than with rod ends. please remember we are talking locost geometry pesific here as ime pretty shure that nascars and oz v8's dident look at a locost for their insperation. The fact that we are stuck on such a short lenght arm is the main problem and ive already said that the likes of mallock dont have this problem with longer arms.

As for my reputation well i dont care what you think but i do have many customers who will testify that they are hapier whith thier setup after switching to a complient rear end after using rodends.
You dont seem to want to agree to disagree or explore the other side of the fence.

as i said im'e not going to get into a slanging match.

hoever if you would like to join us when we are at a trackday i will gladly let you drive one of the cars and you can make up your own mind.

this is the last post im'e going to make on this thread as their is a big fence with two sides and a wide ledge on top and this could carry on till the cows come home.

cheers matt

US racers use something called a birdcage to eliminate binding in their rear suspension setups, ever heard of these Syd?

[Edited on 29/3/06 by MikeRJ]

not remotely expert on this stuff but they would appear to solve the problem by allowing one end of the 4 link to rotate on the axle tube????

Don't really understand that as you must get strange torque reactions, but as you say, it seems that this is used to prevent bindup.







[Edited on 29/3/06 by NS Dev]

At times like these I feel like getting the old Weegie out and calling up the spirit of Fred Lanchester.


To quote Newton "We stand of the shoulders of giants"

Ok, just modelled it with string and short lengths of 1" x 2" wood.

My first result was upsetting - it binded!!
However, this was found to be due to my string catching on the edge of the wood, making two of the "arms" shorter.
Once this was fixed, then the system was free (ie the string remained taught when the pieces of wood were twisted), even when taken to extremes.

anyone care to knock one up using rigid links?

If this is not convincing, then I could make a bigger example with wider bits of wood. I could even try videoing the experiment if someone would llet me post it somewhere

[Edited on 29/3/06 by JonBowden]
Rescued attachment test.gif

Yep, needs trying with solid links, the bind (if indeed it exists, as I say I am totally confused now!) will be VERY small in magnitude on a small model (as it would on the car) but nonetheless a problem with solid links.

Just running through it with a colleague here who I usually turn to with geometry stuff, and he's musing over it too.

Conclusion at the moment is that if the roll centre is dead centred within the 4 link brackets on both the axle and the chassis then it will roll perfectly with no binding............

however, as soon as any roll actually occurs, with a panhard rod, the roll centre migrates and the links no longer share a common centre and will then tend to bind up. On the other hand this makes no sense to me as the links are all the same length and spherically mounted so as Syd said, should be fine, but as I see it, one link needs to be able to change length for the system to work????

With a watts linkage, the roll centre remains in the axle centre, so in pure roll the linkage will not bind, but roll will almost certainly be combined with pitch in the real world, or more often just be single wheel bump, in which case again the system will bind up???

God I am confused.

Found a great quote which I have decided sums it up great, by one Colin Chapman which roughly stated any suspension system works fine until it moves!!!

about right here!!

I can move the bits of wood all over the place (simulating moving the roll centre) but there is always a way to keep the string taught.

Thinking about it some more, I am convinced that 4 equal length parallel trailing arms will not bind. But :

- Non parallel arms will bind.
- Non equal length arms will bind (or at least constrain the roll centre).

Thus such arms if using rose joints must be adjusted to exactly the same length and the mounting points must be such that the arms are exactly parallel.

Further, any chassis or axle flex will lead to "dynamic" binding.

Since any binding due to the above causes would place large loads on the bearings, I think it would be best to allow some compliance in the system.

I think that must be the answer John, as you say a system with no elasticity in the links will only need a very small movement of a mounting point to cause a binding situation.

I'm going to have to forget this one now or it will drive me insane!!

Matt.
There is no 'fence'. You made a blanket statement that is blatantly wrong, and you do not appear to be able to admit it. I will gladly take up your offer, should I ever be in the position to do so. However, I'll more than likely be spannering my youngster racing against you before that happens.

quote:

---

Originally posted by MikeRJ
US racers use something called a birdcage to eliminate binding in their rear suspension setups, ever heard of these Syd?
[Edited on 29/3/06 by MikeRJ]

---

Hi syd i did say that my first post was probably worded wrong but i will stand by a compliant set up on a locost will be better than one with rod ends.
after that we will have to agree to disagree ime affraid.

i have never met you but i would still shake your hand if we met one day

cheers matt

Equal lenght and parallel trailing arms will not bind. Using a Panhard rod or Watts linkage won't change that statement.

Unequal length OR non parallel OR unequal and non parallel will bind. Something must twist or bend somewhere.

The Fury kitcar in live axle form uses a Watts link on each side to locate the axle fore-aft and a penard rod to locate sideways. It will bind BUT it is rumoured that this is deliberate as it allows the rubber bushes to function as a progressive anti roll feature. Either that or it's a design mistake that's been made to work well in the real world.

Other linkages are possible. Satchel links won't bind, can handle big power, and need four links as opposed to five. A three link and Panhard or Watts will also work but strangely no one uses that either. Two links and an A frame, as used on the Caterham Deon, will also work perfectly with rod ends without binding.

So I'm with Syd on this.

Ok, I'll admit to not reading this thread (too late at night i'm afraid) but i've got one question.

In the real world perfectly equal length is near enough impossible to do, so what is an acceptable amount of difference between your trailing arms.

1mm, 2, 5 ?

No idea Mike, my brain has reduced itself to porridge with too much thinking about bloody links!!!!

Happy now that I was wrong though, but also happy that Matt was right but for various reasons other than the ones that I thought!

Ah well, just as well my 7 isn't live axle!!

yeah but mine is and one of my trailing links is a smidgen out!

(2mm as i recall)

You would be suprised at how many vehicles have "bind" in their suspension and it is considered acceptable. Pretty well all OEM bushings are rubber which is the great compensator for binding. The rubber will deflect enough to deal with it. For example, on a fox body mustang if you install solid or poly bushings you will only get about 1/2 the wheel travel. Lots of older 3link GM cars are the same.

Ideally you will be able to remove the coilover on one side, and jack the diff up into the bump stop w/o compressing the other coilover. This is how you check for bind. If it compresses the other sides spring, or lifts the vehicle, its bound up.

IMO, for a fun street car you need at least 3" of bump and 3" of rebound to avoid the dreaded bumpsteer. That means you can have up to 6" of difference between the 2 rear wheels. Do you see that in normal driving with a solid axle? Id bet never.

Long story short, the rubber bushings will deflect enough to deal with minor binding at the extreme ends of travel. If built right you will tear the rubber before fatiguing the steel. Rose joints can not, I repeat NOT bind. They will come apart if they do. You must check for bind if you are using a rose joint. I think most only have about 20degrees of rotation in the radial plane before binding.
Also, From my experience poly is totally unacceptable for use as a bushing unless it is greasable and you grease it often. Poly is a sticky materiel and will bind when loaded. It also cold creeps. Stick to rubber, or solid bushings (ideally lined with delrin or similar)

The sprint car bidcages are used because there are no axle tubes. They ride directly on the spinning axles .

For the 4 links using a panhard bar you have to remember that the panhard moves in an arc. Since its attached to the axle, that means the axle is moving sideways while moving up/down. This is where a lot of 4links end up with bind. This is where the rose joints come in handy.

I am no expert, but I do tinker with suspension on a regular basis .
Hope I have helped a little.
Cheers.

quote:

---

Originally posted by procomp
i have never met you but i would still shake your hand if we met one day
cheers matt

---

Now shoot me and leave me for dead if I have got completely the wrong end of the stick, but I'm trying to think how it could bind and try to word it and get a picture up so as to show it from procomp's point of veiw. Now after much head bending and convincing myself that it is impossible for it to bind being equal lengths etc as everyone has already said i resorted to paint as simplicity sometimes overcomes complexities such as autoCAD.

Now, the only way I can see that it can is as shown in the diagram i provide. I think everyone is thinking too much on the big scale of torsion and its quite simply the fact that if you look at the rear axle as if looking down the prop shaft you will see the axle and spherical joints as shown on the axle. Now when one side is up and the other down, the rod part of the joint will rotate, until it gets to the bolt that goes through the spherical part, at which point it will stop rotating. Now assuming this is happening at both sides of the car at once, "binding" will occur. This is assuming it is also happening at the other end of the rod aswell, and that there isn't a rotational bearing halfway down the rod aswell, or the nuts will just undo and free rotation will occur resulting in the bar falling off.



Does that make ANY sense or have I just grabbed the shitty end of the stick that no one was talking about? Been a good think for me though because I've always thought "this is how it's done, it works, it's not broken, don't fix it!"

Mark

thats the mechanical binding that led to the 11 degree figure popping up. Matt thinks it will bind sooner for a geometric reason.

Fair enough, so I was on the right trail to his thinking. Think a 3D model with spherical joints included is the only answer to this conundrum.

i'm getting sorely tempted to get my technical lego out and see what happens.

If i go i'll take some pictures and post them.

(but not tonight)

Confession Time

This had been playing on my mind for too long...I decided to re-model the setup.

I made a mistake in my first model, (basically I didn't allow for the fact that vertical separation of the trailing link attachments at the axle effectivly reduces). When this was taken into consideration the torsion in the axle dissapeared.

So.........I appologise for my previous, (misleading) post.

Mick

The moral of the story................my spatial awareness and ability to use CAD are adversly affected by lager...

You've got it right, Mark.

The system will only bind at the limit of rodend rotation, as I said earlier. At this point, the system will 'lock' abruptly.

A poly bush setup will be putting torsion into the arms, and bushes, with the first movement.( The rodends won't, until the lockup point.) This torsional load is there constantly, but varies, obviously, with normal car movement. The poly bush setup will be softer and progressive (just as Matt says), but WILL lock up, just as rodends will. Simple geometry.

Cheers,
Syd.

Hi,
I resisted the temptation to post before reading the whole thread and I'm glad I did since it seems to have been resolved more or less.

I have a full size working de-dion axle setup on my car with equal length parallel links and a panhard rod and I can confirm that it does not bind throughout the limts of travel at both sides. I did take great care when building it to get the links exactly parallel and exactly the same length and having read this thread I'm glad I did!!

Out of interest, my links are adjustable and have a left hand threaded rod end at one end and a right hand threaded rod end at the other (with locknuts both ends). This means I can adjust the length in situ by loosening the locknuts and turning the arm itself. To get them all the same length I just made a jig from a piece of square tubing with two 1/2" bolts welded to it the desired distance apart. It is then just a case of adjusting the length of each link until they slide on to the jig. Any subsequent adjustment is done on a 1/2 turn at a time basis equally on each link.

Cheers,
Craig.

Hi Craig,

Thanks for the info. What size/thread rod ends did you use?

Thanks in advance ,

Martin

Hi Martin,
I used 1/2" UNF (both bore and shank) which I bought from Rally Design. I did some calculations at the time as to required strength because, as you may already be aware, there is a huge variation in strength amongst rod ends of the same physical size.

Have a read through the following thread:

http://www.locostbuilders.co.uk/viewthread.php?tid=4635

This is a thread I started when trying to decide whether to use rod ends and how strong they needed to be. You will see that I ended up opting for National MSM8T rod ends for my trailing links. The panhard rod needs something a bit stronger but RalDes didn't have the one I wanted in stock so I just got a pair of their Carborace ones just to get it put together and SVA'd. I will be changing these before I put the car on the track though for something with a load rating of around 16000lbs.

I hope this helps,
Craig.

Hi Craig,

That's great. Thanks for the quick response!

Martin.