Tubular Wishbone Project

[Cameron 16]

Hello! Im currently in the middle of a little project to make some tubular wishbones for my 205. I've been posting in my project thread on another site so a couple of you might have already seen some of this. I think its about time I posted on here about it as you may find it interesting, and I could do with a bit of constructive criticism and some fresh eyes to find points I may have missed.

 

So, I guess my project starts with a look at the standard wishbones. I took the 309 GTI wishbones off my 205 and set about measuring every part with my vernier so I could make an accurate CAD model of them. In places where the dimensions were slightly obscure, such as corner radii and some thicknesses I have been generous with the dimensions in hope that if the model differs to the original it should only be stronger.

 

Here's an image of the CAD model:

 

 

A couple of details have been left out, the ball joint being the main one, and a small upward kink near the balljoint.

 

Next, analysis of the standard wishbone. The wishbone appears to be made from forged steel, so would most likely be made of medium carbon steel with a yeild strength of just over 500MPa. In trying to figure out the loads I have made a few assumptions, and it is here that I could do with a bit of guidance. Firstly the loads carried by other components (eg, strut) have been ignored, and load is assumed to be carried by the wishbone only. Secondly, I haven't got to the stage of shock loading and fatigue so this is just for normal loading.

 

I ran some FEA simulations to find the maximum load the standard wishbone could take. I started with a 500kg load at the balljoint pin, as if the car was braking. That equates to roughly 1.1G deceleration if the car weighs 900kg. Again some assumptions are made here, 1 - that the weight distribution is 100% front, 2- that the weight distribution is 50% Left / Right.

 

The results:

 

Stress distribution:

 

Location of peak stress:

 

As you can see the maximum stress is about 314GPa, which equates to roughly a 30% safety margin. (314 / 500) * 100

 

Then I ran a few more simulations with different loads to find the maximum braking load it could carry.

 

 

The load in the image above was 750kg, and a maximum stress of 512GPa which exceeds the yield strength of the material. So it seems a load of 500kg would be a ball park figure.

I could do with a couple of bits of info though; firstly, does 1.1G braking seem like a reasonable number? I tried to find a figure, and it seemed that most road cars would struggle to reach 1G, but obviously a race prepared car could well reach higher levels. Secondly, what sort of cornering G would the car be able to manage? In the next part I have used 500kg again (1.1G assuming 60% front weight distribution). I'll post the details and results of that later, being on a Mac I have to switch operating systems to run Catia.

 

Also, could a mod put this in the projects forum? It might be better suited in there.

 

Cheers.

[Anthony 1,003]

How does the sort of loading the wishbone is subjected to in an "impact" compare, as I would assume that would put significantly more load through the wishbone that braking or cornering? By impact, I mean things like speedramps, potholes, and curbs - stuff you'd expect cars to deal with on a frequent basis - rather than hitting armco and other vehicles.

[GLPoomobile 958]

Whoa! From a technical level this is all way over my head! But even as a complete numpty I can appreciate the detail that has gone in to this.

 

It's just nice to see someone demonstrating that they've properly researched the product rather than just welding a bunch of pipes up to the correct diameters and thinking it's fine because they trust their welding. And that is not a snide comment aimed at anyone

[Cameron 16]

How does the sort of loading the wishbone is subjected to in an "impact" compare, as I would assume that would put significantly more load through the wishbone that braking or cornering? By impact, I mean things like speedramps, potholes, and curbs - stuff you'd expect cars to deal with on a frequent basis - rather than hitting armco and other vehicles.

 

Well, I don't think those sort of "impacts" really put as much stress on the wishbone as cornering and braking. If you imagine how the whole system works the wishbone is essentially just a tie-rod in that situation, the majority of that bump load is carried by the strut and the wishbone just stops lateral movement. It will see some loads but I don't think they are as high as in braking and cornering (lateral loads) where much more of it is focused at the wishbone.

[JamesLumley 4]

Nice research Cameron, wish I could draw stuff like that in CAD!

 

Question - are std wishbones not cast?

 

I'm sure that the price they are would not warrant forging them.

 

If so what does this do to the loading forces?

 

From my experience when rallying - wishbones are quite brittle and will snap clean off if impacted at certain angles, when examining the snapped ends, the metal looks cast to me - could be wrond mind!

[omega 82]

i thought they were cast also,looks nice

[Cameron 16]

Oh really? To be honest I don't know. At first I thought they were cast but when I examined them they seemed to have some tell-tale forge marks so I thought I'd be cautious and assume they were. I tried to search for some info but couldn't find anything other than something saying the 106 rallye had forged wishbones. I thought that if they had bothered to forge wishbones for that then they may well have done it for the GTI.

 

Does anyone know weather they are forged or not?

 

That's interesting about the cast wishbones breaking like that. Does it happen quite often in rallying then?

Edited November 25, 2008 by Cameron

[B1ack_Mi16 67]

I'm pretty sure they're forged as the parting line seem quite thick, also casting would be much more brittle.

 

See you are using Catia here

The stress concentration you see there is what it is, a concentration, most likely the stresses you see there are more than what's the case in real life.

 

Also remember wishbones are coupled with the bushes which will reduce peak forces quite well.

 

About the load scenario I think it's quite good, been doing some analysis on wishbones earlier myself, and the way to go is to first determine the G-forces involved and take it from there.

 

Proably max 1.2-1.3 G with very good tires, and like 85% weight on front.. as I've never really seen the rear to go air-borne either.

 

You could test cornering too.

 

Maybe you could send me the catia file?

[welshpug 1,658]

I've never really seen the rear to go air-borne either.

 

I have!!

 

Quite common on rough surface and tarmac rallies for the rear to come up over severe bumps, even in 4wd cars!

[swordfish210 20]

Is that Auto CAD you're using there?

[Cameron 16]

Narp, its Catia.

 

Update will have to wait for tomorrow as I ran out of time.

[swordfish210 20]

ah ok. We have to use Solid Works at uni and once you pass the tuition stage it's pretty wank. It wont let you do some things (like butt sheet aluminium togeather ) and the COSMOS software on it is fairly poor as well as it wont let you chose a loading direction.

[B1ack_Mi16 67]

I have!!

 

Quite common on rough surface and tarmac rallies for the rear to come up over severe bumps, even in 4wd cars!

 

Surely, but if the rear comes up because of uneven road surface it does not mean that 100% of the weight are on the front wheels, which is the important factor when setting loads on the analysis.

If rear came up on perfectly flat tarmac because of very hard braking it would be more of what I was thinking of.

[Cameron 16]

Ok, so here's the rest of the first part of the project.

 

Ran a couple of analyses on the wishbone under cornering, first with a 500kg load:

 

 

The peak load in this is 165MPa which is pretty low, so upped the load to 750kg:

 

 

The peak load increased to 247MPa which is pretty reasonable and within the strength of the material, so it seems the wishbone is far stronger in this direction.

 

So in summary it seems the wishbone has been designed to take roughly 500kg braking, with a fairly reasonable factor of safety (30%), and at least 500kg in cornering. I think for the new wishbones I will try to make them capable of withstanding 750kg + safety factor so that they are nice and strong. Which moves me on to the next part.

Edited November 26, 2008 by Cameron

[Cameron 16]

I've gone through several different designs but I won't worry about showing those, I'll just show the latest one as it is the strongest and simplest. I have decided to make the wishbones out of 4130 tubing as it has a very high tensile strength of over 800MPa and is well suited to components like these. This means the tube can be relatively thin walled and save weight.

 

Here's the design:

 

 

Handily its within the boundaries of the standard wishbone too which means interference with the ARB shouldn't be a problem.

I've modelled the rose joints to try and suss out weather they would be strong enough as they use the standard bolt sizes. Front being a 10mm joint and the rear being a 12mm joint. Lucky I did really as you'll see below.

 

 

Above is the wishbone with a 500kg braking load, the location of peak stress has been hilighted and as you can see it is the rose joint that is the weak point. The stress seen here is 828MPa which far exceeds the strength of the rose joint (bearing in mind they are just mild steel). So larger joints are called for, or at least joints with a larger diameter thread. I'll re-design it with some 14mm joints and see how that helps.

 

The next simulation is with a cornering load of 500kg.

 

 

The wishbone is far stronger in this direction, like the standard one, and has a peak load of 104MPa which is well within the limits - equates to a safety factor of 66% (since the peak is in the rose joint) which is really good! Still, a larger rose joint will be used for safety.

 

So I'll be re-designing that wishbone slightly so that the rose joints are a larger diameter and then I'll try again. Will hopefully get on with that tomorrow.

[number2301 1 1 Cars]

Unfortunately I'm lost when it comes to the technical stuff but your design looks an awful lot like a 106 wishbone. That might be something to look into for comparison?

[base-1 17]

Glad to see you've modelled it with the rosejoints, remember the boss for them will need to be pretty damn hardcore aswell, will be interesting to see how it fares with the M14 joints.... will the (assuming cast) joint bodies be as strong as the material you're using there though I wonder? You might need to go up to say M20

 

And wishbones do indeed get destroyed on rallies as James said. At least these will be more likely to bend than snap than a cast equivalent, can get back to service on a bent 'bone and swap it to carry on

[welshpug 1,658]

IIRC its 3/4" joints that are used on the outer ends of the modified arms you can get, which is 19mm, so M20 as mentioned would probably be the most suitable size to use.

[Spiky 37]

so you not going to run an ARB with these then?

 

love all the technical info on here. keep it up

[Cameron 16]

I think M20 might be a bit overkill, but I'll see how I get on. The inner ones should hopefully be ok with M14. The bosses will be fairly substantial, they're 30mm deep in those images, that's about half the thread depth so should be ok.

 

Yeah I will be using the ARB on mine which is why Im designing it like that. If I wasn't then I could just make it A-shaped and save a whoooole lot of effort! I'll look into attaching the roll bar to the wishbone too which could solve any clearance problems.

[Spiky 37]

well some advice from me,

 

i'd measure the size of your wishbone and how far it sticks out from the subframe, and how close it gets to the ARB

[Cameron 16]

That's what I've done.

I tried to keep the wishbone within the dimensions of the standard one so there shouldn't be any problems.

[Cameron 16]

Done a little more work tonight, I made a new design that uses 14mm rose joints for the inners, and an 18mm for the outer. It also has smaller diameter tubing as an experiment. There is now a brace between the Y section to try and remove some of the load from the inside of the bend and from the rear rose joint.

 

The design:

 

 

Then ran some simulations, first with a 750kg braking load:

 

 

 

Now there's a problem as the max stress is now 1.16GPa which is a fair bit over the material strength (about 40%) so I think I will have to revert to the original material size. Also under braking the stresses in the rear rose joint were too high, so I may have to try using a 16mm one for that. The front is happy with a 14mm.

 

It does much better in cornering though, same load of 750kg:

 

 

 

Peak stress was only 94.6MPa, all of the rose joints were fine.

 

So I'll have to try again, again. Will try a few things before increasing the size of the rose joints as that will start seriously driving up the cost!

[welshpug 1,658]

Here's a pic of a 306 RJ wishbone, note rear mounting position on the subframe has been modified, ARB also attaches to wishbone.

 

 

I'm sure I have one for a 205 somewhere, very similar, ARB again bolts to the wishbone with a short link.

[Cameron 16]

Cool, that's a handy pic!

I think attaching the ARB to the wishbone could be a good idea, an A shape wishbone would be a lot stiffer. Not sure about modifying the rear mount position though, I'd like to try and keep that standard so I have a better chance of being able to sell some once they're proven.