Strange Non-Stock swaybar configuration and front suspension issues

[autox320]

By all means man been meaning to PM you about it and not just stab on the forum. No harm I promise; PM if want to chat, i'd even give my number cause sure we'd talk shop.

[Retoropak]

I agree. Its not that the 7/8” is not strong enough. It will be the fatigue from the stress concentrations at the threads that will eventually cause it to fail.
It will take some time for the cracks to develop and even when inspected after every race session you might not be able to see it without magna fluxing it.

FEM analysis would show concentrated stresses at the both crests and roots of the threads. I mean stresses probably will be below the yield stresses of the material however repeated loading might work harden some areas which might promote formation of the cracks.


I decided to go with the cup and a bearing for that reason.



Note that with the cup and bearing solution you need to be really careful while welding it.

Most of the cups are chromoly or other high alloy steel that might need to be normalized after welding.

I suggest inserting the brass or copper bar the size of the bearing into the cup before welding to act as a heat sink. That way you reduce Haz ( heat affected zone) on the cup and wont have to normalize it.




Sent from my iPhone using Tapatalk

[Somjuan]

Clever with the bar there on the HAZ! Cool to see another FEA guy in here, what do you normally use it for? I always hate verification and validation in threaded rods. Singularities everywhere.

More on topic - The bar is a beefy chunk the way you have it made, but they are right about long term viability concerns. Just probably on the order of millions of cycles mixed in with harder stops and starts. Just something to be aware of. I’d hate it if someone who put so much work into design and fab got injured because of something like that. I say run it as is for a while and run a study on the sucker. Make two so you can swap it out and see for yourself.

[Erik Dlux]

I guess I just dont get it. I dont see any possible way for it to fatigue.

The stock E21 control arm is smaller and the material is weaker and the stock sway bar comes in at a further distance creating more stress. How could you imagine a ball joint that is bigger and stronger would fail before before the stock control arm? Maybe if somebody can get me past that I can understand....

Also, Autox, how long is the stickout on that bolt that goes into your strut and what diameter is it? I would guess that sees way more stress than what we are talking about with it mounted in that fashion?

- - - Updated - - -

Autox, I think thats youre previous setup that you have above in the pics that you say "dont do this"?

If it is, did you run that setup? Did the heim bend? What size was the heim?

[Erik Dlux]

Retoropak, what size tubing did you use that butts up against your uni ball?

[Erik Dlux]

Autox, same question as I asked Retoropak.. what size tube are you using for the control arm and what wall thickness?

I am not sure but I am guessing the tube you guys have might only be as strong as the tie rod I am using. If not, I bet its close in strength. Could be even less strength than my tie rod though too, which really makes a point IMO.

[autox320]

I guess I just dont get it. I dont see any possible way for it to fatigue.

The stock E21 control arm is smaller and the material is weaker and the stock sway bar comes in at a further distance creating more stress. How could you imagine a ball joint that is bigger and stronger would fail before before the stock control arm? Maybe if somebody can get me past that I can understand....

Also, Autox, how long is the stickout on that bolt that goes into your strut and what diameter is it? I would guess that sees way more stress than what we are talking about with it mounted in that fashion?

- - - Updated - - -

Autox, I think thats youre previous setup that you have above in the pics that you say "dont do this"?

If it is, did you run that setup? Did the heim bend? What size was the heim?

True most control arm ends are cast steel, but they don't have threaded ends at the hub that's the main difference. Factory ball joints at the hub then take the brunt especially brake force when combined with large slick tires and high brake torque compounds. In this scenario BMW's of all ages have seen high ball joint failure especially noted when e36's hit tracks. It's even in most rule books now that control arms due to the ball joints have to be replaced every 2yrs or sooner depending on events etc. I remember Dave's e21 racer had a serious outer balljoint failure using stock e21 arms once. A few friends have had the outer balljoints snap off e36/e30/e46 arms. It's common when up the ante.

Factory e30/e36/e46 ball joint stem or shaft is tapered but the meat part is 5/8" that it rides on. Our setup bearings are 3/4 with 5/8 high angle sleeves to allow a 5/8 grade 8 bolt. All bolts in our rig are cut to length to keep shank through the sleeves and threads only for the top lock metal nuts. True the bolt in this fashion see's load including the bottom of the strut where it has 4x10mm bolts as E30M3's. Brake forces try to break the bolt (sleeved shank), tear the bolted 4x10mm (e28,e30M) strut casting off the bottom of the strut, and the end of the control arm off. Sidebar on the strut casting that bolts in your case and the m3 btw It's common in Germany to just lay a few weld beads to make the strut one with these. It was only designed to be bolt on due to rules in early dtm racing they made spacers to go under the strut for RC correction which also moved bumpsteer equal amounts. Can fail too so best to weld the piece on and use spacers on individual link ends at the hub. I know, I know right.

Yes top pics of what NOT to do are the early part of our build. Front end heims are 5/8 qa1 endura slot loaders. We decided not to run the REIB setup pictured and it was scrapped soon after(next page I think) in my build thread. A few engineer buddies also wouldn't stop nagging me about it It cut into our beer drinking time in the shop when peeps looked over the rig in progress. Including posts with small talk to Retoropak (thanks man).

Tubing on our rig control arms are moly 7/8 with .156 wall I believe but don't quote me. The M3 arms are 1" moly .120 wall. The gusset around the front edge of the cup including to the clevis mount strengthens a lot. That's the part I'm suggesting could just add a gusset steel from front edge of the ball joint to the clevis mount tab making one piece. But that would warrant the ball joint hard to service unless can press a new on in that same casing.

To the OP; apologies of straight up hijacking your thread. Owe ya beers or something.

[Retoropak]

Ill add that the ball joint shaft is strictly in tension and shear it doesn’t see any bending.

Stresses in the “spindle to arm”bolt due to the tension or shear might not be as severe as stresses due to the bending also the distribution of these stresses might be different. Of course geometry and point where the load is applied plays a major role in the equation.

I guess with threaded rod big enough you can minimize the stresses to the point when you wont have any problems.

Maybe one of the reason you don’t see it done in the racing applications is the added weight.


Either way custom arms should be inspected on a regular basis to avoid any unpleasant surprises.


Sent from my iPhone using Tapatalk

[Erik Dlux]

So, threaded does not make things weaker, its the minimum diameter that makes it weaker. Say my 7/8 threaded end has a minimum diameter of .800. It will have the same strength as a .800 non threaded piece. Having said that, if you have a control arm that is .600 and my 7/8 minimum diameter is .800, the 7/8 will be stronger because its bigger. Not to mention the quality of the material which means a ton.

Of all these ball joint failures, I am guessing that they are all of the actual ball joint and none of the actual arm, correct?

So, youre using a 5/8 bolt. Thats a shear strength of about 20k. On top of that, its single shear but not only single shear but also single shear with a unsupported length of what, 1.5 or more inches? That is a stress riser for sure also.

Mine is a 7/8 bolt. Thats shear strength of about 42k (over double your 5/8 bolt).

Ok, so even in chromo your tube is significantly weaker at around 25k shear strength. I agree that the gusset strengthens it but just making a point.

[autox320]

So, threaded does not make things weaker, its the minimum diameter that makes it weaker. Say my 7/8 threaded end has a minimum diameter of .800. It will have the same strength as a .800 non threaded piece. Having said that, if you have a control arm that is .600 and my 7/8 minimum diameter is .800, the 7/8 will be stronger because its bigger. Not to mention the quality of the material which means a ton.

Of all these ball joint failures, I am guessing that they are all of the actual ball joint and none of the actual arm, correct?

So, youre using a 5/8 bolt. Thats a shear strength of about 20k. On top of that, its single shear but not only single shear but also single shear with a unsupported length of what, 1.5 or more inches? That is a stress riser for sure also.

Mine is a 7/8 bolt. Thats shear strength of about 42k (over double your 5/8 bolt).

Ok, so even in chromo your tube is significantly weaker at around 25k shear strength. I agree that the gusset strengthens it but just making a point.

I'm not disagreeing with you the 7/8 is strong enough. But thats part of the point. The 7/8 is larger and heavier than needs to be to avoid bending. A lighter welded cup bearing or ball joint will do the same job without having the end in bending load.

The main arm tube is gusseted with plate steel significantly changing how much shear strength there is. The entire gusset and tube has to shear in that plane which is more than a heavier larger rod hanging past the clevis mount point. Would have to shear in our rig's case a chunk of 7/8 tube and 3/8 plate gusset.

The shear of the 5/8 bolt is encased in spacer sleeves making it 3/4" thick diameter.

Yes failures of factory ball joints are the joint or stem itself. Because the arm is solid. I've seen both ball shatters and some with the stem shank broken clean off shear. Single shear yes that's how factory is as well unfortunately.

I feel like we should be drinking beers discussing this stuff in the shop.

[Erik Dlux]

The weight isnt an issue IMO. I just weighed a spare tie rod and complete with castle nut, cotter pin and jam nut is 1.2lb. I dont think your overly long bolt, uni ball, nut, larger gusset could weight less, maybe even more?

Well, I guess if you get to count that your 5/8 bolt is encased in a 3/4 sleeve, I get to count that my 7/8 threads are encased in a jam nut that is 1.25" diameter? Either way, I would put money on the 7/8 rod holding up over the 5/8 rod. Its 2x the strength... None of this really matters though, but my point is that I think if you look at mine and say its weak, I think you need to take another look at yours. On top of that, I am guessing yours is getting results so does any of it really matter? I dont think so. I am still stumped that you think my tie rod could have an issue.

Yeah, beer in a shop sounds good. I hate text. Its always way too hard to understand the true meanings.

[captain awesome]

Every time I see you guys throwing up pics of the custom control arms, it makes me want to go down that rabbit hole. Then I come to my senses, and realize I'm probably not that great of a driver. At least one that is capable of seeing whatever benefit they may have on the car. Maybe when I tear down next winter I'll revisit it though.

[Retoropak]

Quick simulation of the 7/8” rod 1020 mild steel in pure shear and in bending. 2000lb applied in the y direction


Yield strenght of this material is 30,000psi
Ultimate tensile strength is 50,000psi
This is the shear in the zx plane



2500psi max stress at the fixed end of rod in bending that would be probably at the jam nut location



This is the shear in the xy plane
2710psi max







Shear stress in the zx plane as expected was larger than other planes at 7500psi





So shear forces are relatively small however they are mainly concentrated at the jam nut location.


Now the von misses stress is quite different.

For the rod in bending it exceeded yield strength of the material and almost reached the ultimate strength. Maximum stress was 44,0000 psi at the jam nut location



Rod in shear did much better with only 1500psi where the bearing pressed the bolt.























Sent from my iPhone using Tapatalk

[Erik Dlux]

I dont understand your fancy program nor your nomenclature so I cant respond. LOL.

Are you comparing his control arm tube (7/8") to something?

[Retoropak]

I am just trying to show that given same amount of force the rod put in bending develop much higher internal stresses than the same rod in shear. It means that your ball joint shaft might not the be the weakest link any more it might be the threaded rod end.




Sent from my iPhone using Tapatalk

[Erik Dlux]

Take a step back again....

Do you think that my .875 rod is weaker than the stock E21 cast piece at .640?

Looks like the VACmotorsports model is using either a 5/8 or 3/4 rod end. Give them the benefit of doubt and say they are using the 3/4. Still smaller and weaker than what I am using. They have been selling these for a long time to my understanding. Do we assume that they are failing? Did they upgrade the kit to something else but never change the pictures? Do they not sell any? David Mcintyre claims to have been running it for years with no issues. He said he was the first to run it.

[Somjuan]

Are you running that in a blend of tets and bricks? I think you might be getting an error at the base due to the modeling constraints in the program. I mostly run in Ansys 18.1

[autox320]

paper I reference in my build thread from Aurora

http://www.aurorabearing.com/pdf/rod-ends.pdf

[Erik Dlux]

Theories are theories . You could sit here all day and make stuff up about how it could break.

What about real world results?

As posted above, look at the vacmotorsports, kit and explain to me why it works when your theories say it shouldn't.

[Erik Dlux]

If you can't/don't want to explain the vacmotorsports kit, then what about the repeated question of the stock arm being weaker? How does that hold up?

An even better one? What about skis build where he cut the stock arm and welded it back together?! Lol.

[autox320]

Theories are theories . You could sit here all day and make stuff up about how it could break.

What about real world results?

As posted above, look at the vacmotorsports, kit and explain to me why it works when your theories say it shouldn't.

The vac kit is also in bending. They ignore the fact of the end is bending load completely. Also they are using approx 1/2 heim which IMO is undersized. Just because they "made something up" and retail to racers as a kit doesn't mean they have liability of failures. Doesn't make it best engineering practice. It's typical racing aftermarket where even guys selling bushings incorrectly sized just want your money cause your willing to pay.

Not making stuff up, not theory, it's just facts. Bending load is not best practice. Can you get away with it using large oversized parts sure. Real world results just look for REIB baja sae and Fsae failures. Those guys do it all the time and still debated to this day. Does it work sure until it doesn't.

[Erik Dlux]

I understand the vac kit is in bending. My point is that it's been for sale for a really long time .Do you believe that all or even 1 of their kids are failing?

I agree it is in bending and I agree it's not the best practice but the point becomes moot if it's strong enough . It's like arguing your single shear bolt at the strut . Not the best idea but who cares if it holds up?

The theory I am taking about is the theory that mine is going to bend .Imo there is proof that lesser parts are holding fine. So saying I will have an issue when weaker parts don't is all theory .

To be clear, if I ran e21 struts, I would do the same as you guys . Makes sense and the way you do it is stronger. Since I have e28 and can take advantage of my situation I feel like my setup is more beneficial and think it's silly to lose advantages when I believe it's plenty strong .