Front Ball Joint Discussion Thread and Potential Safety Concern

[FFRSpec72]

Bill,
Do you know what ball joint is used on the roadsters?

Uppers are from 79-82 Chrysler (Moog 772)
Lowers are from 87-93 Mustang (Moog 8259)

[Jaime]

It doesn't matter if it's cantilevered. The only thing holding the hub from floating off the LCA is the ball joint. If you connect the LCA to the hub, that connection will have to have a ball joint, since the hub moves in so many axes relative to the LCA. If effect, all you would be doing is making an alternative ball joint mount - not that it would be a bad idea, just an not a different idea. You would still need to switch to a load bearing ball joint.

[Buzz Skyline]

If you look at post #5 by xxguitarist in this thread, you can see how the ball joint in the WRX is designed to be used. The bracket I'm proposing would restore things to the way the WRX ball joint is supposed to be used.

As things currently stand, the entire weight of the car in front hangs off the ball joint. With the added bracket, the ball joint only locates the LCA, while the weight of the car goes through the hubs, as is true of the WRX.

It doesn't matter if it's cantilevered. The only thing holding the hub from floating off the LCA is the ball joint. If you connect the LCA to the hub, that connection will have to have a ball joint, since the hub moves in so many axes relative to the LCA. If effect, all you would be doing is making an alternative ball joint mount - not that it would be a bad idea, just an not a different idea. You would still need to switch to a load bearing ball joint.

[Buzz Skyline]

This is what I'm proposing, based on xxguitarist's sketch.

sketch2.png

[Junty]

Very interesting thread, I wish I had the conclusive answer - FFR might help? When I first ordered my kit and started with the registration process in NZ, scratch build kit car... I submitted an application to Low Volume Vehicle compliance NZ. They initially raised these same concerns - relating to the push or pull tension on the lower ball joint. This was very quickly concluded by FFR - who have provided LVVCA NZ with detailed drawings and Subaru information relating to the ball-joint. Although I don't have these, I understand that it relates to Subaru building all WRX and STI's same as rally cars with a ball joint design for both push and pull tension. In short although the appearance of the joint is that it is now upside down, it can still hold almost as much force in either direction...
Low Volume Vehicle compliance engineers in NZ are convinced, from the number of images for WRX cars fully airborne with all brakes, wheel and suspension weight pulling on this joint - also under heavy braking with sway bars pulling on one side to push on the inside - I too am convinced that these joints will more than sufficiently do their job for all 818's - just as FFR designed it.

[STiPWRD]

I couldn't swing $5000, but I'm definitely interested in a solution.

It seems to me that it wouldn't be that hard to make a bracket that takes the load off the ball joint and puts it back on the knuckle. Here's where I could see it attaching.

Attachment 34740

Are you proposing at attach the spindle to the lower shock mount with an L-bracket? Because this would greatly reduce the rotational degrees of freedom of the ball joint and mess up the suspension travel. Even if the L-bracket attachment was able to pivot at the lower shock mount, the ball joint would still be over constrained.

Edit: Just saw your latest sketch, your going to de-couple the lower shock mount from the LCA. That might work.

[Buzz Skyline]

Thanks, Junty. I'd love to see the specs for the Subaru ball joints.

I'd still pay for a different knuckle or a bracket that would restore the ball joint to a non load bearing configuration. Even if these things are tough, it would be nice if failure could be a safer mode than would occur if joint in its current configuration gave way.

(I'm hoping to give the car to my daughter someday to race autocross, so I want to do everything I can to avoid "fail dangerous" designs.)

[Buzz Skyline]

Edit: Just saw your latest sketch, your going to de-couple the lower shock mount from the LCA. That might work.

Yes, that's what I have in mind. The main issue would be how high the perch would have to be to make sure the end of the shock never contacts the LCA.

Of course, I can't actually make anything because I don't have any shop equipment to speak of. But I would buy them if one of our enterprising vendors would make them.

[Jaime]

This is what I'm proposing, based on xxguitarist's sketch.

sketch2.png

Why would your new brackets not need ball joints themselves? You can't simply bolt something to both the LCA and hub because they rotate in two different planes relative to each other.

In a MacPherson design, the body rides on the hub on a load-bearing ball joint at the top of the strut and is located at the bottom with a non load-bearing ball joint. In a wishbone design, the LCA hangs from the lower load-bearing ball joint and the top of the hub is located by a no load bearing ball joint. In both cases, the tie rod connects to the hub with a ball joint. Everything that connects to the hub needs to connect via a ball joint and exactly one of those ball joint needs to bear the load.

[Buzz Skyline]

Why would your new brackets not need ball joints themselves? You can't simply bolt something to both the LCA and hub because they rotate in two different planes relative to each other.

The ball joint is still there in the same location, but the shock is now attached to the hub, as is true of McPherson struts. It's not a great sketch, so it may be hard to tell that I've deleted the bracket that connects the shock to the LCA. This might be a bit clearer
sketch3.png

The first sketch is effectively the WRX set up. The second is the 818, and the third is what you would have with the bracket I'm proposing. In both the first and third sketch, the ball joint only fixes the relative position of the LCA and the hub (and carries very little force). In the second case (i.e. the 818) the ball joint feels the weight of the whole car pulling on it.

[Jaime]

But, what about the ball joint that attaches the shock to the hub? What about the side-load that is created on the lower ball joint that should be about 60 or 70 percent of the magnitude of the vertical load?

[Buzz Skyline]

The ball joint at the end of the shock is designed to handle the forces that we're talking about. Changing from a LCA mounted bracket to a hub mounted bracket won't change the forces on the shock's ball joint.

The ball joint in the control arm (probably) isn't designed to handle high pull-out loads because it would never have seen them in a WRX.

The side loads on the ball joint in the LCA are going to be comparable to the side loads it would have experienced in a WRX, so I'm not worried about that.

[DMC7492]

Okay Buzz your on to something here. I just. Looked over the knuckle and the only time it would be close to the lower control arm is when the front end was off the ground.
The issue is the lower mount of the shock/spring will now make the swing of about 2.5" radius as the wheels are turned and the shock will have to pivot within itself.
The other issue is the springs will both windup or unwind as the suspension compresses and relaxes and also when turning the steering, Toyota had this issue in the early Camrys as both springs were wound the same direction, when loaded it would steering would naturally pull in one direction... More deciphering but the new bracket along with some slight machining to the knuckle would end up being about a new bracket from upper ball joint mount to the lower spring mount. 14 inches by three inches wide and some properly sized spacers to as sandwich everything.
Once again will the shock/spring be able to handle the pivoting action, it does have a spherical mount on both ends to help with the misalignment.
Food for thought?

[Buzz Skyline]

Good points. You're ahead of me now. I haven't had a chance to go out to look at my 818 after realizing there was a potential problem with the ball joint. It's certainly not a simple to solve as I had thought. Hmmm.

[tirod]

Being I am a parts counterman, no, we can't spend days sorting thru all the boxes measuring them to discover what spec's are out there. First, the books don't say. Second, we usually stock less than 5% of what is out there. Auto Parts stores only stock what sells - not what might speculatively work for a one off kit builder in that region.

The MOOG rep does that - when you are talking 250,000 unit's. Not only will they send you a list of what joints might work, they send drawings, too. And if you like, they help design one for your car. It's what they do. You were politely rebuffed as there is simply no money in it for them. C'mon, "I haven't got the specs" is like a Colt rep saying "I can't tell you how long the barrel is on the M16." In the case of the MOOG rep, tho, he's got thousands to choose from, and they generally replicate what other engineers specify.

Use the HSLD Subaru joint and call it good until it does blow up, but use it in a way that promotes taking the load - oriented the way it's supposed to be used. I suspect that fabricating a new spindle would offer other more tangible benefits, like having steering arms set up to reduce bump steer, optimize the Ackerman, tilt the joint to center it's working envelope, incorporate a bracket for larger calipers, get the SAI to work with the scrub radius of wider tires, etc.

There IS a lot of stuff going on up there, if and when someone starts pulling out ball joints then the market will respond with a newer spindle design because demand will increase. That underwrites the expense of development and a small shop will take the plunge. At which point the escalation in 818 racing will have started and the donor runners will be left behind as money starts sorting out who takes first.

Be careful what you ask for.

[Buzz Skyline]

After a little more thought, the bracket I proposed is a TERRIBLE idea. It would couple the shocks into the steering and probably make the car un-drivable. Assuming you could even get it to go where you wanted, a bump mid-corner would rip the steering wheel out of your hands. Ah well. I think we need a redesigned (and expensive) knuckle.

[Buzz Skyline]

Use the HSLD Subaru joint and call it good until it does blow up, but use it in a way that promotes taking the load - oriented the way it's supposed to be used.

It's not possible to use the joint the way it's supposed to be used. It's not supposed to carry any significant load at all. It's not even a good idea to exposed a compressive load bearing joint to tension, or the reverse.

So 818s are going to be running around on critical, but fragile, suspension pieces that will leave your frame on the ground or rip off a wheel if one fails. I can't imagine what that would be like if it happened after hitting a pothole at 65 mph.

What's more, I had hoped to share this car with my son and daughters. While I might consider calling it good until it blows up, I'm sure not going to send my kids off in it.

So, I will continue hunting for a solution I can trust.

[Blitzkrieg]

push-rod suspension.

[Bob_n_Cincy]

push-rod suspension.

Or maybe something like this.
33.jpg

[STiPWRD]

So 818s are going to be running around on critical, but fragile, suspension pieces that will leave your frame on the ground or rip off a wheel if one fails.

That's a bit of a stretch. These cars have been raced with no ball joint issues and I'd have to imagine FFR did some bit of homework before choosing this design. I'm not saying these things can't fail but I just won't believe it til I see it on this car and this particular ball joint.

[FFRSpec72]

What's more, I had hoped to share this car with my son and daughters. While I might consider calling it good until it blows up, I'm sure not going to send my kids off in it.

The sky is falling... I have been racing the challenge car since 2005, no issues, challenge car uses same setup as 818, my daughter races the challenge car also, she will race the 818 and I have no problem in letting her do so

[Buzz Skyline]

The sky is falling... I have been racing the challenge car since 2005, no issues, challenge car uses same setup as 818, my daughter races the challenge car also, she will race the 818 and I have no problem in letting her do so

The challenge car re-purposes a non load bearing joint as a load bearing joint? There's nothing fundamentally wrong with the configuration in the 818. But the joints have different designs and different specs based on their applications.

[Buzz Skyline]

This looks like a load bearing ball joint to me (I got it from this thread http://thefactoryfiveforum.com/showt...er-control-arm)
roadstersuspension.JPG
Do you know the part number for the roadster's lower ball joint?

[Buzz Skyline]

The lower ball joint that comes with the Roadsters is this

http://www.summitracing.com/parts/me...FFwaAnQZ8P8HAQ

It is a load bearing type. I'm not surprised that they don't fail catastrophically. They are designed to be used that way. The 818's are not.

Edit: The roadsters lower ball joint may be this one, also a load bearing design
http://www.oreillyauto.com/site/c/de...0317&ppt=C0106

[Blitzkrieg]

Or maybe something like this.
33.jpg

That would fix the problem also. I don't think it would be to hard to build a push-rod suspension for the 818.

[Buzz Skyline]

That would fix the problem also. I don't think it would be to hard to build a push-rod suspension for the 818.

It sure would. But I don't think I can manage it in my home shop. It would probably cost as much as the whole kit to have someone else do it.

[Frank818]

I don't understand why we compare with racing different cars. How many potholes have you guys hit on a race track?

[Blitzkrieg]

It sure would. But I don't think I can manage it in my home shop. It would probably cost as much as the whole kit to have someone else do it.

The big cost would be the pivot arm but if you did a group buy you would have someone like Boyd make some up. Basicly they are a box of steel with a bearing and you could use off the self adjustable rods.

[matteo92065]

We (my boss and I) are seriously thinking about building front knuckles made for the 818. Not just different Subaru knuckles. And not because we thought the ball joint was a problem, but because, have you lifted one of those things'?!

To fix concerns about the lower ball joint we could maybe flip the A-arms, to put the ball joint into compression.

I would like to start with a knuckle from a car that has aluminum knuckles, with the 5x100 bolt pattern, to use bearings and hub. After a quick search, the 03 350z looks close enough to start with.

Any other thoughts on this?
350 z knuckle.jpg

[Buzz Skyline]

We (my boss and I) are seriously thinking about building front knuckles made for the 818. Not just different Subaru knuckles. And not because we thought the ball joint was a problem, but because, have you lifted one of those things'?!

That would make me very happy! But when you design them, please make sure that they use load bearing ball joints, not the donor car's joints.

[D Clary]

The Red FFR car is going 150+ mph and pulling over 1g. They have been beating the crap out of the blue car for a year. I would hope if it was a problem they would let us know. If you call Moog and expect them to tell you that can use their parts for anything other than their application they will lawyer up and say no. But this is a kit car and everyone has the right to re-engineer or change the design. That been said if you redesign to much the car will never be completed some of the compromise has to be lived with.

[johngeorge]

How heavy is a stock subaru? 3100lb? so if we remove 1000lb from the car doesnt that mean all parts designed for the stock subaru overbuilt for the 818? And thats not even taking in account the engine is no longer on the front axles.

[Buzz Skyline]

How heavy is a stock subaru? 3100lb? so if we remove 1000lb from the car doesnt that mean all parts designed for the stock subaru overbuilt for the 818? And thats not even taking in account the engine is no longer on the front axles.

But the ball joints in question hold almost zero load when used on the WRX. Now they have to hold up 450 pounds while you're standing still, and lots more force when you hit a bump.

Unlike many quality load-bearing ball joints, they have no lube nipples or wear indicators. Even in their designed use on a WRX they eventually wear out. I would guess 80k miles is a reasonable lifetime of a ball joint on a WRX. But they WILL fail at some point, especially with so much more load on them. How long will it take on the 818? I don't know, but I'd bet much less than the lifetime on a WRX. How much less is anyone's guess. Will you be able to tell they're done in before the ball pulls out and you skitter across the track or the highway? Who knows? Why take the risk if you can find a fail-safe alternative?

The bottom line is you're risking your investment (and yourself) on waaaaay under-engineered components. It's like holding up the car with the steering ball joints! That would be nuts. And this is pretty close.

There are lots of things that can be improved on the 818. This is the most important one I can see because so much is at stake in the event of a failure.

Edit: To put numbers on it, when a WRX is sitting still the LCA ball joint holds up the weight of the LCA. Let's overestimate it at 10 pounds. When the 818 is sitting still, the LCA ball joint is holding up a quarter of the weight of the whole car, or about 500 pounds.

When you go over a bump, the ball joint on a WRX has to handle the acceleration of the LCA. When the 818 goes over the same bump, it has to handle the acceleration of a quarter of the weight of the car.

In other words, the 818 ball joint will routinely handle 50 times the loads it sees on a WRX. Even if you believe the components are over designed, are they really over designed by a factor of 50?!?!

[myjones]

Or maybe something like this.
33.jpg

Funny choice for a substitute;
That set-up uses an upper ball joint that is in compression from the impact loading of hitting a pothole so It's the 818 problem turned upside down.
I replaced my 33 uppers with an early mustang joint that was designed for compression loads like the 33 generates. No advice for the 818 sorry.
DB

[MrDude_1]

We (my boss and I) are seriously thinking about building front knuckles made for the 818. Not just different Subaru knuckles. And not because we thought the ball joint was a problem, but because, have you lifted one of those things'?!

To fix concerns about the lower ball joint we could maybe flip the A-arms, to put the ball joint into compression.

I would like to start with a knuckle from a car that has aluminum knuckles, with the 5x100 bolt pattern, to use bearings and hub. After a quick search, the 03 350z looks close enough to start with.

Any other thoughts on this?
350 z knuckle.jpg

Yes I have been looking into it myself. The aftermarket and WRC subaru uprights are not what I am looking for. The odds of me getting to use the Ti-3d printer at work are near nill... so no point designing them here... I keep eying all the aftermarket lotus stuff... but no measurements yet. If the locations are close for that spindle you posted, that could be a better starting point. Whats the distance between the upper and lower?

[C.Plavan]

Yawn..... Yep, put me to sleeeeeeeepppp.

[HCP 65 COUPE]

Dont compare the WRX ball joint to the fox or sn-95 ie roadster lower ball joint the WRX ball joint was not designed to be in tension the
mustang one was, the Fox and sn-95 mustangs have a tension ball joint from the factory the front springs sit in a pocket in the a arms the
struts just locate the top of the knuckle and contain the shock.

I have had 2 of the mustang ball joints fail 1 time in a fox and 1 time in a sn-95 while in motion after hard pot hole hits
and both times the suspension colapsed on its self.
I was lucky both times cause the car was moving slow in a parking lot but if it had been on the high way or on a race track the car would have
ran over its own wheel causing a major disaster. I would suggest to all 818 builders to replace every ball joint with a better or tension designed
ball joint and dont risk your life or your investment over a 15-30$ part its just not worth it ,plus alot of the donor cars have been in an accident
the wrx ball joints may have taken a hard hit and been damaged. DONT RISK IT yours and others lives are at stake.

[HCP 65 COUPE]

Why cant you use the FFR spindle the only compromise I can think of is the need for STI wheels aren't they 5x114
I don't think the .0118 difference in bolt circle would be to much of an issue. I'm not sure about the suspension pickup points but they may be better.

[Buzz Skyline]

How about replacing the ball joint with this
ball joint rod end.png

The high strength one from McMaster-Carr with 1/2" hole and 1/2"-20 threads is rated for 16,000 pounds.

http://www.mcmaster.com/#end-links/=u74bj8

Edit: Forget it. It couldn't take the lateral loads. It's gotta be a ball joint.

[CHOTIS BILL]

I am not sure I fully understand the issue because I am not familiar with these parts but from the pictures it looks like the same setup as used on the roadster and type 65 coupe and I have never heard of a problem on those.

Bill Lomenick