That fracture seems to have propogated at the root of a cut thread. Cut threads are stress risers. Get bolts with rolled threads, and keep the threads OUT OF THE SHEAR PLANE.

Over tensioning/torquing will exacerbate the problem with cut thread fasteners.

The weakest point of the bolt is at the root of the threads... less section there..

Having bolts in single shear is usually a no no... but I dont know of a single car that has tie rod bolts in double shear? Alot of real live auto engineers have designed these things.

Maybe you got a bad fastener that was subsequently overtorqued.. then subject to many fatigue cycles

Glad your OK and it can be fixed. The REAL question is what was it like to drive that beast home and on Hooisers? I imagine you still might have been a little tight in the b-t muscle department still after that off.
Steve

Double post

Why does it scare you. I know several people including myself who have run this setup for years, problem free.


Regarding the sheared bolt thats not that uncommon. My mechanic warned me about that possibility when we bumped steered my car. He showed me three cases where that had happened in the past. Although I'm very happy with other ERP products, and use them on my car. I don't use that bump steer kit on my car. With the bent arms on my spindle, Spacers on the rack were all that was required to bump steer my car. That failure going thru turn 8 at Willow could be fatal.


Glad your okay.

Hi Craig,

I'm sorry to hear of the misfortune of your failure -- and the good fortune that you suffered no serious consequences. I sure enjoyed the ride you gave me at Pacific Raceways this summer in that car. I thought I'd chime in on what happened because the study of failures is my business.

According to my partner, a metallurgist, your failure is a classic case of reversed bending fatigue. Not overtorque. Overtorque tends to cause immediate failure (during overtorque) and displays fracture characteristics very different from what's shown in your photos. It appears that only about one fourth of the bolt (across the diameter) remained before final fracture, and that the fatigue had been progressing, perhaps for a long time, as a crack grew in from either side.

Here's the kicker: it's most likely that the failure was the result of the bolt being too loose -- not too tight. If proper torque is maintained in the bolt, it should never be subjected to bending (at least, not sufficient to cause a fatigue damage). The primary loading it sees is tensile. If torque is lost (i.e., the bolt is left loose or becomes loose), the bolt can then be unfavorably loaded in bending sufficiently to grow a fatigue crack. This is the most common mode of failure when wheels fall off a car -- the lug nuts are left loose and the wheel studs break off. What's interesting and different about the bolt in your case is that unlike a wheel stud, your steering bolt appears to have been loaded (back and forth) in only one direction.

Did you happen to keep track of the indexing of the bolt before you removed it? It would be interested to know whether the the final fracture zone was aligned with the axis of the car (which is what I would expect). I would think that the highest stresses would be wheel load forces that would act transverse to the car, thereby starting cracks on the left and right sides of the bolt. But then again, there may be other forces at work that I'm not taking into account.

One thing for sure: I'd check the other side. In fact, I'd replace the other side bolt, just to be sure. Then I'd torque to spec and periodically check the torque.

I also take an interest in this because I have the very same ERP bump steer kit on my 930!

Below are some photos of reversed bending fatigue from the book "How Components Fail" by Donald Wulpi. Looks like a textbook case you have...

http://forums.pelicanparts.com/uploa...1100310553.jpg http://forums.pelicanparts.com/uploa...1100310571.jpg

Look where the failure is in those photos... right at the root of the threads... you can even see the inner diameter of the thread in the pics...where stress will be max..

Overtension+ less section+ some fatigue cycles=failure

Get rolled thread fasteners, and keep the threads out of the sher plane

bending failure can come from the bolt being to loose. or to tight. The photos provided dont show the shank of the bolt...so you comprehensive answer cant really be had!!

when was the last time you checked your tie rod ends?

Just like a paper clip.

Glad you're OK Craig! It could have been much worse.

Craig,

I'm glad things turned out as well as they did considering your speed and track location.

I was coming through Turn 8 and saw a RED flag waving furiously. I rolled slowly through turn 9 to see who was involved. When I saw your car sitting sideways I was expecting the worse.

I pointed Rob at this thread because I knew he would be interested both personally and professionally. We all have the same ERP Bump Steer kit and Rob and I both run 275/40 Hoosiers up front. That size tire has to exert even more force compared to what you were running.

Dr. Rob, great info!

I guess I'll be adding one more item to my winter projects list...

Lyall

TimT,

Failure appears to have emanated from the thread root, just as one would expect. But reversed bending fatgue (which this certainly was) is not caused by overtension. It's caused by pure bending, which in this case was likely enabled by a condition wherein the bolt was undertensioned. If there's adequate bending stress (in this case, from wheel loads reacting through a loose bolt), it doesn't matter whether the threads are rolled or cut, or whether there are any threads at all. In fact, if you look at the fatigue photos I posted, they're of unthreaded components. Obviously, the larger the stress riser (a notch, flaw, thread, or whatever), the lower the bending stress (and/or the fewer the number of bending cycles) required to propagate a fatigue crack. Without careful analysis, it won't be known for certain which of many possible factors in this fatigue failure was most significant in raising the bending stresses above the endurance limit of the material, or whether the endurance limit of the material was compromised in some other way. My opinion is that a greatly higher bending load, due to normal operating forces imparted on the bolt while it was undertensioned, is the most prominent culprit here, not the thread profile, material specs, or any of a host of other secondary effects.

That said, I'd be looking very carefully at the other bolt...

This kit has a tapered sleeve allowing a smaller diameter bolt to fit into the steering knuckle shown in the diagram below. The factory part has no sleeve, it is one solid piece of steel. Though still in single shear, the factory tie rod cross section in the shear plane is much thicker than the kit.

The increased length of the bolt creates a moment about the steering knuckle (and base of the bolt) that is much greater than stock. This increased moment combined with a smaller cross-section bolt looks suspect.

http://forums.pelicanparts.com/uploa...1100314544.jpg

Note in the diagram, the threads and nut are on the bottom. On yours it appears the bolt was installed upside down with threads on the top. I don't know how long the threads are, but if they are into the shear plane this could have been key.

Incredible. Great information. Thanks Lyall for having Rob call me - and thanks Rob for the info! THX to Chuck too - come up with something else soon, Chuck.

I'll go look at the other side in the am.

Man that could have been ugly. She-it.

I'd definately check to see if those bolts were put in upside down.

The "sway bar" on my Dodge has a type of ball joint with a long shaft through the axle mount. A mechanic at some point before I purchased it put them on upside down and it sheared both in just the same way... Too much side to side torque at the thread to allow failure...

It was a rush watching from the grandstands as you tore up the gravel and locked em up. I know what you were going through... just a passenger waiting for the madness to end!

We all let out a collective sigh when the car stopped, and Lyall stopped as he saw the red flag. Lucky for you (and Lyall) there was a large crew in the stands, as they flew the red flag only after everyone in the stands screamed and waved at the flagger to do so. It could've been UGLY! Did you get your lottery ticket?

Let us know if anything else is broke! I would be surprised if any other component broke, as these cars are tough as nails!

Later,

2 BIG dents underneath
Scratched (big deal) B&B header heater boxes
Thrashed sway bar sliders and mounts - they functioned as gravel plows ;-)
Dents in 2 fenders from gravel storm
Did I mention I removed a 1X1ft section of concrete and pushed it behind the left rear wheel (see gouged track pic above)?

The other (left, unbroken..yet) side.........bolt appears UPSIDE DOWN.

http://forums.pelicanparts.com/uploa...1100379350.jpg

You must be living right Craig. Glad you and the car survived.

Beyond lucky. The weird thing?
When it was happening, I thought.... "What the h*ll caused that? WHY am I heading in his direction?"

Here is a pic of the bolt pieced together. The end piece (threaded only portion) was on the car, the longer piece was found on the track during gravel cleanup.

http://forums.pelicanparts.com/uploa...1100387790.jpg

Analyze what happened here compared to the stress points in the above photo - Rob, others, -- what do you think?

THX

IMO, the spindle arm probably allowed the bolt to rattle around because the threaded portion is smaller in diameter than the non-threaded portion nearest to the bolt head. I suspect the original design was not for the bolt to be installed upside down. The fracture point was at the lower surface of the spindle arm.

The threads look hammered where the spindle arm bumped around.

The upside down bolt is the cause. IMO. Grr.

Bingo.....

If the bolt had beeen reversed I suspect you would have never had a catastrophic failure. You know what are the local guys are asking ourselves right now.....

Craig,

Glad to hear you survived without a scratch, and the car is reparable.

I've learned a valuable lesson - double check everything to ensure it is installed properly, particularly if the job was done before you purchased the car.

A relatively cheap lesson.