Fatigue Testing

[Denny Trimble]

UW had a fun little failure today. The lower front a-arm tube failed under braking, the rodend and insert pulled out, and the upright proceeded to wrap itself around the upper a-arm, pullrod, and tierod. Inspection of the right lower a-arm showed a crack in the same place.

This car has about 180 hours on it, and the tube that failed is 5/8x.035 4130 (normalized, not hardened!). There was about a 1.0" moment arm from the rodend center to the point where the tube failed, which is right at the root of the a-arm tube joint. Also, the WWU guys pointed out that the rodend in bending was not the failure mode http://fsae.com/groupee_common/emoticons/icon_smile.gif

http://students.washington.edu/dennyt/a-arm_2.jpg

http://students.washington.edu/dennyt/a-arm_1.jpg

Redesign, manufacture, and better heat-treatment are happening in the near future http://fsae.com/groupee_common/emoticons/icon_smile.gif

University of Washington Formula SAE ('98, '99, '03, '04)

[Denny Trimble]

UW had a fun little failure today. The lower front a-arm tube failed under braking, the rodend and insert pulled out, and the upright proceeded to wrap itself around the upper a-arm, pullrod, and tierod. Inspection of the right lower a-arm showed a crack in the same place.

This car has about 180 hours on it, and the tube that failed is 5/8x.035 4130 (normalized, not hardened!). There was about a 1.0" moment arm from the rodend center to the point where the tube failed, which is right at the root of the a-arm tube joint. Also, the WWU guys pointed out that the rodend in bending was not the failure mode http://fsae.com/groupee_common/emoticons/icon_smile.gif

http://students.washington.edu/dennyt/a-arm_2.jpg

http://students.washington.edu/dennyt/a-arm_1.jpg

Redesign, manufacture, and better heat-treatment are happening in the near future http://fsae.com/groupee_common/emoticons/icon_smile.gif

University of Washington Formula SAE ('98, '99, '03, '04)

[Brent Howard]

Could you posta picture of the other wheel so that everyone can see the wheel assembly before it broke?

Brent

www.ucalgary.ca/fsae (http://www.ucalgary.ca/fsae)

[Denny Trimble]

http://students.washington.edu/dennyt/a-arm%20before%20failure.JPG

This is the same view of the same corner. The failure occurred where the little white arrow is, just outboard of the weld.

University of Washington Formula SAE ('98, '99, '03, '04)

[ben]

A rod end in reverse cyclic bending.

May I ask you what you expected to happen?

As I mentioned in a previous post - this stuff matters. Make the wishbones again, and use a spherical bearing rather than using a bar full of stress raisers.

Ben

University of Birmingham
www.ubracing.co.uk (http://www.ubracing.co.uk)

[Jarrod]

the rod end didn't break, but just as bad as a rod end in bending is a thin weld bead under serious bending stresses. If you cut the tubes at an angle to the threaded plug and increase the cross section of the weld it may be better, but sphericals are a much better solution.

[alfordda]

How was the plug welded in? Was it just butt welded at the end of the tube, next to the jam nut? If so, welds aren't very strong in tension. You could try notching the tube like Jarrod suggested, or, using what I believe, is called a rosette weld. Basically, drill a hole in your tube before the plug is inserted, and then fill that hole with weld, connecting the plug and the tube though a weld that is in shear.

[Denny Trimble]

It wasn't the rod end, and it wasn't the weld, but yes, we've already redesigned it for a spherical bearing, slightly larger tubes (higher stiffness is the main goal there) and a better heat treatment.

There's about 1/8" of tube between the tube-joint-weld and the insert weld (which is only welded around the end of the tube). The tube failed in between the welds; it was just too small, with too much of a bending moment.

I was surprised when last year's guys kept the rodend in bending design, and it's sure not to make it into the '04 car. It was one of the things the judges didn't like in the semi's.

In the picture below, you can see the rodend and insert dangling near the brake line, and the lower a-arm fracture surface.

http://students.washington.edu/dennyt/a-arm_failure_3.jpg

But Ben, I have to ask, how many hours are on your most recent car? Any interesting failures you'd like to share with the group?

University of Washington Formula SAE ('98, '99, '03, '04)

[Brent Howard]

Are those tubes that you are using chrome moly? If so the reason behind the brittle failure is likely chromium carbides in the HAZ. The area just to either side of the weld will stay within the correct temperature range for teh right amount of time and the formation of chromium carbides can occur. These lessen the amount of slip planes within the material, resulting in a brittle material and brittle failures. Replace the other A-arm as well as you won't be able to really detect an impending failure on the other A-arm either.

Brent

www.ucalgary.ca/fsae (http://www.ucalgary.ca/fsae)

[Denny Trimble]

Brent,
The parts are 4130, and they were "stress relieved" after welding, at a professional heat treating shop (not the oxy/acetylene torch method). Also, 4130 welding rod was used. Do you still think chromium carbides are the problem?

And yes, the other a-arm is cracked in the same place.

University of Washington Formula SAE ('98, '99, '03, '04)

[Brent Howard]

I think that they could still be a problem as stress relieving a weld will not dilute the concentration of carbides in the HAZ because you are not heating it to the point where the metal can diffuse. Let me check one of my text books tommorow (at work of course) and I'll let you know more about them, how they form, and what you should change to make it better.

Brent

www.ucalgary.ca/fsae (http://www.ucalgary.ca/fsae)

[Brent Howard]

Hey Denny,

I have a paper that I wrote about weldment corrosion in stainless steel that discusses carbide formation, but only really talks about how it affects corrosion. If you want it let me know your e-mail and I will send it.

Basically the carbide that I think is forming is Cr23C6. This could be caused by carbon in the atmosphere around the weld (oily rags, touching the weld area, etc...). Basically what is happening is Cr is being grouped together to form much larger molocules which lessen the slip planes and then you end up with a much more brittle material, one which is stronger but not able to take the fatigue loading (as you saw). Using a lower carbon steel should also help.

Let me know if you want my paper...it isn't exactly your situation, but I think it applies.

Brent

www.ucalgary.ca/fsae (http://www.ucalgary.ca/fsae)

[ben]

[QUOTE]Originally posted by Denny Trimble:
But Ben, I have to ask, how many hours are on your most recent car? Any interesting failures you'd like to share with the group?
QUOTE]

Point taken that the rod end didn't actually fail, but I think a spherical installation done right has better load paths and is less likely to fail in the welds as well.

As for failures, nothing structural on the suspension, wishbones are all 5/8" mild steel 1.6mm wall thickness with sphericals at all the joints. Previous car had 0.9mm wall thickness and has done a competition and a year of testing with no structural problems eiher, we did loose a wishbone on that car in an accident, but a spherical popped a ball out before any welds failed.

We did bend an in-house brake balance bar because it was made out of a piece of shit bar.

Ben

University of Birmingham
www.ubracing.co.uk (http://www.ubracing.co.uk)

[Denny Trimble]

Ben,
You're right, the spherical installation has better load paths and is stronger. Sorry if I seemed to defend this design; it's not the best.

But, interestingly, the previous car (2002) had the same design, same tube size, but probably not as many hours on the car, with no cracks in sight (yet). We'll keep inspecting it, but I'm wondering if this is a welding/heat treat problem like Brent has mentioned.

University of Washington Formula SAE ('98, '99, '03, '04)

[madhay]

Hey guys,
I don't mean to jump in and beat a dead horse but it sounds like it is in aggreament that the design is not the best. (Weve done it also). We have never had a failure but we have put maybe a fourth the amount of drive time also. I would suspect that the stresses at that point are going to be well above the fatigue limit so the failure is bound to happen. I also don't think that infinite life should be a goal on many of our components either. The vehicles would weigh a ton if we did! I don't have much of a point but I just joined the forum and I got excited about posting something

Good Luck!!! http://fsae.com/groupee_common/emoticons/icon_smile.gif

[Mechanicaldan]

The reason it broke is because the tire was trying to rotate itself back to the left rear. A good lesson for us all to learn. Never mount rear tires on the front, especially on the wrong side of the car. http://fsae.com/groupee_common/emoticons/icon_smile.gif

Cyclone Racing
www.cyclone-racing.com/fhome.htm (http://www.cyclone-racing.com/fhome.htm)
Iowa State University
Project Director

[Sam Zimmerman]

Thanks Dan. http://fsae.com/groupee_common/emoticons/icon_smile.gif I will pass your words of advice to our team. There might even be one or two that I could convince to check the tires before we drive to make sure they don't cause accidents like these. http://fsae.com/groupee_common/emoticons/icon_biggrin.gif

Sam Zimmerman
Vandals Racing (http://www.uidaho.edu/~racing)

[Brent Howard]

Ummm, it is the left front Daniniowa.

Brent

www.ucalgary.ca/fsae (http://www.ucalgary.ca/fsae)

[gug]

wow, im now scared of the same problem happening to our car! Dan, do you have the name of a good tyre psychologist, to make sure none of our tires are having displacement issues?

and do you know the name of an engine psychologist, cause we had a valve who thought he would do better living in the piston...

- if it isnt coming, you need a bigger tool.

[Mechanicaldan]

Hmm, yea, I goffed on that one. It's the left rear, in the picture, not on the car.

Why is the tire marked for left rear? Is there a concern for cord seperation since it is on the wrong wheel? Or, is the tire marked wrong?

Goodyear specifies the tire location based on where the serial number is on the tire. Both front have to have the serial number on the same side, and opposite side for rear, because of forces generated during acceleration and brakeing. We have a set that was not heat cycled properly, and yes the cord did seperate because the tires were on the car wrong. It was a competition move because we had rim clearance issues with our uprights.

Cyclone Racing
www.cyclone-racing.com/fhome.htm (http://www.cyclone-racing.com/fhome.htm)
Iowa State University
Project Director

[Denny Trimble]

Those are our cheap driver training tires, Toyo RA-1's. They were rotated fronts to backs for wear reasons after a day of dirt-trackin' it. A diagonal rotation might have been best, and that's the only thing that works for Goodyears, otherwise they separate as you've seen.

If anyone runs the toyos, we've found 20psi is the only way to get similar slip angles to Hoosiers. This will wear the center of the tread, but the car feels like a Winston Cup car when the pressures are low enough for even temps across the tread.

University of Washington Formula SAE ('98, '99, '03, '04)

[Jackson]

Denny and Brent;

The Cr23C6 is the main reason why carbon is known as "Tramp" in stainless steels.

The formation of these eats up all the Chromium (look at that stoichiometry!) which causes a loss of the "stainless" properties as well as brittleness.

Not sure if this is a major problem in 41xx and 43xx CrMo steels though-- I have little experience with them. I just know that stuff near a weld is always what seems to break. (I do have some experience with that... http://fsae.com/groupee_common/emoticons/icon_smile.gif)

Brian
WSU Formula

[Brent Howard]

Brian,

Yeah, my paper and the research I have done is also on stainless steels (316 SS to be exact), but looking at the failure, a brittle failure near a weld in a materail that contains Cr...I figured it might have something to do with it.

Here's my best guess. The heating caused by the welding caused the area just adjacent to the weld to stay in the 700-750 C range for about 2-5 minutes or so. This picked up some carbon from the weld area due to grease from peoples hands as they set-up the a-arm for welding. Then this reacted to form at least some Cr23C6, which are larger molocules and would form in the brain boundry area. These molocules blocked slip planes that a ductile materail needs to be ductile (the smaller the grains the increased ductility). Then through fatigue it experianced a brittle failure, because brittle materials cannot cope with cyclic loading anywhere near as well as ductile ones.

I am probably wrong in more than one spot, but I think the general idea is correct.
To stop this problem from happening lower carbon steels can be used and also the weld area must be kept free from grease or oil.

Brent

www.ucalgary.ca/fsae (http://www.ucalgary.ca/fsae)

[Jackson]

The "brain boundary" huh http://fsae.com/groupee_common/emoticons/icon_smile.gif

12:58 AM must be your peak operating time http://fsae.com/groupee_common/emoticons/icon_smile.gif

Brian (Well rested and all smart-assed up)
WSU Formula

[Brent Howard]

I actually meant the Brian boundry...sorry, you must get that messed up all the time by people.

Brent (at work and slacking)

www.ucalgary.ca/fsae (http://www.ucalgary.ca/fsae)