What caused this to fail 1

[karthur]

Would anyone want to list their best guess as to why this shaft failed?

Shear, tension or compression? Was it a cyclic load or an abrupt failure?

What can you tell from just this picture?


Thanks

 

[unclesyd]

Can you post the picture/s on one of the image storage sites so we can get a better view of the fracture surface?

 

[tygerdawg]

pic is very small so the details are almost impossible to see.

GUESS: Based on the different colors I would guess first that it is too much overhung load on the shaft. This causes excessive bending loads at the flange. Cyclic loading during rotation kill fatigue life. Probably a stress riser of some sort initiated a small crack which propagated, then catastophic torsional shear failure across the remaining diameter of the shaft.

It's just a guess based on my experience of the most common failure mode of shafts. Need bigger picture.

TygerDawg
Blue Technik LLC
Advanced Robotics & Automation Engineering

http://www.bluetechnik.com

 

[monkeydog]

Yes very bad photo.
Can you take a close-up of the fracture surface?

 

[karthur]

Here is a larger pic. Hopefully it will come thru.

Thanks

 

[MiketheEngineer]

Is it a threaded shaft??

 

[karthur]

Yes. The shaft is threaded Ø1 9/16-12UN.

This shaft is the working end of a punch ram where the tooling is connected. Under "normal" conditions, the shaft would only see axial compression loading.

 

[byrdj]

don't know if this applies, but i've seen a hydrualic cylinder rod to clevis failure, where the treaded rod had a small amout of thread clearance in the clevis. A jam nut was specified on the assembly to preload the threads so they would not see a cyclic load

 

[sbozy25]

Well I was looking at this thread and it just so happens that my Metallurgical failure engineer was near. I had him look at it and here was his response....


That is a weird one. It has the appearance of a rotary bending fatigue failure--does the part rotate? The off center nature of the center, final separation is one clue; the eccentric shaped propagation of the outer periphery is another clue that suggests rotary bending fatigue as the fracture mode. There is insufficient definition in the photo to identify an origin, other than to say that it was in the thread root. Multiple fatigue origins are common with these types of failures.
The apparent step at the 6 o'clock postion suggests the fatigue started in more than one thread.

 

[rob768]

This is definitely a rotary bending fatigue fracture. I assume the stiffness is not symmetrical, causing the remarkable shape. The absence of a clear initiation may be caused by a sharp notch effect along periphery.

 

[metengr]

Mechanical fatigue failure under bending stresses based on the appearance and orientation of beach marks (radial lines emanating from the OD surface of the shaft).

The appearance and size of the final fracture area (off center) suggests high stress concentration in this location of the shaft.

 

[deanc]

More:

http://www.tribology-abc.com/default.htm

 

[deanc]

Couple more I have found.

http://www.fatiguecalculator.com/

http://www.roymech.co.uk/

 

[karthur]

I got the broken part back from our customer today. This part was on a new 125ton press that has only been in service for a couple weeks. The part was supposed to be heat treated to 50Rc. Somehow this part got past our QC dept without being heat treated. Since it was not heat treated, it could not withstand the 250,000 lbs axial load, thus failed in compression fatique.

Here is another picture that I made. You can see the beach marks on each side (pear shaped pattern around the center). The spacing of the beach marks indicate that it didn't withstand the loading but a few times before it failed completely.

It was intersting to see what everyones thoughts were about the failure.

Thanks.

 

[rob768]

I just hope the heat treatment will give you enough margin..

 

[karthur]

I need it to withstand 130ksi. With the heat treatment, S= 200ksi to 230ksi. Without the heat treatment,it would would be less than 100ksi.

 

[rb1957]

it looks like you've found out why it failed. a couple of points ...

200ksi bolts tend to be very brittle. if you need 130 ksi (presumably with all your safety factors included, why not 160-180 ksi ?

you say that the part failed in "compression fatigue" ... ?? maybe, since the bolt was under strength to failed in simple tension, what sort of preload do you apply (a lot I'm guessing) ? maybe this caused the bolt to yield during installation, then the operating loads caused the bolt to fail in tension fatigue ??

 

[karthur]

This is not a bolt. It was never under any preload. The threaded portion on the end was to secure tooling to the end of this (a punch).

This part is made from tool steel, heat treated and then drawn down to 50 to 55 Rc.

 

[rb1957]

sorry, it looked like a bolt ... are you machining the thread onto the rod ?

not quite sure about the build-up ... is the thread used to guide/control the position of various pieces of the tool ? otherwise, if the thread is used to join two pieces together isn't there going to be a torque (and so a preload?)

 

[deanc]

http://news.thomasnet.com/fullstory/482183/1274

This may be of intrest. Do a search on Punch Head Failure