Broken screw conveyor “live bottom” screw drive shaft - thoughts on weld failure 1

[USAeng]

Hey guys

We had this screw drive shaft break at the welds today. It was installed less than a year ago and I’d estimate only 1500 hours on it.

Just curious what you guys think on the failure. The drawings do state the plate can be welded to the shaft to develop full strength. I wonder if this means the shaft should have been welded with a much bigger weld?

Anyhow just wanted to see what you guys think

Thanks!

 

[USAeng]

 

[hydtools]

The weld failed. It did not develop full plate or shaft strength.
I would have made the shaft extend to or beyond the far side of the plate rather than welding shaft edge to the near plate hole edge.

Ted

 

[dhengr]

USAeng:
That is not a particularly good weld detail for that application, and the .5” radius is most likely not adequate. The external fillet looks small and is certainly not shaped to a .5” radius at the circumferential reentrant corner. The stresses in that weld will oscillate six times with every revolution due to a varying end plate stiffness, and of course the bending normal stresses will change from max. tension to max. compression, at any location on the circumference, once every revolution. I would need much more info. on the entire system to comment much further. For starters, the drawing is to small for me to be able to read. How is this screw conveyor arranged in the bigger picture and what are the loads on it, operating speeds, bearings, power supply? It appears that one of the six bolts may not have been fully tightened. It appears that the weld penetration was kinda poor at some locations. What is the small hollow tube shape in photos 3 & 5? What are the materials of construction and their dimensions, etc., what is the material being moved by the conveyor?

 

[Jboggs]

Change your design to distribute the torque load over a greater surface. For example, what is the functional radius of the actual weld joint? 3"? 3-1/2"? Whatever it is, do what you can to make it bigger. This does two things for you. One, it decreases the actual force on the joint by moving it farther from the center axis of the applied torque. And two, it allows a longer bead which effectively distributes that applied force over a larger weld joint.

Second point - ANY misalignment will cause a cyclic loading on the joint. In fact, that would be the FIRST thing I would look for. If there is a potential source of misalignment, do whatever you have to do to eliminate it completely. That welded joint will evenly distribute applied torque and force ONLY when the source and the screw are in perfect alignment, otherwise you're creating high points of loading. Leading to cracking, fracture, and failure.

 

[dvd]

What is the material? I suspect improper preheat.

 

[desertfox]

Hi
If the shaft can be welded to develop full strength, then where is the load information that would allow an engineer to calculate size and type of weld to enable correct function?

I don't think its a good idea to weld it because from what I can see it would end up being a fillet weld

“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein

 

[Strong]

Image/photos 1 and 5 appear to show wire embedded into the weld. I can't think of a good reason for this, unless the wire became wrapped around the shaft during operation. If previous service time was much longer than 15-months, then the weld/repair is most suspect. If repeated failures occur in short time interval, then the overall design is most suspect. I have seen live-bottom screws get beefed up in one location only to fail at another location.

Walt

 

[Tmoose]

Even if the shaft and flange were integral, the corrosion would clobber the material's fatigue strength.

The "plate can be welded to the shaft to develop full strength."
As others said, that weld detail is far from "full strength" compared to an integrally formed flange with a proper 1/2" radius.

On top of all that, the poor old fillet welds lack the proper detailing to give them a fighting chance.
For instance, Fig 3 here -

https://www.twi-global.com/technica...sign-rules-for-welded-structures-january-2000