Preheat weld required for 1045?

[JLoxterkamp]

We have a spindle assembly that developed a fatigue failure and broke where the spindle shaft steps down from 1.25" dia to 1" diameter for the bearings.
One of the options we are considering is to replace the 1018 shaft with a shaft made from 1045. The spindle would be welded to a 1/2" thick plate with an additional 3/8" bolster plate directly underneath.
Would we be likely to achieve satifactory results if we do a multi-pass weld rather than preheating the spindle shaft?

 

[CWIC]

JLoxterkamp:
Your application would require preheating to achieve a satisfactory weld. If you decide to step up to the 1045 material (an increase in C), then preheating would be required more so. As the carbon content increases, so does the preheat requirement.

The change from 1018 (a low carbon steel) to 1045 (a medium carbon steel) would result in a decrease in weldability, the omission of preheat would only add to your problems IMHO.

 

[EdDanzer]

You should check the shoulder radius. This may cause any material to fatigue break. We have several parts that the radius at the step cause premature failure. FEA was used to determine the best geometry solution. 8620 is a better material to weld than C1045 or 4140. 4140HT would be the highest strength. Any of these materials should have a 200-300° F preheat, and cooled slowly. Try to keep the maximum temperature to 500° F. You may find preheating the C1018 prior to welding might help reduce breakage. Multi-pass welding will not solve a heat effected zone problem.

 

[JLoxterkamp]

Thanks for your responses.
EdDanzer - We currently have a 0.03/0.05" shoulder radius called out, which is about as large as we can go with the bearings that come against that shoulder. The outside of the shoulder is chamfered to facilitate seal installation. If the chamfer is left as is, the shoulder radius can go no larger than 0.08".
We also wonder about making the radius larger and using a spacer with a beveled I.D. to fit over the larger radius. That way, the chamfer is applied to the spacer.
We also wondered if doing an undercut radius into the face of the shoulder would be beneficial. The only problem we see with that is there is very little shoulder face area between the undercut and the edge of the chamfer, resulting in more of an edge that might cut the lip of the seal during installation. Making the shaft bigger might help that, but then a lot more machining is required to get back down to the 1" diameter for the bearings.
Any more thoughts or suggestions would be appreciated.
Thank you in advance!

 

[EdDanzer]

To fix one of our problem parts we used an undercut radius and the other a large radius covered by a seal ring/spacer. Increasing the shaft diameter may not fix the breakage problem. If you do not have FEA and this is a costly problem, subcontract out the redesign. The lowest cost fix is to undercut a larger radius, but a large radius covered by a seal ring/spacer is the highest strength if there is space available.

 

[ccw]

Hi JLoxtercamp,

This may be a stretch, depending on the fab./mfg. expense vs. a special welding setup: Redesign the 3/8 bolster plate and 1/2 plate to a construction that allows a drilled and reamed center hole to be formed, staying with the 1018 material. The length of the hole will need to be sufficient to react the bending moment and shear load of the axle. Press fit or shrink fit a pre-machined, tempered, high strength, high alloy shaft(such as 1045 or 4140) into the reamed hole for the bearing axle.