Topper Machines had a customer who told him that a part with a 10 inch circular base and a 14" tall 2" dia shaft that had been welded together broke after a short time at the weld. Topper’s solution was to get a 10" diameter cylinder of steel and over many hours create a new part with no welds. I think topper said he had 400 lbs of steel waste in the form of shavings. This struck me as a tremendous waste of time and material so I started to think that a great video would be to duplicate what Topper had done with a single 10" cylinder of steel and then stress test the shaft to the point of breakage It may be the rotary torque on the shaft that broke the part, but you could test both point of failures. Once you have that data on the 1piece part, it would be great to see what stresses could be put on the same part that was welded together just like the ones that broke. You could make a 3 part series where you and the audience propose and test different strategies to build a part that is as string or stronger than a 1piece part. For example what about coring out the shaft for maybe 4 inches and machining a 10" circular base plate with a 4" tall 1.5" diameter shaft that the cored pipe would pressure fit on. I am sure with your vast experience and comments from the audience we could get some really strong alternatives to the scrapping of over 400lbs of shavings.
Here is the link to Topper’s video which will make this all clearer than my words. Thanks and I would love to see you attempt a series of stress tests.
This is a great topic. I watched the video and was left with more questions about the part. It seems like this part was requested to be built out of one piece. This seems silly to me from a manufacturing perspective. Without all the information on what the part is for or is end use it had to say the one piece part is needed. If this was a prototype, then maybe. We have this same issue with our Hardtail vise screw. The meatball is 3” on the end of a 1 1/4 shaft. The two parts are combined with a friction weld. To me they are virtually one part when it’s finished. The other way to do it would be forge the end like a car axel flange. Or maybe even spline or key the plate to the shaft. This would be an interesting testing. I’m not sure I have the equipment to test all the methods but anything can be built.
In the video he comments that he’s working off the original vendor drawings, which specify it to be a single piece, and that he’s had other customers with similar parts requirements. To analyze the part for conversion to a two piece, welded assembly, one would need to know all the engineering data behind the design and would have to analyze the same part when made as a two piece assembly. Depending on the material requirements, welding, heat treating and stress relieving all add complexity and cost to the fabrication.
Hogging out a blank like that, once set up has a lot of machinist down time, so he’s probably working on something else during each pass, but still keeping watch over it.
If that blank was 1018HR, it was only about $1100.
Jason, I’ve been thinking that you would only have to test the breaking point of a 18" steel bar 2" in diameter; you wouldn’t have to duplicate Topper’s part because a test of a circular steel bar with the same length as Topper’s part would be the same stresses as Topper’s original part. It would be great to see what sort of forces it would take to either bend or break the bar if you pulled on the top and even more interesting to see what turning i.e. torque forces would spin break the solid bar and then compare the forces that break an 18" circular pipe welded to a base. You would probably have to use a big hydraulic press to attain the breaking forces. It would be an interesting video no matter what you do. Thanks for even entertaining the idea Jason.
I don’t think it would be that easy. For one how do you know the bar was the failure point?
I would think you would need know how it gets used, how it failed when made in two pieces.
If this part fails does it cause catastrophic damage? if so than its a no brainer in my book
On the part fabrication, if a shop were capable of heat/cryo fit, the large disk could be heated and the shaft ultra-cooled (liquid nitrogen) to a slip in fit and then the two pieces would virtually fuse. Add a keyway and key. Thermodynamics of steel is a huge and complex engineering field, but a shop that does this type of process all the time would treat this like it was a toy. Then whatever the machining for the disk face could be completed. The question remains as to use and forces, and whether this prototype is on a tight budget.
That just screams to be a forging. Then it would be even stronger than the part machined from bar. The problem is Topper is out in the middle of nowhere so not many options to get something like that forged.
I was just interested in seeing several methods of securing a steel bar to a plate and then seeing how the rotary and bend forces to break each method stack up. I understand that in some cases a solid part is absolutely necessary, but I would like to know “how necessary” it is. That way. others could decide if the welding rather than turning a solid part would suit their needs and eliminate the waste.
It’s easy to say this or that method is the strongest, but without any stress testing, it’s mostly speculation. Thanks everyone for your interest and input.
Where do the welded versions break? What kind of di$a$ter does a failure cause?
I wonder what the designer(s) estimated the stress at the shaft/disk interface to be under service load… Increasing the radius will decrease the stress.
