Bending Stainless 440C

[toes112]

We have come across a situation in which we have 2 bends on a 5/8" dia. bar of 440C stainless, annealed and recorded at 225 Brinell, where we have compressive failure on the one bend and not the other. The bar is 20" in length with 2-90 degree bends located 5 inches apart. The bar also gets machined to an external 3/8"-16 thread 1" in length. The bend closest to the machining is consistently failing, while the bend further from the machining is fine. We are using a tube bender with the same setup and clamp area for both bends, 1 7/16" inside bend radius. It also does not matter as to what order these bends are made, same results occur. Is there any good reasoning why the bend closest to the machining is consistently breaking?

 

[swall]

How far from the threads is the first bend?

 

[israelkk]

How close?

 

[toes112]

The threads end 2" from the start of the bend.

 

[btrueblood]

When the bar is being threaded, presumably on a lathe, how is it being chucked/clamped? How is the free end of the bar supported?

I envision a long bar, with approx. 2" protruding into the working gap of a lathe. The remaining 18 in. or so is hanging off the back of the lathe chuck and running out thru the tailstock. If not supported, the 18 in. tail may be whipping/vibrating during machining. This may work-harden the material near your bend, or possibly cause fatigue (but you don't say the material is cracking, but failing by buckling?)

 

[swall]

Here is my guess about what could be happening. When a metal bar is bent, the outer fiber of the metal has to flow to accomodate the length increase at the outer portion of the bend and the shortening on the inside of the bend. There has to be a certain amount of metal available to flow to accomplish this. Where you have machined the 3/8" thread, you have removed metal from the outer fiber of the bar that would otherwise flow during bending. Without this metal, there is not enough plasticity at the bend and fracture occurrs. You could prove this out by taking a longitudinal section through the good bend and doing a macro etch to reveal the flow lines. One you have revealed them, measure and see how far from the bend they extend. If they go further than 2", you will know that metal removal due to the thread cutting is the problem.

 

[arunmrao]

In order to go along with swall's suggestion which appears quite possible,toes 112 please drop the threading operation and then bend at both the ends. Notice if there are any cracks. If no then you need to perhaps stress relieve once before you perform the bending operation.

 

[toes112]

arunmrao,
You are thinking in the same path as we were. We took a piece that was not threaded, and bent the end. We matched the distance from the bend to the end of the bar, to be the same as the distance from the threads to the bend. Result: a clean bend with no cracks. We actually even bent this to a 110 degree angle with no cracks.

 

[swall]

Good! Glad we're on to something. An additional fix might be to change the geometry--machine a nice smooth transition from where the threads end to about halfway to where the bend starts. Also, shorten the threaded area as much as possible.

 

[israelkk]

You may try to roll the thread instead of machine it

 

[arunmrao]

thread rolling is a better idea than cutting,I suppose.

 

[Kenneth]

Recognize you will need a differnt blank diameter (at least in the area to be threaded) for threadroll than for cutting threads..