Minimum bend flange width (not length) for mild steel (ie. ASTM A36)?

[lfunk]

We are getting cracking on the outside radius of 3/8" thick plate bent at 90°. The bend radius is specified at 1/2" but could be as low as 3/8" if the operator doesn't check the drawing.

The width of the bend is only 1/2" in two locations with a hole feature in between. The direction of the grain varies and is not determined by our laser vendor. According to our fabrication department, the part always cracks regardless of grain.

I'm thinking it might be the narrow flange width that is the problem. Wider flanges do not exhibit this behavior.

Has anyone ever heard of a minimum bending flange width based on material thickness?

 

[berkshire]

If you have a narrow flange that is cracking when you are bending with a radius that is larger than min bend radius, check the edge condition of the material.
You may find you have micro cracks at the sheared or punched edge of the material, that should be dressed out or de-burred before the bending operation is done. If the part is plasma cut look for ripples in the edge. Anything that can act as a crack initiator.
B.E.

You are judged not by what you know, but by what you can do.

 

[racookpe1978]

How physically are you bending the plate?

I assume cold bend (no heat), but is the 1/2 radius the inside radius (the formed rod you are bending the plate around), or is it the OD of the bend?
How long/wide is the 3/8 plate?
How long are the straight tangents on either side of the bend?
How are you physically supporting (AND restraining) the plate before, during, and after the bend process?
Where are you restraining the plate during the bend?

 

[lfunk]

I'm just looking for a rule of thumb. Consider an extreme case, where the width of bend is much less than material thickness, say only 10% of thickness (0.050" wide for 1/2" plate). We would expect separation on the outer radius even if you follow the minimum inside radius rule (thickness of material). Or, maybe not. My original question: Is there a recommended minimum width in proportion to thickness even in ideal conditions? Don't be afraid to say "no".

@berkshire, the edges are laser cut, so there is a bit of HAZ. It is conceivable that the hardening of the edge prevents formability at the edge, causing crack initiation due to the brittle scale. However, this seems to have no visible effect when the bend flange width is very large. We never prep the edges prior to bend and rarely have a problem. I guess the HAZ/width ratio increases with a reduction in bend width. Maybe more factors to consider...

Thanks

 

[berkshire]

Ok so let me get this right, you are bending an 0.05" thick plate half an inch wide the hard way.
You are going to get an extreme amount of metal thinning outside of the neutral line, with corresponding thickening on the inboard , ( Compression side ). You would be better off hot bending this part , at the very least you should polish the outboard edges before the bend. I am presuming you are using some kind of bar bending machine , like a Kennedy bender, or Diacro bender.

You are judged not by what you know, but by what you can do.

 

[lfunk]

@berkshire, that was a hypothetical extreme case. I guess I'm wondering at what point the thinning becomes the cause of cracking on the outside radius. Here is the long-awaited specimen:

We have noticed that extending the slit so that there is only 1/2" width material left at each end, we get cracking on the outside radius. Material thickness is 3/8" and inside radius is 1/2". I am just looking for information on a minimum width/thickness ratio for bending mild steel, if such a thing exists. Just a thought, maybe the slit doesn't need to be on the bendline, but closer to the hole?

 

[berkshire]

Ok, you already know that bending the part will stretch the outside of the bend and compress the inside of the bend.
The shorter the bend line the more extreme this effect is. One of the aggravating factors is that unless that part has the bend line supported by a sacrificial plate, the bend is not going to look like your drawing. it is going to be a straight line on either side of the slit, with an extreme amount of thinning, and tearing around the ends of the slot.
If you can support the bend , you may be able to make this part, otherwise put the slot in after.
If the slot is there to make the part easier to bend , get rid of it.
B.E.

You are judged not by what you know, but by what you can do.

 

[lfunk]

The slit is there to prevent deformation of the hole and is supposed to be welded shut afterward. I am pretty sure this is a common practice. Now I'm wondering if the slit should be closer to the hole and not right on the bend line. The fab shop is pushing to make it a 3-piece welded part. Right now they are grinding out the crack and welding the full length anyway.

I like your sacrificial plate idea. Thanks for sharing your thoughts.

 

[hendersdc]

Another option would be to remove both the slots and the hole to get good bends then go back and make the hole with a secondary machining process. Or cut a much smaller locating hole with the laser and go back after bending to drill to finished size. This would save welding and grinding, but you will have to talk with your fab shop to see which route ends up being cheapest.

 

[berkshire]

Ifunk,
I agree with hendersdc , Or cut a much smaller locating hole with the laser and go back after bending to drill to finished size.
Get rid of the slot and put a Pilot hole in. It will take you far less time to drill a hole than re weld that bend line.
If you want to punch the hole and you have enough of them , you could grind up a punch with a pilot pin for an iron worker to enable you to punch that.
B.E.

You are judged not by what you know, but by what you can do.

 

[tomsmith01]

The width of the twist is just 1/2" in two areas with an opening element in the middle.

I'm simply searching for a dependable guideline. Consider a compelling case, where the width of curve is a great deal not as much as material thickness, say just 10% of thickness (0.050" wide for 1/2" plate).

in any event you ought to clean the detachable edges before the curve. I am assuming you are utilizing some sort of bar twisting machine

Liftech Handling

http://www.liftech.com.au