Bending Extrusions 2

[Stezza]

I have an extruded L angle and I am going to bend one leg to create a U shape. Material is 7075-T73511 L angle leg thickness 0.050 inches.
Our forming specification says that for extrusions less than 0.5 inches thick it is acceptable to bend them using the same bend radi as sheet.
Thus I requested a bend radius of 0.25 inches.
But now I have a Co-worker that has stated that extrusions cannot be bent at all, because the bending axis is in the same direction as the Longtidinal direction of the extrusion, and will cause cracking in the radius. Showing him the forming specification was not enough to convince him that this is allowed.
Thus I would like a sanity check to confirm that bending a extrusion is generally permissable where the leg thickness is less than 0.5 inches or is there some rule of thumb that I am not aware of?

 

[rb1957]

without heat treating (to O- or W- condition) the problem in bending an extrusion is the standing leg (ie one flange of the extrusion forms the perimeter to the bend, and the other is bent).

without heat treating you won't be able to bend on a 0.25" rad ... look at what you'd be doing to the extreme fibers

 

[Stezza]

I think I didn't make what I want to do 100% clear

Here's a my attempt at a simple picture

| | |
|______ to |__|

So the new bend is in the leg of the extrusion.

 

[Stezza]

second try...

|
|_____ to

| |
|___|

 

[debodine]

i think what rb1957 was trying to say (and if i misunderstood i apologize and accept any correction from rb1957) was that if you were using the bend radius you specify on typical sheet aluminum you would be fine if you were bending in the LT direction, which is how you would make the part from sheet metal. However, if you are attempting to bend along the L axis (as you would be with the extrusion) you are going to risk cracking issues.

if i am wrong, i look forward to correction by rb1957 as you can tell from my handle (electrical) i am out of my depth here. )

 

[rb1957]

oh, you want to bend the flange ! (sorry, i thought you were bending the extrusion !)

it looks like you're bending along the grain direction.

bending with 5t will be difficult with 7075T7 and with the reduced access. if "they" take their time you should be able to do it, if they can get the radius block in.

it'd be much easier if you heat-treat.

 

[Stezza]

Thanks for the responses. I went and saw a materials expert today and got the following information.

It is acceptable to bend along the grain direction and at 5t we are on the minimum bend radius for the 7075 in T7 condition. We are going to do this way as its the simplest method to meet the requirements.

Cracking is possible problem because we are not heat treating and at min bend radius. We have the NDT inspection called out to check for cracking in the radius.

If we have cracking problems we can heat treat or 2 new options that I wasn't aware of are (1) shot peen to introduce residual compressive stresses before bending to reduce the chance of cracking or (2) bend at 350F, as you can go to this temp for about 2 hours without losing the 7075-T7 temper.

 

[debodine]

Thanks for the feedback on what you found. Since I am NOT a materials expert, I am delighted to pick up this additional knowledge by reading your note. )

 

[edmeister]

Has anyone looked at a cost analysis of this?

Assuming that flange thickness is .125" or more ..
The longest bend possible on basic leaf brake would be a couple feet long. A press brake would not work.
- If the inside gap is less then 1' spacing - you will not be able to 'close the gap' with a leaf brake either.

If only a couple items of a couple feet long are required - just machine it from a solid block or attach an angle using fasteners. The cost of fab / NDT / & the notch sensitivity of 7075 makes this a poor design.

If a prodution run is expected - just obtain the proper extrusion.

 

[berkshire]

Edmeister,
Your assertion that a press brake will not work for this is incorrect. You would need to use a dogleg punch and you could go the length of the press, The major obstacles you will face, are that with a .25" nose radius, you will need a sacrificial strip of greater thickness than the part you are bending under the formed part to keep it against the radius, or the formed part will pull away from the rad in the middle of the bend and crack.
Also with a .25 nose radius the minimum thickness of the nose of the punch will be .5" So you will not be able to bend the back of your channel narrower than 5/8"
It is also necessary to bend in two operations up 45 degrees then ninety degrees or the 7075 will crack.
B.E.

 

[edmeister]

I stand corrected ..
Should a shop have specialized punches (dogleg punch for example) - it should also have the capability of machining - neverthelees everthing is hypothical - what is the length ? what is the quantity? how critical is the component? why not just fab from a u channel?

 

[RPstress]

Messing with heat fuly treated extrusion has a reputation for problems and scrap.

Make sure you design to maximise the bend rad. (What are your design constraints on the rad? How it's being made is probably a big one, see above.)

Do the op warm if possible (but below 150 deg C and to a controlled process).

Be *very* sure that the post-forming crack detect op is carried out thoroughly.

If you must try shot peen before bending, find some sort of reliable spec to carry it out to. (I can see some merit in the idea, but it's not a pre-bend process I've ever seen before. And a surface compression implies a sub-surface tension...)

 

[rb1957]

"If you must try shot peen before bending ... "

to put it succinctly ... don't. by all means shot peen after bending, won't hurt, but not beforehand. you Could try to shot-peen form the bend, but i doubt you'll get there

 

[OldME]

Hot Joggle dies are a common solution used by both Boeing and Bombardier, there are specs available that cover this process.

 

[pirko]

You must use technology Hot Joggling. This is bending by teperature 150-190 degree Celsius. If you are inrterested in this technology send reply. Regards libor