Aluminum extrusion bending problem

[gsyrett725]

We are having difficulties creating a consistent radii on our new mandrel bender. the material and 6063 T4 and the parts have about a 110mm radius and about 40mm wide x 15mm tall. it has 4 bends all 90 degrees. The part has a +- tolerance of 1mm on both the radii, length and width. We are having a hard time getting a consistent radii and it will vary from 107 to 110 in a single part. This is making it very difficult to make consistent parts. As for every mm the radius changes it effects the leg length equally.They are not staying consistent as which bend is small or large either. One time bend 1 will be perfect all other small. Next time bend 3 will be good and all others small. The angles all seem to be good within .3 degree.

Thanks in advance.

 

[desertfox]

Well it's even harder to visualise what your component looks like, all we know is that it has four bends and its radi and what appears to be its cross section.
Do you think it might help if you provided a sketch or a drawing of the overall component?
In addition if it's a new mandrel bender have you considered contacting the supplier of the machine?

“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein

 

[IFRs]

Inconsistencies in the extruded metal on a grain level, hardness, temper, amount of oil from previous processes, etc will make it difficult to get the radius correct the first time every time.

 

[chicopee]

It seems to me that have spring back after the formation of the bends. I have not done any metal bending for several decades but I do remember that when you form bends in sheet metal, you should increase the angular deformation by a few degrees to get the correct angles. How many degrees should you increase the bends, well it depends on a number of factors some of which are mentioned by IFRs. I suggest that you research this spring back phenomenum as there is literature and even college textbooks on the subject. Also practice on some sacrificial pieces to get acquainted with the material.

 

[IFRs]

In my experience rolling aluminum (mostly extrusions, mostly 6061-T6 and 6063-T52) I was never able to set the machine, run them through once and get consistent roll radius bends. Sometimes it worked well but you had to check every one against a template or gauge. The best we could do was get it close, avoid over-rolling and run it through a second time if needed. Being able to reverse the roll sometimes helped. Setting it up for always two passes sometimes helped. Our stuff was usually custom shapes, 5" channel the legs in, 4" x 12" tee the easy way, 1 1/2" pipe, etc. You may be able to specify a tighter tolerance on the temper ( and pay more ).

I just read your original post and realized I'm way off base - you are mandrel bending, not rolling. Some of my wasted breath may be useful but not much.

Are you air-bending? Are you stretch bending? You could try heating the material as you bend it.

Tell us more about your shape and equipment for better answers and less BS.

 

[gruntguru]

Is it possible to purchase the tubing in the annealed state and temper it after your bending process?

je suis charlie

 

[GregLocock]

Yup, T4 is probably the issue.

One possibility would be to redesign the part around what is attainable with your bending process. There are others.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[gsyrett725]

We are bending window frames. The part is bend all the way around at 90s and then welded together to create a 1 piece frame. we are using a tube mandrel bender that has been retrofitted to bend extrusion. The retrofitter expressed they felt the tolerance was tight for the process considering the size of the cross section of the extrusion but should be obtainable. They said they have worked with others that bend extrusion and warned that the material may change properties over time as the age hardening process some times takes years to stabilize in certain types of aluminum. The noted that their other customer bends all their extrusion within 2 weeks of being extruded. Unfortunately at this current time we can not take this route. The current orders are too small to bend all the material we get from the our minimal orders from the extruder. As the customer has been changing revs left and right

 

[EdStainless]

With the property variation all that you can do is over bend a little and then set it back to the correct angle.
If you slow the bender down a little it might help also.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[EdStainless]

Isn't the radius formed against hard tooling? Either they are both correct of they are both wrong.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[gruntguru]

The radius is subject to spring-back after forming and the degree of spring-back will vary depending on the temper (elastic limit) of the material.

je suis charlie

 

[MikeHalloran]

... and the temper changes as the material ages.
Maybe grouping the blanks by estimated age would minimize the amount of fiddling you have to do with the tooling.

Also, I'm wondering if the bender could be adapted to heat the area to be bent just before bending, how much of a temperature change would be needed, and if it would work at all....




Mike Halloran
Pembroke Pines, FL, USA

 

[GregLocock]

I can't belive you get radial spring back without angular spring back.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[chicopee]

With spring back effects, you'll get different radii.

 

[gsyrett725]

We get a bit of Angular spring back but its a minimal difference between angles. we are over bending about 2.5 degrees some have 3. and from angle to angle we maybe see + or - .2 .3 degrees off of 90 at most.

 

[IFRs]

I'd be happy with + or - 0.2 to 0.3 degrees out of 90.0 - that is a very small percentage !!!

 

[gsyrett725]

and we are as we have a .5 degree tolerance. The linears are whats the problem. But they are caused by the radii being out.

 

[3DDave]

Just to get a picture - these are bent the hard way?

I suspect the neutral axis is shifting due to variations in the material properties and that the bender is not controlling the bend at the surface where the radius is to be controlled.

 

[gsyrett725]

bending them like the 2 on the right not the one on the left. at 4 90s to make a rectangle