Camber by welding

[kennewton]

Is it feasible to generate a 5% camber (height of curve vs length, ie 100" length and 5" height) in a 3" Sq 1/8" wall 6061-T6 aluminum tube, by welding beads along the side inside the camber? The idea is, the welds would distort the tube into an acceptable curve. This is proposed to make a mild large radius curve without heavy machinery. I have no idea if such a thing would work. Residual stresses may be an issue, since the welded side will be in tension.
Any thoughts or ideas would be appreciated.

 

[gtaw]

The weld will degrade the mechanical properties of the 6061-T6. Tensile strength will be reduced by about 40%.

If that isn't a concern, then yes, the residual stress introduced by welding will act in tension and camber will result.

Best regards - Al

 

[MiketheEngineer]

Usually done with a rolling device that "bends" the tube or beam to the shape you want. Sometimes simple heat is used on one side to "warp" the beam into shape. Not as accurate but is often done.

 

[rmw]

In my experience, it is done more as a last resort than a first resort.

rmw

 

[weldtek]

In my experience it's first discovered right before the screaming of epithets, hammer throwing and heads banging on walls.

 

[boo1]

Cold work it..

 

[paddingtongreen]

Ask someone like this:

http://bending.net/structural-bending.html

They might send you on to someone who does do aluminum.

Michael.
Timing has a lot to do with the outcome of a rain dance.

 

[drawoh]

kennewton,

I don't think your weld concept works.

h = R - sqrt(R^2 - 0.25L^2) (you can work this out)

where h=height, L=length, and R=bend+radius. I am guessing R=252".

h = 252in - sqrt((252in)^2 - 0.25[×](100in)^2) = 5.010in.

252in is close enough.

From my mechanics of materials book by E.P. Popov...

1/[ρ] = M/EI, where [ρ] is the bend radius, which I am calling R.

M = EI/R.

For your 3"[×]1/8 square tube I=1.98in^4.

For aluminium, E = 10[×]10^6lb/in^2.

M = (10[×]10^6lb/in^2)(1.98in^4)/(252in) = 79000lb.in.

Stress [σ] = My/I = (79000lb.in)*(3in/2) / (1.98in^4) = 59000lb/in^2

For aluminium 6061-T6, [σ]_y=40ksi. [σ]_u=45ksi. As noted above, aluminium 6061-T6 is heat treated, prior to you welding it.

You cannot elastically bend your 100" tube mechanically by 5".

I remember doing the experiment in college, where we applied a weld bead to one side of a piece of steel and we measured the resulting deflection. It bent into a circle. Aluminium is a better heat conductor. I have a hard time visualizing this one.

JHG

 

[Ron]

Have no idea what happened to my post, but I posted right after rmw...here's my take

The camber for the length is excessive and would be difficult to accomplish with welds. Besides, welding on aluminum, as gtaw noted, will significantly reduce the allowable stress.

I would roll it.

 

[MikeHalloran]

It may not take _heavy_ machinery, either.

Any bender that can bend a 3"x1/8" aluminum angle "the hard way" can probably do it, maybe even a big mechanical bender. E.g. Hossfeld.

Some kind of flanged roller on the bender might be enough to prevent the walls from bulging out. If they bulge in, you are screwed, unless you prefilled the tube with sand and plugged the ends.

Mike Halloran
Pembroke Pines, FL, USA

 

[fegenbush]

MikeHalloran,

I have had the best luck filling with either salt or pitch (rosin) instead of sand. I find that sand compresses too much and I still get some collapse of the tube.

 

[kennewton]

Appreciate the input about a cambering a square aluminum tube.

Taking another tack on forming a shallow camber in a square aluminum tube with minimum machinery. What would the prospects be, for making a series of small narrow linear indentations across one side the tube, to form a chamber in incremental steps. Obviously making a big dent the tube would crush it and make a kink. Would a series of carefully controlled smaller indentations make a suitable incremental camber?

 

[berkshire]

Are you thinking of bumping this in a press ?
It can be done, however at 5" in 100", you are pushing the limits of formability.
You would still need to pack the tube and use a distance limiting stop.
I would recommend putting the bottom supports 18" to 24" apart.
And leave at least that much waste on each end of the bar.
B.E.

The good engineer does not need to memorize every formula; he just needs to know where he can find them when he needs them. Old professor

 

[MikeHalloran]

Well, sure, if you don't need a nice continuous camber, it gets easier. Just look at some of the extreme forming done to aluminum tubes in cheap lawn/camp furniture.

Basically, if you use a punch to plastically deform the inner wall at intervals, you can get a pretty extreme overall camber. The punch impressions effectively shorten the inner wall. The punch, in this case I'd guess 2" square with a somewhat rounded nose and maybe rounded corners, _partially_ buckles the inner edges, so they get shorter overall, and bingo, there's your curvature.


Mike Halloran
Pembroke Pines, FL, USA

 

[kennewton]

Thinking more along the lines of an arbor or power press , with no or minimum spread between bottom support(s), using limiting stop or load, and touching it every 5 to 10" with a relatively small radius tool of 0.10 R or less. The idea is to shorten one side with full width cross wise indentations, with a bit of bend at each touch. Would guess the indentations would be in the order of 0.01 to 0.02" deep, with a minimum of buckling in the sides.

 

[berkshire]

I have bent tubes that way, and you can make it work.

You had better have lots of scrap pieces, because this is very much a trial and error method.
I do not think at 0.01" to 0.02" depth you will get what you want.
You may have to go to 0.04" to 0.05" . with some flaring on the sides of the tube.

An alternative to sand, rosin, or salt in the tube is greased strips of sheet metal to pack it out.

Your best bet is to find somebody with an angle iron roller And get it done outside.
B.E.

The good engineer does not need to memorize every formula; he just needs to know where he can find them when he needs them. Old professor