WELDING AROUND TANK

[imagineers]

I am wondering the best practice or common practice for welding around the edge of a box, which will be a reservoir tank. In the picture you can see two concepts. 1. the side plates are undersized slightly to do a fillet weld all around, or 2. keep the faces flush and doing a fillet weld around. I would imagine the one with the undersized side would offer a nice weld area to get a good bead. I am not sure how they would weld the other one with faces being flush. I have seen laser welds dont on flush faces, but I am also looking for the strongest weld possible. Thanks

 

[metengr]

For this type of application, I would prefer the fillet weld over a square edge butt joint. Why? Because one can control the size of the fillet relative to the thickness of the material, versus a square edge butt joint.

 

[imagineers]

or maybe a weld inside with a flush face side plate??

 

[metengr]

No, not my favorite because of the possibility for root cracking.

 

[imagineers]

root cracking from what I was reading is caused from using bismuth gas shielded flux cored wires with bismuth-bearing compounds, but usually occur at high temperatures. Are you sure this would apply in this case? It mentions ensuring your flux cored wire contains < 20ppm bismuth. Anyhow you think it is best to make the outer plates should be slightly smaller dimensions to get a nice fillet weld?

 

[Tmoose]

Is the material stainless steel?
How thick?

 

[imagineers]

It is stainless and is 14Ga (1.9mm) thick 316SS

 

[imagineers]

I guess the thing is..Is it possible to do it either way>?

 

[tbuelna]

If you want a weld joint between the tank body and end caps that provides nice stress distribution, then I would suggest machining the end caps with a short flange that matches the profile of the tank body and can be butt welded around its perimeter from the outside. With the type of pressure vessel arrangement you show, there will be stress concentrations where the tank walls connect to the end cap panels. So this is not the best place to locate a weld joint. Better to move the weld joint away from this point of stress concentration, and use a butt weld that will be more evenly stressed from the internal pressure forces.

 

[MikeHalloran]

Tbuelna's idea has merit, but I wouldn't machine the end caps if any quantity is involved.
Instead, I'd form a short flange all around, e.g. by hydroforming, or with a simple punch and die cut from steel plate. That will work better if you increase the bend radii of the tank shell a bit.

Actually, at the 10 psi design pressure you've revealed separately, you might be able to use much thinner stainless, and roll crimp the ends using tin can tooling, which would work better if you eliminated the reversed curvature and made all the shell bends in the same direction, just like an antique Log Cabin Syrup can. The crimped ends would be justified only if your quantities are high, facts not in evidence.





Mike Halloran
Pembroke Pines, FL, USA

 

[imagineers]

I don't get how you can butt weld sheet metal. You have the end face edges of the long body mated to the large surface face of the end caps? what would the weld look like exactly? I have seen laser weld that i believe can do that but not sure about a butt weld , example?

 

[gtaw]

If one was to envision the vessel being made from soft material, it is easy to visualize how the metal would deform when subjected to a load. Were the fillet weld placed along the outside of the vessel, the walls would bulge placing the root of the fillet welds in tension as the walls rotated.

Placing the fillet welds on the inside of the vessel would eliminate that concern. as the vessel walls bulge outward, placing the face of the fillet welds in tension. A better situation than what is described in the previous paragraph.

The minimum amount of weld metal would be required if fillet welds were placed both inside and outside the vessel. the weld size would be dependent on the design assumption. Using the allowable stresses permitted by AISC, a 70ksi filler metal, 36 MSYS carbon steel base metal, and double fillet welds (inside and outside the vessel), the fillets only have to equal to 3/4 the thickness of the vessel wall to develop the full strength of the base metal.

Best regards - Al

 

[imagineers]

Here is what I have so far. The thing is that i can weld all of this internally, then weld the cover last with an outside weld, which may or may not be the best. Simulation looks ok at 30psi even except a little bowing on the top cover near the middle, but otherwise everything appears to hold up. Still not sure about the three ribs

 

[MikeHalloran]

It seems incredibly complex for what seems to be a simple low pressure manifold.

Also, it's not easily cleanable, if that's important.
No such requirement was stated, but is suggested by the sanitary flanges.


Mike Halloran
Pembroke Pines, FL, USA

 

[berkshire]

imagineers (Mechanical)

Based on what you have drawn so far, your best bet would be to weld the ends and lower body of the tank together from the outside using 316L filler rod. Then Spotweld your baffles in. I would also suggest turning the flanges upward on the lower baffle to enable you to weld those with a spot welder. When the baffles are in place you can weld on your lid. The pipe bosses would be better welded before assembly to allow for any cleaning and hand straightening if the product distorts on welding.
Also if the working pressure is just 10psi this tank looks like over kill.
Just what liquid are you placing in this tank?
B.E.


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

 

[tbuelna]

I don't get how you can butt weld sheet metal.

Here is a sketch showing how the pressure tank you describe would usually be fabricated. The center body is brake formed in two sections, and there is a flanged end cap at each end. All of the tank body joints are external butt welds. The internal baffles can be welded in place before the tank panels/caps are butt welded. Butt welding 14ga cres sheet should present no problems.

 

[imagineers]

tbuelna (Aerospace)--- it looks like you have formed the outer panels, not sure if that would cost more or to just use the flat panels I have now. Still 4 pieces however.


MikeHalloran (Mechanical)---it has several filters throughout the system and should not require cleaning. It is using glychol water mix


berkshire (Aeronautics)----i was planning the same idea on weld steps except thought I could add spot welds even if the flanges are pointed down??

Ok so maybe I should clarigy this setup. We have relief valves set to 20psi so I am simulating the assembly using around 30psi, you can see the results in image attached. It looks alot better with the ribs in place than without. The reason for the lower baffle is to prevent foaming of the liquid from the fluid which tends to happen when it hits the back, as I have tested on a previous tank I designed. the upper ribs are to stiffen the outer walls.

Another question for you guys is...my steps would be to weld the the bottom baffle to the outer box then the top cover. I would then slide the ribs into place through the opening and try to reach in and spot weld them to the lower baffle, top cover and sides, is this possible?? I would put the end caps on last with a outer fillet weld. Does this make sense and can the welder reach into this box and weld the ribs with ? main reason is to prevent the top cover from bowing from any high pressure incident. which appears to help in simulation.

 

[imagineers]

Simulation screenshot

 

[imagineers]

any thoughts on reaching into the box after the bottom and top is welded together and tack welding those ribs to all four surfaces?

 

[berkshire]

Imagineers,
You can do it by changing your weld sequence. Weld one end on your tank, add a top flange to the ribs. Now spotweld with a spotwelding machine, not a Tig torch filling a hole. Spot in the lower tray, then put on the first rib with the flanges facing to the open end .Add the top, tack it in place. Add the second rib spotwelding to sides, and top, then the third rib. Now with these parts spotwelded in place you can add the last end cap. I would also suggest reversing the flanges so that you can weld the end with the spigots in first and do any required hand straightening. The ideal would be to weld the plain end on last. Doing it that way you can also delete the positioning flange you have bent on in favor of using a welding fixture.
B.E.

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