Vessel Support Bracket - Punching Shear Calculation 1

[strainstress]

Hi,

I am trying to figure out what supports to use to mount the vessel. The vessel has a 54" diameter and thickness of 5/16" (See attached rough sketch). The vessel is to to be mounted to customer beam platform at 68-1/2" spacing using 4 mounting legs.

The legs are made of 3x3x1/4" tube welded to the vessel. The requirement is that the vessel cannot dislodge during a seismic event and has to maintain its structural integrity. If the tube and weld size are adequate, what modes of failure in the vessel at the tube-vessel connection should I check ? Since the requirement is that the vessel cannot dislodge, is just checking for tube punching through the vessel wall adequate . How does the local bending stresses developed at the tube-vessel connection effects the strength of the joint ?

 

[WARose]

As far as the vessel itself, probably the best place to look would be a pressure vessel handbook. Another good reference would be a paper I’ve used in the past that combines the various forces acting on a vessel and compares that to an allowable……that paper is: ‘The Design of Vertical Pressure vessels Subjected to Applied Forces’ (by: Bergman, et al; ASME, 1955).

How to get some of the forces to plug in (from those localized conditions) would probably best be obtained by some sort of FEA analysis.

But I’ll be the first to admit I’m no expert in this regard (as I am structural). The only reason I’ve messed with sort of thing at all was a feasibility study I did for a vessel supported platform…..ultimately an ME/manufacturer made the final design.

 

[pflow]

See “Pressure Vessel Design Handbook” by Henry Bednar, Section 7.4.

 

[dhengr]

Strainstress:
My first impression is that your sketch shows an awful way to support that tank. The 3x3 HSS kicker and its base angle must carry the loaded weight of the tank, plus any dynamic loads, lateral or vertical. And, the bottom flat plate of the tank is also unsupported. The HSS kickers will induce very high concentrated loads at four points on the thin tank shell, not a good condition, without a stiffener ring at that load distribution level. A better arrangement similar to what you show would be 4 bot. pls. (foot pls., your base pls.) which could be welded directly to the bot. pl. of the tank, and would be about 8" wide and cantilever out to about a 66" bolt circle. Then put 4 triangular shaped stiffener pls. btwn. these foot pls. and the tank shell about 18-24" high. These stiffeners are welded to the tank shell to take most of the loads into the tank shell in a more favorable orientation. Maybe they still need a stiffening ring at their top, or at least some small part of that ring to distribute that stiffener punching action.

Better yet, why not space your support beams at 54" o/c both ways? Then the tank sits right on the support beams, and most of the loads are taken right up into the tank shell. Then 4 shorter ears (the foot pls. above) are extended from the tank bot. pl. and bolted to the support beams outside the beam webs, as a hold down means. Maybe some small triangular stiffeners are needed here too. While either of these schemes sacrifices a little in overturning moment lever arm, they take the loads into the tank in a much more favorable way, even if they are a bit higher.

 

[JStephen]

If the support beams are fixed at that spacing, can you put horizontal beams at 45 degrees, making a diamond shape, rather than the kneebraces as shown?

 

[JStephen]

In response to the original question- punching shear is almost never an issue on a steel tank or vessel, but bending at the connection is, thus the recommendation to check pressure vessel handbooks.

 

[strainstress]

@ All : I may have used the wrong terminology. The vessel I talk about is actually fan casing of a vertical vane axial fan.

I can obtain local bending stresses from the pressure vessel handbook, but what should be allowable bending stress. The pressure vessel limits the bending to 2*Sy for shakedown purpose. But I do not have any cyclic loading in my case.

By limiting bending stress to 2*Sy will I be over designing ? My only requirement is that the fan cannot dislodge during a seismic event. In that case, should I still be worried about local bending stresses. Would'nt the ductility of steel spread the stress ? At what stress state will the vessel start to crack by tube punching through. Is'nt this the only mode of failure I have to prevent ?

@ dhengr : Unfortunately, I am constrained to use 68.5" dimension. Thanks for your suggestion of using gusset plates though. I agree, that would be a more favorable design. We may go that route.

 

[strainstress]

@JStephen : Can you please explain more about the horizontal beams at 45 deg design. I am not able to visualize the design. Do you mean just the horizontal beams and no diagonal beams at all ?

Thanks

 

[dhengr]

Strainstress:
Why on earth would you call a vertically mounted axial fan housing a vessel and use pressure vessel handbook methods to design these fan housing supports. Why not use nuclear reactor vessel codes, rock mechanics theory or spacecraft design criteria?? The gist of my suggestion was to get the 4 bearing points as near the round shell as possible, to minimize the reaction moments on the shell and to impart the loads into the shell, in the plane of the shell. Not as 4 point loads with large components perpendicular to the shell, which are difficult to deal with. Are the beams spaced at 68.5" existing, and they can’t be changed or moved. How do you know they are strong enough for this load? Why not clean up your sketch a bit, so dimension lines actually go to something of import, 12"+54"+12" does not equal 68". What are the loads, weights, C.G. locations, etc. What are the existing framing sizes? Your support system must support the operating weights, pay attention to fatigue since it’s a fan, and carry all lateral loads. And, I have no idea what “ limits the bending to 2*Sy for shakedown purpose” means, do you? I suspect that JStephen’s suggestion was to add four diag. beams, atop your customer’s existing beams, at about 45̊, but right up against your 54" shell, again, to minimize the cantilever out to the reaction points. You shouldn’t be doing this design on your own if you don’t have a vague idea what you are doing. You should get help from your boss on this problem, so he knows what you know and what you don’t know, and he can help keep you out of trouble. Shame on him if he assumes you can do this design with no experience and guidance from someone with engineering judgement and experience.

 

[JStephen]

See the attachment for what I had in mind. This is assuming the beams you have shown are existing beams that can't be shifted around. Add in the red diagonal beams, then use a lug mounted on the shell for support. The diagonal beams could go on top of, or in the same plane as, the other beams, depending on what worked out best.

 

[WARose]

Why on earth would you call a vertically mounted axial fan housing a vessel and use pressure vessel handbook methods to design these fan housing supports.

If it’s a thin walled tube, about the only place I can think of that he would find allowable stresses (for a localized loading) is a pressure vessel handbook. I went through something similar to this some years back and figuring the allowables and the forces developed for such a situation was best facilitated by such a handbook. (I certainly couldn’t find anything in Roark’s handbook on it.)

 

[dhengr]

WARose:
I’m not suggesting that some of the analysis methods and details used in pressure vessel design might not be used here, but this is certainly not what most of us would consider a pressure vessel or even a vessel. How those methods and any allowable stresses would apply to this (the O.Per’s) problem are highly dependant on the shell thickness/diameter ratios, and the operating pressure and temperature, etc. and because of these variables the pressure vessel approach may be misleading unless there is a full understanding of how these variables play into the analysis. A good fundamental understanding of plates and shells, shells of revolution, Engineering Mechanics, Strength of Materials, Theory of Elasticity, buckling of shells, etc. are probably every bit as important and useful. He shouldn’t just be plucking allowable stresses, details, and design methods, out of a pressure vessel code or handbook if he doesn’t have a vague idea where they come from and what he’s doing. If it’s a circular fan housing, why not call it that and show a good sketch of the important details, loads and dimensions. Studying the various details and methods used in the design of steel storage tanks, smoke stacks, tubular structures, etc. might be just as useful.

 

[WARose]

I’m not suggesting that some of the analysis methods and details used in pressure vessel design might not be used here, but this is certainly not what most of us would consider a pressure vessel or even a vessel. How those methods and any allowable stresses would apply to this (the O.Per’s) problem are highly dependant on the shell thickness/diameter ratios, and the operating pressure and temperature, etc. and because of these variables the pressure vessel approach may be misleading unless there is a full understanding of how these variables play into the analysis. A good fundamental understanding of plates and shells, shells of revolution, Engineering Mechanics, Strength of Materials, Theory of Elasticity, buckling of shells, etc. are probably every bit as important and useful. He shouldn’t just be plucking allowable stresses, details, and design methods, out of a pressure vessel code or handbook if he doesn’t have a vague idea where they come from and what he’s doing. If it’s a circular fan housing, why not call it that and show a good sketch of the important details, loads and dimensions. Studying the various details and methods used in the design of steel storage tanks, smoke stacks, tubular structures, etc. might be just as useful.

That’s an excellent point…..and I’m not saying anyone should take the allowable in the pressure vessel handbook at face value. But the problem I ran into for is this: for shells, cylinders, and so forth, the only allowable I could find [when I was up against this problem way back when] were in some academic texts (I can’t remember if Roark’s was one of them or not). And the problem with these texts is they gave allowable but then they turned around and gave a range of recommended safety factors to use (based on initial imperfections and so forth; which apparently is critical for such things). So the problem became: what SF to use? You can go real conservative……but it was still vague. And that’s where the pressure vessel handbook came in handy: you were working with real world values. (And that’s not to mention the fact that almost all the afore mentioned texts [including Roark’s IIRC] didn’t treat the localized case.)

But your point is well taken.