Pipeline structural support design.

[jonathanjc]

Hi all,

I am designing a pipeline support and need some help. I am trying to find out the load that the support will need to carry and have been finding some unconvincing solutions. See diagram attached. The solutions have been:
Fa = 65kN
Fc = 48.48kN
Fb = 114.kN

I am convinced that I need to find Fb but the solution I have is more than the axial force acting on the pipe, I have also tried via the principle of superposition but considering Fx only gives me Fb = 46.765kN and I know Fy must be somehow incorporated and it is confusing me, as simply adding Fb and Fy gives me 81kN and this doesn't seem right.

So I am alittle stuck, would be great if i could talk it through with anyone willing to take a look at what I have done so far.

Thanks in advance
Jonathan

 

[dhengr]

Jonathanjc:
Post your diagram as a pdf file, I can’t see png files, or at least don’t know how to bring them up. Maybe I can give you a bit of a hand, although I’m not sure exactly what you’re trying to do without seeing a sketch, etc.

 

[jonathanjc]

Thanks for the reply.

See pdf attached. I have had another look at this with a fresh mind and have come to some more acceptable solutions, I now have Fa = -52.649kN Fc = 70.812kN and Fb = 88.239kN.

I have gained Fb by Pythagorean theorem. I know the self-load of the pipe will also contribute to the load of the support.

Do you think my solutions are reasonable?

Thanks in advance

 

[jonathanjc]

Also I forgot to mention - would I be right in saying that the moment, Mz, wouldn't necessarily be a factor for loading the support.

Or is that an incorrect assumption?

 

[MartinShane]

I believe that the design intent of the bracket will alter your loading, no?
Do you intend to constrain the pipe to prevent "sliding" along / rotation about the axis? (Clamped?) or "Loosely" supported such that rotations and sliding are allowed?

 

[jonathanjc]

There is a reinforcement sleeve at the bracket section. I don't think this sleeve will be welded to the pipe at the ends but the support will hold the sleeve and pipeline in place. My assumption is that the movement will be restricted in the horizontal and vertical directions. So, I would assume that it is loosely supported.

If the assumption were that it was fixed and restricted sliding/rotational movement - would we then consider the moment Mz to be an affecting factor.

 

[dhengr]

Jonathanjc:
Come-on Jon, put your thinkin cap on. Have you had any courses in Statics and Strength of Materials? You certainly should have if you are doing these kinds of design problems. This should not be a tough problem, but you haven’t got anywhere enough detail to even start on the problem, and you have to think through it step by step. For starters, if you mean 60̊, show 60̊ on your sketch, the computer does it for you, so don’t be so lazy. Also, correct proportions, some dimensions, thicknesses, loads and correct orientation are important on your sketch. When an experienced engineer looks at it, and he/she is being asked for help, they immediately start making judgements and impressions, based on experience, about the situation from what you have shown. So the more accurately you show it, the more likely it is that they will not be mislead, and you’ll get meaningful advice and suggestions. I grant you that the general concept doesn’t change for a few degrees, but the outcome sure can change.

As MartinS suggests you must know much more about the support; the saddle, the U-bolts or straps, the reinforcement sleeve, can you tolerate pipe movement if there is slip at the saddle? You do make assumptions about these things at the outset, but you must pin them down before you can really address your actual problem. Sometimes this is an iterative process, can the pipe take any added stresses caused by slippage, it’s going to bend or be loaded someplace to counteract this. And, 60̊ means slippage is more likely, than the 45̊ you show; one of those first impressions. If you restrain this action at this support, it will mean bending in the saddle along the axis of the pipe. The pipe will also bend and have reaction stresses across this saddle/support, and they are additive to the internal pipe stresses. Can you restrain longitudinal movement by bumping up against the flange? All you have right now are a compressive load on the pipe (Fx), a shear load (Fy) and a moment/torque (Mz) and the pipe and flanges must be able to take these. But, you have to have some understanding of where these came from, and they probably do include the self weight of the pipe. All, of each of these loads, may not even be taken out at this support, and you must sort that out. Then you need to start developing this support detail, right down to the foundation, by getting answers to the above type questions. Then you finally start sizing members and looking at final stresses in the various parts. Show us some more sketches of how this develops. There is some good help here at E-Tips, but it should probably not be your primary help and support in your development. And, don’t try to hide what you don’t know from your boss, that can be dangerous, in many ways.

Do you have a mentor, or a senior engineer at your work place who you can go to? Develop these relationships. It’s almost always better if you can be looking at the same plans and specs., be drawing explanatory sketches, asking questions back and forth, see the expression on my face when I kick you in the butt, etc. Usually, they understand that you won’t know all this stuff the first time you face a new problem, and they should know what you do and don’t know, so they can help keep you out of trouble.

 

[jonathanjc]

Hi Dhengr,

Thank you for the honesty. I am learning, I am not a professional engineer and I have completed a statics and strength of material course with a good result. However, answering questions and examples from textbook material with all the parameters in place is worlds apart compared to applying the methods and techniques to a totally new real world problem. I have never analysed a pipeline structure support before. I have though analysed trusses, pressure vessels, cantilevers etc and these certainly do all apply to some degree to this particular problem. This is ultimately my first time in attempting to apply the methods and techniques that I have learnt to a new and unfamiliar problem and it is a skill that removes any comfort given by examples and questions in a textbook, which I am quickly finding out.

I have read carefully everything you have said (sorry for the laziness) I certainly will take note. My firsts thoughts when looking at the problem was, 'how would we know what loading the support would need to carry' and that brought me towards the force equilibrium method. My thinking was to find the horizontal and vertical components and that would (in theory) enable me to find the perpendicular support load. The information I have for the structure is of course the riser pipe loads you see on the diagram. I also have:
Pipe diameter 12in
Wall thickness 17.5mm
Working internal Pressure - unknown
Youngs modulus 207000/mm^2
Pipe Steel SMYS - 414 MPa (Pipe material)
Pipe Steel SMTS - 517 MPa (Pipe material)
Structural Steel SMYS - 355 MPa (Structural material)
Poissons ratio 0.3
Current Velocity @ 1m Above Seabed - 0.25m/s
Water Depth Range (LAT) - 38-40m
Hydrodynamic loading on the structural steel is not considered.
Hydrodynamic loading on the piping is incorporated within a safety factor.

I don't have the length of the total pipe section but from the center of the flange it is suppose to be about 2.5m to the seabed. This is all of the information I have. Is this what you would expect to see?

I understand that the steeper the angle the more likely slippage will occur but the load component of Fx would increase the slippage, wouldn't it? , how would this slipping affect the load on the support? when the pipe is under an axial force at a 60deg angle it will deflect vertically and perpendicularly from the 60deg angled section, the further it slips the higher the load on the support. If the sleeve were not welded at the ends to the wall of the pipe the Fx will have a pushing affect through the sleeve. Could you account for slippage by way of looking at the deflection of the pipe from Fx + self-load, using the cantilever theory? Or would it be better to have it welded and stop movement?

I have attached a photo of the structure for your interest. I am thankful for your tips and I certainly don't intend to hide what I don't know from anyone. I am not afraid to say I don't know something, after-all, how on earth would I develop or learn anything if I didn't ask questions. I don't have many people around me to aid my development and learning so I am taking it upon myself and have seeked Eng-tips to try and talk things through with interested and more experienced individuals. Everything I learn from this problem will be brought to my next problem and so on and so forth. I will be better equipped to think about what I have done and try and apply relevant bits of each experience to every new problem that I tackle. The problem right now is that, I am laying my very first brick and I don't know what position it would be best laid. (Apologies for the cheesy analogy).

Thanks for taking your time on this.

Jonathan

 

[dhengr]

Jonathanjc:
I’m just an old Structural Engineer, Mechanical Engineer and a Theory of Elasticity guy, not a Marine or Petrol Piping guy, so I sure don’t have all the answers to your problem. Reread my last para. above, and find a local mentor, maybe from another company if one is not available at your own company, maybe through joining a Professional Engineering group like ASCE, AISC, AWS, ASME, a marine or oil field group, the UK will have similar groups. You buy lunch or the beers once in a while, but you develop a real helping relationship, where he/she cares about your development, and they share bits of 20 years of experience in 15 minute bites which apply to the problem of the moment. You can be looking at the same details of the problem together, be doing sketches which explain the point being made, look at the same ref. material or text book, etc. etc. That is much more rewarding and productive than trying to do the same by long distance. I’m at rwhaiatcomcastdotnet. I am always surprised today, that a company/boss hands you a problem like this without someone with sufficient experience to help you learn the ropes. Either the boss is too dumb to know how to do it and the complexities of it, or he’s trying to get way to much mileage out of a young fellow. This is really being thrown into the deep end of the pool, without any help. Obviously, you want to be a quick learner, and do some of the study on your own time, to show that you care and want to learn. It is good that you are honest about what you know; and you are right, you quickly learn there’s a big difference btwn. school and the text book problem and the real world. Maybe your boss would know someone who might be a good mentor, and know the industry. Can you get specs., drawings and calcs. on the skids/supports you showed, those would teach a lot. I’ve seen those before, but never actually designed one.

I’ll bet the way the photos show those supports to be made: the sleeves you mentioned are welded directly to the short (25'?) bent piece of pipe and are a reinforcement to get the concentrated loads out of the pipe in a reasonable way, without over stressing the pipe. Note that the two U shaped plates are primarily hold downs, and I’ll bet welded to the sleeve, the longitudinal side plates and the base pl. With drain holes near the base pl.? Then there are two side plates which take most of the gravity loads and the longitudinal loads from the pipe system, and I’ll bet they’re welded to the sleeve and the base pl., and sloped in a bit at the base pl. to allow bolting. Then you finally bolt this sloped support shoe to the skid; and the two bent pipes and the capped end manifold pipe/supply pipe or our single pipe, shop fabed. as one piece, are supported, with U-bolts or U-straps and a small saddle only, at the supply end, gravity loading only. You guys call this entire piece of equip. a coupling skid, transition skid or some such, don’t you?

If you are trying to replicate the photo, I suspect all three loads are taken out of the pipe and into the skid base basically through this sloped support shoe. Who gave you the design loads? Have them explain a bit more about where they come from and how they should be handled at the support. Btwn. your first and second posts you have two diff. sets of loads, with two diff. coord. systems, why is that, where did they come from? I’ll bet you can’t really ignore the dynamic or hydrodynamic loading, but the skid probably is allowed to move a bit on the seabed, and that 12" pipe is really pretty flexible when it is 38-40M long. You could get expansion and contraction movement and stresses too. The skid is in the mud, and close to the seabed, and has small exposed area to the water flow action, therefore small lateral loading; but the vert. pipe is long and exposed. You can resolve load components to any coordinate system you wish, just as we learned in Statics and Vector Analysis, and you do superimpose them, just break it down into simple steps for each load component, then sum them. And, be careful when you change coordinate systems, you may have Fx1 & Fx2, where Fx implies axial loading, and 1 & 2 are the two diff. coord. systems; Fy is a shearing force (transverse force), 1 & 2, and Mz is a twisting moment, probably in either direction, and may change with coord. system. That guy who gave you the loads must explain this a bit more. I would guess these forces are applied at the flange and induce bending in the bent pipe, and slightly different reactions at the sloped support shoe and at the gravity support at the left end. The bent pipe is a beam simply supported at the left end, simply supported at the sloped shoe, and cantilevered out to the flange, actually a beam-column. Again, just break it down into its simplest elements or forms and step through the problem. These types of problems and designs are pretty dependant on experience and industry stds., that’s why it would be good for you to try to find a mentor in your own area and industry.

Then you need to learn a whole bunch about weld detailing and design because this can make or break a design like this. With those loads and pipe sizes, you certainly don’t need full pen. welds for stress reasons, they are very expensive to do. Although, the code you are working to might dictate them, I’d argue they’re not needed. There are a number of books from Lincoln Electric, the US welding people which are very good, and they are inexpensive too. Two of them are by a guy named Omer W. Blodgett, he’s a friend of mine and a very practical welding guy: “Design of Weldments,” and “Design of Welded Structures.” Two others are “Metals and How to Weld Them,” and “The Procedure Handbook of Arc Welding.” Go to their web site and have a look, the two books by Blodgett, teach welding and weld design and the exact engineering concepts and simple logic we are talking about on this project. If you’re not a Structural Engineer, and don’t already have them, you might want to get a couple good structural engineering and steel design text books. Always ask, which is the cleanest and most direct load path, that’s probably what the structure will try to do, want to do, unless you really screw it up and force it to do the impossible by your poor design and detailing. And, that almost always leads to trouble someplace.

Looking at the way the short bent pipe on the skid is supported, you have to take the axial load Fx out of the pipe at the support or you would drive the bent pipe right down into the mud at the bend, by buckling it at the bend, the inside radius of the bend will just crumple. Its only other support is a simple support, 15-20' further to the left. The 81kN = Fx is taken as a shear flow, through the two sloped side pls. into the base pl., and down into the sloped skid leg. The 8 base pl. bolts aren’t even in shear, they are only hold down bolts, because it appears in the photo that there is a shear block welded to the sloped leg on the skid, at the lower edge of the base pl. You will have to know the working pressure of the pipe, and have to calc. the pipe internal stresses, so you can combine them with the stresses your three loads cause on the pipe. You might want to ask about empty pipe and buoyancy, or uplift, issues.

Now, I’m gettin a little lazy, I’m used to working in ft. & inches and lbs. & kips, etc. so without conversions I have a little trouble with an intuitive feel for the load magnitudes, etc., I think we have Fx = 18.23k, Fy = 7.88k & Mz = 76.71ft.-kips, right? And, I don’t understand the meaning of Fa, Fb & Fc at this level of the structure. This is getting a little rambling and it’s getting late, but I think there is some food for thought for you in the above.

 

[jonathanjc]

I make Mz ~ 7.13 kip/ft.

You have given me something to think about certainly. I will take on board what you have suggested and will definitely put it to good use. I have looked up the books you advised and they seem like good reference material and no doubt knowledge in welding would come to good use.

The second set of loads came from results from the FBD I attached. The first set was not right. I will look through it again along with the advise you have been kind enough to give and will step through the problem in the simplest way. I have requested some structural drawings so hopefully will get those though soon.

Thanks again it has been a lot of help.

 

[chicopee]

Looking, closely, at your pictures the sleeve appear welded to the pipe line on both ends. Also your analysis should include the moments, shear and direct forces to the bottom portion of the sleeve since fluid will be flowing in the pipe line and external forces will be caused by the underwater current. In the z-y plane, I see that the fluid flow changes 90 degree directions in two segment of the piping system which cause additional moments, shear forces and forces at the bottom section of the sleeve. In the x-y plane, the fluid flow will change its direction 60 degrees and that will cause additional moments, shear and direct forces on the sleeve. You may or may not discount the effects of buoyancy. To solve this problem is messy so you will need to make assumptions such as basing the strength of the bolts from one end of the sleeve because I think that some of the shear forces, direct forces and moments on the sleeve may cancel out, so you'll have stronger fasteners.