I am embarking on creating a Pipe Support Standard for our company.
The standards will include allowable loads, recomended applications, etc.

Does anyboby:

*) know of good books or references i can study for help.
*) have an opinion on how to approach the determination of allowable loads? (Computer or hand or both)
*) Suggest good programs to use?

I am leaning towards purchasing a FEA program to start with, (FE/PIPE)?

I've previously used the kellogs methods in conjunction with roark, but i don't really think this is the way to go for a Pipe support Spec.

Alternatively, i've thought about approaching someone like Carpenter and Paterson for a more detailed load matrix than supplied in there literiture

Come on give me some good ideas.!!!!!!

I find FEPIPE very useful for many detailed looks at things in our business.... vessel nozzles and local attachments BUT like all things FEA has its hidden dangers so buy some training and not just the software.

FEA has its hidden dangers so buy some training and not just the software

Fern are doing an intresting starter course at the moment, looks ok
Training? Nah, i'll just hit the GO button, that'l do wont it


Now paying for all this is another matter...

My condolences.

"Corporate Standards" are a very dangerous thing. You work your behind off, with no budget, for quite some time, at the expense of your billable work. Then you publish them. People ignore them, but somehow that's YOUR fault.

Time passes. Standard practices change. Analysis tools are improved. You have no budget, time, or procedure for updating your standards, so they remain static. Eventually, someone notices that the work you did in 2007-08 is no longer valid; or, worse, something fails and investigation reveals that your work does not meet the 2018 Code. That, too, will be YOUR fault.

Standards require eternal vigilance. Thus they also require, and are seldom given, annual maintenance budgets.

Overdesign the hell out of your standards, and then have a note allowing higher loads "if a more detailed analysis of local conditions permits". That should keep your retired butt out of hot water down the road.

How about,

BS3974:Parts 1, 2 & 3 - Specification for Pipe Supports

It may well be withdrawn but contains dimensions, safe loads, some guidance on design & welding and covers most simple support systems.

Allowable load determination has got to be via first principles + suitable structural steelwork code. FEA would seem more suitable for one off design issues. If you are looking to define your own standard then the design of each component should be resilient enough to cope with the variables likely in fabrication, installation and use. You need to be able to state a SAFE working load. FEA would typically be used to optimize a design. Do you want your standard support to be designed so that there is just enough strength for the expected load?

Following the "If it's not broke..." philosophy can you not use/verify the swl's specified in standards & catalogues and spend more time on the application side?

Just use another companies support standards & change the headers as long as they are a well recognized company with a good track record then you should be OK no need to reinvent the wheel.
If you don't feel happy with some of the loads do some simple hand calcs.

When you work for various companies & see there support standards you begin to notice a lot of similarities usually because they have all copied off each other.

It's a mammoth task to come up with good support standards that can be used for most supports.

So its OK to steal???? Well frankly thats a surprising comment from somebody in the U.K., but we are all individuals I guess!

Companies have these standards. Where possible I take Licad (or similar from other supplier) from www.lisega.co.uk (lisega.de). Makes drawings for me. I just wish they would release EasySteel for steel parts. They have been promising it for 2 to 3 years.

Benefit with these programs is that they have(especially Licad) database. Easy to order parts. Easy to erect also.

Be careful with FEA. Can be costly and wrong results.

SUPERPIPER, you have quite the task ahead of you!

The training effort for FE/Pipe or other FEA will certainly pay off, I bet at least you'll break even on time. If you can take the FE/Pipe course directly from Tony Paulin, do it. He's phenomenal in action.

I've been involved in developing or analyzing sets of pipe support standards, for generic Oil & Gas stuff. A few things I've come to accept is that FEA is not substitution for hand calculations. If there is ever a failure and a court case you have to show that you've done everything you can, within reason, to prove the support design.

This also means you can't just copy another company's supports. I don't think the judge would be to impressed if that was your argument.

I use FEA (FE/Pipe AND Cosmos) to determine local stresses in the pipe wall, along with Roarks AND Kellogs hand calculations.

I use hand calculations exclusively to determine capacity of the welds. Using FEA to determine weld capacity is difficult, if even possible. The weld on/off in FE/Pipe is just to determine the location to calculate the stresses (failures occur typically at the toe of the weld). In solid modeling (Cosmos), if you have full penetration welds the models work okay, but with fillet welds you get a hinge effect where the stresses are infite at the discontinuity. So I'll model with and without the fillets, but model as full pen and take the higher stress. But this doesn't check the weld capacity, just gives a conservative estimate of stress in the pipe wall.

I find the combination of hand calculations and appropriate FEA gives me more confidence in the design when the numbers sort of match. They are never exact, but depending on the geomety, if the hand calc.'s are within 50% of the FEA you are doing good. ie., if you built an FEA model of a Roarks case, you would get much closer results, but in pipe supports the geometry's are actually much more complex and the relative stiffness' have a large effect on the results.

Hmm... well actually I was quite shocked to discover that a FEA analysis showed that a standard support detail from a well recognized company was not, in my view, suitable for use (I had other options, so I didn't use it)... but if my name were going on the standard, I would not just be copying an existing one... now if someone said "what do I use" I might say "this is typical"

SLH

Their are only so many sensible designs of pipe supports, if you tried to do a corporate standard without using any support types that are in another companies corporate standards you wouldn't get very far.

What company first started using bolt on slippers or duckfoots?
Does that mean no other company should use slipper or duckfoot supports because that would be stealing the design?

I am sure lots of designs of pipe supports wouldn't pass FEA analysis but if the same supports have been proven in plants for a lot of years then what is the problem with using them.
The computer & it's software isn't the answer to everything.

I would

(1) Use hand calculations/Roarks as far as possible if I am not an FEA expert. The reason I can give is most of company standards were developed in the era before FEA was ivolved and they are working fine/proven.
(2) Use good factor of safety for allowable loads.
(2) Surely involve Civil/Strucutural engineer in design/verification.
(3) If entirely remaking new standard, use other best company standards as reference to get ideas about good designs/concepts.

Don't get me wrong for pt. 3. I am not talking about copying/stealing something, but there is nothing wrong in getting ideas about good concepts from various standards and taking best concepts out of them. When doing calculations for loads, we can compare allowable loads with proven standards and we can figure out how much factor of safety they used etc. If we see supt. stds. of various EPC compamies we can see that they are quite similar and hence there is quite possible that each company may have developed few concepts and other concepts came from other standards/with people moving from here to there.

I will honor the one which is working and is proven rather than calculated if its more conservative. Afterall all calculations/equations are based on what worked/not worked in a particular conditions.

I agree with all the disapproval of pirating another company's standards.
This is unethical, and may not even be safe. There may be unstated procedures that support "their" method of building supports; when "your" construction procedures are used, you may end up building something that has much less (or no) safety margin.

Use load rated components where possible - riser clamps and shear lugs are less elegant than trunnions, but the design calculations are much simpler. Similar considerations apply to using rod hangers to build trapeze supports vs. welding them up with angles for the load-bearing members.

Keep standard supports as simple as possible - the intent of building a set of standard supports is not to cover everything, it is to provide a collection of basic supports that handle a large majority of cases with a very, very high degree of reliability. To support a line on shoes to knee braces off a column row, for example, I would only include the shoe (and perhaps a slide plate or guide) as a part of the "pipe support" standard. It's not that I don't trust other engineers to design a knee brace correctly, I don't trust a construction foreman with a set of "my" pipe support standards to check the design of the knee brace.

As senior engineer types, we all fall victim to the thought that we can do it better than others. We have to realize that a newbie engineer, faced with a defined set of loads and dimensions, can do a damn good job of designing a knee brace. On the other hand, we may struggle to adequately define the load envelope that applies to a "standard" knee brace of our own design. It's not that we don't know what the envelope is, it's that it may be difficult for us to communicate that information to a construction foreman in a manner that makes it impossible for him to misuse it.

For example, we design a knee brace for "no more than 2 kips at 2'-6" from the face of the column," and the construction guy puts two lines, each with a 1.5 kip load, on the knee brace; the outer one is at 2'-6" from the face of the column and the inner at 1'-6". Boom, it falls down. Or he puts the knee brace on a column that is already loaded heavily, and the column fails. This all ignores dynamic loads, and dynamic load factors, too. Yech. Better to leave that can of worms for your pipe support design team, or the site structural engineers.

My point was that I'd like something other than "well the other guys used them so I copied them" should I ever end up in court after something failed. So, if I'm writing the standard, I will do enough analysis to satisfy myself.

The particular support that I looked at was one that I've never personally used before, I wasn't comfortable with it, I checked, I still wasn't comfortable with it, so I chose a different option. There are a LOT of supports in the standard support list for some companines that are only very rarely used if ever.

I didn't say "don't have standard supports that look like eveyone elses" I just said "if you are responsible for them, then you have to know they are good, and I think something stronger than "because I copied them from so-and-so and he's good" is perhaps reasonable to ask...

FEA wasn't done because FEA is a wonder solution to everything, just because it was a reasonable quick and dirty check... there's a LOT of things out there in the real world that don't meet current design code -- that doesn't mean that the work I am responsible for should not.

SLH

Another good reference to consider:

MSS SP-58-2002

You might want to look at the pipe support standards that are published by PIP. That is Process Industry Practices at the University of Texas at Austin. www.PIP.ORG
Regards, John