Hi Len

The piping engineers in our company used to also spend a lot of time compiling loads and interfacing with the structural engineers re support loads and design of pipe supports.

We build CAESAR II models, often of cross country piping, which may be at any angle and not in a plant co-ordinate system at all which means the structural engineers (or the stress engineer) also had to resolve all their loads into a local co-ordinate system.

To make it easier we developed an inhouse spreadsheet that takes the restraint reactions from the CAESAR II output, resolves all the support loads into a local forward, sideways, down direction and also combines and factors the laods as required by the local country structural codes. The stressing engineers do this prior to giving the loads to the structural engineer.

This results in a consistant set of loads, the structural engineers always get the information in the same format, (and they have then written their own spreadsheets to automatically design/select the pipe supports based on the loads they have received.)

It also makes it very easy to track changes when support designs need to be checked as the piping changes.

We can complile loads for perhaps 150 supports, or even more, in a few hours.

While most of the time we are having the supports designed by structural engineers within our own company, (although often in another country), we have had times where another company has designed the pipe supports and as such having the support loads presented to them in a neat, easy to understand format makes life so much easier for them and us.

I guess if you have no control over the format the support information you are given is in then it is tricky, especially depending on who is the dominant party - ie can you tell them what you want? or can they tell you, this is what you are getting - you sort it out?

While we are not at the point where we can output the support loads such that they can be read by a structural program, because all the pre-processing of the loads has been carried out it is still much easier for the structural engineers to use these loads rather than them having to process all the loads into something meaningful before they can even begin the design of the supports.


We run both AutoPIPE and CAESAR II and our speadsheet (and all the macros) can be used with either package which makes it really good. It might well be worth your while to develope a similar spreadsheet that can automatically process the support loads you get. It does take time to develop but if you design supports day in and day out then it really is worth spending the time to develop a suitable system.
_________________________
Miss Itchy

Hello, Miss Itchy,

Thank you for your response.
If a structural engineer is assigned to design singular supports for one given pipeline, the utility I’m developing is not needed and the database you described can be a sufficient tool.
If a structural engineer is designing let’s say a 200m-long 4-level pipe-rack with average 8 pipes on each level, he is dealing with approximately one thousand support points situated on the same structure (200/6*4*8=1067)
Each pipe belongs to a Piping System and each System stretches across the Plant (or a Design Area) like a gigantic octopus (and each ‘Octopus’ is usually assigned to a different Stress Engineer).
Than, no database would help, the engineer should be able to see his pipe-rack at somewhat Graphic Interface, with all the Piping Systems that pass through it, he should be able to cut out his ‘Area of Interest’ to isolate only those parts of ‘octopuses’ that are confined within his pipe-rack, and view the support points (and maybe pipe center-lines) hovering above or aside the beams of his pipe-rack.
This part is mostly implemented; we still have some issues with bringing Stress Output to Plant Coordinates but the functionality and interfaces are there.

Yet, the ‘Gray Area’ for us now is correlating restraint reactions resulting from Stress to the Structural Loads from Piping (by the way, these loads are formalized not by National Building Codes of different countries but by Industry Guidelines for Structural Design accepted in the given EPC-Design office)

In this respect, the most interesting process you described is when your program “takes the restraint reactions from the CAESAR II output, resolves all the support loads into a local forward, sideways, down direction …”.

Could you, please, elaborate on this?
Restrain Reactions may have three for or more Sets of Load (XYZ), each set resulting from a particular Stress Combination.
Yet, as a rule, Structural Guidelines stipulate only Four Load Cases:
1. Dry Piping weight
2. Piping Content
3. Friction
4. Anchor
What Set do you take and assign as “local forward, sideways, down direction”?
Do you relate to the type of restraint: for example, Directional Anchor >> Forward and Guide >> Sideways would fall into “Anchor” Structural Load Case?
Do you single out ‘Gravity’ Stress Combination to split it onto ‘Dry Weight’ and “Piping Content” or just take the maximal vertical-downward value regardless of what Stress Combination it came from?
Do you exclude Seismic Stress Combinations from your Database processing?
Perhaps, the problem is that Stress Engineers do not produce Combinations suitable for Structural; after all the Stress the main goal is to test the Piping System for the worst case scenario but not to assist with the design of supporting structures? Than, we could look into Stress Load Cases and determine what Combinations would be more suitable for the output to Structural Discipline (though not always valuable for Stress Design as such)?

I hope we’ll continue this interesting discussion…

Cheers,
Len

Hi Len

I wrote a nice long reply to this the other day but must have forgotten to submit it after I checked the preview so here is a response again.

I see your point regarding having multiple lines on a pipe bridge, whereas we typically work on stand alone pipe supports, there might be three or four pipes sharing a support at times but typically we would not have a large number of supports being supported on one structure.

Typically we give the structural engineers all loads that are in the piping model, i.e. gravity, pressure, temperature, seismic (usually 4 directions), wind (usually 4 direction), point loads, etc etc. There might be 4 or 5 different temperature cases as well so there might be 10 or 15 base loads. We extract the base load information from the CAESAR II results and then resolve this into a local axis system, then factor and combine the loads.

We use the following local axis system, forward is defined as standing on the pipe and looking in the direciton of fluid flow, this then defines left and right, thus we resolve the loads into an axial load (forward) and a load perpindicular to the pipe (right) and well as a vertical load.

We use this local axis system for all support types ie guides, slides, anchors, line stops. It is also very useful because once we start talking about pipe supports with a left gap of 50mm and a right gap of 10mm then we all know which way is left and which way is right.

If we are unsure of the required structural code combinations we supple the resolved base loads ie. Gr, P1, T1 etc. If we know the combinations that the structural engineers need we can supply those (i.e. 1.4gr+1.4P1+1.4T1+1.1Ex etc). We might have 50 or 60 different combinations we need to define for each support point.

We prefer to manipulate the restraint loads ourselves rather than just giving a CAESAR II print out of restraint reactions to the structural engineer as we are obviously much more familar with the piping system. We also check the loads to make sure that there are not stupid loads (i.e. loads that are way too high that by opening up some gaps can be reduced from 110kN to 10 kN).

This saves heaps of time interfacing with the structural engineers and avoids loads of questions and also gives a way to keep track of the loads that have been produced and sent to the structural engineers. Once we have the loads we issue them as a deliverable, which also makes it very easy to keep track of changes - we just uprev the support loads document like any other deliverable document.

In the end a lot of it may come down to who is the customer and who is the client? i.e. can you tell the piping engineers I need the support loads in this format, or can they tell you this is the loads your getting, what you have to do with them now is your problem. I would have thought you should be in a position to request that the loads are manipulated and resolved before you get given them. I guess some piping engineers jobs are done once the CAESAR II model has been run and is code compliant, but we certainly like to keep control of the piping and pipe support interaction where possible (including making sure that the pipe supports are long enough and the pipe shoes are long enough to make sure that when the piping displaces it does not fall off it's pipe support).

(hope this provides you with some more food for thought - I have also sent you a personal message view this board if you want to have a look).

Karen