Friends and colleagues, lend me your ear.

In the spirit of reducing man-hours, I have been handed the task of evaluating the possibility of transferring piping files from PDS and PDMS to CAESAR. Does anyone out there have any experience in such matters?

My concern is that, noble though the effort may seem, the savings may not amount to a hill of eraser nubs. Many of my "seasoned" colleagues also doubt whether anything done by pipers in our effort can be trusted. This may be a valid concern, although I am willing to give designers the benefit of the doubt.

It seems to me that in order for there to be any benefit, the minimum that should be included in this transaction are pipe O.D., wall thickness, C.A., branch connection types, support locations (this is going to be the big one), and possibly valve weights. In the "It would be nice if we could get" category, I have included Design temp. and pressure, insulation and specific gravity.

I do know that PDMS requires a third program, StressC, to translate the piping model to a .cii file, while PDS, supposedly, expots it directly.

What is the .cii neutral file able to pass along? What is reasonable to expect from the CAD databases? Are there any other issues to consider?

I would love to hear war stories you may have encountered.

Richard Havard

------------------
Richard P. Havard
Mechanical Engineer
Mustang Engineering

Richard,

I believe I can shed some light on this issue.

At Shedden Uhde in Melbourne we have developed our own way of transferring data from PDMS to CAESAR II. We decided on doing it ourselves after our parent company in Dortmund thoroughly investigated CADCENTER’s PDMS to CAESAR II interface and found that gains in productivity were not significant.
They concluded in the report that stress engineers would be replacing modeling work with time required for correcting transfer errors and/or adding missing information and that the system wasn’t practical to use efficiently on the project.
UHDE India on the other hand bought the interface and in order to achieve satisfactory operation spent some time re-coding it. They claim, that now it works well for them.
Before we decided to proceed with developing our own system we already had experience with CADWorx and its bi-directional link with CAESAR II. You can read about how this interface works and what it does in numerous issues of COADE Mechanical Engineering News. I believe that this interface would satisfy almost everything you are looking for in a transfer. Working with this system gave us the opportunity to learn about the real gains in productivity, taught us what to expect from a transfer and what to aim for if we decided to do something ourselves.
After using this interface for nearly three years I am convinced that on its own it works very well indeed.
Based on this experience, in 1998 when we consider the idea of transferring data from PDMS to CAESAR II, rather than trying direct transfers, it was natural for us to concentrate on a much simpler task of transferring data into CADWorx. You can find about how we did this in COADE Mechanical Engineering News, Vol. 25, June 1998. We have completed the basic task in a very short time and performed testing on various projects improving it on the run, until satisfactory performance was achieved.
One new thing which emerged out from this venture is that we can now cleanly distinguish between three working environments. PDMS where all modeling is done, CADWorx where stress analysis work is prepared, organized and reported and CAESAR II where analyses are carried out.
As you might suspect we had to adopt a whole new organizational structure to cope with the new way of doing things and the increase in overall productivity. At first people didn’t quite understand the need for this and we had problems. In order to achieve even greater gains they started sending huge files over which were taking only a few minutes to do whilst using conventional methods of modeling would normally take up to 3-4 days. Off cause these files were useless until the piping has been grouped into systems suitable for analysis, supplied with all necessary information and somehow documented, because traditional stress sketches have been eliminated from the process by then. Some may argue that there are many potentially problematic areas in this system but as I mentioned, organizational structure and awareness of the new system also needed upgrading and once we accomplished that the gains in overall productivity were immense.
These days apart from piping geometry we can transfer from PDMS into CADWorx any information needed for analysis such as possible support locations, process data, corrosion allowances, etc.. This is done in a form of attributes attached to the piping system. How much of this information gets transferred into CAESAR II depends only on the current status of CADWorx to CAESAR II interface. In my opinion it would be sufficient even if only geometry and supporting locations were transferred correctly and it is arguable if anything else actually needs to be transferred. Another benefit from the system is newly created ability for stress engineers to very quickly assess the actual requirement for in-depth analysis. We found out that many lines previously submitted almost by default for analysis were actually OK, whereas many others which in piping designers opinion could have been solved by simple field routing, were actually overstressed and needed attention.
Recently in an attempt to close the “design loop” we have also created a system to send the calculated restraint loads from CAESAR II back into PDMS, placing them in locations where they will actually occur. Following, structural engineers can pick the loads from these locations and perform their analyses.
I hope this information is going to be of some help.

Misa Jocic
Principal Stress Analysis Engineer

A good question. I would advise you first to look at the CAESAR II documentation files on the subject matter things such as external interfaces and the neutral fiule format are well documented.

We have occasionally tried the PDS interface which does work after a fashion. However because dratsmen (oh excuse me designers) usually control the data base and the design use of the program, the "help" for the piping engineer is usually limited.

For instance to the draftsmen it make sense to creat a single valve which can be used time and time again, however it is far to much trouble to assign a weight to the valve the "stress guy" can just look it up time and time again.

So what am I trying to say?
1) CAESAR II documentation thoroughly outlines the interfaces and neutral file formats.

2)A time saving is very difficult to obtain this issue is a "maybe" in my experience.

3)The draftsmen calling themselves designers if allowed to control the program, its data base, and use, by themselves alone will not end up with a product that will be as useful to you as it could be.

4)If the draftsmen you work with are truly designers problematic issues such as weight supports and thermal displacement strains in the systems you analyze for the most part will be minimized decreasing your work load far more than any interface can.

So thats my soap box opinion of the day, I hope that you and your design crew can work things out to a mutually satisfactory conclusion. Computers are tools for a skilled worker, but they cannot replace the skilled worker.


Best Regards,

John C. Luf

[This message has been edited by John C. Luf (edited August 24, 2000).]

Richard,

The whole interface starts with the issue that you need a correct PDS model. This means that all piping components do need to be modeled completely if you do not comply to this your file transfer will not be successful.

With the present interface between PDS and CAESAR II on which we have worked very close with Coade (Richard Ay) since 1998 we are now able to transfer the following data transfer:
 Piping configuration ( The CAESAR II interface allows you to filter out small bore branches like drains and vents).
 Pipe sizes and wall thickness (No CA this is not specified in PDS)
 Support point locations if modeled by the PDS designer
 Insulation thickness (The interface give you the possibility to follow what has been used in PDS, SI or imperial units).
 Rigid weights if they are entered in PDS.
 Flanged elbows.
 Correct branch types.
 Pressure and Temperatures, this needs to be checked by the stress engineer on the basis of cold / hot situations.
 Plot rotation to find same north directions as on isometrics

At this moment only one item is still missing and that is dummies on elbows but this is now under construction with Coade Inc. Hopefully it will be available soon.
All in all we are very pleased with present interface. However, there is only one issue and Misa Jocic did mentioned it already at the end of his debate, the transfer of restraint loads to structural.
An other option as what is used by Shedden Uhde, would be if we could transfer the restraint loads from CAESAR II into a CADWorx stress isometric, but this issue must be picked up by Coade Inc.

I hope this information will help you.

Henk Vodegel
Supervisory Piping Engineer
Fluor Daniel
The Netherlands

The above posts correctly summarize the state of affairs regarding the interfacing issue. There are basically three requirements for the successfull translation of a CAD model into a Stress model:

• The designer must model the system correctly, including items such as temperatures, pressures, restraints, and weights.
• The CAD software must be capable of generating a neutral file containing this information.
• The Stress software must be capable of properly translating this neutral file.

Once this neutral file is available, the Stress program can read it, and generate its own model.

The CAESAR II neutral file (.CII) which PDMS creates is a CAESAR II file format. This file contains everything in a CAESAR II model, it is 100% complete. The best way to see this is to take any existing CAESAR II model, and generate a .CII file. You can then read this file back into CAESAR II (under a different name) to see how this neutral file interface works.

With regards to the ability to send CAESAR II restraint loads to CADWorx/PIPE, this ability is currently available. The image below shows a small portion of a CADWorx/PIPE stress iso generated from a CAESAR II model.



For a better image, click <a href="http://www.coade.com/ubb/msg_images/cx_c2.htm"> this link</a>.

<font color="0000ff">In all translations between software packages, it is imperative that the stress analyst verify the correctness of the stress model.</font>


------------------
Regards,
Richard Ay (COADE, Inc.)




[This message has been edited by rich_ay (edited August 25, 2000).]

Before the topic is closed I would just like to add something.
For non-PDMS jobs we always use CADWorx/PIPE to produce restraint and other reports from CAESAR II. For PDMS projects we also export from CAESAR II into CADWorx/PIPE to create analysis reports, but in addition send restraint loads directly into PDMS.
The reason for direct export was that structural engineers were not satisfied with the format of CADWorx/PIPE or other standard CAESAR II reports. No matter how good or bad these reports were they always had problems interpreting them. To avoid spending considerable amount of time helping structural group to resolve this issue we created a method of direct transfer into PDMS.

A step by step guide of how the transfer method works is disclosed below.

In addition to the transfer file created for CADWorx/PIPE, PDMS also creates a simple text file for each transferred line constituting of four columns:
Column one with Unique Support Names and three columns with coordinates (East, North and Up).
Following, CAESAR II global coordinate system origin is adjusted to mimic PDMS’s system and when the analysis is completed, from standard CAESAR II restraint report, using a program we developed, a file is created containing summarized restraint report with four columns:
Column one with Node Numbers and three columns with coordinates adjusted to correspond PDMS designation. Additionally, a file is created with our program using CAESAR II coordinates report containing four columns:
Column one, Node Numbers and column two, three and four containing coordinates adjusted to PDMS nomenclature.
In order to replace CAESAR II Node Numbers with Unique Support Names as defined in PDMS, two files with coordinates, one created in PDMS and one with our transfer program, are compared within the same automatic transfer routine and matching coordinates used as criteria for substituting Node Numbers with Unique Support Names in the first column of the summarized restraint report. Following, the resulting report file is read into PDMS and loads distributed to supports at their unique locations. The loads information can then be called up at any stage within PDMS or alternatively a printed report issued for the structural group.
This system works well but I have a “small” problem. Our PDMS group for various reasons doesn’t like the idea of using coordinates as a criteria for substituting CAESAR II Node Numbers with Unique Support Names. I can’t think of any other link at the moment and would appreciate any suggestions.

Regards,
Misa Jocic
Principal Stress Analysis Engineer

Richard, this is a hugely exciting area of development. Over the years in my company we have worked on making this data exchange succeed, both by developing our own software products and by working with the software vendors to enhance their import/export functionality. A combination of work process issues and software technical issues have prevented us from developing the full potential of a CAD/CAE transfer (although we do perform a number of CAD-to-CAE data transfers with the current systems.)

You are fortunate in addressing this issue now because the technology that has come out this year and will (hopefully) come out next year will do much to enable the key factors to a fully successful CAD/CAE electronic data interchange that I have listed below:

a) Stress engineers must be able to read the CADD model data directly. They cannot rely on intermediaries (typically CADD coordinators, IS&T support people, or other high-end database types of personnel) to generate their import data for them. The stress engineer gets the benefit of importing data (i.e. eliminating some of the drudgery of 3D input), therefore the stress engineer is the one motivated to make the transfer happen. The transfer tool must be designed for, and operated by, the stress engineer.

b) The design work process must support doing preliminary design in the 3D space, rather than generating manual or 2D sketches of a potential layout prior to going into the 3D space to model.

c) To realize the full potential of a CAD/CAE linkage, the linkage must be a real-time relational linkage. By that I mean, there is no 'translation' or 'import' of data. The CADD model must be read-only exposed to the stress analysis engine. The stress analysis engine will take a snapshot of the relationally exposed input data as documentation for what was run, and as a container for the stress engineer to add additional data. The relational hook to the live CAD model allows a basic change management functionality between the snapshot input data and the evolving CADD design model.

There are a number of other issues that impact the level of success you can have with a CAE/CAD linkage, but the three issues above are the ones I believe are the most important.

I have been assigned to look into PDS to Caesar II transfer. I was able to generate PDS neutral files. However, when I plotted the isometrics in Caesar II, the routing looks a little different. I was using the original defaults.dat file in PDS and I had to add a couple of components to pdstress.tbl to generate the neutral file. Other than that, I did not make any changes to the PDS files. I believe there are many people out there who have experienced this before and hopefully found solutions. Anyway, please let me know your experiences and fixes. Your help is greatly appreciated!!

The geometry should have come across to CAESAR II without a problem. Issues releated to this are outlined below:

• Make sure you are running CAESAR II Version 4.30. This version has the latest Intergraph interface, which includes code to address vertical dummy legs. (Horizontal dummy legs are not correctly associated with the piping and will appear at the beginning of the CAESAR II model.)
• Intergraph has a check list of 6 steps to perform (to ensure a good model) before generating the neutral file. If the CAD model doesn't have proper connectivity, any software using the neutral file will have a bad model. (Your profile doesn't include an e-mail address, otherwise I could send this checklist to you.)
• Occasionally we encounter files with coordinates that cause <em>precision problems</em>. This can happen when global coordinates are specified with values in the tens or hundreds of millions and element lengths are fractions.

Other than these issues, there is no reason why the model shouldn't transfer correctly. Did the CAESAR II Interface generate any messages (such as "n disconnected segments")? If so, you really need to run through the PDS checklist. After you do this, if you still have problems, e-mail me the neutral file.


------------------
Regards,
Richard Ay (COADE, Inc.)


[This message has been edited by rich_ay (edited April 18, 2001).]

Dear Richard,
I know that from PDMS I can import a file compatible for Caesar ii , but this for piping, now if I want to JUST VISUALIZE the structure (not calculate) which is a pipe rack for facilitate the work for structure engineer on locating the supports with its number, Is that possible?
Or I have to pass through cadworks, if so how i can do this ?
Thank you.

I'm unaware of any link between PDMS and CAESAR II related to Structural Models.

Thank you Richard for your prompt reply.
What are the software or format that Caesar II able to read if we look for visualize the support or structure with selected piping ?
Regards.

PDMS generates a CAESAR II Neutral File.

Yes but this for piping, it can include structure (pipe rack) for visualization ? If yes how we can do this ? I need your reply Richard, Thank you

Sorry, at this time, the only CAD Structure (images) that can be imported into the CAESAR II Piping Input (for visualization) are from either CADWorx or SmartPlant.