Need a book? Engineering books recommendations...

Return to index: [Subject] [Thread] [Date] [Author]

Re: Structural Drawings with PDMS ?

[Subject Prev][Subject Next][Thread Prev][Thread Next]
Thomas,
 
Thanks for your answer. It is very enlightening on the subject. In fact, there was
a managerial decision, in order that this project should go with PDMS. The model
of PDMS will not be used for analysis. The structural analysis is done with SAP
and/or RISA. They say that the structural engineers and drafters provide the data
for the PDMS model, which is working with the rest of the disciplines.  For the
steelwork, the fabricators get their models for XSTEEL erection and fab. drawings,
directly from the PDMS model.  But, I have heard of another way of working : 
from a STAAD out-put, the structural model goes directly to XSTEEL, allowing
the erection drawings to be generated, and these latter are the in-put to the
PDMS model.
I should be very dated, but in all these flings of files from one model/program
to another, do not feel when and where, we can efectively do the proper and
corresponding checks to produce safe and sound structural design.
 
Regards,
 
Raúl Labbé
     
----- Original Message -----
Sent: Tuesday, June 08, 2004 11:10 AM
Subject: Re: Structural Drawings with PDMS ?


Raul,

3-D PDS systems became popular with the big industrial companies back in the mid 1980's.  Two of the major benefits were interference checking and construction viewing capabilities.  With all the disciplines working in the same space it was often hard to check all the bumps and hits between design groups by hand in a 2-D world.  It is critical that the piping and cable trays sit on top of and not through the steel supports.  It is also tedious and hard to check piping which  may have several inches of insulation that may be next to a steel column with concrete fireproofing.  Most of these programs will print out reports of these conflicts so the design team can work these interferences out before constructions starts.  The second item is the benefit of having a "viewing station" at the job site so the construction team can view the structures, equipment, piping, etc. from any angle in 3-D space.  Views can be printed out and given to the various trade foremen.  This is a big benefit for the rigging department for selection and positioning of cranes.  This also allows the construction team to visualize and plan their erection sequencing.

From a structural design standpoint trying to use the same CAD model to perform analysis has two major problems.  The first is that CAD Designers/Draftsmen NEVER model a steel structure the same as an Engineer would.  The CAD person is very concerned with getting the top of steel elevations correct and the bracing/columns starting and end points correct.  The engineer typically models the centerline of all members and all the work points (member intersections) match.  What happens is that often the CAD model has member node points floating in space due to all the small little offsets to get the dimensions to work out.  Another issue is that CAD operator does not work from node to node.   Columns are often modeled as one member from top to bottom and the beams "appear" to frame into the columns but it is only pictorial and there is actually no node there.  Debugging these models for instabilities due to nodes and work points not aligning is a nightmare.

The other main problem of using the structural CAD model for analysis is that there is just too much extra non-structural "fluff".  The CAD model will have all the miscellaneous 2 and 3 foot cantilevers, all the small channels for catwalks and grating support, handrails, stairs, davits, etc.  This non-structural stuff can really get in the way and is time consuming to take out or deactivate for analysis.  You are often better off creating your simplified structural model from scratch than trying to debug and change the CAD model.  In many situations a 3-D model for analysis is really overkill.  For symmetrical structures and long repetitive structures such as pipeways simple 2-D models may be more appropriate.

As far as the actual structural design capabilities of these programs you will have to review each systems to see if it does what you want.  Few if any go beyond AISC unity checks and all the Building Code prescriptive requirements you must incorporate by hand.

3-D PDS CAD is a wonderful tool for large industrial projects but it is unlikely you will find any savings from the structural engineering viewpoint.  You are probably still better off doing separate STAAD, RISA, or SAP models for your structural needs.

Thomas Hunt, S.E.
ABS Consulting



"Rlabbe" <rlabbe(--nospam--at)cadeidepe.cl>

06/07/2004 01:35 PM

Please respond to
<seaint(--nospam--at)seaint.org>

To
<seaint(--nospam--at)seaint.org>
cc
Subject
Structural Drawings with PDMS ?





List,

I would like opinions on this issue. Regarding large industrial engineering
jobs where all disciplines are involved, piping, mechanical, HV&C,
structures, etc. it is coming a usual procedure to handle all of the design
activities through PDS or PDMS modeling tools. It seems to me, that these
fancy models have been conceived for piping / mechanical lay-out issues.
How benefitial is this way of design, from a structural standpoint ? The
vendors are usually mentioning steelwork design, as one of the resources
availabe. The main and critical structural issues, like special seismic
connections,
or details, are they well covered ? The out-put, is it proper to check all
these
issues ?, etc.etc. Anything about the reinforced concrete structures, in
this
environement ?
Any experience, or case study, or opinion about, would be appreciated.

Regards,

Raúl Labbé S.E.
M.ASCE.


******* ****** ******* ******** ******* ******* ******* ***
*   Read list FAQ at: http://www.seaint.org/list_FAQ.asp
*
*   This email was sent to you via Structural Engineers
*   Association of Southern California (SEAOSC) server. To
*   subscribe (no fee) or UnSubscribe, please go to:
*
*   http://www.seaint.org/sealist1.asp
*
*   Questions to seaint-ad(--nospam--at)seaint.org. Remember, any email you
*   send to the list is public domain and may be re-posted
*   without your permission. Make sure you visit our web
*   site at: http://www.seaint.org
******* ****** ****** ****** ******* ****** ****** ********