Need a book? Engineering books recommendations...

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

Re: Seismic/Non-Seismic design

[Subject Prev][Subject Next][Thread Prev][Thread Next]
Petro-chemical process plants, electric power generation plants, and
other similar industrial "nonbuilding" structures share the same
concerns.  Concentric braced frames generally work the best to resist
the earthquake (regardless of zone) or wind forces.

Another major concern is the control of lateral drift.  Although the
pseudo static lateral loads are lower with EBF's or SMRF's it should be
noted that the better seismic performance is due to the ability to
absorb energy as the structure displaces laterally.  This displacement
can fracture piping and attachments, and put the plant out of service.
Evaluate the performance goals of buildings as opposed to the required
performance of a process plant.

The lower R value (indicative of concentric braced frames as opposed to
EBF's) implies more of a required elastic response.  The systems that
dissipate seismic energy better also displaces more.  Generally the best
seismic systems for industrial process structures are concentric braces.
 Within this family of systems you can still select a Special Concentric
Braced Frame (which has good seismic performance with rather easy
detailing design requirements), the traditional X brace, single diagonal
brace, Chevron brace, etc.  Generally, I select from the family of
concentric braced frames based on the process piping requirements and
clearances.

Occasionally I will use a moment resisting frame for pipe racks to
optimize piping layouts, but care needs to be exercised to accommodate
lateral drift and piping anchorage design.  Generally, if the piping
layout requires a moment frame, I will use an Ordinary Moment Frame as
opposed to a SMRF.  The penalty (lower R value) is generally offset by
the lower lateral drift and ease of design.

There are some restrictions currently in the UBC (example: 160' height
limitation on X braced structures) that we have a problem with in
nonbuilding structures.  The 1997 NEHRP Provisions has attempted to
rectify these issues.

Regards,
Harold O. Sprague
Black & Veatch
 ----------
From: James_F_Fulton(--nospam--at)rohmhaas.com
To: seaoc(--nospam--at)seaoc.org
Subject: Re: Seismic/Non-Seismic design
Date: Wednesday, September 03, 1997 8:12AM

     I am interested in this topic also for one and two story industrial
     buildings and chemical process structures that could extend as much
as
     5 or 6 stories high.  We typically design these as x-braced frames
     with simply suppported girder-to-column connections. Most of our
     designs are in non-seismic areas (Texas), but what about this type
of
     structure for other seismic zones ? What are the drawbacks ? This
type
     of steel construction seems to be much more economical to design
and
     construct than moment frames or even eccentrically braced simple
     frames.


______________________________ Reply Separator
_________________________________
Subject: Seismic/Non-Seismic design
Author:  seaoc(--nospam--at)seaoc.org at Internet
Date:    9/2/97 9:38 PM



On 2 Sep 1997, Ghassem Khosrownia wrote:
>      Besides being fun to design and detail, is it worth (economically) to
>      incorporate in seismic zone 2, 3, or even 4? Anyone has any
>      experiences to share on single story industrial type application?

We did some investigations for a typical commerical structure.
(2-1/2 stories, length about 100 m and width about 40 m).  Our
finding was that for this structure, highest seismic zone
provisions will increase the structural cost by only about 7-8%
compared to non-seismic design.  However, as you can imagine,
the analysis and design cost (that you incur) will be significantly
higher for seismic design.  I do not expect the single storey
industrial building to show significantly different figures.

By the way, for highest seismic zone, about 40% of the members
had to be designed for load combination including seismic, while
for the remaining, the dead + live load continued to be critical.

I will also be interested in knowing of other people's experience
on this issue.

Ravi Sinha