Need a book? Engineering books recommendations...

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

Re: SAP2000, RISA3D, and

[Subject Prev][Subject Next][Thread Prev][Thread Next]

----------
> From: Roger Turk <73527.1356(--nospam--at)compuserve.com>
> To: seaoc(--nospam--at)seaoc.org
> Subject: Re: SAP2000, RISA3D, and
> Date: Tuesday, December 16, 1997 3:04 PM
> 
> In response to Bill Allen's reply to my message:
> 
> . > You probably write your own programs to solve your engineering
problems 
> . > (let me guess, DOS right?) or worse, do most of your calculations by
hand
> . > (still!). 
> 
> You are absolutely right, Bill!  At least I know what I am doing.  In a 
> recent posting to this listservice, there was a request for a computer 
> program that would design/analyze a combined footing with the loads
offset 
> from the geometric center in both directions.  In the time that it took
to 
> compose and post the message, not counting the time to read and digest
the 
> responses, the problem could have been calc'd by hand.
> 
> If I am going to be responsible for a program or analysis error, then *I*

> want to be the one to make *that* error.  I don't want to be responsible
for 
> someone else's error that I have no idea was made, nor who made it.
> 

IMHO, only people who are capable of doing the calculation by hand, and
know how the programs work (i.e. assumptions made in the method), should be
using them.  The ease of using these programs is almost scary.  It was
better, in some ways, back when the input files made it look like magic,
and the programs were intimidating to those who weren't competent with
them.

However, it is unreasonable to expect a client to pay for the time it takes
to do some of these calculations by hand compared to the time required to
make a model.  In many ways, computer analysis is the same as using a
calculator.  It's another tool for the engineer to use when appropriate.

When I was in grad school, and a few times since then, I taught some
engineering courses.  Students would make huge mistakes by punching the
wrong button on their calculator, and they would not even realize it.  This
is the same problem, just on a different level.  Engineering judgement and
experience have to be used to analyze the results.

> . > I have not run across any software author that "seals" his
application ...
> 
> Neither have I, but that doesn't mean that they shouldn't!

I don't think the programmer should "seal" his application any more than a
textbook author or a total station manufacturer (in the case of a
surveyor).  They all give a method to reach an answer to a problem.  As the
engineer of record, it is our judgement that applies these tools.  In my
experience, errors caused by improper modeling of structures are more
numerous (by orders of magnitude?) than errors caused by bugs in the
program.  How can you expect the author of the program to take
responsibility for your model?  He does not have the opportunity to check
your results; you do have that opportunity with the program.

> 
> . > This is clearly the responsibility of the structural engineer of
record.
> 
> Why is sealing someone else's work the responsibility of the structural 
> engineer of record?  The person(s) who wrote the program are the ones
that 
> have selected the method/criteria for analysis/design and *they* are
doing 
> the arithmetic to come up with the solution, so why should the the
structural 
> engineer seal *their* work and assume responsibility for that work?
> 


> How do the programmers handle "bugs" and errors?  Do they just provide
the 
> reportee with a work around or correction?  Do they notify all of their 
> licensed users of the "bug"/error or do they do that only for *major* 
> "bugs"/errors?  How do they determine what is a *major* "bug"/error and
what 
> is not?  What are the qualifications of the persons responsible for the 
> program?  What are their names?  Are they licensed structural engineers
or 
> are they just computer programmers that are using formulae out of some
book?
> 

These are some of the reasons I use my preferred program, VisualAnalysis (
http://www.iesweb.com for demo).  I know the authors.  One was my
structural engineering professor, another was a classmate in grad school. 
Both have worked in private practice.  

Recently, they found a bug in one of their programs.  I was unaware of the
bug, since I had not yet used the feature that had the bug.  I received a
letter informing me of the bug, and sending me to their web site to get the
fix.  Their support has been excellent.

I have not used some of the programs mentioned in this discussion.  It was
long enough ago that I used others that I can't offer much insight into
them.  Everyone's needs are different, however I have been pleased with
VisualAnalysis for my uses.  

To give you an idea of my uses:  I work in a small firm (7 employees total,
including 4 PE's - there is no SE license here).  We do primarily
structural engineering;  Typically smaller jobs (commercial or large
residential);  The largest structure I have modeled on this program was a
12 story hotel in seismic zone 4 with 120 psf roof snow loads;  Seismic
zones vary from 4 down, up to 100 mph winds in nearby areas, ground snow
loads from 30 to over 200 psf within about 50 miles.  Porbably similar to
many parts of California, except that our seismic zones get as low as zero.


Brad Ebel, P.E.
Bozeman, Montana