Need a book? Engineering books recommendations...

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

Re: Enercalc's Structural Engineering Library Survey

[Subject Prev][Subject Next][Thread Prev][Thread Next]
Dennis:

I have been following the threads on Enercalc and spreadsheets in general
because all of our calculations for the approximate 15 years have utilized
spreadsheets.  We began with Visacalc on a Radio Shack model 2.  Don't
laugh, we also ran Ed Wilson's SAP80 in this machine - it was a money maker!

Eventually, we moved over to Lotus with an XT machine and began to develop
basic spreadsheet modules.  As time went on, we were no longer handwriting
any calcs but doing them all on the computer, linking page after page down
the template.  Then in 1989, my son and I started preparing a full blown
spreadsheet program that we began to market in 1991.  Some people may
remember it, the RCT System.  Very, very powerful, still useful today, but
not on the market.  This spreadsheet was modified by an add-in call Ice,
written by the Baler people and was run in Lotus123, originally on a 286.
At that time, some recalc was painfully slow, especially searching the
steel data base for beam or column selection; today it's a blink.

I think that there are about 180 templates, some simple, some complicated,
but the difference in presentation of the program is fundamentally
different from Enercalc's.  Previous to this time, I had been chairman of
the SEONC Electronic Computation Committee and had listened to all of the
criticism of the dangers of spreadsheets such as documentation, tinkering
by designers, etc.  Further, as an employer of a number of structural
designers, the last thing I wanted was my people trying to change the
format or other things.  This can be especially true today with Excel or
the windows version of Lotus - the time that can be wasted with drawing
lines and sketches in the spread sheets.

The basic one or two page module was set up and presented as follows:
1.      You can analyze seven cases at once.  (This fit the page on the
screen).
2.      You can play with the input and watch the output at once.  All the
spreadsheets were constructed similarly - always looking for the D/C ratio
which has to be <1.00.
3.      Each spread sheet had the following items within the spreadsheet:
        a.      A guide to it's usage.
        b.      A written explanation of EVERY cell (documentation).
        c.      The actual coding.
        d.      Almost all of the intermediate steps are shown)

The basic concept was a 30 page spreadsheet on which any of the many
modules could be imported.  Answers (output) from one module could be
linked to another.  Essentially, a blank 30 page file with headers, company
name, page numbers, date and time would be loaded.  Then you could bring in
the first 3 pages, title page, index and project description typed in.
Then the next page(s) you can write regular calcs or determine loads or
whatever - just like regular calcs; then the next page could be a RCT to
perform the actual design and so on.  Pages from other projects which you
have a considerable time investment can be imported into you file.  Just
need to know the file number and the page number.

If you are doing many similar structures, this system is a time saver.  For
instance, we are presently designing numerous monopoles and taperpoles.
The original set of calcs probably took one day.  The big differences in
many of the poles are the wind speed, the amount of antennas and dishes and
the soil conditions.
The RCT file for this is about 18 pages, there are only about two standard
RCT templates in this; the UBC non-constrained pole routine and the UBC
wind loadings.  Especially written originally were the loading pages, the
stress analysis page, a rotation section and a deflection section with a
P-delta analysis, a couple of pages for openings in the poles.  Further
there is a page for the base plate and anchor bolt selection as well as the
welding of the pole to the base plate.  

Time to design a new project:    1/2 hour to 3/4 hour - depending usually
upon the interpretation of the soil report.

One of the basic premises of this approach was to have others to be able to
follow the calculations and hand check them.  Further, the presentation
tries to be Idiot Style.  That is, it can be easily followed and verified,
both the plan checker and by me ten years from now.  (Old habit instilled
in me by Hans Mager in 1952).

Bottom line - we get very few plan check rejects.

Unfortunately, we have found it difficult to move this system over to
Excel.  The command language and macros in Lotus do not transfer very well,
if at all.  The program is geared to 1991 UBC, with the wood section no
longer valid.

Writing the templates and modules was easy.  Providing the documentation
and verification was extremely time consuming.  We found mistakes in the
UBC; talked directly with them.  We found errors in the AISC manual;
corrected them.  Our hold-down selection modules for the five different
manufacturer's data bases had to be continually updated.  Code changes and
interpretations were difficult to keep up with and sometimes even understand. 

And we didn't sell enough of this program to make a living.  But, I sure do
make money off the usage of the program and I haven't told you half of it.

Neil Moore
S.E.