Advanced Structural
Designs ACN 097 789 87
92 Vasey Crescent Telephone 61612171 Facsimile 61612170 Email mal@structuraldesigns.com.au
Which is the
best prestressed concrete design software?
I come from a background of designing
prestressed concrete structures for the last 20 years starting with hand
calculations and spreadsheets on anything from concrete slabs to incrementally
launched curved prestressed bridges. It is
amazing how far some judicious load balancing can get you but eventually
everyone has to deal with Secondary (Hyperstatic)
effects and the number crunching starts to become both tedious and complex.
After moving to TTW in around 1995 I
started using their program for concrete floors developed in house called RAPID which was written
by Graeme Deaker and sold commercially. It had recently won an Institution of
Engineers Australia design award and was being marketed by the Standards
Association of Australia. Frankly this was in my view the best program on the market and
was a year or two ahead of the opposition. The disappointment I had with
the program was that no significant development work was carried out in the 6
years I was there and the program was overtaken by other products some of which
had been developed later on a windows platform (RAPID was developed in Fortran
on a DOS platform). I imagine that people who bought into this program commercially were
feeling
like they had been let down by the developers so a sense of commitment it is
certainly something I try to gauge when I talk to program developers. It
is also useful to know whether these developers are in a position to make
commercial decisions about the products future. Rapid has not been available
commercially since the mid 1990's and is now only an in house program used by
TTW.
I took a week off work to
review what prestressed post-tensioning concrete design software I should be using in 2005.
Well, actually it was 2 weeks but I stressed 3000m2 of complicated transfer
floors twice with 2 lots of software so that's a decent weeks
work. I started by asking anyone who’s
opinion I thought was worth something and came up with the following
candidates.
PT3D – 3D program developed by Inducta who’s Slabs programs sets the industry standard for
reinforced concrete.
ADAPT – This one wasn’t recommended but
they advertise enough so I thought “What the hell!”
RAM Concept – 3D program with very
strong support from some quarters. Used
to be called Floor and Floor2 most codes are covered.
RAPT – Industry standard for quite some time most codes covered but only 2D.
2D or 3D
The first question to ask yourself these days is "Is 3D worth the trouble?" Let's face it we have survived without it for years but we have also been guessing a fair amount of the time when it comes to things like which really is the stiffest load path and by how much and how crack control may be being compromised by stiffer support elements such as stair and lift shafts. 3D is certainly slower but it is also more rigorous and treats the building holistically rather than piecemeal. It also keeps the calcs in one file and is less likely to contain geometric errors when it is based on the CAD drawing of the building outline.
The other point worth considering is whether you can you make use of the info that you never had before. For example these programs can generally give you;
-
The total amount of concrete in the slab.
-
The total length of strand used.
-
A full schedule of duct supports.
-
Some indication of the restraint to average floor compression afforded by the support system.
-
Some idea of moments induced in columns due to slab creep and shrinkage.
In addition to this you get an excellent 3D rotatable view of the slab which allows you to quickly weed out any geometrical inconsistencies. The fact that you based the model on the actual drawing also assists in this regard. Another point is that the whole slab is in one file so you are not messing around with curious file names for strips that are at some odd angle to a grid reference.
Of course the other intangible is the wow factor when your client wonders in and you have a 3D model of his structure on the screen showing all the tendons. They leave with the impression that you are on the cutting edge when it comes to prestress. I firmly believe that in five years time all buildings with a substantial footprint will be designed this way.
I should also point out that you tend to get a different answer to what you may be used to and that reo layout may be somewhat more complicated in certain circumstances. Take as an example compatibility torsion in an edge beam. In a 2 D approach many designers ignore it and design their slabs with a simply supported end restraint. This is often done to reduce congestion of shear/torsion ligatures in the edge beams especially where the support beams are narrow. You can fiddle with end restraints in a 3D approach to achieve the same goal but you need to think about it more.
Thinking along the same line you can also have torsion issues in the design of internal beams (or bands) especially if you are trying to use some column stiffness to bring down slab deflections. This has been especially true since 2001 when the authors of AS3600 (see 7.6.4) effectively insisted on pattern loading every prestressed concrete structure while including prestress as a load rather than an internal action.
That is not to say that 3D is producing the wrong result, indeed quite the contrary is true but it will wake a few people up to the shortcomings of their current design approach.
Many people come to these new programs with the expectation that the distribution of moments will be based on cracked section stiffnesses but the reality is that they are straight linear elastic FE programs using gross section properties for all but deflection calculations (much like most 2D programs).
The
developer (Emil) emailed me after reading this and said there program has
had the glitches removed and pointed out that they have an association with Jeff
Lind who is quite an experienced PT designer and is quite well regarded.
Emil invited me to have another go but I will probably won't have another spare
week for a while. www.inducta.com.au
ADAPT The rep that I contacted new nothing about the
opposition and not a lot more (or so it seemed) about ADAPT when questioned on finer
detail, but he was happy to send
some glossy brochures which were basically a repeat of what was available on the
net. The brochures (and net) seem to indicate that the programs were piecemeal affair and that
if you wanted something as necessary as a 2D program to put you in the ball park
it would cost extra. I decided to strike them off my list not so
much because of the price difference but more because their listed contact
appeared to know so little about the product. When I want help I want to
be talking to someone who lives and breathes it. www.adaptsoft.com/software.shtml
RAM Concept’s Jim Trenerry
was incredibly helpful, knowledgeable and buoyed by a spate of recent
sales. I annoyed the hell out of him for
a week solid and the guy was dead set cheery every time he answered the phone
which never rang out once. The program’s graphics are amazingly slick and the
platform is logical, intuitive and rock solid.
You don’t get anything for nothing though and getting the model into the
program takes some effort. My tip is
that care and precision when entering the model pays off in spades. It helps a lot if you have a few CAD skills
but if not expect a learning curve. You
could probably leave a lot of the hack work to a draftsman if you are not
interested in being all hands on yourself.
The shortcomings were, no punching
shear checks (what gives), incorrect modelling of torsion in deep beams and a strip wizard
that was a little too skinny on features for my liking.
They say the first two will be fixed by October. This is normally programmer speak for “Yeah
we dream about that sort of stuff too.” but I saw this program a year ago and it
is moving ahead like a steam train so they may well be right.
The other problem I have with this type of program is that it requires you to define design strips and pretend basically that it is not really a 3D program after all. This is because the moments need to be smeared in accordance with the code and the code assumes you are designing in strips – talk about the cart pushing the horse. In a more logical world moment smearing would be more a function of support stiffness and the program should be working it out for itself. Of course the people on the Code committee are not about to change the rules to keep programmers happy and this is basically an American program and their code has always lagged about a decade behind the rest of the world (unlike their steel code which sets the standard globally). The problem here seems to be that they consider no matter how good the research is if it didn't happen in the good old US of A, it didn't happen.
In a perfect world RAM Concept would be producing long term deflection
contours using cracked section properties but apparently this is some time off. Actually they do go close to this by plotting
cracked deflections along design strips. They could do a little more work on graphical presentation of load
diagrams but their graphics are so good in other areas this is me wanting it all
now.
I loved the program, it is the future for sure, but the question is “Is the future really here yet or am I just being sucked in by the swank video parlour graphics?” If your problems are always complicated and not too fluid it is almost certainly the program for you but if you do mostly dead simple stuff maybe not. www.concsoft.com/
RAPT
What can I say, rock solid, does everything you'd expect a 2D program to do, works like a dream but it’s only 2D. The strip wizard that would have sold the above program a million times over (well maybe not a million).
Gil Brock is knowledgeable, helpful and is an experienced PT designer. www.raptsoftware.com
Recommendations
For most people Rapt will still be first choice
because on a simple job you get where you're going faster but if you have the dollars or you just
stress slabs for a living buy RAM Concept as well for the tough stuff, you'll
love it.
Feedback
I do receive a fair amount of feedback on this article and perhaps I should point out to any European readers that the Swiss software CEDRUS is apparently the one to watch in that part of the world. Dam near produces the drawings for you (or so they say).