FEM Analysis and Design Software
FEM Analysis and Design Software
(OP)
As we move forward in the continuum of structural engineering we are losing engineers that began their career at the beginning cusp of or before the advent of computers. When I started all engineering and drafting was by hand. The profession is losing engineers that have completed designs for structures and components with only hand calculations and have a "feel" for evaluating the output from software packages available today. Engineers with only design experience using software are at a disadvatage in evaluating the validity of output. The codes are being written to require engineering that can only be accomplished by software in order to remain competitive within today's fee structures that were set in place more than 50 years ago and do not reflect the complexity and increased amount of analysis and design that is required today. I believe that the world will see a major building failure related to the misuse of engineering software or a glitch in the software within the next decade or two. Software has become a black box that most engineers have no idea of the programming or inner workings. Currently, for large and complex structures, to help address this issue we have instituted the use of independent tandem models that are input by a separate engineer using a different software package for use as a design comparison check. This is in addition to the standard technical reviews with component design checks, etc. I am wondering if and how other firms are addressing this black box software issue. I have seen articles warning of this danger but am unaware of any concerted response from the structural engineering community.






RE: FEM Analysis and Design Software
Insightful comment. You should post more often here, and I encourage you to do so.
Having been in the structural consulting business for the past 30+ years too, I often do gutchecks regarding computer output, but try to keep my projects as simple as possible structurally to keep those gutchecks applicable.
It is true that there is a risk if we have not written a computer program, but continually use it, not knowing of any nuances. However, the confidence in the larger project results comes with use and gutchecks using the simpler project results.
I, too, think that there will be major failures, but not so much due to what you describe, although it is highly possible. I feel it will more likely come from what we did not know that will affect the buildings that are standing. Classic example is the proximity of any particular project to any unknown earthquake faults. After all, the code is constantly changing with more knowledge and we are continually retrofitting.
Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask
RE: FEM Analysis and Design Software
These articles appeared in Structure Magazine a couple of years ago. The links below should take you to those articles.
http
http
One of the things I liked about these articles is he discussed not just the problem, but also some solutions. Essentially, we need to emphasize different techniques in the teaching of analysis.
I believe he is primarily referring to university education, but I believe the concepts apply well to those who are training the next generation of engineers in the work place. In fact, I personally think this 2nd aspect of education is probably more practical / important.
RE: FEM Analysis and Design Software
I often find inconsistencies in computer programs that would not be discovered unless the results where verified by hand calculations (or as I like to do, with a spreadsheet which in my opinion is just an electronic calc pad).
RE: FEM Analysis and Design Software
Although very tempting, I don't think there is enough time/budget to be doing parallel models.
I entirely agree with you that verifying the validity of the output is of utmost importance. In defense of blackbox software, I am sure those firms are trying their best to make error-free software. But in millions of lines of code, between updates, something is definitely bound to screw up. That's where the "feel" and the "knowing something is off" sense will play a vital role.
We are Virginia Tech
Go HOKIES
RE: FEM Analysis and Design Software
I supervised some young engineers up until I retired, some were interested in knowing while others used the black box. I have seen, on here, signs of young engineers who want to "know", who investigate the questions asked by others, as much for their own benefit as to provide an answer. I am very confident that they will find ways to experiment and gain their own "feel" for the behavior of the structures they design. After all, as I said, we had our "tools, they were slow, and we spent more time with a structure than they can, but the interested ones will still find the way and the time to watch the loads spilling down the load paths.
Michael.
Timing has a lot to do with the outcome of a rain dance.
RE: FEM Analysis and Design Software
But I'd also point to an opposite effect. Having straddled both hand calculations all the way through using RAM and RISA modeling, I can say that the use of computer programs many times ENHANCES the ability of the engineer to understand load paths and the repercusions of various assumptions much better than hand calculations ever could.
For example, in building a model of a structure, I can vary support conditions, material properties, pin assumptions, loads, etc. and get an instantaneous result that shows me very clearly the degree of influence of many of my assumptions.
With hand calcs you could never get this sort of feedback.
So in some cases I'd suggest younger engineers have a better feel for structural analysis than my generation did when coming out of school or early in our careers.
RE: FEM Analysis and Design Software
In later years with the advent of the PC, many early PC's only had 16K of RAM... and the first frame and FEM programs I wrote were in BASIC and 'swapped' to 360K diskettes in order to run with 16K... was fun, but I don't really miss that...
On a good day, you could buy a box of 10 DSDD diskettes for $50...
Dik
RE: FEM Analysis and Design Software
Mario Salvadori too thinks its a good thing in the book why structures fall down.
RE: FEM Analysis and Design Software
"If you can't do the calculations by hand, or depending on the complexity or rigorous nature of the calculation, explain the physical mechanics of why something works the way it does, then your a fake."
Personally, I think it is impossible to use a reinforced concrete design program, or a wood design program, without a classic textbook on either of these materials. If I ever find myself using a program and saying to myself "what does that mean?" thats when I know it is time to do some more research.
For example, say I was to investigate strengthening an existing structure with CFRP laminates so it can accommodate an increase in design load. Because of the time constraints the modern engineer faces, would it be ethical to use an FRP design program from a manufacturer of this product (say Sika or Fyfe). Input the parameters and use the output without any considerations for learning the physical mechanics of externally bonded FRP behavior? Or completing the calculations by hand to verify that the stresses are within the limitations to ensure that all failure modes have been sufficiently accounted for?
When an engineer first develops a degree of "expertise" in using a computer package to aid design, then that is when the engineer really starts to develop efficiency with their work.
I have highlighted the work "expertise" because I have only ever met a few engineers who I would deem experts in their field. Personally I believe the word is too commonly overused as a sales pitch. How many times have you heard someone sell themselves as a finite element modelling expert?
CFRP = carbon fibre reinforced polymer.
RE: FEM Analysis and Design Software
Cheers
Greg Locock
New here? Try reading these, they might help FAQ731-376: Eng-Tips.com Forum Policies http://eng-tips.com/market.cfm?
RE: FEM Analysis and Design Software
As far as expecting a building failure to occur because of this technology, I think back to the many failures that have occured because of bad design through hand calculations. Somehow I don't think it's the method employed that's to fault but the person who's been employed.
Tata
RE: FEM Analysis and Design Software
Of course it's true that many (most?) young engineers don't understand the basics as well as they ought to, and it's true that thoughtless use of "black box" software is dangerous.
But it's also true that use of computer software (including the dreaded finite element analysis software) can often give a better understanding of how structures really behave than applying a standard hand calc method, which can be just as much of a black box if done in a mechanical way.
It's not the tools that are the problem, it's how they are applied.
Doug Jenkins
Interactive Design Services
http://newtonexcelbach.wordpress.com/
RE: FEM Analysis and Design Software
I agree that software can be dangerous, but like others have said, I think it can be very helpful.
Like Pad has suggested, I think engineers can really learn a lot from taking a closer look at their FEM results as far as load paths and relative stiffness....if I change this member size, how does it affect others? How does using moment connections vs simple braced frames affect member sizes? This is a difficult thing to do by hand quickly.
I for one, given the time, enjoy doing simple hand calcs as checks of my FE models.
Where I have seen some dangerous things happening was in the arena of connection design (maybe ConnectEgr will chime in).
I worked with a group of younger engineers (myself included, especially at the time) tasked with using proprietary connection design spreadsheets. It shocked me how haphazardly folks would pull connection design loads from FE models and plug them into spreadsheets without any regard to transfer forces and the like. This can be potentially disastrous.
I think another area for concern is loads in general. I don't think proper determination of loads is covered adequately in school.
If your loads aren't accurate, the most precise analysis in the world won't do any good whatsoever.
RE: FEM Analysis and Design Software
Historically our stuff had ben analyzed with classical methods (hand calcs) and then the 'worst cases' proven by tests.
I think the direction the govt was going in was that rather than requiring as many structural tests, instead equipment had to be validated by at least 2 of the 3 methods - classical calculations, FEA & testing.
My point is that I suppose the important thing is ensuring your method & results are validated one way or another.
What is Engineering anyway: FAQ1088-1484: In layman terms, what is "engineering"?
RE: FEM Analysis and Design Software
I am only out of school 4 years and I believe that I do not understand many aspects of structural engineering was well as I would like to, but this does not mean I do not understand the basics as well as I ought to. It is not possible for an engineer to understand many aspects of the field with 4 years of college and 4 years of experience as well as they could.
I feel it is important to know what you know and understand what you understand. And it is even more inportant to know what you do not know. A engineer with 20+ years of experience better have more understanding of the field than one with 4 years of experience.
It all goes back to only working on projects you completely understand, continueing education, peer reviews when working on projects which push your knowledge and experience in the subject, etc.
The black box software is not danerous, it is the engineer who is sloppy which is dangerous.
I say this with all due respect to those with many more years of experience than myself.
RE: FEM Analysis and Design Software
RE: FEM Analysis and Design Software
Here is an extract from an article on the subject written, sums up the my thoughts on situation fairly well:
"Advanced Finite element design tools have become the selected weapon of choice for engineers. Inexperienced engineers are drawn to finite element modelling (FEM) programs as it gives the engineer the feel of freedom to design almost anything the architect can dream up, from complex floors to unusual loadings without relying on experience or intuition.
Whether it is an approximate calculation to confirm the viability of a concept or the ultimate design analysis, the preferred approach is to use a computer. This itself this is not necessarily an unscrupulous thing and computers can be valuable in understanding behaviour. Nevertheless, if the dependence is such that the engineer loses the confidence to carry out simpler methods of analysis consequently the ability to carry out a self-regulating check of their model is compromised. Conversely this creates an interesting situation for the checking engineer (senior engineer) as it is almost impossible to ensure that a complex FEM model, you have not generated yourself is correct. There are few sources of practical advice on how to model and analyse using this technology, this guide seeks to highlight some of the topics engineers must be aware of when utilising this software.
The advantages of FEA/FEM is the ability to model complex issues such as transfer slabs, large opening, unusual loading conditions, easily update calculations and adjustment of structure if changes occur. The other advantages over the equivalent frame method or similar is the ability to account for irregular column layouts. For example circular slabs with column supports around the outside and one column in the centre, the equivalent slab frame method can be used for this design but unless it is an experienced engineer it will be conservative. FEA models can handle this type of arrangement effectively without conservative assumptions.
The disadvantage of FEA is the steep learning curve involved and the checking is difficult. Recently graduated engineers are normally not fully educated in the analysis concrete hence errors can occur especially with modelling assumptions. Finite element design requires a "feel" and experience for the concrete behaviour hence the user should not treat the software as a black box with all the answers and should seek to understand what assumptions are made by the software in both analysis and post processor."
DWHA,
I would expect a 4 year grad to understand a in the context of FEA for concrete modelling, 95% if not more, the idea that you only learn what you need to get the job done, doesn't past muster in my view. You should understand all modelling b'ndy conditions and assumptions before your first FEA job is completed. I may let you get off with not understanding the best way, but not the theory. Truth is there are some engineers that are keen to learn both the practical and theoretical side, and then there are some that aren't keen to learn much would much prefer to be fishing. I expect all engineers that work for me to be in the first group.
An expert is a man who has made all the mistakes which can be made in a very narrow field
RE: FEM Analysis and Design Software
RE: FEM Analysis and Design Software
I agree with the rest of your post, in fact I think we were making much the same points.
Doug Jenkins
Interactive Design Services
http://newtonexcelbach.wordpress.com/
RE: FEM Analysis and Design Software
There are days when I prefer to be working over fishing. They include, but are not limited to, the following:
-Extremely cold conditions below, say, below 30ºF
-Extremely heavy downpours or 1"-2"/hr
-Extremely high temperatures, say, above 90º F
-Extremely windy days with winds exceeding 40 mph
I kid, somewhat