Variability in Professional Design
Variability in Professional Design
(OP)
Attached is a very interesting article about "Variability in Professional Design" published in "Structural Engineering International" 4/94. I would like to see similar "study" here in the USA. If somebody has it please share it here.
Regards,
IV
Regards,
IV






RE: Variability in Professional Design
Michael.
Timing has a lot to do with the outcome of a rain dance.
RE: Variability in Professional Design
RE: Variability in Professional Design
Michael.
Timing has a lot to do with the outcome of a rain dance.
RE: Variability in Professional Design
The current codes and the elaborate (computational methods) computer analysis which we us today and have had foisted upon us, fool some of us into thinking that we are actually doing better analysis and designing better structures. 30 or 40 years ago most of us had a fair idea of the research which lead to the various parts of the codes. Now we have very little idea due to the complexity, and many variations on the loads and every design stress the code prescribes, if this or if that, and then many interpretations of the multipliers which should be applied to these loads, stresses and capacities, just witness many of the questions which get asked here on these topics. In that earlier era we used the computer to improve or simplify our analysis, not to make it more complex. What percentage of the variation in results was due to the different computer programs used and exactly how they treat the problem or interpret the code?
RE: Variability in Professional Design
One reason I have hated LRFD and usually prefer ASD is that with LRFD I have to track two sets of loads through the building. Both the actual loads and the amplified LRFD loads. So now I have twice the likelihood for error. We increase the risk by 100% so that we can save 5%. I can kind of understand the need for LRFD with bridges for example. But for buildings I don't know why it is so important that we have a statisically correct LRFD which can accurately predict a failure of 1 in 10,000 or whatever is was that they based it on. How about zero failures.
In ASD you basically pretend that your material is weaker to account for load variations, material variation, etc... And while that all leads to a factor of safety none of the resulting factor of safety is actually there to account for poor workmanship. For example the .85 factor (and I'm not talking about the phi factor) for concrete column strength is there to account for the striations of vertical water rise within the column due to hydrostatic pressure. We've all seen how water rises to the top of a concrete cylinder. The point is that for all our accuracy we don't even have a factor for human error in field.
So while our analysis is getting much more accurate, and our codes are becoming more statistically realistic, our analysis is getting that much more complicated which leads to more errors in the calcs themselves.
In school I think engineers would be better off spending a semester in a class that focuses purely on the load paths for different types of structures and how the loads move through various types of connections versus say a and introduction to finite analysis.
For all of the analysis we do now its amazing how many holes you can shoot into our analysis because the load paths are incomplete on the documents.
Sorry everyone I'm really venting this evening.
John Southard, M.S., P.E.
http://www.pdhlibrary.com
RE: Variability in Professional Design
It is the engineers responsibility to ensure that there is no poor workmanship by being diligent in performing site inspections of the various elements and requesting submittals to prove that the materials used are as per design.
There is always going to be some variability of design between engineers but I like to think that there is a line to be drawn when it comes to engineering practice. The line defines a structure which is over-designed or under-designed. Good engineers will produce designs which are just above the design line, producing an engineering design which serves it's function to ensure the safety of the public while maintaining a cost efficient solution.
Cowboys will frequently turn in designs which are under-designed based on poor engineering assessments and lack of knowledge of the applicable design methodology.
Then there are the conservative type who are exceptionally conservative with there engineering judgement.
RE: Variability in Professional Design
RE: Variability in Professional Design
We do not have a responsibility to ensure there is no poor workmanship. It the owner pays us to do so than sure, but if not we have no responsibility to go out there on our own dime. More often than not we do structural observations, not special inspections. These are intended for general conformance, not to ensure there is no poor workmanship.
On a different note,
I agree, far too much use of computers. Every time I see somebody post a response recommending a finite element analysis to get the correct answer I shake my head. Why not use an approximate method more suited for the degrees of uncertainty (load, material, methods) we deal with in design.
RE: Variability in Professional Design
RE: Variability in Professional Design
Marking with a crayon
cutting with a chainsaw.