TheMoonlitKnight
Structural
- May 22, 2013
- 52
I was reading a text on structural stability and found this commentary on codes:
[quote="Stability of Structures" by Dr. ZDENEK J. BAZANT (2010)]With various finite element or finite difference computer programs that have recently come into existence, it is now feasible to design columns and frames on the basis of the maximum load Pmax (i.e., the load capacity). Such computer solutions are in close agreement with test results provided that they take into account: (1) the finite deflections and the equilibrium conditions on the deflected structure (which is, of course, the essence of buckling analysis); (2) the actual nonlinear stress-strain diagram; (3) the residual stresses from hot-rolling or cold-forming, as well as stress concentrations around holes, etc.; (4) the imperfections (inevitable initial crookedness); (5) the actual end restraints, with the elastic as well as inelastic and frictional properties of the end joints, and in the case of frames, the buckling interaction among all members; (6) possibly also load repetitions prior to the final overload up to Pmax (which produce, in the case of inelastic behavior or friction, further residual stresses and residual bending moments in indeterminate structures); and (7) for statically indeterminate structures, the possibility that in some members the axial force-displacement curve may already descend while in others it still rises. Why then, despite the availability of these computer programs, do we still need design codes? We still need them because the aforementioned completely rational analysis is quite involved and requires various extensive data that are not readily available to the design engineer. Cookbooks as they are, codes nevertheless fill four important functions: (1) The procedures, or "recipes," enacted by codes, provide reliable guidance to the designer who neither has the time and will to study the theory in depth and make his own choice among several possible procedures, nor is able to collect input data on crookedness, end joint friction, residual stresses, etc.; (2) by sanctioning certain procedures, the codes lend the designer legal protection (an engineer can get sued for lack of wisdom, but a code-writing committee has not been sued yet); (3) by favoring simple approximate formulas and procedures, the codes make the designer's work easier, more efficient; ( 4) at the time the code specifications are instituted, they speed up progress by forcing the designers to use better methods.
The benefit of the last function, however, gets reversed with the passage of time. It is patently difficult to change a code. Thus, all too often, code specifications tend to impede progress when they get old, when better methods become available. For this reason, and also because ripening of the theory in general obviates the need for codes, the future trend will probably be, and ought to be, away from codes.
[/quote]
I was skeptical of a theorist's view of codes from the start - they obviously get in the way of their inherent love of complexities. However, he began to argue the benefits of codes; and, thus I felt more relieved that I was 'listening' to a practical voice in theory. But, the last paragraph seemed to just give up on his previous arguments (favoring codes). The comment, "ought to be" really, really bugged me.
Does he really think it is a good idea? Serious issues I see are:
1. All structural engineers will need to be mathematicians and material science experts - is this even possible to hold ALL structural engineers to this standard - particularly ones who work with smaller project, i.e. residential?
2. How can structural engineers be appropriately held to legal standards, when the standards don't exist - other than "A building or bridge shall not fall down".
3. The author really makes seem like all loading, material, and geometric variables are determinate to high precision. This is INSANE.
4. The individual engineer is forced to aggregate all the test data from others and formulate a conclusion - essentially become a researcher who has to construct his own tools in order to solve problems.
5. How could anybody afford the rate of a structural engineer who has no building codes to rely on? It would be like paying for Boeing to design and sign/seal your single-family home addition.
Rant over. Feel free to comment or question my perspective (it may be off).
[quote="Stability of Structures" by Dr. ZDENEK J. BAZANT (2010)]With various finite element or finite difference computer programs that have recently come into existence, it is now feasible to design columns and frames on the basis of the maximum load Pmax (i.e., the load capacity). Such computer solutions are in close agreement with test results provided that they take into account: (1) the finite deflections and the equilibrium conditions on the deflected structure (which is, of course, the essence of buckling analysis); (2) the actual nonlinear stress-strain diagram; (3) the residual stresses from hot-rolling or cold-forming, as well as stress concentrations around holes, etc.; (4) the imperfections (inevitable initial crookedness); (5) the actual end restraints, with the elastic as well as inelastic and frictional properties of the end joints, and in the case of frames, the buckling interaction among all members; (6) possibly also load repetitions prior to the final overload up to Pmax (which produce, in the case of inelastic behavior or friction, further residual stresses and residual bending moments in indeterminate structures); and (7) for statically indeterminate structures, the possibility that in some members the axial force-displacement curve may already descend while in others it still rises. Why then, despite the availability of these computer programs, do we still need design codes? We still need them because the aforementioned completely rational analysis is quite involved and requires various extensive data that are not readily available to the design engineer. Cookbooks as they are, codes nevertheless fill four important functions: (1) The procedures, or "recipes," enacted by codes, provide reliable guidance to the designer who neither has the time and will to study the theory in depth and make his own choice among several possible procedures, nor is able to collect input data on crookedness, end joint friction, residual stresses, etc.; (2) by sanctioning certain procedures, the codes lend the designer legal protection (an engineer can get sued for lack of wisdom, but a code-writing committee has not been sued yet); (3) by favoring simple approximate formulas and procedures, the codes make the designer's work easier, more efficient; ( 4) at the time the code specifications are instituted, they speed up progress by forcing the designers to use better methods.
The benefit of the last function, however, gets reversed with the passage of time. It is patently difficult to change a code. Thus, all too often, code specifications tend to impede progress when they get old, when better methods become available. For this reason, and also because ripening of the theory in general obviates the need for codes, the future trend will probably be, and ought to be, away from codes.
[/quote]
I was skeptical of a theorist's view of codes from the start - they obviously get in the way of their inherent love of complexities. However, he began to argue the benefits of codes; and, thus I felt more relieved that I was 'listening' to a practical voice in theory. But, the last paragraph seemed to just give up on his previous arguments (favoring codes). The comment, "ought to be" really, really bugged me.
Does he really think it is a good idea? Serious issues I see are:
1. All structural engineers will need to be mathematicians and material science experts - is this even possible to hold ALL structural engineers to this standard - particularly ones who work with smaller project, i.e. residential?
2. How can structural engineers be appropriately held to legal standards, when the standards don't exist - other than "A building or bridge shall not fall down".
3. The author really makes seem like all loading, material, and geometric variables are determinate to high precision. This is INSANE.
4. The individual engineer is forced to aggregate all the test data from others and formulate a conclusion - essentially become a researcher who has to construct his own tools in order to solve problems.
5. How could anybody afford the rate of a structural engineer who has no building codes to rely on? It would be like paying for Boeing to design and sign/seal your single-family home addition.
Rant over. Feel free to comment or question my perspective (it may be off).
