-
1
- #1
Eng-Tips is the largest engineering community on the Internet
Intelligent Work Forums for Engineering Professionals
ASD Steel Construction Manual... 16
- Thread starter Jambruins
- Start date
JedClampett
Structural
Like some of you others, I'm moving slowly into LRFD, kicking and screaming. But I don't think that totally blaming pointy headed academics is fair. The way I understand it, LRFD was developed to align steel design with the limit based designs used in concrete (and available in CMU). That way, when a building has a steel frame supported on a concrete grade beam or columns you can use the same factored loads. And also, the probability of unlikely load combinations can be weighted, like LL plus full wind.
Now this logic falls apart when the soil allowables are allowable stress values or the CMU is done b allowable stress. But that's the approach that the powers that be are taking.
Now this logic falls apart when the soil allowables are allowable stress values or the CMU is done b allowable stress. But that's the approach that the powers that be are taking.
dcarr82775
Structural
Careful Jed, there is a movement afoot to get Geotechs to start giving ultimate values.
paddingtongreen
Structural
This is indirectly related. At my first job I had National Service deferments and given one day per week to go to the technical college, until I took my Ordinary National Certificate. We tackled statically indeterminate structures during the third year. When I learned Moment Distribution, I used it at work. When we learned Slope Deflection, I used it at work. When we learned one of the calculus methods, I used it at work. Until the boss told me to stick to Moment Distribution and Slope Deflection because nobody else could remember how to check my work when I used something else.
Michael.
Timing has a lot to do with the outcome of a rain dance.
Michael.
Timing has a lot to do with the outcome of a rain dance.
westheimer1234
Structural
there is a pdf of aisc 9th edition.. you just have to be "resourceful".. if you know what i mean.. =)
abusementpark
Structural
+1
I've never understood people who complain about the code being complicated. Are you seriously crunching through all these numbers by hand? If so, then me, LRFD, and my computer programs can run circles around you, ASD, and your pen and pad when comes to cranking out design. For me, the vast majority of the time, the difference between using the two methodologies is a simple click on the mouse.
Also, as it was previously stated, what about the major formulations has changed? The majority of the changes I have seen have been to account for using forces instead stresses (i.e. simple unit conversions).
I don't get the argument for it being desirable to look at stresses. You can tell me a beam is stressed to 25 ksi, but that doesn't mean anything to me unless I know the critical stress, which will vary from beam to beam depending on bracing and compactness. The real telling thing to look at is the stress ratio (i.e. applied load/allowable load). IMO, this is the best way to quantify where you are with regards to design.
paddingtongreen
Structural
abusementpark, don't take this the wrong way, I know from your other posts that you know what you are doing, I just want to use your last paragraph to explain something.
"I don't get the argument for it being desirable to look at stresses. You can tell me a beam is stressed to 25 ksi, but that doesn't mean anything to me unless I know the critical stress, which will vary from beam to beam depending on bracing and compactness. The real telling thing to look at is the stress ratio (i.e. applied load/allowable load). IMO, this is the best way to quantify where you are with regards to design."
For someone like me who wasn't raised on the computer, this looks as though you are using the computer as a black box. I like to see, in my mind, the loads flowing down the load paths, I like to know the stresses, but then I think of critical loads, not of critical stresses, even though they are expressed as critical stresses; the strength of the material of a member doesn't change because a member is slender. If I know what is going on, I can allow local stresses to exceed the "critical stress" as long as it is not in a location that matters. It seems so sterile to just check that the stress ratios are <1.
As I said, I know you are not a black box engineer, but your paragraph gave me a chance to say something that has bothered me for years.
Michael.
Timing has a lot to do with the outcome of a rain dance.
"I don't get the argument for it being desirable to look at stresses. You can tell me a beam is stressed to 25 ksi, but that doesn't mean anything to me unless I know the critical stress, which will vary from beam to beam depending on bracing and compactness. The real telling thing to look at is the stress ratio (i.e. applied load/allowable load). IMO, this is the best way to quantify where you are with regards to design."
For someone like me who wasn't raised on the computer, this looks as though you are using the computer as a black box. I like to see, in my mind, the loads flowing down the load paths, I like to know the stresses, but then I think of critical loads, not of critical stresses, even though they are expressed as critical stresses; the strength of the material of a member doesn't change because a member is slender. If I know what is going on, I can allow local stresses to exceed the "critical stress" as long as it is not in a location that matters. It seems so sterile to just check that the stress ratios are <1.
As I said, I know you are not a black box engineer, but your paragraph gave me a chance to say something that has bothered me for years.
Michael.
Timing has a lot to do with the outcome of a rain dance.
dcarr82775
Structural
Well said Paddingtongreen. I see calc packages that are 3"-4" thick and not a single hand calc anywhere. Just page after page of enercalc/RISA/you name it software.
- Thread starter
- #52
Sorry for getting the ball rolling on this hot topic! My inital argument was never that LRFD was too complicated or too hard to learn. I was educated as an ME and I find ASD more friendly towards heavy industrial/construction design...where it can be worked with the various ASME codes. As an ME we also tend to work more with basic stress analysis with principal stresses and von Mises criteria. AISC steel design has "codified" these design checks.
When it comes to concrete, you are right, I never learned another method so LRFD it was. Likewise when initially learning steel design I was taught LRFD and have used it in practice without issue.
I think the other main argument that takes place is simplification. Some of the arguments above state "ASD 1989 did not incorporate various design checks"...you are absolutely right..it was left to the engineers discretion to verify and check the design.
To tie in with Paddington's rant above this becomes the fundamental question. If all this is a matter of code and software..then structural engineering for basic commerical construction will continue to be degraded. I have found myself saying "...this must be in the code somewhere..." Sometimes I have to stop myself and get back to reality and realize that at one time we could live with less codified design checks and do some real engineering.
When it comes to concrete, you are right, I never learned another method so LRFD it was. Likewise when initially learning steel design I was taught LRFD and have used it in practice without issue.
I think the other main argument that takes place is simplification. Some of the arguments above state "ASD 1989 did not incorporate various design checks"...you are absolutely right..it was left to the engineers discretion to verify and check the design.
To tie in with Paddington's rant above this becomes the fundamental question. If all this is a matter of code and software..then structural engineering for basic commerical construction will continue to be degraded. I have found myself saying "...this must be in the code somewhere..." Sometimes I have to stop myself and get back to reality and realize that at one time we could live with less codified design checks and do some real engineering.
I agree with paddington, as well. I'm a new engineer and was taught LRFD in school. My dad is a PE, so he actually taught me ASD89. As I'm comfortable with both now, I don't really have a preference. Most new engineers I went to school with rely entirely too much on software to run their calcs. At my old gig we had a green one design a moment connection that physically wasn't possible because he just turned around the computer read-outs. Terrible practice. As for myself, I run everything by hand in some way or another. It definitely gives you a better understanding of what's actually happening within a structure. I guess I'm lucky in the regard that my dad initially taught me that way, so now it's reflected in my work. (SIDE NOTE: All of the Senior PE's I work with love that I do most calcs by hand. Maybe because it's much easier to check.)
The design is still left to your discretion but now you have "correct" equations to use whereas you previously did not for many design checks. The creation of LRFD or development of new equations has not taken the design away from engineers. It has taken it away if one would prefer to just cook up untested methods for a larger chunk of his work. We have to do this at times, but the fewer times the better, right?
With ASD89, the engineer would almost inevitably do many checks incorrectly due to lack of guidance. Just as a simple example, say a guy was checking an I-shape with different flanges for axial and/or for flexure. He'd use the basic KL/r approach for Fa, which is incorrect (as most know now and far fewer knew in 89). For Fb, he'd use Fb=170000Cb/(L/rt)^2 or 12000Cb/(L*d/Af), both of which are incorrect.
"Incorrect" meaning that the underlying formulations are not for the type of member being considered. The engineer had no choice but to misuse (a lot of the ASD89 fans want to say that people misuse equations more nowadays...) these equations for lack of better guidance. There are "correct" checks in the 13th Ed. Spec. for those cases.
In my experience, senior PEs and SEs do more manual calcs because their time is too valuable to sit there operating programs for days or weeks on end (and therefore, it's not worth it to learn the programs in detail). They'd do quick and simple stuff manually and have the junior guys crank for days or weeks with the programs. They'd never approve of the junior guys trying to design dozens of composite beams on their jobs manually, for example. The senior guy might use a Manual table for a composite beam or two at random, though.
After all, who do you think approved the purchase of the programs? The senior engineers.
paddingtongreen
Structural
I must be the odd dinosaur, I loved the computer as a tool from the beginning, working off mainframes until the PCs came along. Mostly, we used STAAD on the PCs.
Michael.
Timing has a lot to do with the outcome of a rain dance.
Michael.
Timing has a lot to do with the outcome of a rain dance.
-
4
- #58
I’m with paddington on this one. We aren’t better engineers because our computer prints results out to ten decimal places, when we don’t really seem to have a feel for how the structure acts or reacts to various loads. And, we sure aren’t producing better structures or products, because of all our elaborate calcs. and methods, but it’s all a hell of a lot more complicated. And, that complication has become a self perpetuating end in itself, and the bane of our existence.
We know more about structures and materials than we ever have, in the history of mankind. We have more sophisticated analysis tools and methods and manufacturing and construction methods than we have ever had. I’m all for learning from experience, there is no doubt in my mind that what we have learned from recent earthquakes and hurricanes should be includes in our design codes for human safety and building survival reasons. But, this might be done in a 20-30 page addendum, while my $150 copy of THE CODE starts to show some ware, from use, and I actually learn it. And, even though the testing shows this to be true, no one in his right mind is going show a different fastener schedule for ever 2 sq.ft. area of roof as you move away from the corner, although it is interesting knowledge to keep in mind for special situations. I tend to prefer ASD because that’s what I first learned and have used the most. But, as paddington suggested, I’ve designed closer to yield for years, as long as instability or serviceability didn’t control. I generally agree with the philosophy of LRFD, but this old dog is having trouble learning all the new tricks. They have to make it much more user friendly, instead of so number manipulation intensive.
We are turning this whole business into some sort of a complicated cookbook approach, where any fool must be able to do it, if they only follow the complicated recipe, tooo theee letttter, never mind having any intuition about how things actually work any longer; that’s not necessary, we’ve codified that too. We can just not codify everything and every condition. So, here’s the recipe: it’s for lemon flavored yellow cupcakes with poppy seeds, and it calls for 1200/3600 of a cup of poppy seeds, but I only have 1157/3600 cups of said seeds; oh crap, I can’t make cup cakes today, the recipe doesn’t tell me what to do under these conditions. Well, we get a bunch of committees together, one for flour, another for yellow cake flour, one for lemon flavoring, and the all important one, for poppy seeds. After years of testing and bickering we settle on 1159/3600 cups as being O.K. Finally, we republish the 1200 page cookbook to correct that one recipe, and the publisher is real happy with the new edition, only $69.99 plus shipping, but no one can taste the difference.
I’m not calling for a full blown revolt or uprising, but something close to that, slightly tamer. For starters, we should just quit buying every damn new edition of all of the codes and all of the new computer programs. We pay a premium for all that crap and all of the new bells and whistles really don’t improve our lives or give us better answers, results or designs or really make any of us better engineers. This just keeps enriching the producers of this more and more complicated crap, in effect encouraging them to continue. And each time, we have to stop production and relearn to use this new fangled system or code.
Code writing and publishing and the research surrounding it should not be an end in itself. This has become a large industry unto itself which enslaves us to learning the new version, formulas or format and doesn’t make us better engineers, designing better products. This is the tail waging the dog, and it should stop. We should leave them sitting with a million unsold copies of the latest, and really learn to use and understand the copy we already have. The testing and research is wonderful new knowledge, worth having in my arsenal. I used to get that from keeping abreast of many technical journals. And, I would even use it when it helped me on a design or with an analysis.
We know more about structures and materials than we ever have, in the history of mankind. We have more sophisticated analysis tools and methods and manufacturing and construction methods than we have ever had. I’m all for learning from experience, there is no doubt in my mind that what we have learned from recent earthquakes and hurricanes should be includes in our design codes for human safety and building survival reasons. But, this might be done in a 20-30 page addendum, while my $150 copy of THE CODE starts to show some ware, from use, and I actually learn it. And, even though the testing shows this to be true, no one in his right mind is going show a different fastener schedule for ever 2 sq.ft. area of roof as you move away from the corner, although it is interesting knowledge to keep in mind for special situations. I tend to prefer ASD because that’s what I first learned and have used the most. But, as paddington suggested, I’ve designed closer to yield for years, as long as instability or serviceability didn’t control. I generally agree with the philosophy of LRFD, but this old dog is having trouble learning all the new tricks. They have to make it much more user friendly, instead of so number manipulation intensive.
We are turning this whole business into some sort of a complicated cookbook approach, where any fool must be able to do it, if they only follow the complicated recipe, tooo theee letttter, never mind having any intuition about how things actually work any longer; that’s not necessary, we’ve codified that too. We can just not codify everything and every condition. So, here’s the recipe: it’s for lemon flavored yellow cupcakes with poppy seeds, and it calls for 1200/3600 of a cup of poppy seeds, but I only have 1157/3600 cups of said seeds; oh crap, I can’t make cup cakes today, the recipe doesn’t tell me what to do under these conditions. Well, we get a bunch of committees together, one for flour, another for yellow cake flour, one for lemon flavoring, and the all important one, for poppy seeds. After years of testing and bickering we settle on 1159/3600 cups as being O.K. Finally, we republish the 1200 page cookbook to correct that one recipe, and the publisher is real happy with the new edition, only $69.99 plus shipping, but no one can taste the difference.
I’m not calling for a full blown revolt or uprising, but something close to that, slightly tamer. For starters, we should just quit buying every damn new edition of all of the codes and all of the new computer programs. We pay a premium for all that crap and all of the new bells and whistles really don’t improve our lives or give us better answers, results or designs or really make any of us better engineers. This just keeps enriching the producers of this more and more complicated crap, in effect encouraging them to continue. And each time, we have to stop production and relearn to use this new fangled system or code.
Code writing and publishing and the research surrounding it should not be an end in itself. This has become a large industry unto itself which enslaves us to learning the new version, formulas or format and doesn’t make us better engineers, designing better products. This is the tail waging the dog, and it should stop. We should leave them sitting with a million unsold copies of the latest, and really learn to use and understand the copy we already have. The testing and research is wonderful new knowledge, worth having in my arsenal. I used to get that from keeping abreast of many technical journals. And, I would even use it when it helped me on a design or with an analysis.
I would argue that computer programs give me a BETTER feel for the structure, the load paths, and the effects of my design decisions.
With my program, I can quickly alter various parameters, loads, member sizes, etc. and see INSTANTLY the effect on deflections, moments, etc. I also can see the deflected shape of the structure instantly.
Do some mis-use programs and overly trust the results without checking? Yes. Do some not carefully develop input procedures and take the time to understand the program variables? Yes.
But good engineers don't follow those poor practices.
And the old way (by hand) couldn't offer you the feel of the design like todays programs...if you think you do have a better feel for it...just because you are using graphite and paper....you are fooling yourself.
I used to do them by hand (late 1970's)...no thanks.
Programs have freed us from tedious math and allowed us to become better designers in my opinion.
With my program, I can quickly alter various parameters, loads, member sizes, etc. and see INSTANTLY the effect on deflections, moments, etc. I also can see the deflected shape of the structure instantly.
Do some mis-use programs and overly trust the results without checking? Yes. Do some not carefully develop input procedures and take the time to understand the program variables? Yes.
But good engineers don't follow those poor practices.
And the old way (by hand) couldn't offer you the feel of the design like todays programs...if you think you do have a better feel for it...just because you are using graphite and paper....you are fooling yourself.
I used to do them by hand (late 1970's)...no thanks.
Programs have freed us from tedious math and allowed us to become better designers in my opinion.
As another old-timer, this thread hits too close to home. For 2010 I put my 8th edition with my reference books and the 13th edition is now on my desk. I am trying to use LRFD every day - but sometimes I fail.
My name is Gary and I used ASD.
Part of what I see as the bigger problem is the complexity of the Codes. I used to joke that bridge engineers had to do more statistical analysis than structural analysis (although I only did bridge design in a classroom). But now with 45 degree winds, reduced winds with torsional moments, 4-directions of seismic loading, 3-D models, etc., you find yourself ending up with way too many load cases.
Throw in Direct Analysis Method and my current industrial 3-story process/transfer building has 119 load cases. When I questioned the software rep as to the time it takes to analyze this structure - he said "manually eliminate the load cases that do not govern". Easier said than done. There is no intuitive way to know exactly how the interactions affect different elements.
I can easily see becoming one of those black-box engineers.
gjc
My name is Gary and I used ASD.
Part of what I see as the bigger problem is the complexity of the Codes. I used to joke that bridge engineers had to do more statistical analysis than structural analysis (although I only did bridge design in a classroom). But now with 45 degree winds, reduced winds with torsional moments, 4-directions of seismic loading, 3-D models, etc., you find yourself ending up with way too many load cases.
Throw in Direct Analysis Method and my current industrial 3-story process/transfer building has 119 load cases. When I questioned the software rep as to the time it takes to analyze this structure - he said "manually eliminate the load cases that do not govern". Easier said than done. There is no intuitive way to know exactly how the interactions affect different elements.
I can easily see becoming one of those black-box engineers.
gjc
- Status
Similar threads
- Locked
- Question
- Locked
- Question
- Locked
- Question