As a current user of the AISC ASD 9th edition, I am considering taking the plunge into LRFD. I am curious about other engineers opinions about the third edition. I would like to know from users experienced with it what they consider to be its strengths and weaknesses, and the general opinion about it. Thanks in advance.
Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations cowski on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!
AISC LRFD 3rd Edition
- Thread starter bjb
- Start date
SlideRuleEra
Structural
bjb - Structural Engineer Magazine had an eight page article on this subject in the September 2000 issue. The title is "ASD vs. LRFD". It was mostly letters from users giving reasons and opinions. The article does not seem to be posted on the web any longer. However I have the .pdf version (860 Kb), if you want to post an e-mail address, will be happy to send it.
The 3rd edition is a concise edition of the manual. It does not have each and every shape listed for instance the the 2nd edition or the ASD has but it has the most common. I think I've only had to go look up a shape in a different manual just once so it isn't like it's missing many. Since it is the manual and the connections volume in one book there is some information and charts that had to be omitted. Overall I have liked it. There's a lot of good information there that ASD simply doesn't have because it is not kept current anymore.
Remember too that though LRFD can give you a better more consistent design from a theoretical point of view, many times you will get the same result or a similar result as ASD but with more complex calculations. Always check deflections. That should go without saying but that hasn't always been the case and has given a LRFD a bad start in steel design. Members tend to be a little stronger sized with LRFD in my experience but when deflection controls a design it doesn't really matter.
I have found that connections are usually sized better with LRFD , particularly eccentric bolted and welded connections.
I believe that the 4th LRFD edition however is supposed to be a combined spec for both LRFD and ASD with updates and new information for the ASD specs. You may want to wait for that and avoid buying 2 books.
Remember too that though LRFD can give you a better more consistent design from a theoretical point of view, many times you will get the same result or a similar result as ASD but with more complex calculations. Always check deflections. That should go without saying but that hasn't always been the case and has given a LRFD a bad start in steel design. Members tend to be a little stronger sized with LRFD in my experience but when deflection controls a design it doesn't really matter.
I have found that connections are usually sized better with LRFD , particularly eccentric bolted and welded connections.
I believe that the 4th LRFD edition however is supposed to be a combined spec for both LRFD and ASD with updates and new information for the ASD specs. You may want to wait for that and avoid buying 2 books.
- Thread starter
- #5
I have seen the proposed "unified" spec on the AISC website. The way they're doing it is similar to the light-gauge steel spec. They give equations for the nominal strength, and then you either divide that by the factor of safety and compare it to the working load, or hit the nomiinal strength with the strength reduction factor and compare it to factored loads. From what I saw, the new allowable stress methodology isn't going to be that much different from LRFD. That's why I figure it would be a good idea now to try to become familiar with LRFD.
One thing that bothers me about the LRFD and the unified spec is that they seem to carry material constants like "E" as variables. I think the equations would be simpler if they didn't do that. Another question is, if the end result from LRFD isn't that much different from ASD but is harder to get to, why bother with LRFD? Also, considering that the vast majority of buildings are low rise, and frequently one-story, do we need such a rigorous spec like LRFD for this type of building? Could the theoretically more precise equations be contained in a differnt volume that could be used for buildinigs that benefit from a more accurate approach? That way, designers of low rise relatively simpler buildings could use something similar to ASD in its current form, instead of having to use the more complicated equations. I know that my budgets and schedules won't increase just so I can design with the more complicated LRFD and Unified specs.
One thing that bothers me about the LRFD and the unified spec is that they seem to carry material constants like "E" as variables. I think the equations would be simpler if they didn't do that. Another question is, if the end result from LRFD isn't that much different from ASD but is harder to get to, why bother with LRFD? Also, considering that the vast majority of buildings are low rise, and frequently one-story, do we need such a rigorous spec like LRFD for this type of building? Could the theoretically more precise equations be contained in a differnt volume that could be used for buildinigs that benefit from a more accurate approach? That way, designers of low rise relatively simpler buildings could use something similar to ASD in its current form, instead of having to use the more complicated equations. I know that my budgets and schedules won't increase just so I can design with the more complicated LRFD and Unified specs.
- Thread starter
- #6
Thanks for the offer SlideRuleEra, I probably have that issue in my pile, I will look for it. I also think that before the ASD 9th edition goes out of print I am going to order another copy, as my current one is starting to fall apart, thanks in part to the relatively flims construction of this edition by AISC compared to other editions and all of the LRFD editions.
I think leaving the E as a variable in the equation doesn't really add to the complexity. It's just one number, plus now you can more directly compare the code equations to theory and better understand what you are doing and how different factors affect equations and results. That helped me understand where all these code equations come from. Obviously that's a good thing.
As far as difficult and longer equations, it's harder for a student who has to design by hand. Even checking a beam by hand with LRFD vs ASD doesn't incur that much more time and few people who are concerned with quick turn-around of a design are doing it by hand anyway. There's still plenty of charts and tables and such to use for design. Also, it's just as fast to use software with either method. If nothing else you can always make a spreadsheet. What LRFD does is provide you with a more accurate tool to design your structure which should be what any engineer would want in my opinion.
In addition, you always have a more uniform probability that failure will not occur. With ASD every member has a different safety factor. Once the member with the lowest safety factor fails you're done but you have all this extra left over in the rest of the structure. My understanding of LRFD is that each member when designed correctly will have about the same probability of failure hence a more efficient design. That's the point of the load factors and resistance factors: to calibrate data for loads and strength to arrive at an acceptable probability of failure.
I commend you for wanting to learn LRFD. As much as I think it's better I haven't yet worked in an office that uses LRFD for steel design since I got my degree. So I graduated, got a job, and had to learn ASD in a hurry. I can understand reluctance when you are used to doing something one way but I think burying your head in the sand as my employers have done instead of growing as an engineer and learning is irresponsible and lazy at best. Sometimes when I have time I'll do a design in ASD for the record and then check it on my own using LRFD so that I can compare the two and learn something instead of fall behind technology and live in the past. What I've told you is what I've found based on doing that plus what I've learned in class. I hope it helps.
As far as difficult and longer equations, it's harder for a student who has to design by hand. Even checking a beam by hand with LRFD vs ASD doesn't incur that much more time and few people who are concerned with quick turn-around of a design are doing it by hand anyway. There's still plenty of charts and tables and such to use for design. Also, it's just as fast to use software with either method. If nothing else you can always make a spreadsheet. What LRFD does is provide you with a more accurate tool to design your structure which should be what any engineer would want in my opinion.
In addition, you always have a more uniform probability that failure will not occur. With ASD every member has a different safety factor. Once the member with the lowest safety factor fails you're done but you have all this extra left over in the rest of the structure. My understanding of LRFD is that each member when designed correctly will have about the same probability of failure hence a more efficient design. That's the point of the load factors and resistance factors: to calibrate data for loads and strength to arrive at an acceptable probability of failure.
I commend you for wanting to learn LRFD. As much as I think it's better I haven't yet worked in an office that uses LRFD for steel design since I got my degree. So I graduated, got a job, and had to learn ASD in a hurry. I can understand reluctance when you are used to doing something one way but I think burying your head in the sand as my employers have done instead of growing as an engineer and learning is irresponsible and lazy at best. Sometimes when I have time I'll do a design in ASD for the record and then check it on my own using LRFD so that I can compare the two and learn something instead of fall behind technology and live in the past. What I've told you is what I've found based on doing that plus what I've learned in class. I hope it helps.
- Thread starter
- #8
Thanks for your comments UcfSe
I think it's important not to forget that ASD has a long history of providing safe and economical structures. Also, many beams are governed by deflection criteria, not strength. The fact remains that members sized by LRFD are not much different from ASD, and I also learned in a graduate level steel course that the LRFD has been "tuned" so that it is not that much different in the member sizes that result. I am not a proponent of using the computer for all my design calcs. True, it can be faster, and some really tedious and monotonous equations are well suited for solution by computer. However, I have seen too many engineers who have relied too much on the computer, and their judgement has suffered as a result.
I disagree that leaving "E" as a variable doesn't really add to the complexity. The AISC manual should be a design manual, not a text book. The derivation should be handled in the commentary, and where the equations come from can also be seen in text books. I learned the theory of steel design with textbooks in college at the undergrad and graduate level. Someone who has not learned the theory of steel design using textbooks should not be using the manual to design steel structures.
With that said, I am trying not to have a knee-jerk reaction against LRFD. I have no problem with a more accurate approach. My concerns relate to the user-friendliness of the method. I also question if the supposed increase in accuracy is worth the extra effort for run of the mill buildings, considering the successful history of ASD. Still,I can see where the industry is headed, and I will make the plunge to avoid becoming "technologically obsolete". I think that the Unified ASD/LRFD Manual that AISC is putting together is a step in the right direction. I have no philosophical problem with finding the nominal strength of a member, dividing by a factor of safety, and then comparing that to the working loads. I will finish with this quotation from someone very wise:
structural engineering is the art of molding materials we do not really understand, into shapes we cannot really analyze, so as to withstand forces we cannot really measure, in such a way that the public does not really suspect.
I think it's important not to forget that ASD has a long history of providing safe and economical structures. Also, many beams are governed by deflection criteria, not strength. The fact remains that members sized by LRFD are not much different from ASD, and I also learned in a graduate level steel course that the LRFD has been "tuned" so that it is not that much different in the member sizes that result. I am not a proponent of using the computer for all my design calcs. True, it can be faster, and some really tedious and monotonous equations are well suited for solution by computer. However, I have seen too many engineers who have relied too much on the computer, and their judgement has suffered as a result.
I disagree that leaving "E" as a variable doesn't really add to the complexity. The AISC manual should be a design manual, not a text book. The derivation should be handled in the commentary, and where the equations come from can also be seen in text books. I learned the theory of steel design with textbooks in college at the undergrad and graduate level. Someone who has not learned the theory of steel design using textbooks should not be using the manual to design steel structures.
With that said, I am trying not to have a knee-jerk reaction against LRFD. I have no problem with a more accurate approach. My concerns relate to the user-friendliness of the method. I also question if the supposed increase in accuracy is worth the extra effort for run of the mill buildings, considering the successful history of ASD. Still,I can see where the industry is headed, and I will make the plunge to avoid becoming "technologically obsolete". I think that the Unified ASD/LRFD Manual that AISC is putting together is a step in the right direction. I have no philosophical problem with finding the nominal strength of a member, dividing by a factor of safety, and then comparing that to the working loads. I will finish with this quotation from someone very wise:
structural engineering is the art of molding materials we do not really understand, into shapes we cannot really analyze, so as to withstand forces we cannot really measure, in such a way that the public does not really suspect.
PEStructural
Structural
One thing to consider is that AISC assumed that when they started LRFD, ASD would be "phased" out. So, they spent the last 15 years (since the last publication date of ASD), doing more research and putting more design requirements based on that research in the LRFD.
For example, the ASD says "The effects of torsion should be considered." That's all it says about torsion. LRFD, however, goes into more rigorous explanation and design requirements for torsion and lateral-torsional buckling. Even with simple structures, offset connections can create eccentricity (i.e. torsion) that must be accounted for. Same thing with any curved beam design.
Now with the advent of the combined spec, they took the last 15 years of research and applied it to both methods. I'm looking forward to the new spec. Like many people, I learned LRFD in college, but had to quickly learn ASD in the workplace.
For example, the ASD says "The effects of torsion should be considered." That's all it says about torsion. LRFD, however, goes into more rigorous explanation and design requirements for torsion and lateral-torsional buckling. Even with simple structures, offset connections can create eccentricity (i.e. torsion) that must be accounted for. Same thing with any curved beam design.
Now with the advent of the combined spec, they took the last 15 years of research and applied it to both methods. I'm looking forward to the new spec. Like many people, I learned LRFD in college, but had to quickly learn ASD in the workplace.
bjb
I understand your point and agree with you to a point. Why complicate something that works? I also agree that we shouldn't use a computer to do everything, at least not until we understand everything the computer is doing and verify the results by hand. I always do this with software. Even so, my boss won't buy much software so I have written a great many spread sheets. I still had to sit down and learn everything in order to write them and I think it's a great way to pick up something you don't know not to mention it's much faster and makes short work out more complicated stuff like that derned "E" lol.
The third edition puts as much info as practical into charts and tables. Things like local stability are also in charts for given Fy eliminating any calc, easy or hard. That makes it good for quick design.
Still, I'm willing to bet that ASD 9th sticks around a while. ACI working stress hung around as an alternate until the 02 code right? Maybe the AISC will reference the ASD 9th in their new codes as an acceptable alternative in all or some situations. Then we have our simple method for simple buildings right there ready to go.
I understand your point and agree with you to a point. Why complicate something that works? I also agree that we shouldn't use a computer to do everything, at least not until we understand everything the computer is doing and verify the results by hand. I always do this with software. Even so, my boss won't buy much software so I have written a great many spread sheets. I still had to sit down and learn everything in order to write them and I think it's a great way to pick up something you don't know not to mention it's much faster and makes short work out more complicated stuff like that derned "E" lol.
The third edition puts as much info as practical into charts and tables. Things like local stability are also in charts for given Fy eliminating any calc, easy or hard. That makes it good for quick design.
Still, I'm willing to bet that ASD 9th sticks around a while. ACI working stress hung around as an alternate until the 02 code right? Maybe the AISC will reference the ASD 9th in their new codes as an acceptable alternative in all or some situations. Then we have our simple method for simple buildings right there ready to go.
Great! Soon we will have two ASD methods to go along with LRFD and LFD. If we think it is confusing now, wait until the new methods come out. Everyone will be using a different design method. The only way to straighten this out will be to collect all of our old design books, burn them, and then send us all back to school for the new ASD and LRFD methods. And then we'll STILL have two methods to fight about! Don't fix it if it's not broken!
The way it looks now, LRFD will continue to be the specified (only) design method for highway and bridge design. The FHWA is seeing to that.
The way it looks now, LRFD will continue to be the specified (only) design method for highway and bridge design. The FHWA is seeing to that.
- Status
Similar threads
- Question
