-
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
TXStructural
Structural
It didn't take me very long as a practicing structural engineer to realize two things:
1) engineering is not about calculating precisely the nuances of inaccurate guesstimates. And,
2) a sharper pencil still won't make a weak structure stronger.
I frequently make presentation to groups of SE's on simplifying design. The basic premise is that excessive complication in the beginning of any design is likely to result in a more expensive, more complicated, less robust design.
It is not ASD or LRFD which makes the design complicated or difficult, it is all the things leading up to the first computation, and the degree to which every component is optimized. Sometimes more material costs less, and getting to a design which uses more material often takes less time and results in fewer errors, which saves even more money at every stage of design and construction.
1) engineering is not about calculating precisely the nuances of inaccurate guesstimates. And,
2) a sharper pencil still won't make a weak structure stronger.
I frequently make presentation to groups of SE's on simplifying design. The basic premise is that excessive complication in the beginning of any design is likely to result in a more expensive, more complicated, less robust design.
It is not ASD or LRFD which makes the design complicated or difficult, it is all the things leading up to the first computation, and the degree to which every component is optimized. Sometimes more material costs less, and getting to a design which uses more material often takes less time and results in fewer errors, which saves even more money at every stage of design and construction.
abusementpark
Structural
Anyone shying away from computer programs at this juncture is doing a great disservice to themselves and ultimately their company. For the reasons that were excellently stated by JAE, the computer programs are better at giving you a feel for the load path. Also, there are many simple checks one can do to verify if the input is correct: Do my reactions satisfy equilibrium? Are my moment and shear diagrams as expected? How does the deflected shape look? Is there any unwanted torsion getting into some members? Do the results make sense?.....etc.
I understand that there are a lot of engineers who may abuse a computer program and never fully understand what it is doing. However, there are many of us who know to properly harness the great power of these computational tools and use them to our advantage. It is especially helpful if you have an educational background in the modern methods of structural analysis. Your faith in any computer program has to be tempered with keen sense of where things can stray from reality and where errors can be made.
Maybe some people are emotionally tied to the nostalgia associated with hand calculations. Don't get me wrong, hand calcs still have a vital role in what we do and should still be considered on a case by case basis. A lot of special connections require hand calculations. However, for the vast majority of steel structures of any significant size, doing the main frame analysis in the computer is the way to go. IMO, one the biggest advantages of doing a computer model of a steel structure is the method of convenient bookkeeping. Everything is one place. Changes in the framing are easy implement late in the game. No need to follow a messy paper trail. It is easy to look at your reaction envelope for the purposes of foundation design. Some programs even have the capability to automatically calculate live load reductions for each individual member!
Also, if some of you think that AISC is getting carried away with code complexity. Take a look at the AASHTO LRFD manual. It makes the AISC manual look like a pamplet.
I understand that there are a lot of engineers who may abuse a computer program and never fully understand what it is doing. However, there are many of us who know to properly harness the great power of these computational tools and use them to our advantage. It is especially helpful if you have an educational background in the modern methods of structural analysis. Your faith in any computer program has to be tempered with keen sense of where things can stray from reality and where errors can be made.
Maybe some people are emotionally tied to the nostalgia associated with hand calculations. Don't get me wrong, hand calcs still have a vital role in what we do and should still be considered on a case by case basis. A lot of special connections require hand calculations. However, for the vast majority of steel structures of any significant size, doing the main frame analysis in the computer is the way to go. IMO, one the biggest advantages of doing a computer model of a steel structure is the method of convenient bookkeeping. Everything is one place. Changes in the framing are easy implement late in the game. No need to follow a messy paper trail. It is easy to look at your reaction envelope for the purposes of foundation design. Some programs even have the capability to automatically calculate live load reductions for each individual member!
Also, if some of you think that AISC is getting carried away with code complexity. Take a look at the AASHTO LRFD manual. It makes the AISC manual look like a pamplet.
abusementpark
Structural
I've seen drawings where it is obvious that the engineer just used whatever "optimum" beam the computer spit out for each individual member and there are 25 different w-shape beam sizes on the job. That's certainly not something for which I would advocate.
This brings up a question that I've had for a while. The thread is hopelessly off course now, so I'll indulge in a further tangent LOL.
I've always used frame analysis programs to do the lateral design and design beams somewhat individually using smaller programs. I've not used whole system design programs like the Ram System enough to know the nitty gritty details.
Consider a model of a typical steel-framed floor with a few moment frames as the LFRS. I'm not sure I even know the exactly correct LL reduction to use in such a model. (I know what I do and it's a slight simplification that I think is OK. EJ.) If one was going after the "exact correct" answer, what influence area should be used to determine the reduced LL for gravity beams, gravity columns, lateral beams, and lateral columns? This would seem like an easy question, but it's not IMO. Global P-Delta effects should be for LL over the entire floor area, and these go into the indivdual member moments. However, an individual moment frame beam would seem to need a larger LL for its KLL*At, but I'm not sure that's even exactly correct. Same goes for a moment frame column.
After all that is the question. What does Ram Steel and other similar programs do with this kind of issue? Does it leave the individual element reduced LLs on there and then apply some upward corrective forces as have been proposed in some papers over the years? If someone has a document describing the process used in the programs, I'd be interested to read it.
Abusementpark & JAE:
I agree with you completely, computers and software are indispensable to our work today. There is no doubt that we can study many different alternatives today than we ever could with a slide rule or hand held calculator. For a fair share of our work we could not do without them today. I actually taught a couple of courses in computers and structures, and did a fair amount of programming for consulting firms and industry in the mid to late 60's, while in grad school. But, I am beginning to wonder if the latest codes or latest computer programs, are really improving our designs or making us better engineers. These new versions are coming much to fast for me, and are more and more difficult to keep up with, and still get some productive work done.
I am asking, in the case of the next edition of the codes or the latest version of software, whether we are really just getting a few new bells and whistles which don’t improve our designs and make use better engineers, or are we just enriching the sellers of this stuff, and frustrating ourselves in the process. On both accounts, I submit, that we could slow down quite a bit and really learn to use what we have. Our designs wouldn’t hurt from that, we already have more computing ability than most of us know what to do with. And, there are a bunch of people who probably shouldn’t have access to this computing machinery, because they pretend that this makes them an engineer. In fact this can be dangerous to others.
I agree with you completely, computers and software are indispensable to our work today. There is no doubt that we can study many different alternatives today than we ever could with a slide rule or hand held calculator. For a fair share of our work we could not do without them today. I actually taught a couple of courses in computers and structures, and did a fair amount of programming for consulting firms and industry in the mid to late 60's, while in grad school. But, I am beginning to wonder if the latest codes or latest computer programs, are really improving our designs or making us better engineers. These new versions are coming much to fast for me, and are more and more difficult to keep up with, and still get some productive work done.
I am asking, in the case of the next edition of the codes or the latest version of software, whether we are really just getting a few new bells and whistles which don’t improve our designs and make use better engineers, or are we just enriching the sellers of this stuff, and frustrating ourselves in the process. On both accounts, I submit, that we could slow down quite a bit and really learn to use what we have. Our designs wouldn’t hurt from that, we already have more computing ability than most of us know what to do with. And, there are a bunch of people who probably shouldn’t have access to this computing machinery, because they pretend that this makes them an engineer. In fact this can be dangerous to others.
dhengr -
I've also thought about developing a "Code of Standard Practice" for using software in engineering. Might be interesting. And you might need a different code for different disciplines.
Probably true for everyone
we are enriching someone that's for sure
I think the momentum is too strong. We might start some sort of mass campaign - get signatures, but I'm not too optimistic.
But how would you be able to police something like this?
Mentor the engineers around you - both young and old.
I've also thought about developing a "Code of Standard Practice" for using software in engineering. Might be interesting. And you might need a different code for different disciplines.
JAE:
I wasn’t really suggesting a signature campaign, all we have to do is quite buying that stuff when it doesn’t really benefit us or our practice. Each one of us doing that is worth much more than a bunch of signatures. Between the bunch of us, if we bought only a few copies and critiqued them right here we would discover that they really offer nothing new or better, just a complicated new reformulation. Then talk to your legislators, and building officials, at the state, county and city levels, wherever the adoption decision is made. Their stock in trade is nit-picking on minor code details, and yet, the ones that I have talked to are almost as overwhelmed and frustrated as we are. They can’t keep up with it either, and still get any work done. If we explained our position, and reasoning, to them, they might decide the new edition isn’t worth adoption, in which case we wouldn’t need to buy it. We do probably have to show them that the new version offers no improvements in safety or economy, because that is not normally what they do, without some guidance from practitioners.
The momentum is so strong because the building code writers and computer program writers have become much too comfortable selling us bells and whistles that don’t really improve our operation. In some respects this is just another layer in the economy, which gets its pound of flesh, while not really contributing much improvement to the process or the finished product. In fact, we see right in this thread that they complicate it. Again, I am not suggesting we don’t need building codes or computers and programs to do our work. We just don’t need the two intervening editions in what could be a ten or twelve year cycle, with a few addendums if the change is important enough. SO, just quit buying the new versions, you are just encouraging them.
By now, you certainly must be aware that I am a strong advocate of mentoring. And, when the young guys come into your office saying “they gave me the newest version free,” a slick promotional gimmick, and that’s how I learned to do it. You may want to do some mentoring by splainin that we use the version on generation older so you’ll have to do some slummin if you want to work here. But, we still feel we turn our a darn good finished product, which is specs. and plans that can actually be built with min. confusion and with few RFI’s. Your 4" stack of computer output is not the end result, although it has become a very complicated and at times confusssing end unto itself.
I wasn’t really suggesting a signature campaign, all we have to do is quite buying that stuff when it doesn’t really benefit us or our practice. Each one of us doing that is worth much more than a bunch of signatures. Between the bunch of us, if we bought only a few copies and critiqued them right here we would discover that they really offer nothing new or better, just a complicated new reformulation. Then talk to your legislators, and building officials, at the state, county and city levels, wherever the adoption decision is made. Their stock in trade is nit-picking on minor code details, and yet, the ones that I have talked to are almost as overwhelmed and frustrated as we are. They can’t keep up with it either, and still get any work done. If we explained our position, and reasoning, to them, they might decide the new edition isn’t worth adoption, in which case we wouldn’t need to buy it. We do probably have to show them that the new version offers no improvements in safety or economy, because that is not normally what they do, without some guidance from practitioners.
The momentum is so strong because the building code writers and computer program writers have become much too comfortable selling us bells and whistles that don’t really improve our operation. In some respects this is just another layer in the economy, which gets its pound of flesh, while not really contributing much improvement to the process or the finished product. In fact, we see right in this thread that they complicate it. Again, I am not suggesting we don’t need building codes or computers and programs to do our work. We just don’t need the two intervening editions in what could be a ten or twelve year cycle, with a few addendums if the change is important enough. SO, just quit buying the new versions, you are just encouraging them.
By now, you certainly must be aware that I am a strong advocate of mentoring. And, when the young guys come into your office saying “they gave me the newest version free,” a slick promotional gimmick, and that’s how I learned to do it. You may want to do some mentoring by splainin that we use the version on generation older so you’ll have to do some slummin if you want to work here. But, we still feel we turn our a darn good finished product, which is specs. and plans that can actually be built with min. confusion and with few RFI’s. Your 4" stack of computer output is not the end result, although it has become a very complicated and at times confusssing end unto itself.
- Status
Similar threads
- Locked
- Question
- Locked
- Question
- Locked
- Question