I have been working in construction and engineering field for many years. But I have not done any structural design or analysis extensively for at least 10 years although I would analyze structures very roughly or on concept level occasionally. When I was doing structural design many years ago, my manager didn't want to buy computers and software to complete designs. So I had to do everything manually. Fortunately, the company had a small library contained various standards, codes, engineering manuals, etc. I could always found the design data of various loads (wind load, live load for various structures, etc.), property for common material (weight, strength, elongation rate, proportion, etc.) and common structural calculation formulas. But I know that software are so powerful now. They can do almost all structural calculation for us and are much faster. Do we still need to master manual structural calculation skills?
Eng-Tips is the largest engineering community on the Internet
Intelligent Work Forums for Engineering Professionals
-
Congratulations waross on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!
Is manual calculation still needed nowadays in structural design? 5
- Thread starter ccpe
- Start date
-
1
- #22
milkshakelake
Structural
Some young engineers now use software without doing hand calculations first. They become data entry workers instead of engineers.
I personally find a software, use hand calcs to extensively check the results for a few months, and then use the software based on faith and common sense. Because if I don't use the software, I won't be able to keep up with my competitors.
I personally find a software, use hand calcs to extensively check the results for a few months, and then use the software based on faith and common sense. Because if I don't use the software, I won't be able to keep up with my competitors.
-
1
- #23
Seems the conversation has strayed some from the original post. To me the original post was "Do you need to master "manual structural analysis" when computer programs are so advanced". The original poster seams to have corrected "manual calculations" to "manual structural analysis".
With that in mind, you still need the manual structural analysis ability even with the best software. You may not have to "master it" but you had better be pretty capable of knowing it. With very little manual ability, you will not know when something just looks wrong.
I will elaborate, on this previous quote. A well-known software to this forum I used about 30 years ago. It is a good software even now. But, we noticed the seismic reactions for seismic acting in a certain direction was wrong, seismic to the left, had reactions to the left for example. We contacted them, they found the problem and corrected it. Had we not been capable of noticing it, the software would have run several more days/weeks/months until someone else reported it. They released a new version because of this but never stated why the new version was released. I have come to find out, that is fairly standard in the industry. So anyone who did not notice the flaw, tough luck.
I fully understand that structures are extremely complicated in comparison to the "good ole days" but that does not diminish your need to at least have a feel for what looks "right".
With that in mind, you still need the manual structural analysis ability even with the best software. You may not have to "master it" but you had better be pretty capable of knowing it. With very little manual ability, you will not know when something just looks wrong.
Ron247 said:
I will elaborate, on this previous quote. A well-known software to this forum I used about 30 years ago. It is a good software even now. But, we noticed the seismic reactions for seismic acting in a certain direction was wrong, seismic to the left, had reactions to the left for example. We contacted them, they found the problem and corrected it. Had we not been capable of noticing it, the software would have run several more days/weeks/months until someone else reported it. They released a new version because of this but never stated why the new version was released. I have come to find out, that is fairly standard in the industry. So anyone who did not notice the flaw, tough luck.
I fully understand that structures are extremely complicated in comparison to the "good ole days" but that does not diminish your need to at least have a feel for what looks "right".
Settingsun
Structural
Oldestguy, the Sleipner A offshore platform is the usual cautionary tale about not understanding the computer model and not having effective checking.
Ron247, what software are you referring to? It's not industry standard to hide bug fixes. That's a sign of a shonky company. See the link below: warts and all.
Ron247, what software are you referring to? It's not industry standard to hide bug fixes. That's a sign of a shonky company. See the link below: warts and all.
Sorry, I will not mention their name. They are a very well established company and their policies have most likely changed since the early 90s. So maybe a more correct statement by me is "It was a fairly common at the time". They have been mentioned multiple times in this forum. This was about 25 years ago, but I have found several more companies that send out updates but never really say why the update was released. They may say, "Corrected an input issue with Loads" but do not say "Any force that was input as Force in the Positive X was actually stored in Moment about the Negative Y". To elaborate on the problem, they assumed seismic in the X would be worked first, then seismic in the Y/Z. So after seismic X was worked, they cleared out the array and did Y or Z. I did my project Y/Z first, then X. They software had no requirements either way but all their examples were X, then Y/Z. Well, if you do Y or Z first, the array did not get cleared out, so your X information was added to your Y/Z information.
I have not worked with their software in about 25 years but do remember the release with the correction. The information never stated the precise problem. That is why I started looking more closely at releases, especially the ones with letters like 14b, 12a, 6d. The scheduled releases may be 12, 14, 19 etc but the lettered ones were generally repairs. I was just making a point that software has bugs and will continue to have them forever. I have done a lot of programming myself and know how hard it is to be foolproof. Even though I am from the Fortran days, basic programming concepts are the same.
I have not worked with their software in about 25 years but do remember the release with the correction. The information never stated the precise problem. That is why I started looking more closely at releases, especially the ones with letters like 14b, 12a, 6d. The scheduled releases may be 12, 14, 19 etc but the lettered ones were generally repairs. I was just making a point that software has bugs and will continue to have them forever. I have done a lot of programming myself and know how hard it is to be foolproof. Even though I am from the Fortran days, basic programming concepts are the same.
I think the recent thread on shear deformations, in this forum, is relevant to this discussion.
Applying theoretical procedures in manual calculations can give incorrect results.
Examination of the results of simple computer calculations can give valuable insights.
Asking whether we should use computer calculations or hand calculations is false dichotomy.
Doug Jenkins
Interactive Design Services
Applying theoretical procedures in manual calculations can give incorrect results.
Examination of the results of simple computer calculations can give valuable insights.
Asking whether we should use computer calculations or hand calculations is false dichotomy.
Doug Jenkins
Interactive Design Services
With most software going subscription, I am now doing most of my calculations by hand with my old HP 15 using classical methods. I still have a copy of Risa 3d that I "own" (for now) but I only use it for really indeterminent problems.
I get to build and test break all my stuff and that's where my final answer comes from.
I get to build and test break all my stuff and that's where my final answer comes from.
Same here Buggar, I still have my last purchased Risa 3D but due to subscriptions, I am free to look elsewhere each year. In the past, upgrades were less $$$ than purchasing someone else's product. I think I was with RISA since the mid to late 90s. I have a lot of spreadsheets and other routines to do the bulk of my work.
In 20 years, I've found very few errors in commercial software, but there have been numerous examples of programs performing a calculation in a way that I don't agree with. For that reason, I always check a new (or new to me) feature with manual calculations before I trust it. I think it's malpractice to not do this.
All calculations should be checked, but why does the check need to be a "manual calculation"?
Doug Jenkins
Interactive Design Services
I try to simplify things so that I can run some quick hand calcs on the important parts. Usually I am trying to get a modeler/drafter started long before the final member sizes are chosen and details are complete. Complicated items get the complicated computed models.
If done properly, hand calcs tell the story about the design intent of the structure. Maybe 50% words and 50% numbers. Young guys who eschew hand calcs produce all numbers and tend to neglect the telling of the story. Kudos to those who do it all!
As structural engineers, our job is to tell the contractors what the Owner wants to build, it's not to produce volumes of calculations (unless your AHJ requires it!).
If done properly, hand calcs tell the story about the design intent of the structure. Maybe 50% words and 50% numbers. Young guys who eschew hand calcs produce all numbers and tend to neglect the telling of the story. Kudos to those who do it all!
As structural engineers, our job is to tell the contractors what the Owner wants to build, it's not to produce volumes of calculations (unless your AHJ requires it!).
OK, but you said that not to do the check by manual calculations was "malpractice".
Doug Jenkins
Interactive Design Services
Doug Jenkins
Interactive Design Services
TehMightyEngineer
Structural
IDS said:
Agreed. I believe we're using the term "manual calculation" to mean a bunch of different ways to check software output and it was the original question.
I regularly check software or my spreadsheets against each other. Probably the most beneficial checks I've found are to benchmark software against some textbook examples. This establishes that the software does what you expect it to do and verifies the methods. These checks are generally quick to do but valuable, especially when learning new software.
Ian Riley, PE, SE
Professional Engineer (ME, NH, VT, CT, MA, FL) Structural Engineer (IL, HI)
Good question. IMO, one way or the other, the verification must be grounded in the engineer's knowledge. That usually means manual calculations, comparison with another program the engineer wrote or verified, or a textbook example that the engineer understands.
I am seeing great replies.
Personally I like using Mathcad for manual calculation as well. It is much easier to go back and understand weeks later than Excel, and it is easy to update number when the architect or your project engineer change their mind.
Personally I like using Mathcad for manual calculation as well. It is much easier to go back and understand weeks later than Excel, and it is easy to update number when the architect or your project engineer change their mind.
-
1
- #38
“Do we still need to master manual structural calculation skills?”
I am just now seeing this post – and it is a topic near and dear to my heart.
YES! In my opinion all computer analysis must be validated with simple manual calculations. To blindly rely on software is, in my opinion, negligent. If nothing else, a few minutes spent validating your computer analysis with a calculation or two gives you some “simmer time” – time to stare at the computer results and ask yourself, “Does this make sense?” The blind reliance by on computers by some engineers is in my opinion, a disaster waiting to happen. I have been a structural engineer for over 40 years and I have seen several “near misses” because of engineers having the mindset “…but the computer said it was ok.” In addition, engineers must always ask themselves, “What did the computer not check?” What are the flaws in the software? What are the default settings? Are you even aware of all of the default settings?
Manually validating the results of software analysis and design is something that is not taught is schools - but it should. Young engineers graduating today seem to be proficient in jockeying the software, but if you ask them if the answer makes sense, they usually give you a blank look.
I am just now seeing this post – and it is a topic near and dear to my heart.
YES! In my opinion all computer analysis must be validated with simple manual calculations. To blindly rely on software is, in my opinion, negligent. If nothing else, a few minutes spent validating your computer analysis with a calculation or two gives you some “simmer time” – time to stare at the computer results and ask yourself, “Does this make sense?” The blind reliance by on computers by some engineers is in my opinion, a disaster waiting to happen. I have been a structural engineer for over 40 years and I have seen several “near misses” because of engineers having the mindset “…but the computer said it was ok.” In addition, engineers must always ask themselves, “What did the computer not check?” What are the flaws in the software? What are the default settings? Are you even aware of all of the default settings?
Manually validating the results of software analysis and design is something that is not taught is schools - but it should. Young engineers graduating today seem to be proficient in jockeying the software, but if you ask them if the answer makes sense, they usually give you a blank look.
In response to “oldestguys” question asking if any errors have resulted due to mistakes by the computer which resulted in the EOR being held accountable – I am certain that there have been such claims against engineers. The problem however is that design flaws are often discovered during construction. A flaw is found. The reason for the flaw is determined. A fix is implemented. The structure is made safe. Someone is blamed – and that entity pays for the fix (or their insurance company pays) – and if no one is injured, the public never finds out about the original mistake. I suspect that often “computer errors” are the result of young engineers with little experience being given the task of modeling complex structures without having a full understanding of what they are doing. So it is not necessarily a flaw with the computer, but is instead a flaw in understanding what the computer is doing – and a failure to manually validate the results.
I'm pretty sure that no-one is suggesting that we should blindly rely on software. The question is, is manual calculation the only valid method of verification? I don't think it is.
The blind reliance on books by some engineers is in my opinion, a disaster waiting to happen. I have been a structural engineer for over 40 years and I have seen several real failures because of engineers having the mindset “…but the books said it was ok.” In addition, engineers must always ask themselves, “What did the calculation not check?” What are the flaws in the text? What are the unstated assumptions? Are you even aware of all of the unstated assumptions?
Validating the results of software analysis and design with alternative and independent computer analyses is something that is not taught in schools - but it should. Old engineers who graduated long ago seem to be proficient in jockeying the manual calculations, but if you ask them if the underlying assumptions and simplifications might be better checked with a computer analysis, they usually give you a blank look.
Doug Jenkins
Interactive Design Services
- Status
Similar threads
- Question
- Question
- Locked
- Question