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 MintJulep on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!
Autodesk Robot Structural Analysis 6
- Thread starter ChaliK
- Start date
a "free-free" modal analysis simply means a modal analysis of a structure without any restraints as in a "free-free" beam analysis. There is a zero frequency rigid body mode (with no internal stress/strain) for each unrestrained degree of freedom of the overall structure (usually six for 3D - but can be more if hinge mechanisms are involved). As I said earlier it is an excellent diagnostic tool to find unwanted mechanisms.
Masomenos: I believe johnhors was talking about the use of a modal analysis to debug a model that has inadequate restraints, and possesses rigid body motion.
SAP2000 seems to have no problem with such models although as John points out there are other structural programs that can't do this.
SAP2000 seems to have no problem with such models although as John points out there are other structural programs that can't do this.
masomenos:
First of all, I hope we don't have to get into an argument over this. It's always nicer to stay friends.
I agree with you that in many cases the so called structural softwares are egual or better for certain applications. One reason can be that they often "hold your hand" to a larger extent then some more general softwares. BUT the general software often offer a much larger solvercontrol. Personally I prefer the general software. I like to control the software and not viceversa.
"Structural codes can do nonlinear", again true. But nonlinear is a huge subject. Everything from crash analysis to surface contact and advanced material models and beyond.
What I mean is that regardless of what you want to do with YOUR software you have to make sure that it can perform beform you buy it. To say that "it can do nonlinear" isn't enough because its a huge subject.
Just as dynamic analysis is a huge subject. I have seen a presenter say when asked say "Sure we can du a dynamic analysis". What was available was eigenfrequencies.
You ask why we didn't buy SAP200. The problem was linear statics. We had non-US trafic loads and very complex loadcombinations. We had a very specific request regarding the resultpresentation and it couldn't be done. Might be possible today but at the time it was impossible. We had a software tht could do it but wanted a more modern interface. SAP wasn't it.
I think that what Johnhors is refering to by free-free is a unconstrained model, like a spaceship or an aeroplane.
It's a very efficient way to troubleshoot certain problems in models. Especially when parts haven't been merged together properly. You run a eigenfrequency analysis and see it there are any free parts (with f = 0.0 Hz). Some solvers crash under those conditions others don't.
I won't go into detail about every can and can't in different softwares. The OP was refering to Robot and I think we are OT enough as it is. If Robot is a good choice or not is ultimately a question about what the application/s is and what codes you require. Codes is a big issue for structural FEA in my experience and not an issue at all for general FEA. That is at least one big difference.
Since Autodesk bought it my guess is that Robot will evolve, now they have better resources. Probably as the analysis engine together with Revit. We'll se what happens in the future.
Regards
Thomas
First of all, I hope we don't have to get into an argument over this. It's always nicer to stay friends.
I agree with you that in many cases the so called structural softwares are egual or better for certain applications. One reason can be that they often "hold your hand" to a larger extent then some more general softwares. BUT the general software often offer a much larger solvercontrol. Personally I prefer the general software. I like to control the software and not viceversa.
"Structural codes can do nonlinear", again true. But nonlinear is a huge subject. Everything from crash analysis to surface contact and advanced material models and beyond.
What I mean is that regardless of what you want to do with YOUR software you have to make sure that it can perform beform you buy it. To say that "it can do nonlinear" isn't enough because its a huge subject.
Just as dynamic analysis is a huge subject. I have seen a presenter say when asked say "Sure we can du a dynamic analysis". What was available was eigenfrequencies.
You ask why we didn't buy SAP200. The problem was linear statics. We had non-US trafic loads and very complex loadcombinations. We had a very specific request regarding the resultpresentation and it couldn't be done. Might be possible today but at the time it was impossible. We had a software tht could do it but wanted a more modern interface. SAP wasn't it.
I think that what Johnhors is refering to by free-free is a unconstrained model, like a spaceship or an aeroplane.
It's a very efficient way to troubleshoot certain problems in models. Especially when parts haven't been merged together properly. You run a eigenfrequency analysis and see it there are any free parts (with f = 0.0 Hz). Some solvers crash under those conditions others don't.
I won't go into detail about every can and can't in different softwares. The OP was refering to Robot and I think we are OT enough as it is. If Robot is a good choice or not is ultimately a question about what the application/s is and what codes you require. Codes is a big issue for structural FEA in my experience and not an issue at all for general FEA. That is at least one big difference.
Since Autodesk bought it my guess is that Robot will evolve, now they have better resources. Probably as the analysis engine together with Revit. We'll se what happens in the future.
Regards
Thomas
What I mean is that regardless of what you want to do with YOUR software you have to make sure that it can perform beform you buy it. To say that "it can do nonlinear" isn't enough because its a huge subject.
~~~~
This is the 2nd or 3rd time that you've directly or indirectly attributed statements or opinions to me that I have never written, nor do I believe them. I NEVER suggested, nor do I believe that "nonlinear analysis" is some all-encompassing statement. There are obviously many types of NL analysis, which is why I've gone out of my way to be specific. Neither have I suggested that "capable of dynamic analysis" explains everything.
Thanks for your reply regarding your evaluation of SAP2000, although you didn't mention enough specifics to enable a reply. SAP can generate bridge influence lines, it can automatically export directly to Excel and it offers an API, so output result options have likely improved since you last evaluated. Or not, depending on the specifics. What program did you choose over SAP? SAP has gone through a lot of changes and enhancements, particularly over the past 4 years or so.
I've tried to be specific in my comments, pointing out actual real-life applications where I have seen structural analysis software turn out to be more productive than general purpose FEA in advanced analytical applications. Again, response spectrum dynamic analysis is far easier and more efficient in most structural programs as compared to general purpose FEA programs, most of which I've seen, force the user to fake out the program by defining some mass at the base. That's just one example. Pushover analysis involving nonlinear material plastic hinges is another application where some structural software programs are better suited than, say, Ansys or Algor, for common nonlinear material analysis design requirements. Analysis of concrete foundations modeled with finite elements supporting unbalanced dynamic loads is another application. At the time, Ansys could not even provide a section cut for solid finite elements in order to report integrated shears and moments. Load combinations is another advantage where structural software tends to offer significantly more than general purpose FEA.
How about an honest show of hands.. how many of you were completely unaware that there are structural analysis software capable of nonlinear time history analysis? Specifically, nonlinear TH for NL boundary conditions (gaps, friction, multilinear plastic springs), nonlinear TH for large displacement analysis and nonlinear TH with nonlinear material plastic hinges? In addition, there is NL staged construction and creep and shrinkage available through structural software programs. I'm admittedly being presumptuous, but I'd bet that a lot of you were unaware that programs like SAP2000, Robot, and GT Strudl can perform that sort of advanced analysis.
~~~~
This is the 2nd or 3rd time that you've directly or indirectly attributed statements or opinions to me that I have never written, nor do I believe them. I NEVER suggested, nor do I believe that "nonlinear analysis" is some all-encompassing statement. There are obviously many types of NL analysis, which is why I've gone out of my way to be specific. Neither have I suggested that "capable of dynamic analysis" explains everything.
Thanks for your reply regarding your evaluation of SAP2000, although you didn't mention enough specifics to enable a reply. SAP can generate bridge influence lines, it can automatically export directly to Excel and it offers an API, so output result options have likely improved since you last evaluated. Or not, depending on the specifics. What program did you choose over SAP? SAP has gone through a lot of changes and enhancements, particularly over the past 4 years or so.
I've tried to be specific in my comments, pointing out actual real-life applications where I have seen structural analysis software turn out to be more productive than general purpose FEA in advanced analytical applications. Again, response spectrum dynamic analysis is far easier and more efficient in most structural programs as compared to general purpose FEA programs, most of which I've seen, force the user to fake out the program by defining some mass at the base. That's just one example. Pushover analysis involving nonlinear material plastic hinges is another application where some structural software programs are better suited than, say, Ansys or Algor, for common nonlinear material analysis design requirements. Analysis of concrete foundations modeled with finite elements supporting unbalanced dynamic loads is another application. At the time, Ansys could not even provide a section cut for solid finite elements in order to report integrated shears and moments. Load combinations is another advantage where structural software tends to offer significantly more than general purpose FEA.
How about an honest show of hands.. how many of you were completely unaware that there are structural analysis software capable of nonlinear time history analysis? Specifically, nonlinear TH for NL boundary conditions (gaps, friction, multilinear plastic springs), nonlinear TH for large displacement analysis and nonlinear TH with nonlinear material plastic hinges? In addition, there is NL staged construction and creep and shrinkage available through structural software programs. I'm admittedly being presumptuous, but I'd bet that a lot of you were unaware that programs like SAP2000, Robot, and GT Strudl can perform that sort of advanced analysis.
ThomasH wrote above on June 30:
So I would say, "All FEA tasks"? No definitly not.Structural FEA in terms of steel frames etc? Sure, as long is it doesn't get TO nonlinear or TO dynamic.
You should be honest enough to own up to your earlier mistatement of facts when challenged. Contrary to your assertion, several structural analysis programs offer more than analysis of "steel frames" as well as offeing nonlinear analysis and dynamic analysis, including dynamic analysis. Anyone can read what you wrote. You made a mistatement of fact, and when called on it, you refused to admit your error.
So I would say, "All FEA tasks"? No definitly not.Structural FEA in terms of steel frames etc? Sure, as long is it doesn't get TO nonlinear or TO dynamic.
You should be honest enough to own up to your earlier mistatement of facts when challenged. Contrary to your assertion, several structural analysis programs offer more than analysis of "steel frames" as well as offeing nonlinear analysis and dynamic analysis, including dynamic analysis. Anyone can read what you wrote. You made a mistatement of fact, and when called on it, you refused to admit your error.
masemenos:
First, in your previous post you said that I put more or less put words in your mouth. Sorry about that, it was never my intention. When I wrote "you" in that sentence I did not mean you personally. I meant "you" the analyst as in anyony doing the analysis, but that wasn't clear. Sorry about that.
Then you latest post. You say that I have made a mistatement of facts regarding structural FEA. In what respect?
Havent we agreed on that structural FEA can't do all FEA tasks? That's what I stated, nothing else.
And I wrote "steel frames etc", I was never specific regarding the "etc". That wasn't an accident
.
Finally, when I stated that it shouldn't get "TO nonlinear or TO dynamic". I was thinking about things like contact like multibody-contact of moving bodies. Or material models like hyperelastic. Or ......? Greg mentioned crash analysis.
Note: I don't think I have claimed that EVERY general software can do all of the above mentioned. But a general pre/post processor can usually "talk" to one of the solvers that can.
Sure, we can look specifics like you do and claim that structural FEA is just as good. It might be for some applications. You for example mentioned response spectrum analysis in your previous post. The "fake mass" that sometimes is used. There is nothing "fake" about the large mass method. I think it comes from the aerospace industri. And there are often other options in general FEA if you don't like "fake mass", unlike structural FEA.
If you think I've made an error and unfairly critizised structural FEA. I was just trying to point at some differences. Mayby I have faied miserably.
You have been very specific about the capacity of SAP2000. Since this tread is about Robot I have tried to keep it more general. I have colleges who have tested both of them and they have their pros and cons. At the time I thing Robot was considered better partially due to the connections with Revit. And it shold be possible to get the model from Robot to a general software but that might require some programming of the API.
I'm convinced that whatever software you intend to buy you need to test it and make sure that it fits yur needs. If you end up in the structural or the general "world" is not really important.
As for SAP2000, the reason we skipped it there were two problems if I remember correctly.
First, when we look at the results. For the design we needed max/min moments (envelops) and the associated axial forces and shear. We could only get max/min of everything. It worked if we used the internal design engine but since we had other code requirements it didn't work. We needed everything "on paper".
Second, bridge loads. We needed more general loadtrains then were available. Like for example, two or three concentrated loads (or loadgroups) moving independantly of each other. Now the code has changed so it might be enough with two loads (or loadgroups) independant of each other.
Regards
Thomas
First, in your previous post you said that I put more or less put words in your mouth. Sorry about that, it was never my intention. When I wrote "you" in that sentence I did not mean you personally. I meant "you" the analyst as in anyony doing the analysis, but that wasn't clear. Sorry about that.
Then you latest post. You say that I have made a mistatement of facts regarding structural FEA. In what respect?
Havent we agreed on that structural FEA can't do all FEA tasks? That's what I stated, nothing else.
And I wrote "steel frames etc", I was never specific regarding the "etc". That wasn't an accident
.Finally, when I stated that it shouldn't get "TO nonlinear or TO dynamic". I was thinking about things like contact like multibody-contact of moving bodies. Or material models like hyperelastic. Or ......? Greg mentioned crash analysis.
Note: I don't think I have claimed that EVERY general software can do all of the above mentioned. But a general pre/post processor can usually "talk" to one of the solvers that can.
Sure, we can look specifics like you do and claim that structural FEA is just as good. It might be for some applications. You for example mentioned response spectrum analysis in your previous post. The "fake mass" that sometimes is used. There is nothing "fake" about the large mass method. I think it comes from the aerospace industri. And there are often other options in general FEA if you don't like "fake mass", unlike structural FEA.
If you think I've made an error and unfairly critizised structural FEA. I was just trying to point at some differences. Mayby I have faied miserably.
You have been very specific about the capacity of SAP2000. Since this tread is about Robot I have tried to keep it more general. I have colleges who have tested both of them and they have their pros and cons. At the time I thing Robot was considered better partially due to the connections with Revit. And it shold be possible to get the model from Robot to a general software but that might require some programming of the API.
I'm convinced that whatever software you intend to buy you need to test it and make sure that it fits yur needs. If you end up in the structural or the general "world" is not really important.
As for SAP2000, the reason we skipped it there were two problems if I remember correctly.
First, when we look at the results. For the design we needed max/min moments (envelops) and the associated axial forces and shear. We could only get max/min of everything. It worked if we used the internal design engine but since we had other code requirements it didn't work. We needed everything "on paper".
Second, bridge loads. We needed more general loadtrains then were available. Like for example, two or three concentrated loads (or loadgroups) moving independantly of each other. Now the code has changed so it might be enough with two loads (or loadgroups) independant of each other.
Regards
Thomas
JoshPlumSE
Structural
Relax guys.... I think you're both making a mountain out of a mole-hill.
There are plenty of structural specific applications that advertise non-linear capabilities. I do very much agree that these structural programs (when appropriate) will be much easier to use than NASTRAN or other really high-powered FEA programs. But, that's at least partially because I'm a structural guy.
Even so, some of those program's actual capabilities do not compare well to what is suggested based on their advertising. Therefore, we should all be cautious about using programs without first confirming their suitability for our project.
There are plenty of structural specific applications that advertise non-linear capabilities. I do very much agree that these structural programs (when appropriate) will be much easier to use than NASTRAN or other really high-powered FEA programs. But, that's at least partially because I'm a structural guy.
Even so, some of those program's actual capabilities do not compare well to what is suggested based on their advertising. Therefore, we should all be cautious about using programs without first confirming their suitability for our project.
- Status
Similar threads
- Locked
- Question
- Question
- Question