ZeroExperience
Aerospace
Perhaps you like Ansys or hypermesh?
Eng-Tips is the largest engineering community on the Internet
Intelligent Work Forums for Engineering Professionals
-
Congratulations waross on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!
Why do you like Femap? 5
- Thread starter ZeroExperience
- Start date
- Status
- Thread starter
- #3
ZeroExperience
Aerospace
I am looking into buying a different FEA product, and I honestly trust the opinions here more than I would from the opinions from their sales people.
-
1
- #6
tg,
We've been working on some improvements to non-manifold geometry for the upcoming FEMAP v11, if you're having Free Edge problems with Parasolid geometry, this might be extremely helpful. If you have some sample geometry, please send it to me and I will use it in testing to try to maximize the effectiveness of working with non-manifold shell geometry in FEMAP v11 and FEMAP v11+
sherman dot mark at siemens dot com
Mark.
We've been working on some improvements to non-manifold geometry for the upcoming FEMAP v11, if you're having Free Edge problems with Parasolid geometry, this might be extremely helpful. If you have some sample geometry, please send it to me and I will use it in testing to try to maximize the effectiveness of working with non-manifold shell geometry in FEMAP v11 and FEMAP v11+
sherman dot mark at siemens dot com
Mark.
Mark,
The free edge problems are occurring with geometry that we created in FEMAP, very cleanly, as commented by our re-seller.
The model in question has finally been completed (1/2 of 1 passenger railcar) in longer than I care to divulge here.
We truly thought that this was resolved with V10. We would really appreciate a seminar, course, or webcast on FEMAP best practices for stress-skin structures (in our case trains). i.e. When should we NOT use Geometry Midsurface Intersect, what really happens when using this and\or Non-manifold Add, and how to come back to cleaner geometry after using these commands, if they do not remove all the free-edges.
Why are free edges not even in the vocabulary of your competitors products?
tg
The free edge problems are occurring with geometry that we created in FEMAP, very cleanly, as commented by our re-seller.
The model in question has finally been completed (1/2 of 1 passenger railcar) in longer than I care to divulge here.
We truly thought that this was resolved with V10. We would really appreciate a seminar, course, or webcast on FEMAP best practices for stress-skin structures (in our case trains). i.e. When should we NOT use Geometry Midsurface Intersect, what really happens when using this and\or Non-manifold Add, and how to come back to cleaner geometry after using these commands, if they do not remove all the free-edges.
Why are free edges not even in the vocabulary of your competitors products?
tg
I like Femap because the people who develop it (and I don't mean the company, I mean the people themselves) are serious about making the process of doing FEA more powerful, productive and efficient. I have been using Femap xince 1995 and Nastran since 1987. Femap is the best pre-post for all Nastrans. It is a better pre/post for MSC.Nastran than MSC's own Patran, which I used for 14 years. The difference is in the "ergonomics". Patran was designed and developed by a committee ( a big committee) so it lacks coherence and workflow intelligence. Even in simple things like selecting entities based on their inherent logical association with other things is simple in Femap and clumsy in Patran. And making common geometry edits and have the mesh automatically update is superior in Femap compared to Patran. Patran has a few advantages over Femap, but not many. I think it sometimes does a tidier job of mesh transitioning in both surface and solid meshes with a smaller number of elements. However, based on a value and productivity proposition, Femap is all over Patran.
On Free Edges, any FEA program that doesn't have Free Edges in its vocabulary must be a b it weak. I'm not sure how a complex struture could be modelled without direct and user controllable choices on Free Edges. If the issue is mutliple coincident edges in geometry then "non-manifold add" deals with the bulk of that. Are there modelling packages that take away the thinking on free edges?
On Free Edges, any FEA program that doesn't have Free Edges in its vocabulary must be a b it weak. I'm not sure how a complex struture could be modelled without direct and user controllable choices on Free Edges. If the issue is mutliple coincident edges in geometry then "non-manifold add" deals with the bulk of that. Are there modelling packages that take away the thinking on free edges?
-
1
- #11
Femap ... Overcomplicated to use!!!
I have been a PATRAN user for many years and while it has limitations in features it is easy to use.
FEMAP has way too many options and redundancies ... many ways to do one thing !!! ... clutter !!!!
There are too many mouse clicks and menu popping required to execute an operation ... create a new group/name it/close window/open another window/ activate/go to add entities/ .... etc
Solution ... Open window to create new group, ..... everything required for name, activate, add entities ... etc SHOULD be in this window and NO going around the menu system to do this. If one wants to create a new group they generally want to do it so they can put entities into it and for it to be active.
Still trying to get my head around group reference/evaluate/re-evaluate etc ....... and are they really needed in the way described !!
This is just "groups" ... one could go through many other functions.
Keep the main menu less cluttered and offer more. Minimize the need to continually open windows during the execution of a user operation.
Issue for me is why does the user have to move around the menu system so much to achieve something a single window can do if properly populated with associated options?
It is the usability issue for me and the extent of clutter. This is not an experienced user problem as they already JUST DO IT because. However to attract a NEW user it should be SIMPLIFIED and made easier to use. It has much functionality, that is not my issue here.
A relatively new user !!!! ............. still a work in progress for me!!!!
Is it just me or do others gripe at this issue?
I have been a PATRAN user for many years and while it has limitations in features it is easy to use.
FEMAP has way too many options and redundancies ... many ways to do one thing !!! ... clutter !!!!
There are too many mouse clicks and menu popping required to execute an operation ... create a new group/name it/close window/open another window/ activate/go to add entities/ .... etc
Solution ... Open window to create new group, ..... everything required for name, activate, add entities ... etc SHOULD be in this window and NO going around the menu system to do this. If one wants to create a new group they generally want to do it so they can put entities into it and for it to be active.
Still trying to get my head around group reference/evaluate/re-evaluate etc ....... and are they really needed in the way described !!
This is just "groups" ... one could go through many other functions.
Keep the main menu less cluttered and offer more. Minimize the need to continually open windows during the execution of a user operation.
Issue for me is why does the user have to move around the menu system so much to achieve something a single window can do if properly populated with associated options?
It is the usability issue for me and the extent of clutter. This is not an experienced user problem as they already JUST DO IT because. However to attract a NEW user it should be SIMPLIFIED and made easier to use. It has much functionality, that is not my issue here.
A relatively new user !!!! ............. still a work in progress for me!!!!
Is it just me or do others gripe at this issue?
Outpost11,
FEMAP has many ways of doing the same things, to cater to new and legacy users alike. I strongly suggest either a course or to view the video demos on the Siemens site.
Link
You should look at using the menu tree on the LHS of your screen. Right clicking any item provides many useful options, including for Groups. A really useful feature is Automatic Add. All new entities created after toggling that on will be added to the active group. Other useful Group functions are (in the top menu): Group - Operations - Generate Property or material or other types. These commands automatically create groups of similar entities.
I have to say that since Version 10, I have much less of a need for creating groups, except for viewing or working with an actual slice of my model. This is because most entities in the menu tree have a check box next to them for toggling on or off.
tg
FEMAP has many ways of doing the same things, to cater to new and legacy users alike. I strongly suggest either a course or to view the video demos on the Siemens site.
Link
You should look at using the menu tree on the LHS of your screen. Right clicking any item provides many useful options, including for Groups. A really useful feature is Automatic Add. All new entities created after toggling that on will be added to the active group. Other useful Group functions are (in the top menu): Group - Operations - Generate Property or material or other types. These commands automatically create groups of similar entities.
I have to say that since Version 10, I have much less of a need for creating groups, except for viewing or working with an actual slice of my model. This is because most entities in the menu tree have a check box next to them for toggling on or off.
tg
Hi Tg
I really appreciate your response.
However, my point is it should not have many ways to do something. It just creates clutter and confusion. Experienced users are never the issue as they can adapt to simplifications very easily. I agree that the model info tree is a very nice addition.
However, just as an example, if I create a "new" GROUP I expect to have a single window pop up and input all of name, whether it should be active (toggle on/off), and option to enter entities to group without
having to just input name, then go back to Group and right click to activate, then go back again and right click to add entities .....
this is all excessive clicking ....
this is so with many more window menus, not just group !!!!
thanks
I really appreciate your response.
However, my point is it should not have many ways to do something. It just creates clutter and confusion. Experienced users are never the issue as they can adapt to simplifications very easily. I agree that the model info tree is a very nice addition.
However, just as an example, if I create a "new" GROUP I expect to have a single window pop up and input all of name, whether it should be active (toggle on/off), and option to enter entities to group without
having to just input name, then go back to Group and right click to activate, then go back again and right click to add entities .....
this is all excessive clicking ....
this is so with many more window menus, not just group !!!!
thanks
Fully agree, Outpost11!!
Also not very nice for the occasional user (who of course still understands the theory and what basically needs be done), or users with little time... Carrying along too much legacy (and even all the old code words for this and that).
Regards
Also not very nice for the occasional user (who of course still understands the theory and what basically needs be done), or users with little time... Carrying along too much legacy (and even all the old code words for this and that).
Regards
-
2
- #15
In my experience, you can't do occasional FEA. It has to be a full time job, for a set period. Unless you follow a very detailed, documented procedure each time you fire up the machine. You need to use it every day, day in , day out or you end up forgetting things and re-running analyses. This applies to all packages, not just FEMAP.
That's my 2 cents.
tg
That's my 2 cents.
tg
Outpost11 and gfbotha
For arguments sake I have to disagree.
I think that the possibility to do a task not just one way but several ways is a strength. Also, there is often differences between the similar ways when you come to the details.
Regarding groups I don't use them so much any more. A few years back I used them a lot but today I prefer the Model Info tree. But something that I do use a lot is Layers and combined with Groups (when I do use them) I find to to be very powerful. As for creating groups and just using them, I can do that. Group-Create and Group-Element-ID if I understand the task.
As for mouse clicks and Patran. A few years back MSC.Nastran for Windows disappeared. MSC wanted the N4W users to swap to MSC.FEA so I evaluated Patran. Basically, a colleague with a few decades of experience in Patran showed me how it works. One of my impressions was the huge amount of mouse clicks and he commented on that. Basically he agreed
.
You say you are a relatively new user and I think that is the main concern. You swapped from Patran to Femap and of course you could do things faster in Patran yesterday than you do in Femap today. I would be surprised otherwise ;-). I would probable feel the same regarding Patran.
But if you like Patran and dislike Femap, why do you use Femap? Perhaps that is the question you should ask.
Regards
Thomas
Ps I hope I haven't started in argument because that is not my intention.
For arguments sake I have to disagree.
I think that the possibility to do a task not just one way but several ways is a strength. Also, there is often differences between the similar ways when you come to the details.
Regarding groups I don't use them so much any more. A few years back I used them a lot but today I prefer the Model Info tree. But something that I do use a lot is Layers and combined with Groups (when I do use them) I find to to be very powerful. As for creating groups and just using them, I can do that. Group-Create and Group-Element-ID if I understand the task.
As for mouse clicks and Patran. A few years back MSC.Nastran for Windows disappeared. MSC wanted the N4W users to swap to MSC.FEA so I evaluated Patran. Basically, a colleague with a few decades of experience in Patran showed me how it works. One of my impressions was the huge amount of mouse clicks and he commented on that. Basically he agreed
.You say you are a relatively new user and I think that is the main concern. You swapped from Patran to Femap and of course you could do things faster in Patran yesterday than you do in Femap today. I would be surprised otherwise ;-). I would probable feel the same regarding Patran.
But if you like Patran and dislike Femap, why do you use Femap? Perhaps that is the question you should ask.
Regards
Thomas
Ps I hope I haven't started in argument because that is not my intention.
Hallo trainguy
Well, opinions differ - for say 25 years already I am doing FEA in-between all my other tasks as mechanical engineer. Let's say at least once a month, sometimes for two weeks continually. I started with EMRC NISA - had to keep cryptic notes/tips/guidelines for each type of analysis (Femap reminds me about those days). Then we got Cosmos DesignStar (today only existing as SWorks Simulation) - what a pleasure to use! - the UI is not cluttered and sort of leads one through the workflow required with input boxes relevant to the type of analysis selected upfront. I know what people typically say next..., but, no, it is quite powerful; could do all non-linear types, transient, heat, fatigue, etc. Recently I compared a hyperelastic result done with NX Femap, to same done with DesignStar - for all practical purposes the results were the same!
I belief that doing a wide scope of mechanical engineering, and not only analysis, helps one to better understand the context of a simulation and to make good changes to the design in a more efficient workflow. With all due respect, not all companies can afford an engineer doing only FEA, day in and out.
NX Femap's strong point is a lot of (hidden ;-)) functionality for the price.
Regards
Well, opinions differ - for say 25 years already I am doing FEA in-between all my other tasks as mechanical engineer. Let's say at least once a month, sometimes for two weeks continually. I started with EMRC NISA - had to keep cryptic notes/tips/guidelines for each type of analysis (Femap reminds me about those days). Then we got Cosmos DesignStar (today only existing as SWorks Simulation) - what a pleasure to use! - the UI is not cluttered and sort of leads one through the workflow required with input boxes relevant to the type of analysis selected upfront. I know what people typically say next..., but, no, it is quite powerful; could do all non-linear types, transient, heat, fatigue, etc. Recently I compared a hyperelastic result done with NX Femap, to same done with DesignStar - for all practical purposes the results were the same!
I belief that doing a wide scope of mechanical engineering, and not only analysis, helps one to better understand the context of a simulation and to make good changes to the design in a more efficient workflow. With all due respect, not all companies can afford an engineer doing only FEA, day in and out.
NX Femap's strong point is a lot of (hidden ;-)) functionality for the price.
Regards
- Thread starter
- #19
ZeroExperience
Aerospace
trainguy said:
Trainguy, my friend forwarded this to me. You might like video #2:
Who went to the conference? We need to be enlightened about this.
SteveMills
Aerospace
outpost11 has it right. My company has both FEMAP and PATRAN. The new kids are encouraged to use and learn both with an emphasis on FEMAP. Everyone immediately picks up PATRAN and completely drops FEMAP. That includes the kids that take a class in FEMAP while learning PATRAN without training. The only exceptions to this have been the kids that learned FEMAP in college. When starting fresh with the two codes, PATRAN is always selected.
PATRAN's panels logically step through the model building process whereas FEMAP is back and forth and hidden under layers of menus to build a model. Several experienced engineers that used PATRAN took the 3 and 5 day training classes on FEMAP and came back unable to use the product due to the excessive busyness and complexity of the interface. The class was functionally a waste of time and money since FEMAP was not even attempted.
As outpost11 noted, groups are a huge pain to set up in FEMAP whereas PATRAN it is straight forward. The loads definition in FEMAP I still can't get straight so I just manually create the load sets in a text editor.
FEMAP is robust in general with reading and writing models. PATRAN is broken in several ways especially after MSC bought it with the goal of forcing everyone to switch to MSC products. My version of PATRAN is a bit dated so I'm not sure if this issue continues.
A major thing missing from FEMAP is the ability to report the area and lengths of particular elements. Reviewing someone else's model is more difficult in FEMAP.
The help manual in FEMAP makes your teeth hurt. Although it is complete, it does not report where in the maze of menus and toolbars to find a command.
Also in FEMAP, the mouse button spins the model and a keyboard button has to be pressed to do translation. I do not understand this setup since if I'm working on a model; it is normally being translated to see details. Spinning the #W$% model just blows the area I'm looking at off the screen, which requires significant effort to be recovered. I actually have no use for the mouse spinning feature and they have it locked as the standard.
The interface to FEMAP was obviously created by some IT dude with no understanding of FE analysis. Nothing is where it should be. In strong contrast, PATRAN places things in logical places. Another big issue with FEMAP is the pull-down menus. Frequently 3 or 4 levels of menus have to be opened for a function and then it only operates on a single item. Then this is repeated, over, and over, and over.... PATRAN's use of panels puts all the features needed in one place. Also in FEMAP, when a selection/command is open, the values of the data table or entry info cannot be scrolled to check a value, such as needing the node ID in an element. FEMAP has a superior interface to ABAQUS, but maybe I should just learn CAE. It's got to be easier than FEMAP.
Entity Info in FEMAP is almost useless. It displays the nearest element continuously. There does not appear to be a way to lock it so that the info can be copied or evaluated. PATRAN in contrast has a nice method to query an element and places it into a table for easy comparison with additionally selected elements.
You might be able to get the work done with FEMAP, but PATRAN will not cause dental problems and will finish the task in half the time with more confidence in the quality of the model.
BTW: note that I have been trying to learn FEMAP now for 4years. My ability with it is limited to fairly primitive and superficial items due to the tremendous difficulty in using the product. I waste time on FEMAP because my company is forcing us off PATRAN (consolidation of software and vendors???).
PATRAN's panels logically step through the model building process whereas FEMAP is back and forth and hidden under layers of menus to build a model. Several experienced engineers that used PATRAN took the 3 and 5 day training classes on FEMAP and came back unable to use the product due to the excessive busyness and complexity of the interface. The class was functionally a waste of time and money since FEMAP was not even attempted.
As outpost11 noted, groups are a huge pain to set up in FEMAP whereas PATRAN it is straight forward. The loads definition in FEMAP I still can't get straight so I just manually create the load sets in a text editor.
FEMAP is robust in general with reading and writing models. PATRAN is broken in several ways especially after MSC bought it with the goal of forcing everyone to switch to MSC products. My version of PATRAN is a bit dated so I'm not sure if this issue continues.
A major thing missing from FEMAP is the ability to report the area and lengths of particular elements. Reviewing someone else's model is more difficult in FEMAP.
The help manual in FEMAP makes your teeth hurt. Although it is complete, it does not report where in the maze of menus and toolbars to find a command.
Also in FEMAP, the mouse button spins the model and a keyboard button has to be pressed to do translation. I do not understand this setup since if I'm working on a model; it is normally being translated to see details. Spinning the #W$% model just blows the area I'm looking at off the screen, which requires significant effort to be recovered. I actually have no use for the mouse spinning feature and they have it locked as the standard.
The interface to FEMAP was obviously created by some IT dude with no understanding of FE analysis. Nothing is where it should be. In strong contrast, PATRAN places things in logical places. Another big issue with FEMAP is the pull-down menus. Frequently 3 or 4 levels of menus have to be opened for a function and then it only operates on a single item. Then this is repeated, over, and over, and over.... PATRAN's use of panels puts all the features needed in one place. Also in FEMAP, when a selection/command is open, the values of the data table or entry info cannot be scrolled to check a value, such as needing the node ID in an element. FEMAP has a superior interface to ABAQUS, but maybe I should just learn CAE. It's got to be easier than FEMAP.
Entity Info in FEMAP is almost useless. It displays the nearest element continuously. There does not appear to be a way to lock it so that the info can be copied or evaluated. PATRAN in contrast has a nice method to query an element and places it into a table for easy comparison with additionally selected elements.
You might be able to get the work done with FEMAP, but PATRAN will not cause dental problems and will finish the task in half the time with more confidence in the quality of the model.
BTW: note that I have been trying to learn FEMAP now for 4years. My ability with it is limited to fairly primitive and superficial items due to the tremendous difficulty in using the product. I waste time on FEMAP because my company is forcing us off PATRAN (consolidation of software and vendors???).
- Status