Nodes and comments in Radioss starter engine file
Hello,
I am wondering why my radioss starter input file runname_0000.rad contains all the nodes of my model. Is it possible to hide them in the file? In the picture there are few lines but these are much more and make the file quite large.
In addition, the parameters of the cards I defined in HyperMesh are not displayed in any block of the starter file. Is there a way of activating/deactivating the display of such comment lines? Below are highlighted in an example from Radioss Reference Guide, in my starter file they are missing.
Finally, I organized my nodes groups and elements in an include file. The Model Checker gives a warning about this. In the starter input file I get the following message in the interface blocks:
## HM Message : Set will be exported in correct include file
## HM Message : sets will be exported in correct include file
Is there something I am supposed to fix about this or is it fine?
Thank you for your help!
Best Answer
-
Andrea_20669 said:
Thank you Paul.
Regarding the nodes in the main file under the ## NODES block, I have placed my FE entities in an include file using Right Click>Organize in the Solver browser. Below is what I have at the moment in HM.
But still, the starter input file shows all the nodes.
For the export I forgot to mention that I have to export in the Radioss2019 template, for which the comments are not available. Can I export in the Radioss2021 template and then manually change the version in the engine starter file from 2021 to 2019?
Ah, ok, I see now, you are using the 'classic' HM interface, in that, Nodes are not shown separately in the browser, and I don't know of a way to move them easily independently. So far in Classic, You have moved the sets/groups of nodes, but not the nodes themselves.
When you 'organize' components into an include, you do get the option to move the nodes though
So, only way I can think to move the nodes alone in classic, would be to organize the components to your include (with node ref checked) and then move the components back to your main model (with node ref unchecked)
In the new HW/HM interface, Nodes are shown and you can move them to the include
Regarding version comments, doing as you suggest (export as 2021 and change header to 2019) is not recommended practice, cards change format between versions so changing the number manually risks losing data or misinterpretation of solver input. For your use case, what I would recommend if you need the comments in the 2019 format deck, is to export as 2019, then import/export from HyperCrash in Radioss 2019 format (that will add the comments to the deck).
1
Altair EmployeeAnswers
-
Yes, The nodes are absolutely needed in the starter, the solver needs to know where your nodes are! They define the structure of your model.
You could move them to an include if you don't want them in your main file
For comments, the 'solver field names' option (on export) turns them on/off in the file (this is only recently supported in HM, you say you are using 2021.2, in which case there should be this option to have them written, older versions of HM didn't have the option to write comments)
The warnings about sets are ok, I think you get this before first export only since the sets are only created then in some cases in HM.
1 -
Hi Andrew,
Radioss works with 2 types of file:
1. Starter file (Runname_0000.rad) - contains all Block data (FE information including all entities except Control cards & output requests)
2. Engine file (Runname_0001.rad) - contains exclusively Control cards & output requests
I normally, keep all the data related to FEM in separate includes so that it will be very easy if I modify particularly any specific file. While working with segregating includes files along with nodes and elements, we need to assure that corresponding nodes, elements, systems, etc are moved to corresponding includes. Once you segregate this data, you can export and check, whether everything is moved out of Starter file.
To cross check, if you are working with include files, it will be good idea to assign numbering scheme (with specific numbering range) for each include so that when you open any include file, with the node number, you can identify easily to which part/assembly that include belongs too.
Explanation from Paul is apt related to working.
Hope this helps.
Regards,
Prasad
1 -
Paul Sharp_21301 said:
Yes, The nodes are absolutely needed in the starter, the solver needs to know where your nodes are! They define the structure of your model.
You could move them to an include if you don't want them in your main file
For comments, the 'solver field names' option (on export) turns them on/off in the file (this is only recently supported in HM, you say you are using 2021.2, in which case there should be this option to have them written, older versions of HM didn't have the option to write comments)
The warnings about sets are ok, I think you get this before first export only since the sets are only created then in some cases in HM.
Thank you Paul.
Regarding the nodes in the main file under the ## NODES block, I have placed my FE entities in an include file using Right Click>Organize in the Solver browser. Below is what I have at the moment in HM.
But still, the starter input file shows all the nodes.
For the export I forgot to mention that I have to export in the Radioss2019 template, for which the comments are not available. Can I export in the Radioss2021 template and then manually change the version in the engine starter file from 2021 to 2019?
0 -
Prasad Reddi_20512 said:
Hi Andrew,
Radioss works with 2 types of file:
1. Starter file (Runname_0000.rad) - contains all Block data (FE information including all entities except Control cards & output requests)
2. Engine file (Runname_0001.rad) - contains exclusively Control cards & output requests
I normally, keep all the data related to FEM in separate includes so that it will be very easy if I modify particularly any specific file. While working with segregating includes files along with nodes and elements, we need to assure that corresponding nodes, elements, systems, etc are moved to corresponding includes. Once you segregate this data, you can export and check, whether everything is moved out of Starter file.
To cross check, if you are working with include files, it will be good idea to assign numbering scheme (with specific numbering range) for each include so that when you open any include file, with the node number, you can identify easily to which part/assembly that include belongs too.
Explanation from Paul is apt related to working.
Hope this helps.
Regards,
Prasad
Thank you Prasad! I thought I was doing exactly what you have explained to move FE entities to an include file. Perhaps I am missing something (?). Please, see the reply to Paul's message.
0 -
Andrea_20669 said:
Thank you Paul.
Regarding the nodes in the main file under the ## NODES block, I have placed my FE entities in an include file using Right Click>Organize in the Solver browser. Below is what I have at the moment in HM.
But still, the starter input file shows all the nodes.
For the export I forgot to mention that I have to export in the Radioss2019 template, for which the comments are not available. Can I export in the Radioss2021 template and then manually change the version in the engine starter file from 2021 to 2019?
Ah, ok, I see now, you are using the 'classic' HM interface, in that, Nodes are not shown separately in the browser, and I don't know of a way to move them easily independently. So far in Classic, You have moved the sets/groups of nodes, but not the nodes themselves.
When you 'organize' components into an include, you do get the option to move the nodes though
So, only way I can think to move the nodes alone in classic, would be to organize the components to your include (with node ref checked) and then move the components back to your main model (with node ref unchecked)
In the new HW/HM interface, Nodes are shown and you can move them to the include
Regarding version comments, doing as you suggest (export as 2021 and change header to 2019) is not recommended practice, cards change format between versions so changing the number manually risks losing data or misinterpretation of solver input. For your use case, what I would recommend if you need the comments in the 2019 format deck, is to export as 2019, then import/export from HyperCrash in Radioss 2019 format (that will add the comments to the deck).
1 -
Andrea_20669 said:
Thank you Prasad! I thought I was doing exactly what you have explained to move FE entities to an include file. Perhaps I am missing something (?). Please, see the reply to Paul's message.
Thats nice that you are also following the same way.
Looking at your comments in Paul's reply, sometimes, HM doesn't move automatically nodes & elements to the specific include file even though it says :-). So, its a good practice to move nodes, elements, systems manually to those corresponding include files. You can try this option and certainly it will work.
Cheers,
Prasad
1 -
Paul Sharp_21301 said:
Ah, ok, I see now, you are using the 'classic' HM interface, in that, Nodes are not shown separately in the browser, and I don't know of a way to move them easily independently. So far in Classic, You have moved the sets/groups of nodes, but not the nodes themselves.
When you 'organize' components into an include, you do get the option to move the nodes though
So, only way I can think to move the nodes alone in classic, would be to organize the components to your include (with node ref checked) and then move the components back to your main model (with node ref unchecked)
In the new HW/HM interface, Nodes are shown and you can move them to the include
Regarding version comments, doing as you suggest (export as 2021 and change header to 2019) is not recommended practice, cards change format between versions so changing the number manually risks losing data or misinterpretation of solver input. For your use case, what I would recommend if you need the comments in the 2019 format deck, is to export as 2019, then import/export from HyperCrash in Radioss 2019 format (that will add the comments to the deck).
Now everything is understood! I can correctly export without having all the nodes printed in the starter input file. Import/export with HyperCrash is also working for the comments display! Thank you for your patience and help!
1 -
Prasad Reddi_20512 said:
Thats nice that you are also following the same way.
Looking at your comments in Paul's reply, sometimes, HM doesn't move automatically nodes & elements to the specific include file even though it says :-). So, its a good practice to move nodes, elements, systems manually to those corresponding include files. You can try this option and certainly it will work.
Cheers,
Prasad
Exactly, now I have more clear how this works. Than you for your time and helpful support!
0