OSArch Community

Integration of Code_Aster in an IFC-driven workflow for structural analysis

  1. J

    Thanks @bernd! If you can get something out of AxisVM would be great. I have seen your examples but I would prefer to work on IFC4 versions, as there are some changes and I would prefer tackle with IFC2x3 later. So, it would make sense only if AxisVM supports IFC4.

    I have already tried with ETABs and Sofistik and my problem is with curve connections and supports regarding planar elements, such as slabs and walls. These connections are not exported in the structural ifc files and that's where I am stuck at the moment. If you can try with AxisVM even with a very simple model of two walls connected with support at the base or a model similar to the one in your description folder, that would be great.

    In any case, this week I plan to work quite a bit on this so I could possibly look also into your existing models or try to extract sth with BlenderBIM.

  2. D

    @Jesusbill I've just checked and even though I do have access to Robot there is am 11 euro daily charge my manager might not be too happy about. So for Robot I can't help. Do you want me to ask around to hear if anyone has it installed and uses it? I could maybe ask them for something basic. You say 'a wall with a support at the base' but that doesn't sound very specific to me. Let me know if we should pursue this. I notice a Tekla Structures 2018 licensed is available. There are also a bunch of Strusoft tools.

  3. J

    @Duncan thanks, I have managed to make a few steps forward these days so at this moment it is not indispensable. That said, if you can ask and there is any engineer that uses it, it could be useful to have an export in structural ifc. Any model that contains also slabs, or walls and either connections between them (connection along a line) or a support at the base (always along a line) (support means simply that the displacement and rotation there is blocked) it would be good. The issue is that no software tested until now handles this type of connections, they do handle connections between beams and columns and with slabs and walls (point connections) but not line connections.

    But as I said, it would be good to have but not absolutely necessary, I am going forward with my interpretation and I guess in the future we can modify anything that does not look consistent based on valid export models.

    Regarding Strusoft I am not aware but at this point my bet is that it will not handle this issue as well, given that the big players do not as well.

  4. B

    @Jesusbill said:

    @duncan during the last meetup you expressed your availability to help with the creation of some structural ifc models using commercial software.

    I have access to Graitec Advance Design which exports IFC, but it doen't appear to transfer any connection or anchor information. Attached a test with two steel members, I didn't think to try slabs or walls. It actually has lots of IFC options for different target software, but I'll need to find some time to go through them all.

  5. J

    Thank you @brunopostle for your effort, luckily I have found the way to go on without any reference files from other software. Not sure if Graitec Advanced Design can handle ifc with structural analysis entities. The file you sent does not contain any IfcStructural* objects.

  6. J

    Support for connections between slabs, walls, etc. has now been added in ifc-to-code_aster and two examples added in the analysis-models.

    The examples are created based on a script that does the inverse operation of the original script, taking as input a json file and returning the structural ifc file.

    • A single slab slab_01

    • A simple structure structure_01

  7. B

    @Jesusbill said:

    Thank you @brunopostle for your effort, luckily I have found the way to go on without any reference files from other software. Not sure if Graitec Advanced Design can handle ifc with structural analysis entities. The file you sent does not contain any IfcStructural* objects.

    The IFC files from Advance Design did look incomplete.

    Is your workflow something like this:

    1. author IFC file (including connection information) in blenderbim

    2. use ifc2ca to generate all the input files for analysis, or does the loading regime need to be authored separately?

    3. run analysis

    I see that @moult is working on an ifcsverchok blender add-on. I'm looking forward to the day when I can generate a structure in Sverchok with all the relevant IfcStructural elements ready for analysis.

  8. J

    @brunopostle yes that is plausible workflow. Loads can definitely be in the IFC file as well, except for seismic which are not considered, but other than that typical static loads, loadcases and loadcase combinations I believe are considered. Loads are next in line regarding my work after implementing also surface connections and refining the algorithms for all connections in general as at the moment there are some limitations in what I did for curve connections and I am trying to overcome those for a better and straight-forward usability.

    IfcSverchok looks promising but a structural engineer/developer has to step in to implement ifc structural, Dion cannot do everything, and the same would be needed for BlenderBIM also as at the moment there has not been much progress at the UI level for structural analysis concepts. I would be willing to work on this but I wish I could have also at least another person that helps, my experience in blender is minimal and I do not want to undertake too much and not deliver, given also the work needed for ifc2ca. Let's see how things evolve :)

  9. D
  10. J

    hello everyone. First off, I am fairly new to OSArch and find all of the projects from BlenderBIM to integration with code_aster, etc. quite simply amazing to the opportunities it presents. As a structural engineer (who knows rudimentary Python programming), I'm quite interested in finding a suitable workflow for perparing input files from IFC for structural analysis. @Jesusbill I'm not so familiar with IFC nevermind ifc structural but in my spare time I'm going to start learning what I can on how to use BlenderBIM to author ifc files and I'm sure at some point I'd be able to help out with providing feedback/ideas on structural analysis concepts at a UI level. Not sure how much time I will be able to spend on this at the moment. Also i am thinking that a "IFC to code_aster to IFC" workflow would be amazing but tend to think that rather than writing the whole results for each 1D or 2D element it would seem better to me to also evaluate the analysis results in terms of the plausibility of the geometric parameters within the structural analysis round-tripping "IFC to code_aster to IFC" workflow. As documenting the results of the code_aster analyses seems to be separate from collaboration of let's say the architect and the structural engineer. anyway just want to also comment on reading through the whole thread that the progress so far kind of blows my mind. to start if there are things which I could help with I'd be glad to give it a try despite me being quite new to how these "workflows" work.

  11. J

    Hi @jchkoch ! Great to have you here and to have your support.

    At the moment we are still developing the very basics for authoring structural analysis models in BlenderBIM, I think we will need a couple of months to arrive at a point that we can have a basic workflow (create members, connections and supports) to test with users, but I will definitely post in the forum when this time comes. And one reason is that we are do implementing basic stuff now but I reckon we will have to go over a second round after to polish some aspects that are important to effectively apply structural engineering. I am thinking of showing/implementing units for all quantities, proper orientation implementation of local axes, limits on the property input values and other "details" that we are leaving behind for now .

    Regarding your statement:

    evaluate the analysis results in terms of the plausibility of the geometric parameters within the structural analysis round-tripping "IFC to code_aster to IFC" workflow.

    I imagine your are simply referring to verification of the structural elements right? I certainly think that this would be a great and very important task that completes the cycle of the structural engineering work, apart from drawings and documentation. But for that we would need more resources to do it now, I see it more feasible once we have the possibility to create and to run the model.

    anyway just want to also comment on reading through the whole thread that the progress so far kind of blows my mind

    Well to be honest there has been a setback at some point as there are no valid authoring software to create the models. My initial work was based on the pipeline after the model is authored, which we will find handy in our next step, the analysis implementation.

    if there are things which I could help with I'd be glad to

    I think there will be soon some tasks we could need help, other than user testing the UI when it's the right time.

    Talking with @krande yesterday we do want to start building some profile/section and material libraries, for example for steel profiles and for concrete/steel materials, which would be great to have them as IfcProjectLibrary (right @Moult ?) that can be used to directly select materials and profiles in BlenderBIM. We have to set this up, but I definitely see potential work for non-coders engineers there, in compiling tables with all needed information to create the libraries. We will post more info on this when it gets started

  12. J

    @Jesusbill it is just great to see that there are opensource projects trying to come up with workflows and especially as far as they have come here is amazing to me.

    In terms of what I envision in terms of structural analysis round-tripping workflows is that what I found in real projects is that oftentimes the structural engineering calculations are needed but most other partners (i.e. architect, mechanical, client, etc.) are generally only interested if it works structurally and then what the geometry/materials/cost (e.g. working with architects especially early in the conceptual design phase, they want to know from us, the structural engineer, what wall thickness are needed).

    At the moment we are still developing the very basics for authoring structural analysis models in BlenderBIM, I think we will need a couple of months to arrive at a point that we can have a basic workflow (create members, connections and supports) to test with users, but I will definitely post in the forum when this time comes.

    As far as how I may help, I wanted to start off with easing my way into the OSArch community but I like to think I am somewhat sufficient programming (especially python) to help out some in very early testing of workflows (at the moment struggle sometimes with getting dependencies etc. to work).

    I think there will be soon some tasks we could need help, other than user testing the UI when it's the right time. Talking with @krande yesterday we do want to start building some profile/section and material libraries, for example for steel profiles and for concrete/steel materials, which would be great to have them as IfcProjectLibrary (right @Moult ?) that can be used to directly select materials and profiles in BlenderBIM. We have to set this up, but I definitely see potential work for non-coders engineers there, in compiling tables with all needed information to create the libraries. We will post more info on this when it gets started

    In terms of creating profile/section and material libraries, I could definitely lend a hand. In the past I have done some very rudimentary setup of accessing this type of structural data using Python (i.e. using steel I-sections incorporated within the Canadian Steel Institute of Construction (CISC)). In general, I think these types of organizations (in many countries) already have these types of sections/materials catalogued (albeit in out-dated formats (Excel, PDF, etc.). On the otherhand, I think a balance between having a library of these types of profiles/section and materials and allowing for a parametric definition would be good. Anyway, this might be a good way for me to start helping out here.

  13. J

    @jchkoch great you are based in Canada? It would be great to have you working on material and profile libraries that are following American and Canadian standards!

    Regarding round-tripping yes it is absolutely clear that engineers need to communicate to the architect if current dimensions are ok or which are the ones needed, but the engineer needs to carry out these calculations, even if he does not need to communicate them. So, if we want to talk about a complete open-source pipeline, the verification part is part of the game.

    If you have python skills this could be a nice field of application I would say, in any case I think we will have the chance to talk about it soon. Good to have you onboard in any case

  14. M

    @jchkoch we have a bi-weekly meetup working on structural - would you be interested in joining / watching? That will keep you updated with the progress we are making.

  15. J

    @John I have transferred the discussion in this thread

    @jchkoch @Massimo discussion on profile libraries has now been transferred in this thread

  16. J

    @jchkoch: there is work currently done also on costing with @SigmaDimensions @Moult @iosvarms @dav80 and others, so I believe eventually we will be able to use it also to estimate costs of preliminary designs maybe even directly from the structural analysis elements (?). I agree that compiling and delivering this information back to architects especially in early stages of design are super important, thanks for bringing this up.

  17. J

    @Moult I would be interested in joining(participating) in the bi-weekly meetup working on structural if at all possible. Would love to be able to keep updated with the progress.

  18. J

    @jchkoch Meetings are held every Tuesday and Thursday at 9 am CET, but this time is only indicative, it depends then on the availability we may or may not have, so some times we meet the night before (CET time) or when we can eventually. The best place at the moment to know when and where (which specific link of the shared video platform) is to follow the OSArch chat.

  19. C

    @jchkoch said:

    hello everyone. First off, I am fairly new to OSArch and find all of the projects from BlenderBIM to integration with code_aster, etc. quite simply amazing to the opportunities it presents. As a structural engineer (who knows rudimentary Python programming), I'm quite interested in finding a suitable workflow for perparing input files from IFC for structural analysis. @Jesusbill I'm not so familiar with IFC nevermind ifc structural but in my spare time I'm going to start learning what I can on how to use BlenderBIM to author ifc files and I'm sure at some point I'd be able to help out with providing feedback/ideas on structural analysis concepts at a UI level. Not sure how much time I will be able to spend on this at the moment. Also i am thinking that a "IFC to code_aster to IFC" workflow would be amazing but tend to think that rather than writing the whole results for each 1D or 2D element it would seem better to me to also evaluate the analysis results in terms of the plausibility of the geometric parameters within the structural analysis round-tripping "IFC to code_aster to IFC" workflow. As documenting the results of the code_aster analyses seems to be separate from collaboration of let's say the architect and the structural engineer. anyway just want to also comment on reading through the whole thread that the progress so far kind of blows my mind. to start if there are things which I could help with I'd be glad to give it a try despite me being quite new to how these "workflows" work.

    Hi everyone! another structural engineer with Python capabilities over here. I discovered BlenderBIM at the end of last year and am keeping an eye on its progress because it looks amazing. I believe that we (structural people) have quite some room for improvement on our BIM workflows, and the kind of work that you guys are doing over here will definitely help to bridge that gap. So a ton of thanks for that!

    At the moment, I am still struggling a little with understanding the IFC Schema, everything which is possible with ifcopenshell and lastly BlenderBIM itself. Up to now, I have not had too much time for it but am slowly making some progress.

    Out of curiosity, and pardon for the beginner level on IFC, but how does structural design link to geometry within the Schema? for instance, if we have an IfcBeam (with its corresponding IfcRels to whatever 3D geometry representation) for the beam itself and an IfcStructuralCurveMember corresponding to its structural idealization (with its corresponding IfcRels to whatever 1D geometry representation):

    • are they linked between them?

    • specifically, are they parametrically linked in some way? (meaning for instance that a change in length updates both the 3D geometry and the structural model)

    The latter would be awesome, but perhaps I am confused and it is on the actual structural software to assure that linkage, rather than on relations within the Schema.

    Thanks in advance,

  20. M

    @cvillagrasa They are linked between them using another rel. This allows architects and structural engineers to have the "back and forth" coordination process.

    They are not parametrically linked. There are proposals for parametric linking extensions to IFC, but these are not yet a reality. In the meantime, there are a variety of workarounds, which if enough of us get together and show it works, we can put together a well thought out proposal to buildingSMART.

  21. J

    Ten days ago, Aether Engineering gave a presentation in the ProNET meeting entitled "IFC-Driven Code_Aster Analyses for Buildings", where we presented a summary of the "openBIM Structural Pipeline" work that has been done by the Open Source community until now.

    ProNET is a professional network of companies that use Code_Aster and we are part of it since 2017.

    I have attached the presentation pdf file, it is a general summary where I mainly talked about the BlenderBIM Add-on, IFC2CA, and a bit about adapy, as the three "projects" that I see fundamental in this workflow. I also presented OSArch in the last slides as I thought it was important to take this opportunity, as well as the opportunity to talk about the structural workflow with Code_Aster with the companies of the network, although a large number of them are from the mechanical sector (aviation, automotive, etc.). Of course, just the slides may not tell the whole story as they were also "minimal" but I wanted to share this here and will do also in the social media.

    I would say that the work done until now in the BlenderBIM Add-on is focused, and continues to develop, in specific areas of the structural analysis domain, but I would consider that it is not yet "organized" in a way that a user can be clearly guided in knowing what has been included until now. So, one of my goals would be to "centralize" the UI for the structural analysis domain and produce some material/tutorial that explains how a user can build a structural analysis model in Blender. Another goal is to write the roadmap, I know I should have done this already but in May I have been bloody busy and didn't work much on this. I am positive though that June will be different ;)

    Cheers and I will update the forum when there are news.

  22. M

    Looks like a beautiful presentation! Perhaps one for sharing on social media? :) I will definitely share within my company.

  23. D
  24. B

    Where do you guys meet every thuesday 9:00 CET? I have been part of osarch for years, may be before it even exists ;-) and I never came accross. Shame on me ... I should join the meeting too ...

  25. M

    @bernd join the OSArch live chatroom via either IRC or Matrix :) That said, @Jesusbill and I haven't met in a while on a structural hacking session. Perhaps it's time to restart the weekly meetups and do some hacking :)

  1. Page 1
  2. 2
  3. 3
  4. 4

Login or Register to reply.