plastics materials database

Dear Sir,
I am totally new to Mecway . And interested in knowing whether Mecway has a plastics materials data base which can be used for assigning the material properties and carry out the large deformation analysis .

Comments

  • Hello drssp. It doesn't currently have a materials database though this is a commonly requested feature so hopefully in a future version. For now, you would have to get material data from elsewhere, such as matweb.com .

    For large deformations, you need to install the CalculiX CCX solver available here http://bconverged.com/download.php . Then connect it to Mecway through Tools -> Options -> CalculiX.
  • Thanks Victor for quick response.
    I am a researcher working on plastics parts behaviour under forced vibrations at elevated temperatures. I am curious to know how Mecway can help in optimising shapes / Sections of plastic molded assembly of parts. Normally their are three to four parts for a assembly . For example multimeter box or door handle etc. In case you have a example or case study pl let me know .
    Regards
    Drssp
  • It doesn't have automatic shape optimization, so you would need to do that iteratively. 3-4 part assemblies are perfectly OK. I don't have any examples that specific sorry but there are examples of assemblies and contact in the manual.
  • Thanks Victor for this quick update.
    I have one more question.
    Is the mesh generation module ngmesh is 64 bit version?
    Also in case mesh nodes are more than 1 million what happens to deck while solving.
    Regards.
    Drssp
  • You can check on my own site some assembly examples, all were meshed, pre and postprocessed in Mecway

    http://www.4pingenieria.com/pipping/

    Once you catch the workflow will find easier than other FEA packages.

    Regards
  • The built in Netgen mesher is 32 bit. There's no hard upper limit on the number of nodes. The 1 million listed on the website is just what I've tested Mecway's internal solver with. CCX can probably solver larger models.
  • Impressive meshes on your site Sergio
  • Thanks prop_desing! I have the luck of made superb training with very smart guys. As I start long time ago when workstation were very modest compared to today ones, we need to be very precise and keep control of quantity of elements because times to solve would be very long. Even worst, my simulations include nonlineal material, self contact and big deformations all together, so you must be very carefull to make a good mesh.

    In my industry (rubber parts) good meshing is a must for achiving accurate results in razonable time to solve. That's why I feel a little confuse when I heard about such terrible quantity of elements. New people in FEA tend to replace good meshing skills (as geometry simplification, feature deletion, simetries...) with simple direct element size reduction, but normally that´s not the answer or the best for a good analysis. Worst thing sometimes is when somebody try to estimate the accuracy or power of a solver or FEA package with only the number of nodes he can manage... I don't know, after years in the matter there are other things more important than just to know if X solver can manage 10^6 elements or if Y solver can manage 10^7. Best paramaters to me are if can I apply all the boundary condition that I need? Support multiple steps, multiple contact, specific material models? And the more important... can I pay for that software???

    Best Regards!
  • thanks Sergio / Victor
    Your tips will help me to pick up faster.
    can I use multi threading during number crunching while using Mecway solver?
    This will help to reduce the total time during getting trained on some of the complex plastic parts.
    I observed that in my system, the cup usage is always less than 13 % and memory is less than 17% .
    Thanks and regards.
    Drssp
  • Use CCX solver (don't remember about Mecway solver) that allow nolineal material plastic definition, contact and multiple CPU. Tomorrow I will finish a plastic snap belt assemby (two parts) with all of that and a realistic good mesh for the application to show you.

  • hi sergio, i agree with everything. except i like to know how many nodes can be solved in core with a given amount of memory. that's important in my case. mecway has all the features i need at a more than reasonable price.
  • For the smart guys of the forum, I just realize that the blue part is inverted :-)

    Have finished succefuly a first run on the insertion operation, will try to run also the opposite direction to see how it performs. Tomorrow will post some nice pictures.
  • Thanks for update Sergio.
  • Dear Sergio, Any update on snap belt assembly case ?
  • edited May 2016
    Drssp, this is the example.

    I has removed all the small radius and features that don't affect the simulation, and then cut to have an equivalent quarter of the model. With this two simple things you can decrease the size of the model by maybe ten, small finishing radius would take a lot of elements and add distortion to the mesh. I have let only radius in the stressed areas (have done a preliminar test before to see were I need to refine the mesh), also mesh refinement there and in contact zones.

    I have finish with a very well shaped mesh (didn't find a way to check the quality yet, but looks well), with 49360 elements and 11949 nodes.

    I have applied a PA66GF35 material (Ultramid, a very good reinforced polyamide from BASF), contact and drive the simulation by displacement (normally solvers perform better aplying a displacement than a load). Two load steps were defined, one for assembly and the second for traction force.

    See how well CCX 2.10 manage the contacts, he even compute the case were the belt assemby is being under excesive force (and then fails and release without the normal lateral press-to-release).


    From the simulation now we know clipping/unclipping forces, stress under normal or abuse operation, stressed areas, if the part will have some interference... We can tune easily then the geometry or material to improve those values as we need and run again the simulation to see the improvements

    Having the posibility of ploting the stress vs time graph in the part in real time is also great, is easy to understand/asociate the peak stress before fail and releasing or the constant stress when being pushed (points for Mecway!). Is incredible to be allow to perform such complex simulation with an affordable software.

    What next? Maybe adding the belt, and apply the traction load by means of that (as compression is made by hand appying forces directly in the plastics parts I didn't include), to test also the belt attachments.

    Regards
  • That's a fantastic example Sergio. Thanks a lot for working it and posting it.
  • Thanks Sergio for this update .
    drssp
  • Hi everyone, don't mean to barge in on a thread but this seems like a relevant post.

    Is the Plastic (CCX) material tab not active until a later version? I take it all plastic material properties for large deformation needs to be done at the moment through a manual case for CCX?

    Thank you in advance.
  • You have to switch to nonlinear analysis type for it to be enabled.

    That tends to be quite confusing especially since CCX automatically turns on geometric nonlinearities when you use a plastic material anyway, so I hope to clear this up in the next version.
  • Sergio would you mind sharing your plastic clip example analysis file? I'd love to dig into the details of that analysis!!
Sign In or Register to comment.

Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!