Forum

ABAQUS-SwiftComp GUI is available

  1. Bo Peng

    Somehow I cannot find your new questions in this discussion panel, so I paste them first as follows:

    After neglecting the pop up window showing error, I am getting the following figure. I am not sure if the mesh is being generated or not. Because whenever I am trying to get homogenized property for 1D beam, it is saying that wire model can't be used for beam analysis. Then again whenever I am trying to get homogenized property for 2D shell, it is saying that file is not being found.

    Answer:

    1. I am not sure what macroscopic model you want to use. If you want to model the original structure as a beam, you can use beam model in Homogenization. However, a 1D SG cannot give all the required information to get effective beam stiffness matrix, that is why you see the warning message.

    2. For the second problem, you may check the black command window ( which was opened when you click the short cut of the Abaqus-swiftcomp GUI) and the .echo file.  But since I cannot see your attached file, I am not sure about your problem. 

  2. Anonymous

    Hello,

    I have recently started using TextGen4SC. This is very helpful. I have several questions regarding this.

    1) Is it available only in the cloud? Is there any offline version of it to use with ABAQUS? 

    2) I have calculated elastic properties of flax using SwiftComp. Is it possible to calculate elastic properties of individual yarn using TextGen4SC? Later I will use the calculated elastic properties of yarns to calculate elastic properties of woven fabric. 

    3) To calculate elastic properties of woven fabric, only the elastic constants of the yarns are significant. Fiber diameter, fiber density don't matter. Am I right? 

    4) For weft direction yarns, the x-y-z axis of TextGen4SC coincides with 1-2-3 axis of the fiber. So plugging in the elastic values of yarns is quite easy. Buy for warp directin, the x-y-z axis of TextGen4SC don't coincide with the 1-2-3 axis of the fiber (or yarn). In that case how can I plug in the elastic constants in the TextGen4SC accurately? 

    5) While I was trying to plug in the Poisson's ratios, I found that, they are designated as Poisson's ration X, Poisson's ration Y, Poisson's ratio Z. Do they mean Poisson's ration v_12, v_13 and v_23 respectively? 

    I apologize for such a big note. Thank you in advance! 

  3. Xin Liu

    Hi,

    Thank you for using TexGen4SC. 

    1. Currently, it is only available in the cloud, but you can found all source codes in the supporting documents, so you can compile it on your own system. If you are using Linux system, it is not very easy to compile and I would suggest you to use HUB version. If you are using windows system, it is relative easy that you can just download all the source codes and follow the instructions provided by TexGen official website. We have compiled the installation package for windows, but it is highly depended on the system and python libs, so the safest way is to compile the code on your own computer.

    2. To calculate yarn properties, you need to use common 2D SG "Square Packing" or "hexagonal packing" in the SwiftComp on the Hub. This is very easy to get yarn properties as we just treat it as UD. TexGen4SC can take care of the yarn undualtion automatically when you calculate woven fabric.

    3. Yes, you are right.

    4. The local coordinate of each yarn is generated in the code, so you just need to input the properties in the material coordinate and TexGen4SC will do the transformation during the calculation.

    5. Yes.

     

    Please let us know if you have any questions.

     

    Thanks

    Xin 

  4. Anonymous

    Thank you Mr. Liu for your reply. 

    I have another quick inquiry. Do I need to plug in the values of elastic constants of yarn and matrix in MPa unit while calculating elastic properties of woven fabric? I plugged in the values in GPa unit and I got the following output (file attached). It looks like the E1 value is 17.53 GPa, although from the output file it is not easily readable (probably because I didn't plug in the values in MPa). 

     

    Thank you in advance! 

     

    1.PNG

  5. Xin Liu

    Hi,

     

    The default unit for the inputs in TexGen4SC is MPa, and the unit will convert to Pa in the output. Although SwiftComp doesn't have any restrictions, the unit is automatically converted inside the TexGen. So you should input the properties with MPa as unit and the results are in Pa unit. Sorry for the confusions.

     

    Best,

    Xin 

  6. Anonymous

    Thank you again!

    I didn't take any class on Plate and Shell Theory. That's why I am facing hard time to figure out the following problem. 

    I am trying to find elastic properties of a laminated beam using SwiftComp Cloud Version. For now I am using two layers, both of them made of same material and both of them are alinged at [0/0] stacking sequence. The thickness of each lamina is 5mm and the width is 180mm. The elastic properties of the material that I am using for each layer is given below - 

    E1 = E2 = 1753 MPa, E3 = 7180 MPa, G12 = 2120 MPa, G13 = G23 = 1959 MPa, nu12 = 0.203, nu13 = nu23 = 0.295

    After homogenization, I am getting the attached output. Here S11 = 3.8E-08. So E1 = 1/S11 = 26315789 Pa = 26.32 MPa. 

    I was wondering if this value is right or I am doing something wrong. Shouldn't I get E1 = 1753 MPa (or at least close to it?), because I am using same material with same stacking sequence? 

     

    Thank you in advance! 

     

    4.PNG

  7. Wenbin Yu

    The 4x4 stiffness matrix is beam stiffness matrix, the S11 is 1/EA with A as the cross-sectional area. ---- Emailed forum response from wenbinyu@purdue.edu
  8. Su Tian

    A new way to install the Abaqus-SwiftComp GUI is included in the manual now. User can now start the GUI and do the analysis from any directory without having to make more copies of the code or keep all files in the same place.

    Su

  9. Wenbin Yu

    Su, thanks a lot for the update. I guess that the details are given in the updated manual. Wenbin ---- Emailed forum response from wenbinyu@purdue.edu
  10. Anonymous

    Mr. Xin Liu,

     

    Good morning

     

    Currently I am using TexGen4SC to calculate elastic properties of a unit cell of yarn/epoxy composite. I had a few inquiries regarding TexGen4SC, like following - 

    1) In the GUI, I need to input the values of "Yarn Spacing", "Yarn Width" and "Fabric Thickness". Should the units of these parameters be in milimeter? 

    2) To input the elastic properties of yarns, I need to calculate them first, which can be done using a regular hexagonal unit cell. How can I know which volume fraction of fiber should I use in the hexagonal unit cell? Is it the same volume fraction of yarn that I want to make the yarn/epoxy unit cell of?

    3) How can I control the volume fraction of the yarn in the unit cell of yarn/epoxy that I am going to generate using TexGen4SC?  

    4) Normally the stiffness of weft and warp yarns are not the same. Are they by default considered to be the same in TexGen4SC? 

    My cordial apology for such a long note. Thanks in advance! 

  11. Wenbin Yu

    I will let Mr. Xin Liu answer 1), 3), 4). 

    2) The volume fraction of the fiber in a tow should be given. If not, assume it is fully packed. This value is not the same as the yarn volume fraction. 

     

     

  12. Xin Liu

    Hi,

    Thank you for using TexGen4SC. My answers to your questions are:

    1 & 4. The geometric units are not fixed just like in ABAQUS. However, by default, geometric parameters are assumed to be in meter to calculate effective properties by SwiftComp. Note that if you keep all the default units (geometric and materials), the SC results is in Pa. The weft and warp yarns are the same by default, but you can easily assign different yarn properties through GUI.

    3. You can not control the volume fraction of yarns in a unit cell, but you can check the volume faction by options through GUI after you create the unit cell .

     

    Thanks

    Xin

  13. Xin Liu

    Hi,

    Thank you for using TexGen4SC. My answers to your questions are:

    1 & 4. The geometric units are not fixed just like in ABAQUS. However, by default, geometric parameters are assumed to be in meter to calculate effective properties by SwiftComp. Note that if you keep all the default units (geometric and materials), the SC results is in Pa. The weft and warp yarns are the same by default, but you can easily assign different yarn properties through GUI.

    3. You can not control the volume fraction of yarns in a unit cell, but you can check the volume faction by options through GUI after you create the unit cell .

     

    Thanks

    Xin

  14. Xin Liu

    Hi,

    One thing I forgot to mention that if you only want to use TexGen4SC to compute effective properties by SwiftComp, then the geometric units will not affect the material properties.

     

    Thanks

    Xin

  15. Su Tian

    Version 1.2.2 Release Notes

    New Features

    • Added a mesh control input so that mesh size and element type can be set in the control file for generating cross section with airfoil shape.

    Bug Fixes

    • Fixed the viewing problem when generating 1D structure genome.
    • Fixed the error when generating airfoil cross section with multiple fillings.
    • Fixed the error when import materials file only without layups file in the 'New Layups' tool.

    Optimization

    • Improved the parsing process for converting Abaqus input file to SwiftComp input file.
  16. Anonymous

    Hi,

    What type of mesh element is used in swiftcomp online version (like plain strain, shell etc.)? How many nodes per element are there (like 3 or 4)? What is the function of Set Order 2 option during meshing? 

     

    Regards

  17. Xin Liu

    Hi,

     

    SwiftComp online version uses Gmsh as mesh generator, and you can choose different element types according to your needs. Please refer to Gmsh4SC manual in the supporting documents to see how to set different element types. Here, I guess you are talking about triangle (3 nodes) and quadrangle (4 nodes) elements, and these elements will be linear elements by default. "Set order 2" will give you second order triangle (6 nodes) and quadrangle (9 nodes) elements.

     

    Best,

    Xin

  18. Anonymous

    Hi

    I was trying to play with some random shapes. Fig. 1 and Fig. 2 have same material property (E = 200 GPa, v = 0.3) with same volume of hole in the middle (4%). But after homogenization, two figures are giving completely different E1 values. The first figure is giving E1 = 192 GPa which makes sense according to rule of mixture. But the second figure is giving E1 = 150.75 GPa which doesn't make sense according to rule of mixture. 

    Could you please tell me why is there a difference?  

    fig.png

  19. Ning Liu

    Can you also provide input files of both cases?

  20. Anonymous

    Please go to the following link to download the .cae and .jnl file. 

    https://drive.google.com/drive/folders/0BzDQpu7yML7dUnEyOU9tMEhhZkE?usp=sharing

    Please let me know if the link doesn't work. 

    Thanks!