Stuck? 
support@mecway.com
support@mecway.com
Mecway
FEA
Couplings type Kinematic combined with Distributing surface moment
There are few example files checking the Coupling cards response in ccx. Its performance has been reported as suspicion sometimes and I thought this could be interesting.
I have selected a textbook example with exact solution with three pipe sections with different diameters. Pipe sections are connected with *Coupling + *Kinematic.
The bar of length 1m is fixed at both ends. It has tubular sections with outer diameter of 300mm in section AB, 200mm in section BC and 150mm in section CD. All sections are 5mm thickness.
This bar has an applied axial moment of 20 KN applied by means of a *Coupling + *Distributing at an intermediate section C. Dimensions are as shown in the picture.
The aim is to check Couplings type Kinematic and Distributing combined and response regarding:
-Loads and displacement connection capabilities between the different sections.
-Overall reaction moments and loads response on each support.
-Maximum tangential stress on each section and its uniformity across the model.
Final Expected / Result / Deviation is as follows:
Additionally, all load and moment reactions are zero at A and D (except for torsional moment) as expected.
Stresses and displacements are clean and uniform between sections and across bodies as expected.
I know the file is limited and looks just an specific load case but,…this configuration is relatively simple and flexible providing a base for other configurations (for example using DCOUP elements or combining loads + moments).
I'm using it to go deeper in the understanding of the coupling cards and their limitations.
I would appreciate if anyone wanted to suggest a better approach or any possible drawback of my set up.
NOTE: I have found much more convenient to compute moments directly from the nodal forces. SOF and SOM are inaccurate for such mesh, and I wonder if this could be one of the reasons why these cards have sometimes been questioned. (Ccx manual warns about the accuracy of SOF SOM).
I have selected a textbook example with exact solution with three pipe sections with different diameters. Pipe sections are connected with *Coupling + *Kinematic.
The bar of length 1m is fixed at both ends. It has tubular sections with outer diameter of 300mm in section AB, 200mm in section BC and 150mm in section CD. All sections are 5mm thickness.
This bar has an applied axial moment of 20 KN applied by means of a *Coupling + *Distributing at an intermediate section C. Dimensions are as shown in the picture.
The aim is to check Couplings type Kinematic and Distributing combined and response regarding:
-Loads and displacement connection capabilities between the different sections.
-Overall reaction moments and loads response on each support.
-Maximum tangential stress on each section and its uniformity across the model.
Final Expected / Result / Deviation is as follows:
Additionally, all load and moment reactions are zero at A and D (except for torsional moment) as expected.
Stresses and displacements are clean and uniform between sections and across bodies as expected.
I know the file is limited and looks just an specific load case but,…this configuration is relatively simple and flexible providing a base for other configurations (for example using DCOUP elements or combining loads + moments).
I'm using it to go deeper in the understanding of the coupling cards and their limitations.
I would appreciate if anyone wanted to suggest a better approach or any possible drawback of my set up.
NOTE: I have found much more convenient to compute moments directly from the nodal forces. SOF and SOM are inaccurate for such mesh, and I wonder if this could be one of the reasons why these cards have sometimes been questioned. (Ccx manual warns about the accuracy of SOF SOM).
Howdy, Stranger!
It looks like you're new here. If you want to get involved, click one of these buttons!
Comments
I think it's worth to update the liml file taking that small difference into consideration. Accuracy is generally improved. (File updated).