I have a 3D model with extended bar. At the end of that bar, I have to model a torsional spring to adjust the rotational mode frequency.(about X axis of model) I am using ANSYS apdl for modelling and trying to incorporate combin 14 spring element I have done following steps: 1. Created 4 nodes at the same location (zero spring length) on the bar. 2. Defined combin 14 as following: /prep7 ET,3,14 keyopt,3,1,0 keyopt,3,2,4 !1 d torsional spring allowing rotx r,1,100,, real,1 Type,3 E,981,nno+1 !Defined element in the bar node and corresponding new node E,982,nno+2 !981-984 are the node numbers on the bar where I connected torsional spring E,983,nno+3 E,984,nno+4 3. defined all dof = 0 on the new nodes. But when I tried to run the modal analysis, the effects of the springs are not shown. The structure shows the frequencies and mode shapes as if it is free-free condition. I also tried 1.3D spring (KEYOPT3=1) with some finite length. 2. Creating nodes and coupling it with bar nodes Still it has the same problem. What can be the problem? does nodal coordinate system affect the result?
It is possible to introduce torsional spring and any other kinds of stiff nesses by using “user matrix (stiff matrix)” element. This element has two nodes and 6 degrees of freedom in each node (12 degrees of freedom at all). Therefore its stiffness matrix is (12*12) with 78 independent parameters due to the symmetry of matrix. To model a torsional spring, the mentioned element should be selected and a spring constant should be introduced by replacing its value as a 51th and 78th coefficients by performing “Real constant” command.
Thank you M.Reza Shirazizadeh for your input.
I tried using matrix 27 element, still I am confused about the real constant definition. Can you tell me how to define real constant for this in more detail?
Also, what about the boundary conditions? I need one end of springs to be completely fixed. Do I just need to define D command on those nodes?
You need to fix the end of springs by using DK command
DK, NODE ,ALL,0
When modal results shows mode shapes as if it is free-free condition, the most common mistake are: a) unassigned bc; and b) lack of continuity on the structure. The first of them means that you have a structure like on a bfd, the structure must be restricted in at least one direction of one node, the second mistake, It occurs because there is a lack of node-node connectivity, causing in disconnected remaining structural parts.
I recommend you at first instance, and take it like a safety first filter to make you sure that your structural system is well proposed, performing a static analysis. If the results match your predictions, it is a good indicator to continue, but if is not the case, the results will help you locate that lacked detail to consider. There will be no problems with the loading conditions of the static analysis, if you think that will affect your results in modal analysis, remind that modal analysis is result of solve the equation of motion, thas it is a function of m, c and k.
Animas,
Thank you for your reply. But the details of combin14 element says that, it is defined between two nodes, so I tried DK command but that is showing me an error.
In my model, since I needed some nodal connectivity, I have coupled the nodes so I am getting desired mode shaped but at wrong/undesired frequencies. And those frequencies I need to change by changing the stiffness by introducing a spring element.
The way I am modelling combin14 gives me desired results for 2D areas but not for the 3D objects.
I am using 3D spring only.
What can be the reason?
How to add torsional spring in ansys apdl? If anyone knows kindly reply
- Badges
- Science topic