-
-
Notifications
You must be signed in to change notification settings - Fork 36
/
Copy pathDISTRIBUTING_COUPLING.html
41 lines (40 loc) · 3.57 KB
/
DISTRIBUTING_COUPLING.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
<html><head><link rel="stylesheet" type="text/css" href="style.css"/></head><body> <H2> <BR> *DISTRIBUTING COUPLING </H2> <P> Keyword type: model definition <P> This option is used to apply translational loading (force or displacement) on a set of nodes in a global sense (for rotations and/or moments the reader is referred to the mean rotation MPC, Section 8.7.1). There is one required parameter: ELSET. With the parameter ELSET an element set is referred to, which should contain exacty one element of type DCOUP3D. This type of element contains only one node, which is taken as the reference node of the distributing coupling. This node should not be used elsewhere in the model. In particular, it should not belong to any element. The coordinates of this node are immaterial. The distributing coupling forces or the distributing coupling displacements should be applied to the reference node with a *CLOAD card or a *BOUNDARY card, respectively. <P> Underneath the keyword card the user can enter the nodes on which the load is to be distributed, together with a weight. Internally, for each coordinate direction a multiple point constraint is generated between these nodes with the weights as coefficients. The last term in the equation is the reference node with as coefficient the negative of the sum of all weights. The more nodes are contained in the distributing coupling condition the longer the equation. This leads to a large, fully populated submatrix in the system of equations leading to long solution times. Therefore, it is recommended not to include more than maybe 50 nodes in a distributing coupling condition. <P> The first node underneath the keyword card is taken as dependent node n the MPC. Therefore, this node should not be repeated in any other MPC or at the first location in any other distributing coupling definition. It can be used as independent node in another distributing coupling (all but the first position), though, although certain limitations exist due to the mechanism by which the MPC's are substituted into each other. Basically, not all dependent nodes in distributing couplings should be used as independent nodes as well. For example: <P> <PRE>
*DISTRIBUTING COUPLING,ELSET=E1
LOAD,1.
*DISTRIBUTING COUPLING,ELSET=E2
LOAD2,1.
*NSET,NSET=LOAD
5,6,7,8,22,25,28,31,100
*NSET,NSET=LOAD2
8,28,100,31
</PRE> <P> will work while <P> <PRE>
*DISTRIBUTING COUPLING,ELSET=E1
LOAD,1.
*DISTRIBUTING COUPLING,ELSET=E2
LOAD2,1.
*NSET,NSET=LOAD
5,6,7,8,22,25,28,31,100
*NSET,NSET=LOAD2
8,28,100,31,5
</PRE> <P> will not work because the dependent nodes 5 and 8 are used as independent nodes as well in EACH of the distributing coupling definitions. An error message will result in the form: <P> <PRE>
*ERROR in cascade: zero coefficient on the
dependent side of an equation
dependent node: 5
</PRE> <P> <P><P> <BR> <P> First line: <UL> <LI>*DISTRIBUTING COUPLING </LI> <LI>Enter the ELSET parameter and its value </LI> </UL> <P> Following line: <UL> <LI>Node number or node set </LI> <LI>Weight </LI> </UL> <P> Repeat this line if needed. <P> <PRE>
Example:
*DISTRIBUTING COUPLING,ELSET=E1
3,1.
100,1.
51,1.
428,1.
*ELSET,ELSET=E1
823
*ELEMENT,TYPE=DCOUP3D
823,4000
</PRE> <P> defines a distributing coupling between the nodes 3, 100, 51 and 428, each with weight 1. The reference node is node 4000. A point force of 10 in direction 1 can be applied to this distributing coupling by the cards: <P> <PRE>
*CLOAD
4000,1,10.
</PRE> <P> while a displacement of 0.5 is obtained with <P> <PRE>
*BOUNDARY
4000,1,1,0.5
</PRE> <P> <P><P> <BR> Example files: distcoup. <P> </body></html>