Cyclic Boundary Condition with MPC/CYCLSYM

[MatthiasP]

Hello,
I am new here and registered because I habe an Abaqus-Problem I cannot solve and for which I don't really find any useful information on the internet. Hopefully someone of you uses the command that I can't get to work properly.

I have a 12° ring-segment of a compressor casing and I wanted to define a cyclic boundary condition. The solver used is Abaqus/Standard. Research in the manual and on the internet directed me to the *MPC CYCLSYM - command.
I defined a node set for each of the two surfaces: The node set at 0° is called "master", the one at 12° is called slave. Each have 118 Nodes. Then I used the commands:

*NODE
99998, 0.0, 0.0, 1.0
99999, 0.0, 0.0, 0.0
*MPC
CYCLSYM, master, slave, 99998, 99999

There as no problem for the solver and he calculated without errors, however the results were not physically useful. What the solver did was: arrange the nodes with respect to their names, then use CYCLSYM between the nodesets.
The original node sets were:

Master:
Node
1299
15227
3421
11323
1300
...
Slave:
Node
1252
15078
3325
16283
1251
...

He should now enforce cyclic symmetry between nodes 1299 and 1252, 15227 and 15078 and so on.

What he did was:
Master: Slave:
Node Node
1068 = 1176
1069 = 1177
1070 = 1178
1071 = 1179
1072 = 1180
...
1299 = 1250
1300 = 1251
...
15227 = 16251
...

So he arranged the node sets and then enforced cyclic symmetry between the #1, #2, #3, ... elements of the nodesets.

It is frustrating to see the command work, but not in the correct way. Can I somehow force the solver to use the node set as I have defined it, without changing the order of the nodes? Is it possible to rename nodes, so that the node set will be in order automatically? Is there maybe another way?

I attached the Input-file:

http://files.engineering.com/getfil...4-8e0943733bed&file=Aux3h_Direct_Symmetry.inp

Node sets and the command are in lines 29874 to 29906

I would greatly appreciate your help,
thanks in advance
Matthias