"2"^^ . . "Selection strategy for fixing nodes in FETI-DP method"@en . "21110" . . "2"^^ . "Kruis, Jaroslav" . . "P(GD103/09/H078)" . "12"^^ . "[403C454642FC]" . "HU - Ma\u010Farsko" . "FETI-DP method; fixing nodes; domain decomposition method; finite element method"@en . . . . "Bro\u017E, Jaroslav" . "Pollack Periodica, An International Journal for Engineering and Information Sciences" . "RIV/68407700:21110/10:00169180" . "2" . "This paper deals with selection strategy of fixing nodes in the FETI-DP method. The FETI-DP method is one of non-overlapping domain decomposition methods. The method was published by Farhat and co-workers in 2001. Selection of the fixing unknowns in the FETI-DP, TFETI or BDDC domain decomposition method has strong influence on the method behavior. The selection itself is not straightforward in the case of irregular domains and subdomains. Three-step algorithm for the selection of the fixing unknowns is presented and some numerical examples are shown. The algorithm is based on nodal multiplicity, which is the number of subdomains sharing the node. If the three-step algorithm selects unsatisfactory number of fixing nodes, additional geometrical conditions are applied. These conditions lead to selection of additional fixing nodes. Moreover, higher number of the fixing nodes usually results in better convergence of the coarse problems."@en . "Selection strategy for fixing nodes in FETI-DP method"@en . . . "1788-1994" . . "286796" . . "Selection strategy for fixing nodes in FETI-DP method" . "Selection strategy for fixing nodes in FETI-DP method" . "This paper deals with selection strategy of fixing nodes in the FETI-DP method. The FETI-DP method is one of non-overlapping domain decomposition methods. The method was published by Farhat and co-workers in 2001. Selection of the fixing unknowns in the FETI-DP, TFETI or BDDC domain decomposition method has strong influence on the method behavior. The selection itself is not straightforward in the case of irregular domains and subdomains. Three-step algorithm for the selection of the fixing unknowns is presented and some numerical examples are shown. The algorithm is based on nodal multiplicity, which is the number of subdomains sharing the node. If the three-step algorithm selects unsatisfactory number of fixing nodes, additional geometrical conditions are applied. These conditions lead to selection of additional fixing nodes. Moreover, higher number of the fixing nodes usually results in better convergence of the coarse problems." . . . "RIV/68407700:21110/10:00169180!RIV11-GA0-21110___" . . . . . . "5" .