This paper proposes an analytical method to improve the accuracy of the dynamic response of classically damped linear systems, as given by a standard truncated modal analysis. Upon computing the first m undamped modes of a n-degree-of-freedom system, two sets of equations in the Rn nodal space are built, which are uncoupled and govern the contribution to the response of the m computed modes and the remaining (n-m) unknown modes, respectively. The first set is solved in the Rm modal space by using the m available modes; the second set is solved in a reduced R(n-m) nodal space, without computing additional modes. Specifically, it is shown that the particular solution of the second set of equations may be obtained by a series expansion involving repetitive time derivatives of the first-order static solution. The convergence conditions of such a series are discussed and proved on a rigorous basis. Numerical applications are also presented to demonstrate the effectiveness of the proposed method
A correction method for dynamic analysis of linear systems / DI PAOLA, M; Failla, Giuseppe. - In: COMPUTERS & STRUCTURES. - ISSN 0045-7949. - 82:(2004), pp. 1217-1226. [10.1016/j.compstruc.2004.03.001]
A correction method for dynamic analysis of linear systems
FAILLA, Giuseppe
2004-01-01
Abstract
This paper proposes an analytical method to improve the accuracy of the dynamic response of classically damped linear systems, as given by a standard truncated modal analysis. Upon computing the first m undamped modes of a n-degree-of-freedom system, two sets of equations in the Rn nodal space are built, which are uncoupled and govern the contribution to the response of the m computed modes and the remaining (n-m) unknown modes, respectively. The first set is solved in the Rm modal space by using the m available modes; the second set is solved in a reduced R(n-m) nodal space, without computing additional modes. Specifically, it is shown that the particular solution of the second set of equations may be obtained by a series expansion involving repetitive time derivatives of the first-order static solution. The convergence conditions of such a series are discussed and proved on a rigorous basis. Numerical applications are also presented to demonstrate the effectiveness of the proposed methodI documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.