Numerical Methods for Calculating Component Modes for Geometric Mistuning Reduced-order Models

Author(s):  
Joseph Beck ◽  
Jeffrey Brown ◽  
Alex Kaszynski ◽  
Daniel Gillaugh

Abstract Geometric mistuning models formulated from a component mode synthesis methods often require the calculation of component modes, particularly constraint and fixed interface normal modes, during substructuring. For Integrally Bladed Rotors, these calculations are required for each sector. This paper proposes methods that reuse information garnered from solving the constraint modes of a single sector on the remaining sectors to reduce memory requirements and solution times. A mesh metamorphosis tool is used to ensure finite element models match geometry obtained from a 3D optical scanner. This tool also produces a common mesh pattern from sector-to-sector. This is exploited to produce common permutation matrices and symbolic factorizations of sector stiffness matrices that are proposed for reuse in solving subsequent constraint modes. Furthermore, a drop tolerance is introduced to remove small values during constraint mode calculation to reduce memory requirements. It is proposed to reuse this dropping pattern produced from a single sector on the remaining sectors. Approaches are then extended to a parallel processing scheme to propose effective matrix partitioning methods. Finally, information gathered during the constraint mode calculations are reused during the solution of the fixed interface normal modes to improve solution time. Results show reusing permutation matrices and symbolic factorizations from sector-to-sector improves solution time and introduces no error. Using a drop tolerance is shown to reduce storage requirements of a constraint mode matrix, while reusing the dropping pattern introduces minimal error. Similarly, reusing constraint mode information in calculating normal modes greatly improves the performance.

2021 ◽  
Author(s):  
Joseph A. Beck ◽  
Jeffrey M. Brown ◽  
Alex A. Kaszynski ◽  
Daniel L. Gillaugh

Abstract Geometric mistuning models formulated from a component mode synthesis methods often require the calculation of component modes, particularly constraint and fixed interface normal modes, during substructuring. For Integrally Bladed Rotors, these calculations are required for each sector. This paper proposes methods that reuse information garnered from solving the constraint modes of a single sector on the remaining sectors to reduce memory requirements and solution times. A mesh metamorphosis tool is used to ensure finite element models match geometry obtained from a 3D optical scanner. This tool also produces a common mesh pattern from sector-to-sector. This is exploited to produce common permutation matrices and symbolic factorizations of sector stiffness matrices that are proposed for reuse in solving subsequent constraint modes. Furthermore, a drop tolerance is introduced to remove small values during constraint mode calculation to reduce memory requirements. It is proposed to reuse this dropping pattern produced from a single sector on the remaining sectors. Approaches are then extended to a parallel processing scheme to propose effective matrix partitioning methods. Finally, information gathered during the constraint mode calculations are reused during the solution of the fixed interface normal modes to improve solution time. Results show reusing permutation matrices and symbolic factorizations from sector-to-sector improves solution time and introduces no error. Using a drop tolerance is shown to reduce storage requirements of a constraint mode matrix. Additionally, it is shown that reusing the same dropping pattern introduces minimal error without degradation in solution times. Similarly, reusing the information from constraint modes for calculating fixed interface normal modes greatly improves the performance in a shift-and-invert technique for solving eigenvalue problems.


Author(s):  
Alain Batailly ◽  
Mathias Legrand ◽  
Patrice Cartraud ◽  
Christophe Pierre ◽  
Jean-Pierre Lombard

The study of interactions through direct contact between bladetips and outer casings in modern turbomachines may be very time-consuming when the classical finite element method is used. The construction of reduced-order models using component mode synthesis (CMS) methods generally allows for dramatic increase in computational efficiency and may be used in order to improve the knowledge over these interaction phenomena. Among the available approaches, both a fixed-interface method and a free-interface method are considered here in an original manner to reduce the size of a realistic two-dimensional model. The equations of motion are solved using an explicit time integration scheme with the Lagrange multiplier method where friction is accounted for. This method offers energy momentum conserving which is a critical point to ensure the convergence of the algorithm. Moreover, it is shown that even in a non-linear framework the reduced-order models converge to the finite element solution as the number of modes included in the models increases. Considering the fixed-interface method of Craig-Bampton (CB) and the free-interface method of Craig-Chang-Martinez (CCM), it is shown that a method with fast displacement convergence may be less efficient in terms of motion convergence.


Author(s):  
Duc-Minh Tran

A new fixed interface component mode synthesis method using partial interface modes is presented. Partial interface modes are the structure normal modes which result from the static condensation of the structure to the interface between the substructures and which are clamped at a part of this interface. This method is the generalization of the classical component mode synthesis method which keeps all the interface physical displacements in the assembled reduced system and the method using interface modes which eliminates all of them. It allows one to reduce the number of the interface generalized coordinates and at the same time to keep some of the physical displacements at the interface. This latter capability is very useful to build reduced order models in which the presence of physical displacements are essential, for example in order to impose prescribed motions or to take into account local non-linearities. The new method is applied to a bladed disk in both tuned and mistuned cases.


Author(s):  
Polarit Apiwattanalunggarn ◽  
Steven W. Shaw ◽  
Christophe Pierre

This paper describes a methodology for developing reduced-order dynamic models of nonlinear structural systems that are composed of an assembly of component structures. The approach is a nonlinear extension of the fixed-interface component mode synthesis technique developed for linear structures by Hurty and modified by Craig and Bampton. Specifically, the case of nonlinear substructures is handled by using fixed-interface nonlinear normal modes. These normal modes are constructed for the various substructures using an invariant manifold approach, and are then coupled through the traditional linear constraint modes (i.e., the static deformation shapes produced by unit interface motions). A simple system is used to demonstrate the proof of concept and show the effectiveness of the proposed procedure. Simulations are performed to show that the reduced-order model obtained from the proposed procedure outperforms the reduced-order model obtained from the classical fixed-interface linear component mode synthesis approach. Moreover, the proposed method is readily applicable to large-scale nonlinear structural systems.


Author(s):  
Mikel Balmaseda ◽  
G. Jacquet-Richardet ◽  
A. Placzek ◽  
D.-M. Tran

Abstract In the present work reduced order models (ROM) that are independent from the full order finite element models (FOM) considering geometrical non linearities are developed and applied to the dynamic study of a fan. The structure is considered to present nonlinear vibrations around the pre-stressed equilibrium induced by rotation enhancing the classical linearised approach. The reduced nonlinear forces are represented by a polynomial expansion obtained by the Stiffness Evaluation Procedure (STEP) and then corrected by means of a Proper Orthogonal Decomposition (POD) that filters the full order nonlinear forces (StepC ROM). The Linear Normal Modes (LNM) and Craig-Bampton (C-B) type reduced basis are considered here. The latter are parametrised with respect to the rotating velocity. The periodic solutions obtained with the StepC ROM are in good agreement with the solutions of the FOM and are more accurate than the linearised ROM solutions and the STEP ROM. The proposed StepC ROM provides the best compromise between accuracy and time consumption of the ROM.


Author(s):  
Robert J. Kuether ◽  
Peter B. Coffin ◽  
Adam R. Brink

Structural dynamics models with localized nonlinearities can be reduced using Hurty/Craig-Bampton component mode synthesis methods. The interior degrees-of-freedom of the linear subcomponents are reduced with a set of dynamic fixed-interface modes while the static constraint modes preserve the physical coordinates at which the nonlinear restoring forces are applied. For finite element models with a highly refined mesh at the boundary, a secondary modal analysis can be performed to reduce the interface down to a truncated set of local-level characteristic constraint modes. In this research, the cost savings and accuracy of the interface reduction technique are evaluated on a simple example problem involving two elastic blocks coming into contact.


Author(s):  
José L. Escalona ◽  
Juana Mayo ◽  
Jaime Domínguez

Abstract In this paper, the floating frame of reference approach is applied to the dynamics of the impact of flexible bodies, while component mode synthesis is used to describe deformation. The influence of the reference conditions, that indicate the type of attachment between the body fixed frame of reference and the flexible bodies, is investigated. Rigid and free attachments allow the use of fixed interface and free interface normal modes, respectively. A finite number of fixed interface modes does not fulfil the natural boundary condition at the attachment point. Free interface normal modes cannot describe the compressive forces at the contact surface. However, it is shown that both set of modes are able to describe the impact-induced elastic waves. In the evaluation of the kinematic coefficient of restitution, these two approaches differ significantly. When free attachment is considered, the derivatives of the reference co-ordinates coincide with the equivalent rigid body velocities of the flexible bodies, remaining constant after the impact. However, if the body frame of reference is rigidly attached, the equivalent rigid body velocities of the flexible body have to be evaluated as a linear combination of the derivative of reference and elastic co-ordinates. The axial impact of a rigid body on a flexible rod and the transverse impact of a flexible pendulum with a fixed stop are simulated to illustrate these facts. Hertzian contact forces are assumed to occur during impact.


Author(s):  
Ilaria Palomba ◽  
Dario Richiedei ◽  
Alberto Trevisani

Resonant system design and optimization is usually supported by finite element models. Large dimensional models are often needed to achieve the desired accuracy in the representation of the vibrational behaviour at the frequency of interest. Unfortunately, large dimensional models are frequently too cumbersome to be actually useful, mainly at the optimization stage. On the other hand, the choice of the most appropriate reduction strategy and dimension for a reduced-order model is generally left to designers’ experience. Having recognized the effectiveness and spreading of the Craig Bampton reduction technique, the aim of this paper is to propose a rigorous ranking method, called Interior Mode Ranking (IMR), for the selection of the interior normal modes of the full order model to be inherited by the reduced order one. The method is aimed at finding the set of interior modes of minimum dimensions which allows achieving a desired level of accuracy of the reduced order model at a frequency of interest. The method is here applied to a resonator widely employed in industry: an ultrasonic welding bar horn, which is usually designed to operate excited in resonance. The results achieved through the application of the IMR method are compared with those yielded by other ranking techniques available in literature in order to prove its effectiveness.


1989 ◽  
Vol 22 (1) ◽  
pp. 109-148 ◽  
Author(s):  
Madanpal S. Verma ◽  
James C. Romig

Sign in / Sign up

Export Citation Format

Share Document