Development of a Coupling Metric to Assess the Shakedown Limits for a Contact Interface
Abstract not provided.
Abstract not provided.
Abstract not provided.
Abstract not provided.
The Third International Workshop on Jointed Structures was held from August 16th to 17th, 2012, in Chicago Illinois, following the ASME 2012 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference. Thirty two researchers from both the United States and international locations convened to discuss the recent progress of mechanical joints related research and associated efforts in addition to developing a roadmap for the challenges to be addressed over the next five to ten years. These proceedings from the workshop include the minutes of the discussions and follow up from the 2009 workshop [1], presentations, and outcomes of the workshop. Specifically, twelve challenges were formulated from the discussions at the workshop, which focus on developing a better understanding of uncertainty and variability in jointed structures, incorporating high fidelity models of joints in simulations that are tractable/efficient, motivating a new generation of researchers and funding agents as to the importance of joint mechanics research, and developing new insights into the physical phenomena that give rise to energy dissipation in jointed structures. The ultimate goal of these research efforts is to develop a predictive model of joint mechanics.
Tribology International
Abstract not provided.
Abstract not provided.
This report presents an efficient and accurate method for integrating a system of ordinary differential equations, particularly those arising from a spatial discretization of partially differential equations. The algorithm developed, termed the IMEX a algorithm, belongs to a class of algorithms known as implicit-explicit (IMEX) methods. The explicit step is based on a fifth order Runge-Kutta explicit step known as the Dormand-Prince algorithm, which adaptively modifies the time step by calculating the error relative to a fourth order estimation. The implicit step, which follows the explicit step, is based on a backward Euler method, a special case of the generalized trapezoidal method. Reasons for choosing both of these methods, along with the algorithm development are presented. In applications that have less stringent accuracy requirements, several other methods are available through the IMEX a toolbox, each of which simplify the fifth order Dormand-Prince explicit step: the third order Bogacki-Shampine method, the second order Midpoint method, and the first order Euler method. The performance of the algorithm is evaluated on to examples. First, a two pawl system with contact is modeled. Results predicted by the IMEX a algorithm are compared to those predicted by six widely used integration schemes. The IMEX a algorithm is demonstrated to be significantly faster (by up to an order of magnitude) and at least as accurate as all of the other methods considered. A second example, an acoustic standing wave, is presented in order to assess the accuracy of the IMEX a algorithm. Finally, sample code is given in order to demonstrate the implementation of the proposed algorithm.
Abstract not provided.
Abstract not provided.
Abstract not provided.
Abstract not provided.
Abstract not provided.
Abstract not provided.
Abstract not provided.
Abstract not provided.
Abstract not provided.
Abstract not provided.
Proposed for publication in Journal of Sound and Vibration.
Abstract not provided.
Abstract not provided.
Abstract not provided.
SIAM J., IJNME, CMAME or JCP
Abstract not provided.
Vibration Institute Annual Training Conference 2012
Impact is a phenomenon that is ubiquitous in mechanical design; however, the modeling of impacts in complex systems is often a simplified, imprecise process. In high fidelity finite element simulations, the number of elements required to accurately model the constitutive properties of an impact event is impractical. Consequently, rigid body dynamics with approximate representations of the impact dynamics are commonly used. These approximations can include a constant coefficient of restitution, penalty stiffness, or single degree of freedom constitutive model for the impact dynamics that is specific to the type of materials involved (elastic-plastic, viscoelastic, etc.). In order to understand the effect of the impact model on the system's dynamics, simulations investigate single degree of freedom and two degrees of freedom systems with rigid stops limiting the amplitude of vibration. Five contact models are considered: a coefficient of restitution, penalty stiffness, two similar elastic-plastic constitutive models, and a dissimilar elastic-plastic constitutive model. Frequency sweeps show that simplified contact models can lead to incorrect assessments of the system's dynamics and stability, which can significantly affect the prediction of wear and damage in the system.
Proceedings of the ASME Design Engineering Technical Conference
Solutions for analytical models of systems with nonlinear constraints have focused on exact methods for satisfying the constraint conditions. Exact methods often require that the constraint can be expressed in a piecewise-linear manner, and result in a series of mapping equations from one linear regime of the constraint to the next. Due to the complexity of these methods, exact methods are often limited to analyzing a small number of constraints for practical reasons. This paper proposes a new method for analyzing continuous systems with arbitrary nonlinear constraints by approximately satisfying the constraint conditions. Instead of dividing the constraints into multiple linear regimes, a discontinuous basis function is used to supplement the system's linear basis functions. As a result, precise contact times are not needed, enabling this method to be more computationally efficient than exact methods. While the discontinuous basis functions are continuous in displacement, their derivatives contain discontinuities that allow for the nonlinear forces to be accounted for with the assumption that the nonlinear constraints are able to be modeled in a discrete manner. Since each nonlinear constraint requires only one associated discontinuous basis function, this method is easily expanded to handle large numbers of constraints. In order to illustrate the application of this method, an example with a pinned-pinned beam is presented. © 2011 by ASME.
Abstract not provided.
Proceedings of the Royal Society A-Mathematical Physical and Engineering Sciences
Abstract not provided.