This presentation was made at CAASE18, The Conference on Advancing Analysis & Simulation in Engineering. CAASE18 brought together the leading visionaries, developers, and practitioners of CAE-related technologies in an open forum, to share experiences, discuss relevant trends, discover common themes, and explore future issues.

Resource Abstract

When linear perturbation methods are used to determine the natural frequency of a preloaded assembly, it must be recognized that numerical methods inherent to contact formulations can impact the stiffness matrix that is passed to the modal solver. Contact settings that are ideal for convergence of stress analyses may not always lead to accurate prediction of natural frequency. For example, default software settings can lead to contact stiffness values greater than that of the base material, typically causing the resulting natural frequency to be overestimated. Prior published research will be referenced to illustrate that tuning the contact stiffness is an accepted approach that can be used to improve accuracy. In addition to directly managing the contact stiffness to values which are physically appropriate, tuning can also represent physical effects of surface finish, microslip, and so forth that cannot be practically modeled. Capturing such effects at contact interfaces is one reason why it is important to include test fixtures and a portion of the ground reference instead of assuming a fixed support. Users of tuned models must consider that modifications in proprietary algorithms may change the accuracy of their results as commercial codes evolve over time, or that re-tuning may be required when changing software. Likewise, factors such as element size or material properties may be inputs to proprietary formulations that limit scalability of tuned methods across different simulations. Examples of simulations on slender beam and frame structures with welded joints will be compared to published test results. Simulation work from an industrial control valve actuator with bolted joints will also be presented with corresponding lab test data. The work presented will illustrate the sensitivity of the results to how and where the boundary conditions are imposed as well as how selection of different contact settings and mesh sizes influences the predicted natural frequencies.