Simulation of nonlinear coupled physics problems with neural network aided model order reduction in RF MEMS
Bibliographic record
Abstract
The interdisciplinary MEMS field has become increasingly complex and applicable to a wide range of industries. The devices have massive potential, especially in the lucrative field of RF microelectronics, promising better performance, wide applicability and ease of integration. These devices are designed to meet strict standards, and though the design techniques have become more sophisticated, there has not been an overarching method of inclusion for such models into large scale systems. A systematic process which harnesses the original simulation accuracy and generates a macromodel can help develop more adaptable models that bridge the current gap between device and system level simulations.This thesis develops a nonlinear coupled physics solver for electromechanical RF type MEMS devices. A well-established staggered solution is implemented and the use of a modified local coulomb virtual work (LCVW) method is used for the coupling. A new remeshing strategy, mainly used in aerospace is used to deal with the complex deformations of the domain. Structural nonlinearity, large inputs, and dynamic concerns are implemented and handled effectively in the solver. A reduction approach for such staggered coupled physics problems is established ensuring applicability at the system level. The developed reduction technique is based off classical projections methods that are aided with neural networks. Nonlinear projection methods based on SVD are used and later adapted for discrete parametric systems (the strategy, numerical considerations and adaptive mesh are also defined), while neural networks help in the coupling and are used to speed up and improve the computational efficiency. Classical examples such as RF microswitches and a host of resonator structures are tested and proved to match both the results from commercial solvers as well as the experimentally determined values. A reduced model is generated and shown to match the results of the original system. In addition, a newly developed and manufactured micromirror example is implemented and tested in the solver.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".