Internship in MEMS Packaging Simulation - Credibility and Numerical Uncertainty Quantification

Descripción del puesto:

Job Description

The packaging of MEMS sensors, important components in automotive safety systems and consumer electronics, plays a significant role in ensuring mechanical stability, performance, and long-term reliability. The heterogeneous material composition of these packages leads to thermal expansion mismatches, which can cause high stresses at material interfaces under thermal loading. Although thermo-mechanical finite element simulations are an essential method for understanding stress formation and warpage, the results are influenced by numerical modeling decisions. Factors such as mesh density, element formulation, and solver algorithms can affect the predicted stresses and deformations, introducing model-form and numerical uncertainties that are often not systematically analyzed.
* Your work during this internship will be based on an existing FEM model of a MEMS packaging structure, which you will use for a structured investigation.
* You will start by preparing and refining the existing thermo-mechanical FEM MEMS model.
* Using a structured virtual Design of Experiments (vDOE), you will then systematically vary the numerical modeling parameters.
* Your analysis will require you to evaluate simulation results, focusing on reliability-relevant quantities like package warpage, interface stresses, and strain distributions.
* A significant part of your project will be to construct surrogate models (such as Gaussian Process, Kriging, or Random Forest) to approximate the simulation responses.
* To identify the most influential factors, you will conduct a sensitivity analysis on the dominant numerical parameters.
* Finally, you will quantify the propagation of uncertainty to assess the simulation's overall robustness and credibility

Requerimientos del candidato/a:

Qualifications

* Education: Master studies in the field of Mechanical Engineering, Computational Engineering, Artificial Intelligence, Data Science or comparable
* Experience and Knowledge: strong knowledge and practical experience with Ansys Mechanical; solid knowledge of PyMAPDL; proficient programming skills in Python; firm understanding of statistics; sound knowledge of machine learning methods; fundamental understanding of simulation credibility and uncertainty quantification concepts
* Personality and Working Practice: you are a committed and communicative team player with an independent and structured working style
* Work Routine: your on-site presence is required
* Languages: very good in English and beginner in German