TY - BOOK

T1 - A Modeling Framework for Enhanced Reliability of Printed Circuit Boards and Power Electronics

AU - Gschwandl, Mario

N1 - embargoed until 02-07-2025

PY - 2020

Y1 - 2020

N2 - Microelectronics have become ever more powerful in the last decade. Hand-held devices turned from a nice-to-have feature to an inevitable must-have companion. Modern power electronics shape the way to highly efficient and environmentally friendly transportation and power generation. The two main constituents of the applied microelectronic system applications are printed circuit boards (PCBs) and and electronic devices, such as power packages. Both parts entail complex geometries, consist of multi-material set-ups and underlie multi-physics electro-thermo-mechanical loads. Therefore, the understanding of the employed materials is crucial and enables the precise determination of temperature dependent material properties. Additionally, for novel PCB designs traditional virtual modeling approaches become an immensely time-consuming task and are clearly contradictory to the current time-to-market goals in the industry. When pairing the PCB with power electronics high temperature loads need to be addressed, since power packages generate heat due to the active electrical loading. However, the heating occurs only in a small part of the entire device. Accordingly, it is necessary to describe the generated heat precisely, time-dependent and locally resolved, thus establishing the basis for the prediction of operational reliability. Due to the tight margins in the design of these components, manufacturing induced stresses need to be assessed. Especially polymers are prone to residual stresses induced during manufacturing. To enhance the operational reliability, the influence of the arising residual stresses during manufacturing must be examined prior to operational lifetime assessments. In order to cope with these challenges, a virtual framework relying on Finite Element Analysis (FEA) is presented in this thesis. The developed framework consists of three parts: (i) A highly efficient and accurate FEA modeling approach for PCBs; automatically compiling FEA models of scaleable complexity for any given design. (ii) An approach to predict residual stresses within predominately used duromeric encapsulation materials of power packages. Thereby, allowing to assess different manufacturing processes as well as comparing the performance of different power package designs. (iii) And finally, a multi-physics and multi-domain approach to spatially predict the generated heat of silicon semiconductors within power packages. Most importantly, the presented methods are based upon precise material characterization techniques. Especially for thermal properties of the polymeric constituents - thermosets - specialized and accurate measurement methods are presented. Merging the single parts of the framework together, combined with the presented material characterization strategy, an enhanced, timeefficient and accurate reliability assessment of microelectronic systems can be carried out.

AB - Microelectronics have become ever more powerful in the last decade. Hand-held devices turned from a nice-to-have feature to an inevitable must-have companion. Modern power electronics shape the way to highly efficient and environmentally friendly transportation and power generation. The two main constituents of the applied microelectronic system applications are printed circuit boards (PCBs) and and electronic devices, such as power packages. Both parts entail complex geometries, consist of multi-material set-ups and underlie multi-physics electro-thermo-mechanical loads. Therefore, the understanding of the employed materials is crucial and enables the precise determination of temperature dependent material properties. Additionally, for novel PCB designs traditional virtual modeling approaches become an immensely time-consuming task and are clearly contradictory to the current time-to-market goals in the industry. When pairing the PCB with power electronics high temperature loads need to be addressed, since power packages generate heat due to the active electrical loading. However, the heating occurs only in a small part of the entire device. Accordingly, it is necessary to describe the generated heat precisely, time-dependent and locally resolved, thus establishing the basis for the prediction of operational reliability. Due to the tight margins in the design of these components, manufacturing induced stresses need to be assessed. Especially polymers are prone to residual stresses induced during manufacturing. To enhance the operational reliability, the influence of the arising residual stresses during manufacturing must be examined prior to operational lifetime assessments. In order to cope with these challenges, a virtual framework relying on Finite Element Analysis (FEA) is presented in this thesis. The developed framework consists of three parts: (i) A highly efficient and accurate FEA modeling approach for PCBs; automatically compiling FEA models of scaleable complexity for any given design. (ii) An approach to predict residual stresses within predominately used duromeric encapsulation materials of power packages. Thereby, allowing to assess different manufacturing processes as well as comparing the performance of different power package designs. (iii) And finally, a multi-physics and multi-domain approach to spatially predict the generated heat of silicon semiconductors within power packages. Most importantly, the presented methods are based upon precise material characterization techniques. Especially for thermal properties of the polymeric constituents - thermosets - specialized and accurate measurement methods are presented. Merging the single parts of the framework together, combined with the presented material characterization strategy, an enhanced, timeefficient and accurate reliability assessment of microelectronic systems can be carried out.

KW - Printed Circuit Boards

KW - Power Electronics

KW - Polymers for Microelectronics

KW - FEA-based Reliability

KW - Leiterplatten

KW - Leistungselektronik

KW - Kunststoffe in der Mikroelektronik

KW - FEM-basierte Lebensdauerberechnung

M3 - Doctoral Thesis

ER -