Read more
This book explores how physical laws and data can be combined to model complex systems in fluid mechanics, heat transfer, multiphase flows and turbomachinery. First principles provide a strong foundation, yet they are often not sufficient on their own when dealing with real systems. The text builds from physics toward data driven approaches and presents the full spectrum of hybrid modeling. It moves from white box models rooted in conservation laws to grey box and black box models shaped by empirical data and machine learning. The fundamentals of physical modeling are introduced through dimensional analysis, governing equations and simplified flow regimes. From a data centered perspective the book presents methods for uncertainty quantification, statistical inference and machine learning aimed at model calibration and prediction. Examples range from canonical flows studied in controlled settings to complex industrial systems operating under real conditions. These cases illustrate how hybrid models can combine interpretability with predictive strength. The discussion highlights the importance of a clear modeling purpose, an appropriate model structure and the role of context in giving meaning to data. Context is described in physical, operational and diagnostic terms and is presented as a key ingredient for constructing useful models.
List of contents
Hybrid Modeling in Engineering.- 2 Dimensional Analysis.- Governing Equations.- Governing Equations.- 5 Basic Flow Cases.- 6 Probability Theory.- Statistical Methods for Data Analysis.- 8 Machine Learning and Prediction.- Modeling Isolated Physical Processes.- Industrial System Modeling.- Conclusion.
About the author
Hassan studied aerospace engineering and earned his PhD in the field of aerospace thermodynamics from the University of Stuttgart, Germany, where he researched multiphase compressor flows using a combination of first-principles modeling and data-driven methods. His engineering experience includes roles at BMW Motorsport, focusing on aerodynamic design tools, and at Siemens in gas turbine development, where he contributed to the thermal design and methodology development of hot gas path turbine components.
Over the past decade, Hassan has worked with industrial data across sectors such as manufacturing, chemicals, and oil and gas. His work focuses on industrial data science, integrating engineering knowledge with data analysis to support practical decision-making and operational insights at scale.
Outside of work, Hassan enjoys spending time with his family and scuba diving, with a particular interest for sharks.
Summary
This book explores how physical laws and data can be combined to model complex systems in fluid mechanics, heat transfer, multiphase flows and turbomachinery. First principles provide a strong foundation, yet they are often not sufficient on their own when dealing with real systems.
The text builds from physics toward data driven approaches and presents the full spectrum of hybrid modeling. It moves from white box models rooted in conservation laws to grey box and black box models shaped by empirical data and machine learning. The fundamentals of physical modeling are introduced through dimensional analysis, governing equations and simplified flow regimes. From a data centered perspective the book presents methods for uncertainty quantification, statistical inference and machine learning aimed at model calibration and prediction.
Examples range from canonical flows studied in controlled settings to complex industrial systems operating under real conditions. These cases illustrate how hybrid models can combine interpretability with predictive strength. The discussion highlights the importance of a clear modeling purpose, an appropriate model structure and the role of context in giving meaning to data. Context is described in physical, operational and diagnostic terms and is presented as a key ingredient for constructing useful models.
