The lectures of the course in question will be recorded and made available to the enrolled participants.

Dividing a complex problem in smaller sub-problems is a typical approach in modern engineering. In numerical substructuring, commonly used to perform model order reduction, components of a large structural model are analyzed separately, reduced according to their fundamental dynamics, then re-assembled.
The same approach can be very useful in experimental dynamics, where components are measured separately, and their experimental model assembled for simulation with experimental (or numerical) models of other components. Such substructure-based experimental (or hybrid) strategies can enable efficient modeling in cases where components can hardly be modeled numerically. Unfortunately, rather simple in theory, using experimentally characterized components is highly challenging because measured dynamics always include errors that in most practical cases make the assembled model useless.
Already in the 70’s, first attempts to use substructuring for experimentally measured parts were realized, but with little success. Later, the appearance of affordable and accurate sensors and acquisition systems made it possible to acquire precise transfer function of components at their interface. Over the last ten years, new formulations of the assembly process and advanced signal processing have made measurements suitable for building accurate component models. Also, the concept of blocked forces (known in acoustics for several decades) has recently been properly adapted to structural dynamics and enables a consistent characterization of excitation sources. Proper identification and coupling of substructure dynamics and transfer paths are also essential aspects for hardware-in-the-loop applications. In the field of earthquake engineering, a real-time co-simulation of an experimental and a numerical part is known as "Real-Time Hybrid Substructuring”. In a course of 2018, different aspects of substructuring, both as numerical and as experimental technique, were presented in a graduate course at CISM. In this new course, we will concentrate on experimental substructuring and transfer path analysis, starting from basics and explaining recent advances such as uncertainty quantification, extensions for joint identification (including non-linear joints), specific approaches for high frequency problems (Impulse-based Substructuring) or treatment of complex interfaces in challenging industrial applications. The course will also discuss industrial applications and some software for experimental substructuring.
The course is intended for graduate students either working in the field of experimental dynamics or interested by the field for their future research. The content is also very appealing for engineers in industry who are involved in challenging NVH problems where substructuring-based TPA can help in efficient analysis and troubleshooting. Since the course will start with the basics of substructuring and summarize important experimental techniques, it can be followed with a good general pre-knowledge in structural dynamics, but no specific expertise in substructuring is required.
The course will include the following parts: 1. Introduction, 2. Experimental identification of substructures, 3. Frequency-based Substructuring (FBS), 4. Modal-based Substructuring, 5. Transfer Path Analysis (TPA), 6. Dealing with real data and uncertainties, 7. Non-linear substructuring, 8. Special methods and outlook. It includes academic and industrial examples and computer (industrial code and an open-source code.