"The Role of Behavior Psychology in Affecting the Perception of Risk in SHM Decision-Making"
Structural Health Monitoring (SHM) systems provide actionable information that stakeholders use to make decisions about operations, maintenance, performance, etc. However, stakeholders have a risk profile, a perception of how they evaluate the outcome of the decisions they must make. Thus, the consequences of decisions do not just depend on posterior probabilities of information (such as a structural health state), but also on an individual’s (or even an organization’s) perception of decision consequences, and this presentation will consider a way to model this effect in the SHM decision-making process.
Monitoring and auscultation
From sensors to useful signals for concrete evaluation and monitoring
Before getting any observable from a measurement, signals have to be recorded after a wave or field has been interacting with concrete. The participants will get an overview of signal sampling and acquisition main features together with an insight of signal processing in the time and frequency domains. Notions about interferometry will be provided in relation to ultrasonic and fibre optic measurements.
Lecturer: Dr Odile Abraham (Université Gustave Eiffel, Ifsttar)
From signals to useful parameters: combination and data fusion
NDT parameters are not necessarily directly related to material or structural parameters. They have to be translated. An overview on calibration, combination of methods/data fusion to improve the results will be given.
Lecturer: Dr Ernst Niederleithinger (BAM)
Probability of Detection (PoD), Receiver Operating Characteristic (ROC)
How to connect parameters from NDT and monitoring to reliability and probability assessment? Studies on structural reliability or asset management are based more and more on risk assessment. To include NDT and monitoring data into this framework, their reliability must be known. The basic tools to do so, probability of detection and receiver operating characteristic curves will be explained.
Lecturer: Dr Ernst Niederleithinger (BAM)
Demonstrations and exercises on monitoring and auscultation
Group on Fibre Optics: How to size a crack ?
The advantages of fibre optics to monitor temperature and deformation as well as crack opening will be illustrated through various on-site and laboratory experiments. Knowledge and understanding of the transfer function required to recover quantitative information about crack opening will be provided. A lab demonstration will be carried out.
Group on CWI: Coda wave interferometry a new very sensitive monitoring tool for concrete.
The scattering and reverberation of ultrasonic waves in concrete structure, often viewed as noise in conventional NDT/SHM, is shown here to be of great value to get information on early age damages invisible otherwise. The participants will get to know the basic of CWI signal processing as well as a description about the present state of the art of experimental set-ups with embedded or external sensors. Environmental bias will be highlighted and ways to remove them addressed. The participants will process of the data from a controlled experiment on a beam under loading test.
Structural and action models
Assessment of extreme values of effects in structures
The assessment of extreme values of effects in structures will be based on extreme value theory and use of temporal series (coming either from monitoring or calculations). A short overview of statistics of standards will be made (review periods and return values), explanation of extreme value theory and some examples of application. The exercices will be made with R project, based on monitoring data.
Lecturer: Franziska Schmidt (Université Gustave Eiffel, Ifsttar)
Computational methods for transient diffusion problems in concrete structures
The lecture contains an overview on computational methods for differential equations, numerical solution strategies, explicit and implicit discretization, boundary conditions and implementation of physical processes that frequently occur in concrete structural materials. Emphasis will be on the finite difference method applied to transport processes, active in porous materials such as concrete, insulation materials, etc. The finite element method will be briefly introduced as well for the sake of comparison.
Typical problems that will be addressed are thermal, chloride ingress modelling, and even moisture diffusion. The course provides demonstrations and exercises on different computational methods for simple differential equations, numerical solution strategies, explicit and implicit discretization, finite difference method, boundary conditions and implementation.
The course is to train students who are beginners and/or have no modelling experience, on how to solve partial differential equations and to become familiar with numerical solution strategies for common physical problems in construction materials.
Key topics: steady state problems, transient problems, heat flow, chloride ingress – implementation in Octave free software.
Reliability, risk and decision analyses
Uncertainty and structural reliability assessment
The lecture of uncertainty and structural reliability assessment contains the fundamentals of uncertainty modelling in civil engineering and the methods of structural reliability and structural system reliability analysis. The uncertainty modelling provides the types of uncertainties and the approaches to model these in a structural reliability analysis. Methods for structural reliability analyses like Monte Carlo Simulation and FORM are introduced and their application in a structural system reliability analysis is demonstrated.
Lecturer: Prof. John Dalsgaard Sørensen (AAU)
Decision and structural information analyses
The lecture on decision and structural information analyses provides a summary on decision scenario development, decision modelling and analysis methods encompassing prior, pre-posterior, posterior decision and value of information analysis in the context of structural health information for civil engineering structures. Building upon the structural reliability methods and uncertainty modelling methods, the lecture provides models for structural health information and actions such as non-destructive testing results, monitoring, load testing, load control and repair.
Lecturer: Prof. Sebastian Thöns (Lund University)