PSI - Issue 62

Sebastian Thöns et al. / Procedia Structural Integrity 62 (2024) 259–267 Sebastian Thöns and Ivar Björnsson/ Structural Integrity Procedia 00 (2019) 000 – 000

264 6

^ĐĞŶĂƌŝŽƐ

džƉĞĐƚĞĚ ƵƚŝůŝƚLJ

/ŶĨƌĂƐƚƌƵĐƚƵƌĞ ĞdžƚĞŶƐŝŽŶƐ

ZĞƉůĂĐĞŵĞŶƚ ŽƉƚŝŽŶƐ

^ĞƌǀŝĐĞ ůŝĨĞ ĞdžƚĞŶƐŝŽŶ ƐƚƌĂƚĞŐŝĞƐ

Time

Start of service life

Information management

System state management

Utility

Fig. 3: Depiction of service life management, replacement and extension strategies

Chance Choice

Choice Chance

The model basis for decision support requires the formulation of a decision scenario consisting of the system state model, the utility model and the information and action models (see Fig. 4 and Fig. 5). This main decision theoretical classification facilitates the integration of the individual information and system state management models into the analysis.

Information management

System state management

Utility

Chance Choice

Choice Chance

System states Actions

Outcomes

Information

Utility u

, i j Z

i i

k a

l X

Fig. 4: Decision tree for a decision scenario

The system state model of the infrastructure system can be based on limit states and contains such as, e.g., intact, damaged, and failed states. Bridge system modelling can be very comprehensive and may constitute an own research and engineering topic encompassing several methods such as e.g., nonlinear finite element analyses (e.g., Lonetti and Pascuzzo (2014), Sarmiento Nova, Gonzalez-Libreros et al. (2022)) or complex probabilistic models (Schneider, Fischer et al. (2015)). However, such complex system models are for purposes of detailed analyses and do not normally facilitate a holistic service life management. A utility model encompassing, e.g., benefit or negative utilities is connected to each system state. These include direct and indirect consequences and costs and are often expressed as monetary values. The discount ratio is then required to convert the monetary values to the decision points in time. With both the system state model and the utility model, the system performance (or the system risk in case of solely consequence modelling) can be quantified. The system performance can be adapted with changes in the system states or the utility. Such changes are denoted as actions (e.g., repairs). The system modification with actions can further be supported with information, which are defined as observations of system model parameters (e.g., inspection or testing results). Specifically for service life extension, actions (or measures) may encompass e.g., repair, strengthening, and their implementation support by monitoring and inspections. The main decision theoretical characteristics are type, longevity, precision or execution quality (uncertainty), and costs. Actual reviews of monitoring approaches can be found, e.g., in Abdallah, Atadero and Ozbek (2022) and Rizzo and Enshaeian (2021) and, with the help of unmanned aerial vehicles, in Kapoor, Katsanos et al. (2021) and Greenwood William, Lynch Jerome and Zekkos (2019). These monitoring strategies should be aligned with the structural integrity management of the infrastructure in order to facilitate a monetary and risk reduction value (Thöns (2022), Zhang, Lu et al. (2021)). Information Outcomes System states Actions i i , i j Z k a l X Utility u