Optimal maintenance planning of existing structures using monitoring data
In the last decades maintenance optimization of civil engineering structures has gained increasing attention since the number of ageing structures is large and the available budget dedicated to maintenance is constrained in European countries. Currently, most advanced management systems, in terms of optimal maintenance, combine models of the ageing structures together with a probabilistic representation of the uncertainties both on the initial state (e.g. material properties, crack length) and on the kinetics of the degradation phenomena. NDT techniques allow the analyst to gather additional information all along the lifetime of the structure. In this context, it becomes important to optimize the maintenance strategy using a cost/benefit analysis.
Advantage of the type of monitoring information obtained from the sensors developed in WP1 will be included in the structural health model of the structure using Bayesian techniques. Various types of criteria related to the results of the monitoring will be proposed together with an estimation of the associated costs. The overall procedure will then be optimized, in order to reduce the cost of maintenance all over the life time of the structure. The approach will be illustrated on the structural components of bridges and/or wind turbines.
The ESR will take advantage of the type of monitoring information obtained from the sensors developed in WP1 and will include this additional knowledge in the structural model using Bayesian techniques. Various types of criteria related to the results of monitoring will be proposed together with an estimation of the associated costs. The overall procedure will then be optimized so as to minimize inspection/monitoring/maintenance costs all over the lifetime of the structure.
The approach will be illustrated on bridge structural components and also on wind turbine substructures. Fatigue is a major design driver for welded details in the steel substructures (tower and foundation consisting of monopiles, jackets, ...) and fatigue reliability assessment is essential. In wind turbine, blade manufacturing defects will always be present to some extent. Therefore, the use of information from monitoring systems to update the reliability level taking into account all uncertainties in a consistent way is an important contribution to minimize the cost of energy (COE).
Development of cracks from the defects can be critical and a reliability-based approach using information from monitoring can be a useful method to ensure that the reliability of the blades remains at a satisfactory level during the whole lifetime.
During the first part of the thesis the ESR will work at Phimeca in order to perform a literature review and state of the art on maintenance planning methodologies and to experience probabilistic methods for reliability assessment of structures. During this period at Phimeca, the ESR will also benefit from existing experience at Ifsttar in Marne-la-Vallée (with Dr André Orcesi). The ESR will then perform a secondment at BAM where the ESR will work on uncertainties of monitoring techniques. Back to Phimeca, these uncertainties will be taken into account inside the Bayesian approach. The ESR will then perform a secondment at EPFL to work on admissible damage level which constitutes a crucial point for maintenance optimization. The ESR will then finish its work at Phimeca with, once again, the benefit of working also with Ifsttar.
- An innovative methodology will result from the thesis.
- 2 accepted peer-reviewed papers.
- Presentations at national/international conferences.
- Presentations at workshops and for potential end-users.
- Successfully defended PhD thesis.
- Ahmadivala M., Mattrand C., Gayton N., Dumas A., Yalamas T., Orcesi A.
Application of AK-SYS method for time-dependent reliability analysis
CFM 2019, 24e Congrès Français de Mécanique, Brest, France, 26-30 August 2019
- Ahmadivala M., Sawicki B., Brühwiler E., Yalamas T., Gayton N., Mattrand C., Orcesi A.
Application of time series methods on long-term structural monitoring data for fatigue analysis
SMAR 2019 - 5th Conference on Smart Monitoring, Assessment and Rehabilitation of Civil Structures, Postdam, Germany, 27-29 August 2019
ESR 10: Morteza Ahamadivala (PHIMECA)
Local industrial supervisor: Dr Thierry Yalamas (PHIMECA)
Academic co-supervisor: Dr André Orcesi (IFSTTAR)
PhD director: Dr Nicolas Gayton (SIGMA Clermont)