Horizon 2020 Marie Skłodowska-Curie Innovative Training Network

ESR8

Probabilistic modelling of bridge damage based on damage indicators
ESR8 2017-12-02T05:20:33+00:00
ESR8: Probabilistic modelling of bridge damage based on damage indicators

While in the 90’s, Structural Health Monitoring (SHM) projects were concerned only with the construction phase, nowadays, the main purpose of monitoring the structural response is to confirm design assumptions in new structures, and the controlled lifetime extension of existing structures. SHM systems provide indirect indicators of damage in bridges – for example, a change in a mode of vibration may indicate damage.

There is an extensive literature in SHM, particularly in methods which use accelerometer signals as damage indicators, however, the uncertainty associated to these damage indicators is unknown and will be investigated by combining Phimeca’s advanced uncertainty models with University College Dublin’s expertise in bridge analysis. It is known that damage in bridges due to, for example, corrosion, is spatially correlated, i.e., if there is a high level of damage at a point, the probability of a high level of damage at an adjacent point is increased. This theory has been developed by O’Connor (Dublin) (who will advise in this project), Stewart and others.

Damage indicators provide information on the location of the damage but this relationship between indicator and location is uncertain. Nevertheless, the probability density function for damage indicator often contains implicit information on damage location.

In this project, Bayesian Updating will be used to infer probabilistic information about the spatial location and level of damage throughout the bridge, given measurements of damage indicators. The results will include an allowance for the uncertainty in the damage indicator/location relationship as well as prior knowledge about the correlation between damage and location. Spatial correlation of traffic loading will also be considered and the effect of combining correlation information on load and resistance.
Methods to more accurately evaluate bridge safety based on traffic load and bridge condition measurements.
This project involves a secondment of some months to University College Dublin (UCD) (supervised by Prof. OBrien). The UCD supercomputer facilities will assist the ESR in carrying out simulations and probabilistic calculations required for the project. UCD will advise on how to relate the spatial distribution of bridge safety to measured damage indicators in bridges.

You can follow blog and publications by ESR8 below.

ESR8 blog
ESR8 publications