Modelling High Cycle Fatigue and Assessing Failure Probability for Reinforced Concrete Bridges Using Site Specific Fatigue Load Model

This model analyzes the fatigue failure of low strength reinforced concrete bridges using a site-specific fatigue model. Since AASHTO truck does not reflect the extent of damage caused by actual traffic load traversing a bridge, a new fatigue load model has been proposed on the concept of equivalent damage using Miner hypothesis. Thus incorporating an axle width and the distance between the tyres, a single axle load of 210 KN was considered in the model. The outcomes of this model are vital in term of in identifying the severity of Fatigue Damage

Abaqus Modelling

A low strength concrete mode with maximum compression strength of 16 MPa is used in material modelling. Moving load was applied using a user subroutine VDLOAD. This subroutine enabled us to define the magnitude of wheel load as a function of position, time and velocity. From the stress profile at the most critical node at the mid span section of the bridge, It was analyzed that the stress generated due to self-weigh is 147 MPa and this stress start increasing gradually as the fatigue load travel toward mid span of the bridge. This peak value of the stress which is below the yielding limit of the steel is used further in FE-SAFE for calculating the Fatigue failure life cycle.

Fe-SAFE

After importing the results in fatigue, S-N curve of the material is specified and Un-axial stress life approach is used as a fatigue evaluating criteria. Profile of the passing truck is specified in the amplitude section which estimates the fatigue failure cycle equal to 4.09 million trucks. In the next section the performance of structure is assessed by monitoring its failure probability under different load variance percentage. This is performed by defining Weibull distribution to the material strength (S-N Curve). From the results as in Figures it was analyzed that at early stages of fatigue life, the reliability of the bridge with high load variation is low.