FE-ANN Based Modeling of 3D Simple Reinforced Concrete Girders for Objective Structural Health Evaluation

Project Details
STATUS

Completed

START DATE

03/01/15

END DATE

05/31/17

FOCUS AREAS

Infrastructure

RESEARCH CENTERS InTrans, CTRE, MTC
SPONSORS

Kansas Department of Transportation
Kansas State University
Midwest Transportation Center
USDOT/OST-R

Researchers
Principal Investigator
Hayder Rasheed

About the research

The structural deterioration of aging infrastructure systems and the costs of repairing these systems is an increasingly important issue worldwide. Structural health monitoring (SHM), most commonly visual inspection and condition rating, has proven to be a cost-effective method for detecting and evaluating damage. However, the effectiveness varies depending on the availability and experience of personnel performing the largely qualitative damage evaluations.

The artificial neural network (ANN) approach presented in this study attempts to augment visual inspection through a crack-induced damage quantification model for reinforced concrete bridge girders that requires only the results of limited field measurements to operate.

Using Abaqus finite element (FE) analysis software, the researchers modeled simply supported three-dimensional concrete T-beams with varying geometric, material, and cracking properties. The ANNs achieved excellent prediction accuracies, with coefficients of determination exceeding 0.99 for both networks. Additionally, the ANNs displayed good predictions accuracies when predicting damage levels in beams not included in the database. Results indicate promise for this application of ANNs.

Utilizing the two top-performing network architectures, the researchers developed a touch-enabled software application for use as an on-site bridge member damage evaluation tool in the field. The application was given the acronym BRIDGES, for Bridge Rating for Induced Damage in Girders: Evaluation Software. The application’s outputs were validated as matching the ANN predictions.

The researchers developed a similar software application for the reverse problem/damage detection and use as an on-site damage prediction tool. This application tries to predict the crack configurations using ANN, based on the geometrical and material parameters, as well as the nine nodal stiffness ratios. This touch-enabled application was given the acronym DRY BEAM, for Damage Recognition Yielding Bridge Evaluation After Monitoring.


Funding Sources:
Kansas Department of Transportation ($34,153.00)
Kansas State University ($15,784.00)
Midwest Transportation Center
USDOT/OST-R ($49,937.00)
Total: $99,874.00

Contract Number: DTRT13-G-UTC37

TOP