Prediction of Transfer Length of Prestressing Strands Using Neural Networks

This study investigates the efficiency of an artificial neural network (ANN) in predicting the transfer length of prestressing strands in concrete beams. The transfer results from various research projects have been collected to train and test the ANN model. Each parameter affecting the transfer length of prestressing strands (ratio of the area of prestressing strand to the area of concrete, surface condition of prestressing strand, diameter of prestressing strand, percentage of debonded prestressing strands, effective prestress, plateau in strain profile, and concrete strength at the time of measurement) was arranged in an input vector and a corresponding output vector that includes the measured transfer length of prestressing strands. Findings showed that the ANN model developed was able to predict the transfer length of prestressing strands for both input patterns used in training and testing processes. Although the performance of the developed ANN model is limited to the range of input data used in training process, the model can easily be retrained to expand the range of input variables by providing an additional new set of data. Sensitivity analysis showed that all the selected parameters used in the model have an effect on the transfer length of the prestressing strand.

Language

  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 01046561
  • Record Type: Publication
  • Files: TRIS
  • Created Date: Apr 17 2007 9:39PM