The actual tensile force in the pre-stressed (PS) tendons of a pre-stressed concrete (PSC) girder is an important factor for evaluating the performance of PSC girder bridges. To measure the tensile force in a PS tendon, an artificial neural network (ANN)-based tensile force measurement method is proposed in this study, incorporating a fibre Bragg grating (FBG) sensor and an elasto-magnetic (EM) sensor. The FBG sensor measures the strain change in the whole of a single tendon while the EM sensor measures the local permeability changes in all tendons. An FBG-encapsulated tendon is fabricated by installing an FBG sensor onto a perforated tendon and EM sensors are fabricated by embedding the EM sensor into the girder. An experimental study is performed to verify the capability of the sensors using a material testing system (MTS) and a down-scaled girder model. The FBG sensor measures the change of strain due to the tension variation, while the EM sensor measures the magnetic flux change. The ANN is used to improve the accuracy of estimation. The measured strain and permeability are used to train the ANN to estimate the tensile force in a PS tendon. To verify the capability of the trained ANN, the long-term tensile force is estimated using the ANN, the result is compared with that from a conventional regression model and the reference tensile force is measured by a load cell. The results show that the proposed method can monitor the pre-stressing force in the PS tendon of a PSC girder with high accuracy.
展开▼
