Dynamic Neural Network for Predicting Creep of Structural Masonry

One of the inherent modeling problems in structural engineering is creep of quasi-brittle materials (e.g., concrete and masonry). The creep strain represents the non-instantaneous strain that occurs with time when the stress is sustained. Several creep models with limited accuracy have been developed within the last few decades to predict creep of concrete and masonry structures. The stochastic nature of creep deformation and its reliance on a large number of uncontrolled parameters (e.g., relative humidity, age of loading, stress level) makes the process of prediction difficult, and yet accurate mathematical model almost impossible. This study investigates the potential use of Dynamic Neural Network (DNN) for predicting creep of structural masonry. The main motive of use DNN is that DNN could memorize the sequential or time-varying patterns while training process. Thus, DNN becomes more capable of capturing the time-dependent of creep deformation than the static networks. The results showed that the developed DNN models are able to predict the creep deformation with an excellent level of accuracy compared with that of conventional methods and the static networks models.