Abstract:
The quality monitoring technique for Cement stabilised earth
blocks (CSEBs) is so challenging that it is often neglected. This
study has investigated the possibility of using machine learning to
predict the compressive strength of CSEBs based on cement content,
electrical resistivity and Ultrasonic pulse velocity (UPV) as
a potential way to enhance quality control. The study considered
three types of soil and different cement content in the preparation
of CSEBs with 10 different cement-soil mixtures. Various machine
learning models were proposed to predict the compressive
strength of CSEBs. The models were evaluated using 180 experimental
datasets, and the best model for predicting the compressive
strength of CSEBs was selected. The ANN and BTR models performed
better than the other machine learning models tested in
this study for predicting the compressive strength of CSEBs. The
results show that a combination of cement content, electrical resistivity
and UPV can be used to assess the quality of CSEBs more
accurately, which can contribute to the knowledge base and be
applied in the real world. Materials scientists and engineers can use
reliable predictive models to assess the strength properties of both
new and old brick structures without damage or loss of use.
