An experimental investigation to determine the failure load of optimal hammer head pier cap and interpolate using univariate splines machine learning algorithm

• Published in: I-Manager's Journal on Civil Engineering Vol. 13; no. 4; p. 1
• Main Authors: Chandrasekhar, K. N. V., C. Sai, Ankitha Varma, D. Indu, Reddy, Eshwari, D., Venkatsai, G., Dhanush, N., V. Niranjana, Chary, Samyuktha, C., Bharathi, M., P. Arun, Kumar
• Format: Journal Article
• Language: English
• Published: Nagercoil iManager Publications 01.10.2023
• Subjects: Algorithms; B spline functions; Concrete; Design optimization; Failure load; Foundry sands; High strength concretes; Industrial wastes; Machine learning; Optimality criteria; Steel fibers; Strain energy; Topology optimization
• ISSN: 2231-1068; 2249-0779
• DOI: 10.26634/jce.13.4.20281

The application of structural optimisation is fast changing the thinking of Design Engineers. With the rapid development in high strength concrete and other materials which can resist higher loads, the task of reducing the weight of the structure can be addressed with ease. The present study is an application of topology optimisation of continuum structures. The design domain is modelled using first order basis splines and optimisation is performed using optimality criteria minimising strain energy as the objective function. A model pier cap is chosen with the standard dimensions of 3 feet x 9 in x 4 in. The size and location of openings are determined using topology optimisation and drawn in AutoCAD® software. The casting is done using concrete with different percentages of replacement of cement and fine aggregate. Cement is partially replaced using Alcofine and fine aggregate is partially replaced using waste foundry sand. The foundry sand is an industrial waste obtained from the foundry industry located at Balanagar, Hyderabad. Four specimen beams are cast and tested in the laboratory. Steel fibres are used to care for the tensile stresses produced within the beam. The analysis done here can be applied to any material other than concrete as well. The failure load is determined in the laboratory for each sample. The interpolation of failure load is done using python code and run on Anaconda Jupyter ® platform to determine the value of failure load for any percentage of replacement between 0 to 10%.