Multi-model Investigation on Dynamic Response of Machine Foundations Resting on Geogrid Reinforced Foundation Bed

This paper explains the behavior of harmonically loaded machine foundations supported by geogrid-reinforced soil beds using multi-model approaches, such as experimental, numerical, analytical, and artificial neural network (ANN) modeling techniques. Field block vibration tests are performed on a foundation bed measuring 1.47 m × 1.47 m × 0.65 m, prepared on the soil at IIT Kanpur, India [N26° 30′ 59.0892″ (latitude), E80° 13′ 51.6888″ (longitude)]. Two different reinforced cement concrete (RCC) footings of sizes 0.55 m × 0.55 m × 0.2 m and 0.65 m × 0.65 m × 0.2 m are employed in this investigation. The tests are performed at three different eccentric force settings considering three testbeds: unreinforced, single-layer and double-layer geogrid reinforced beds. Geogrid layers reduce resonant amplitude and dynamic shear strain while simultaneously improving resonant frequency, system characteristics and dynamic soil properties. Further, the experimental results are compared with those obtained from the finite element and mass-spring-dashpot analysis. The displacement amplitude of footings is also predicted at different frequencies using ANN modelling to endorse the authenticity of the study. Encouraging agreements can be observed among various modeling approaches considered in this study.