Dr. Vivek T. Ramamoorthy is a Lecturer in Mechanical Engineering at the School of Physics, Engineering, and Computer Science (SPECS) at the University of Hertfordshire. He holds a master's degree in Mechanical Design from the Indian Institute of Technology Madras, where he specialized in Vibration and Acoustics. Following his master's, he was awarded an AGFS fellowship and served as a Scientific Officer at the Atomic Energy Regulatory Board. In this role, he conducted research on seismic margin estimation for Class-I piping systems subjected to large earthquake loads. In 2018, Vivek earned a Marie Sklodowska-Curie Early-Stage Researcher Fellowship under the No2Noise EU project and completed his PhD in Computer Science at the University of Nottingham. His doctoral research focused on developing heuristic and metaheuristic methods for the topology optimisation of acoustic porous materials. Vivek's research is at the nexus of mechanical engineering and computer science, with key interests in structural topology optimisation, artificial intelligence, and the acoustics of porous materials.

Keywords: Metaheuristics, Acoustics, Topology Optimisation, Structural Dynamics, Artificial Intelligence

Summary

Structural Topology Optimisation is at the cutting edge of contemporary design techniques for creating lightweight and high-performance mechanical components and structures. These methods are highly effective at identifying near-optimal shapes or layouts that maximise structural strength while utilising a fixed amount of material. The intricate shapes produced through this process are particularly suited for fabrication via additive manufacturing. For example, the use of topology-optimised lightweight components in the Airbus A380 has resulted in a significant weight reduction of 1,000 kg per aircraft (Krog et al., 2004). Given the substantial potential for weight and cost savings, there is significant value in advancing these techniques and broadening their applicability across various domains. Vivek's research is focused on improving the efficiency of topology optimisation methods by employing artificial intelligence to reduce computational resource requirements and addressing the challenges related to the manufacturability of optimised shapes.

Metaheuristic approaches, such as genetic algorithms and covariance matrix adaptation evolution strategies, have proven effective in solving complex optimisation problems that are otherwise impractical to address using exact methods. In a case study on optimising the shapes of acoustic porous materials, Ramamoorthy et al. (2021) demonstrated that combining gradient-free heuristic techniques with conventional gradient-based methods, such as Solid Isotropic Material with Penalisation (SIMP), results in superior Pareto-optimal shapes by avoiding local optima, all within the same computational budget. His future research interests include further exploration of fitness landscapes across different problem domains and the development of domain-specific optimisation strategies.

One of the particular areas of interest is the application of topology optimisation in passive noise control. By strategically introducing cavities or air pockets within sound-absorbing foams, it is possible to create resonances that enhance sound absorption, particularly at low frequencies, which are notoriously difficult to mitigate.

Module support

Semester A (Oct-Dec)

4ENT1163 Applied Design (CDIO) (Auto & Mech)

7ENT1145 Data Analytics and Artificial Intelligence 

Semester B (Jan-April)

4ENT1165 Programming for Aircraft Engineers 

5ENT1125 Industrial Mechanics (CDIO)

6ENT1159 Acoustics

Interactive web demos 

Conventional learning methods, such as books, lectures, and tutorials, remain essential for formal education. However, given the rapidly expanding breadth of human knowledge, these traditional approaches may limit one's ability to master a wide array of topics within a single lifetime. In this context, unconventional learning methods, such as physical demonstrations, interactive demos, or games, can significantly reduce learning time and provide a deeper understanding.

Vivek specialises in developing interactive web demos that enable students to grasp complex concepts quickly and thoroughly. Examples of his work include:

1. Genetic Algorithm tutorial: A simple interactive web application that allows users to understand the fundamentals of genetic algorithms in five minutes or less.

2. Damping ratio demo: An interactive web app designed to help users intuitively grasp the concept of damping ratios in vibrating spring-mass systems.

3. Co-variance Matrix Adaptation Evolution Strategy: A demo showcasing the CMA-ES optimisation strategy.

For more information on Vivek's work and access to these interactive demos, please visit his website at vivektramamoorthy.github.io.