Navigation

Research

Investigation of the mechanical behaviour of graphene nano-composites | Professor Drosopoulos, Dr. Radebe, Professor Foutsitzi, Professor Stavroulakis, Professor Adali
Ideas related to the study of the mechanical behaviour of graphene nano-composite materials are under investigation. Analytical and numerical models are developed and optimization approaches are adopted for the evaluation of the optimum distribution of graphene nanoplatelets. 
   
     
Investigation of band gaps for auxetic metamaterials  | Mr. Koutsianitis, Dr. Tairidis, Professor Drosopoulos, Professor Stavroulakis
Main concept of this research is to apply wave propagation theories on periodic media. Goal is to design composite metamaterials depicting band gaps in chosen frequencies. These metamaterials can then be used in acoustic applications or as a seismic isolation for structures. To implement the research, the performance of composites under vibration is studied, in the framework of optimization algorithms.
     
Multi-scale computational homogenization / Localization of heterogeneous materials | Professor Drosopoulos, Professor Stavroulakis
Multi-scale computational homogenization stands for the exploitation of the microscopic scale of heterogeneous materials (masonry, concrete, composites) and the projection of average quantities in the macro scale, in the framework of a concurrent analysis of both the macro and the microstructure. Within this method, the macroscopic constitutive behaviour is determined during simulation, after solving the microscopic problem and transferring the necessary information on the macroscopic scale.

  Multi-scale computational homogenization 
Recently, some sophisticated efforts for investigating localization phenomena, appear. The ongoing research focuses on applying unilateral contact in the microscopic scale of discrete composite materials, in a multi-scale framework. Damage across the boundaries of the Representative Volume Element (RVE) and the tensile resistance of the considered interfaces are taken into account. In the macroscopic scale, the Extended Finite Element Method (XFEM) has been chosen as the numerical tool for incorporating the microscopic damage, as a discontinuity, at the structural level. 


   
Investigation of crack modelling design in water retaining structures | Dr. McLeod
The research is related to the design of concrete, liquid retaining structures. Main goal is the evaluation of Eurocode’s specifications for this type of structures, by conducting deterministic as well as probabilistic analysis. Both approaches are based on Eurocode’s design equations regarding cracking of liquid retaining structures, under serviceability limit state and cracks due to restrained deformation. Comparison with British design code is considered towards determining an optimum design solution for South Africa. 


 

   
Thermo-mechanical analysis of steel structures  | Professor Drosopoulos, Ms. Singh
The research focuses on the structural evaluation of steel structures, under elevated temperatures. Emphasis is given on the behaviour of the connecting parts under elevated temperatures and the possibility of reinforcing steel structures under fire conditions. Investigation of delamination between the structure and the fire protection material is among the goals of this research.

  Thermo-mechanical analysis of steel structures


   
Topology optimization and homogenization | Professor Stavroulakis, Professor Drosopoulos, Dr. Kaminakis
The design of mechanical microstructures having auxetic behaviour is studied by using techniques of topology optimization for compliant mechanisms. A robust hybrid algorithm based on evolutionary algorithms and local search steps is used. The result may need verification in order to accommodate needs not taken into account in the topology optimization. Therefore, a numerical homogenization scheme is used in order to show that the final design still has the wished negative Poisson’s property. Future research includes investigation of vibration reduction for several applications (base isolation, etc.), by using methods of topology optimization and homogenization.
Topology optimization and homogenization 1
Topology optimization and homogenization 1
     


   
Effectiveness of optimized fuzzy controllers on partially delaminated piezocomposites | Dr. Tairidis, Professor Drosopoulos, Mr. Koutsianitis, Professor Foutsitzi, Professor Stavroulakis
Delamination between the layers of smart composite beams is a possible failure type. The effectiveness of active vibration control, when delamination appears, is investigated. In particular, optimized fuzzy controllers are developed and tested on partially delaminated piezocomposites. A finite element model based on layerwise theory, which incorporates the electromechanical coupling and the adhesive layer, is first developed. Delamination is considered by nonlinear structural analysis techniques and is used for the creation of realistic partially delaminated structures. Fuzzy controllers are built and applied to the smart structure. The fine-tuning of the parameters of control is done using genetic algorithms. Based on the results obtained from the numerical investigation, the applicability of fine-tuned fuzzy controllers, even in the case of partial delamination, is shown. According to different delamination scenarios, which were tested in this work, the design of adaptive fuzzy controllers, robust against delamination, is feasible.
Effectiveness of optimized fuzzy controllers
Effectiveness of optimized fuzzy controllers
     


   
Inverse analysis for the evaluation of damaged conditions of structures | Dr. Stavroulaki, Professor Drosopoulos, Professor Stavroulakis
A novel approach for the investigation of pathological problems in masonry arch bridges using inverse analysis procedures is proposed. A real case study located in Kakodiki village on the island of Crete (Greece) is considered for the application of the proposed computational scheme. The method uses the damaged condition of the bridge as starting point and seeks the potential load cases that led to its development. The damage identification is transformed into a parameter identification problem which is studied by comparing numerical predictions and existing damage pattern. A global optimization approach by means of a genetic algorithm is adopted within the inverse problem, aiming at adjusting the parameters of the mechanical model of the structure so that an error function that measures the differences between the real and the numerically predicted damage pattern is minimized.

The outcome of the methodology might provide a valuable information regarding to the planning of maintenance actions as well as to the design of retrofitting measures. Future research includes the consideration – prediction of existing cracks’ position, on the assessment of the condition of stuctures.
 

Inverse analysis of damaged structures

   

Contact Webmaster | View the Promotion of Access to Information Act | View our Privacy Policy
© University of KwaZulu-Natal: All Rights Reserved