Projects

Ongoing projects....

 

Master Thesis by Elena Giacomazzi: Numerical study of biomimetic strain sensors (HS19)

Strain sensing is a capability, essential for survival in the animal kingdom. Insects, such as spiders developed strategies based on holes (campaniform sensillum) whose shape changes in response to load. Arrays of sensilla can be used to increase sensitivity or to sense strain components. Within the thesis, models with Finite Elements (FEM) of various configurations of sensilla are programmed to study the role of the material orthotropy and piezo-electric filling material on the feasibility of applying such systems as strain sensors.  

Proposed projects....

 

Modelling of Mechanosensing with Tactile Hairs

Sensors e.g. for monitoring or control of buildings are often rather complicated, high maintenance measurement devices. A new generation of bionic sensors could facilitate a number of measurement tasks in building automation significantly. Within a research consortium a group of experts in mechano-sensing in biological systems, micro-robotics and numerical simulation aim at revealing the function of biological mechano-sensors e.g. of Venus Fly Traps or spider hairs to discover unifying principles of tactile hairs.
Within your project work, you will apply a non-linear Finite Element Code to simulate the deflection of tactile hairs from contact to triggering from a purely mechanical point of view. You will provide insight into the mechanics and general principles applicable to bionic sensors.

Mechanical characterization of delignified wood perpendicular to grain

Wood densification is a classical method of wood modification aimed at extending the use of wood for high-performance applications. When wood is delignified prior to densification, the required energy can be drastically reduced, resulting in compacted cellulose, ideal for further processing into high-value product. However the mechanical behaviour of the delignified wood is most relevant for all successive processing steps.

Within your project, you will mechanically characterize the behaviour of spruce wood, delignified at various degrees with the help of a micro-stage. You will make in-situ observations of the damage, failure and compaction behaviour and identify relevant phenomena.

AI-based fracture criteria for granular packings

Simulating the compaction of granular packings with grain fractures, is strongly limited by evaluating stress or energy based fracture criteria for all involved grains. Using AI based criteria, that are trained on predicting failure states of single grains from previous experience could overcome this problem and allow for the simulation of much larger granular assemblies. 

Within your project you will use the Material Point Method to generate the training and validation data sets for a large number of polyhedral grain shapes, sizes and will apply the data to train shallow neural networks. You will test various network topologies and learning strategies to find out the best way to realize AI-based criteria for grain failure. 

JavaScript has been disabled in your browser