Helping You To Use IBSim
Our group’s main goal is to see wider use of the IBSim technique, particularly for industrial applications. Please contact us if you would like to work with us for cases such as:
- if you have an interesting R&D project that would make a good IBSim case study
- you’d like a customised IBSim solution developed for your industrial application
- you’re already working with IBSim and would like a new capability developed
- you’d like to integrate or test your software with our IBSim workflow
Through our partner research groups within the College of Engineering, Swansea University, we have access to all the required experimental and computational capabilities for the IBSim workflow. When working with us we can arrange access to these facilities, assist in developing a workflow around your in-house facilities or a mixture of both.
In addition to this, we have access to an extensive suite of facilities for alternative non-destructive testing methods, materials characterisation, rapid prototyping etc., and are happy to discuss bespoke solutions for your application with you.
As a university-based research group we can offer to host visiting students / researchers / interns for short-term projects. Alternatively, we can provide mentorship to those wishing to work with us remotely e.g. as part of your final year research project or a career development exercise. Our group is a highly multidisciplinary team and therefore are keen to work with you on a wide range of projects. Below are some suggested topics that can be used for a project of varying scope depending on the time you have available. Contact us to discuss these in more detail. Also, feel free to contact us with your own ideas for a collaborative project.
Create a CAD to voxel conversion mini app
A tool is needed to auto-generate ‘classifiers’ to input to a machine learning training algorithm. A step in the proposed approach is to create a ‘ground truth’ 3D image from the computer aided design (CAD) drawing. This task will involve developing a mini-app in python to import an FEA mesh and convert it into a 3D voxelised image.
Create script capable of handling large datasets in processed XCT image to FEA mesh pipeline
Generate python script to take output from SuRVoS (XCT greyscale image segmented by machine learning) into Iso2Mesh to create a finite element (FE) mesh for Code_Aster (all are open-source software). This will be used to automate the workflow pipeline from image to simulation.
Create script to prepare FE mesh to run on high-performance computing system
Meshes generated in Iso2Mesh need post-processing so that they’re optimised to run on a parallelised HPC system. An open source tool such as METIS may be used to partition the mesh so that it may be efficiently divided between processing cores to minimise cross-node (MPI) communication. A script is required to automate the process of taking the output from Iso2Mesh into METIS to optimise the FE mesh for HPC.