3D Image Data in CAD/CAE
Working as PR and Marketing Officer for Simpleware, with specialties in: marketing communications, social media, press, quality management, copywriting.
Calculating effective material properties from 3D images
September 28th, 2015 by Gareth James
Anyone working with 3D image data (such as CT, micro-CT, SEM…) in Materials Science faces the challenge of reconstructing and obtaining useful results from scans. Computing material properties such as solid, flow and electrical properties is a particular challenge, as this often requires a lot of computational resources, or a workflow using multiple software packages.
At Simpleware, we’ve developed some novel software techniques for calculating effective material properties that show the potential of 3D image data for characterising samples. This process involves using finite element-based homogenisation, implemented through the Physics Modules +SOLID (effective stiffness/elastic moduli), +FLOW (absolute permeability) and +LAPLACE (electrical conductivity/permittivity, thermal conductivity, and molecular diffusivity).
Homogenisation involves approximating a complex heterogeneous material by a homogeneous material with effective properties chosen so that its response to external loads resembles as closely that of the original material. Complex structures such as multi-phase composites can be analysed with this method following visualisation, image processing and segmentation in Simpleware software. A built-in FE solver is used to calculate properties by post-processing the fields induced in a cuboidal material sample by a sequence of appropriate boundary conditions, with options available in +SOLID and +LAPLACE to calculate upper and lower bounds for effective properties directly from image data.
While FE-based homogenisation isn’t new, the approach at Simpleware has a major advantage over alternative approaches as it uses smoothed FE-based meshes and a built-in solver; FE mesh surfaces for a given resolution are more accurate than voxel mesh surfaces and their area converges with increasing resolution to the actual surface area of the sample. Segmented domain accuracy can be preserved whilst reducing mesh size, improving on grid or voxel-based methods that produce stepped surfaces and over-estimate the surface area between material phases.
Included in this approach is a method for reducing the potential effects of boundary conditions on FE homogenisation by calculating effective material properties from field averages taken over a domain smaller than the region of interest used in simulations. Graph tools have been created to display these properties, while results from FE simulations can be visualised, animated and exported as text files and images.
Materials scientists and analysts that want to speed up their workflows can use these solutions when working with complex image data, reducing often lengthy workflows in areas like composites, digital rock physics and many others.
See http://simpleware.com/ for more details of what you can do with 3D image data, as well as for free trial versions of the software.