Newsletter: January 2021
Computational Modelling for Cultivated Meat Bioreactor Scale Up
One of the great challenges facing human civilization is the need to feed a growing number of people. By the year 2050 it’s estimated that the number of people on earth will total more than 10 billion. A diversification of approach is possible, with developed countries shifting to plant based foods and developing countries increasing livestock. An alternate approach is gathering momentum; cultivated meat, where animal muscle is grown in the laboratory rather than on the animal.
The techniques involved have been around for a while, and recent progress has suggested the potential from an economic viability perspective (the first lab grown burger needed around $2.5 million of investment to produce in 2013). The environmental benefits could be game changing, with recent analysis suggesting that switching to cultivated meat could reduce land use by 90% (allowing more vegetation to be planted), reduce water use by 90%, and gas emissions by 40% (cow's fart a lot of methane).
One of the key challenges facing the eighty plus companies worldwide working in this space is the scaling up of bioreactors from the laboratory to production scale. One of the key areas where modelling can help is predicting and modifying the designs of bioreactors where the meat cells grow, and applying those predictions to production scale facilities to get an idea of performance before significant financial and operation commitments are made. One of the challenges faced by the nascent industry is the scale up of these reactions from the bench-top in the lab to facilities capable of generating sufficient quantities to realize the environmental benefits. Computational modelling can play a strong role here, with previous experience in the chemical and pharmaceutical industries of modelling mixing and reactions.
There is one new significant challenge though; cells can be damaged and productivity limited if they experience significant hydrodynamic stress. Shear stress and dissipation of turbulence energy are particular culprits. Measures of these can be output from computational fluid dynamics (CFD) models with relative ease, but the real question is how much of an effect on the cell biology does these have, with regards to the cells moving around a bioreactor experiencing different levels as they go? Some of the answer can be found from developing and applying models of the cell culture itself.
Biocellion is doing just that, combining biological and physical models in to their C++ simulation platform. These living systems models are being combined with CFD models to give a more complete picture of whats going on in bioreactors. You can see an example using a conventional stirred tank reactor at the Cultivated Meat Modelling Consortium. This new level of modelling capability combined with the development of new bioreactors specifically for cultured meat applications suggest that simulation may have a significant part to play in the development of this industry.
You can read more about the wider challenges and state actions that could influence the success of the sector in a recent Nature article, and explore some of the challenges of current livestock farming in Channel 4's fantastically named Apocalypse Cow.
blastFoam from Synthetik Applied Technologies
blastFoam is an OpenFOAM based library for single and multiphase compressible flow for applications involving high-explosive detonation as well as general compressible flows. It’s developed and released opensource by Synthetik Applied Technologies, an Austin, Texas based high technology start-up.
Explosive detonation modelling considers the dynamics of pressure waves and the associated chemical reactions; this requires thermodynamic equation of state models not found in the standard OpenFOAM release, as well as higher order explicit time integration and flux schemes. The core solvers offer both Eulerian and Langrangian multi-fluid methods, and provide additional solvers integrated with the OpenQBMM based number-density function transport coupling technique.
The recently released V4.0 also includes particle solvers. The user guide is comprehensive (put your maths hat on) and the validation with academic solutions looks very good. The versatility looks great too – you can find a variety of applications on the web, with the most notable a model of the recent airport explosion in Beruit.
The code capability is extensive, with specific enhancements to adaptive mesh refinement (AMR) extending it to allow refinement on blastFoam specific outputs. At present it compiles under the Foundation version of OpenFOAM, version 7. You can find the code and instructions on GitHub, and more about Synthetic Technologies here.
Cadence Design Systems Inc. Acquires Numeca International
There has been a fair amount of consolidation in the simulation software sector over recent years, with ANSYS' continued acquisition of specialist outfits, and Siemens bringing both CD-Adapco and Mentor Graphics into their fold. The Siemens acquisitions were announced in 2016 and there have been few remaining CFD specialist companies around, and fewer large multinational engineering and simulation services companies who don't already have a CFD string to their bow.
One of those few CFD specialists is Numeca, which was spun out from Vrije Universiteit Brussel in 1993. They offer a complete suite of tools across the CFD process spectrum, as well as a particular optimization offering, and Omnis which manages the whole workflow from a single environment. Cadence is an electronic design software specialist, with simulation tools covering 3D electromagnetics, coupled electro-thermal problems, and finite difference time domain (FDTD) systems simulation. Adding coupled fluid-thermal dynamics to their offering fills an obvious hole, and leveraging Omnis for multiphysics chip and electronics system design makes good sense.