Laboratory for Scientific Computing and Modelling COVID-19 questions
Introduction
In the laboratory for Scientific Computing an Modelling LSM we are involved in four themes regarding COVID19 questions:
- promoting random testing to manage a safe exit from the COVID19 lockdown (group involved: CMT)
- combined High Fidelity Simulations/CFD-Grade Experiments of Particle-Laden Flow in Airborne Infectious Isolation Rooms (group involved: CFD)
- establishing a bridge between basic pharmacologists and clinicians (groups involved: CMT & HPCE)
- providing HPC resources and analytics to COVID19 (HPCE) researchers (group involved: HPCE)
1. Using random testing to manage a safe exit from the COVID19 lock-down
This is an ongoing effort spearheaded by M. Müller, P. Derelet and Ch. Mudry. They argue that random testing is central to controlling the COVID19 epidemics and to optimizing the response to it. Random testing is crucial to acquire currently lacking quantitative information on how various restrictive measures affect transmission rates. This knowledge will:
- significantly improve the predictability of the epidemics
- allow for informed, optimized decisions on how to modify the set of restrictive measures
- enable the real-time assessment of the efficiency of new means to reduce transmission rates (such as new tracing strategies).
It is very important to start random testing for COVID-19 infections immediately and to rapidly increase the testing capacity, the more frequently one samples the population the more reliable and geographically refined will be the data. Preliminary results indicate that for a country with the population of Switzerland, even a few thousand random tests per day suffice to obtain valuable data about the current number of infections and their evolution in time. This is crucial to assess in real time the quantitative effect of restrictive measures. It further allows one to detect geographical differences in spreading rates and thus formulate optimized strategies for a safe reboot of the economy.
2. Combined High Fidelity Simulations/CFD-Grade Experiments of Particle-Laden Flow in Airborne Infectious Isolation Rooms
This research proposal was submitted to COVID-19 call of SNF by A. Dehbi. Patients infected with pulmonary viruses such as Covid-19 are placed in airborne infectious isolation rooms (AIIRs) which are premises held at negative pressure as a first protective measure against contamination in the hospital. A second protection is efficient ventilation to minimize the risk to Health Care Workers (HCWs) attending to patients. Ventilation standards differ from country to country and even from hospital to hospital, and therefore, it is of high value to have an accurate understanding on how infected droplets are transported in AIIRs, given the adopted ventilation protocols and physical and geometrical specifications. The best way to achieve this understanding and help achieve ventilation schemes that reduce infection risk to HCWs is through validated Computational Fluid Dynamics (CFD) simulations. We propose to conduct a dual computational/experimental investigation with state-of-art models and tools. The simulations will be done with the most accurate CFD modeling approach that is available for general flows, namely the Large Eddy Simulations (LES). To validate these simulations, an array of three-dimensional flow and particle measurements will be performed in a large-scale AIIR mock-up facility, comprising the patient and HCW manikins, as well as the patient bed and necessary furniture.
The test matrix for the experiments and accompanying LES simulations will include the variation of the main parameters controlling droplet transport, namely: the droplet size, the expiratory ow rate out of the patient (cough/breathing), the supply air flow rate and the exhaust air location. The central objective of this proposal is not to optimize any particular AIIR design to minimize infection potential. Rather, we strive to provide the medical community with a validated methodology using the most accurate simulation models and sub-models available. As such, the community can elaborate future design modifications of their AIIR, and test the implications of these changes with respect to infection risks using a trustworthy simulation tool.
3. Providing HPC resources and analytics to COVID19 (HPCE) researchers
LSM is providing Merlin computing resources to a team of scientists from the Crick institute in London. This e ort is coordinated by E. Slack (ETH). From LSM D. Feichtinger and M. Caubet are involved.
4. Bridging basic pharmacologists & clinicians
Together with colleagues at BIO, X. Deupi (coordinator) & S. Bliven from LSM, are setting up a database on available drugs that can be used to treat
COVID19. They are getting advice from members of the Swiss Tropical and Public Health Institute and the Drugs for Neglected Diseases initiative on what is the relevant data that medical doctors need to have on these drugs. The goal is to provide clinicians the information required to make decisions on further testing, testing possible combinations of drugs, and also provide a background to talk to clinicians who may want to test new drug candidates. It also gives them an overview on where to set their focus.