The Pawsey Supercomputing Centre

Content

Canonical Simulation of Respiratory Airflow

Aerosol science plays a developing role in research areas in the health sector. Professor Ben Mullins from Curtin University’s School of Public Health is a chemical and environmental engineer who applies aerosol science to his biomedical research projects. In collaboration with colleague and mechanical engineer Dr Andrew King, Prof Mullins is conducting research that compares airway penetration between liquid and powdered medications.

The Challenge

“Chronic Obstructive Pulmonary Disease (COPD) and pneumonia are two of the top five killers of the elderly, with COPD claiming one patient every 10 seconds,” Prof Mullins says.

“Asthma affects approximately 15 percent of the Australian population and $358 million is spent on asthma medication alone per year, yet more than 95 percent of it is wasted.”

No other WA research groups have had prior experience of simulating an expanding and contracting lung that reflects on realistic breathing. In order to attain comparison results, a 3D model had to be created. The model was developed to simulate respiratory airflow and particle deposition during realistic lung expansion and contraction.

The Solution

The solution to attaining results was to couple a ‘moving mesh’ algorithm with custom airflow and particle models.

“We developed the most advanced and physiologically realistic lung model to simulate airflow, which will revolutionise aerosol drug delivery (such as Ventolin inhalers) and also assess exposure to airborne pollutants,” Prof Mullins says.

“This research is the first of its kind and allows us to get better information on what is happening at highly localised regions in the lung. This will, in turn, allow us to better target drug delivery to where it is needed most.”

The Outcome

The image shows a high resolution simulation performed on the Magnus supercomputer of airflow and expansion (breathing) of a lung. The lung geometry used for the simulation is obtained from a 3D, computed tomography (CT) scan

The image shows a high resolution simulation performed on the Magnus supercomputer of airflow and expansion (breathing) of a lung. The lung geometry used for the simulation is obtained from a 3D, computed tomography (CT) scan

Supercomputing is considered an essential element of the research project as it makes the process function a thousand times quicker. The Pawsey Supercomputing Centre’s solutions were extremely helpful to this research project as it is impossible to effectively conduct complex computational fluid dynamics (CFD) without high powered computers.

Both the particle code and the moving mesh algorithm are highly computationally-intensive to resolve, and require a huge data transfer between cores in order to reach a solution within a reasonable time.

“HPC means our experiments are a thousand times faster,” says Prof Mullins.

“Simulations like this are possible not only because of the expertise within the Fluid Dynamics Research Group, but also because of our access to the Pawsey Supercomputing Centre.”

Comments are closed.