MUCAA Project Report.pdf

Investigate air flow in the presence of deformable viscoelastic materials (mucus) in a confined geometry (trachea bifurcation)
Understand and model the interaction between air and viscoelastic fluids
Determine the influence of mucus on air flow rate distribution in trachea bifurcation
Develop a Learning Deep (LD) procedure to model the mucus-air interface using Deep Learning (DL) algorithms
Establish a relation between interface breakup/dynamics and rheological properties of fluids in a confined geometry

Methodology:

Experimental and numerical investigations of air-ways flows in the presence of mucus
Use of CT images to create a physical domain for numerical simulations
Rheological characterization of viscoelastic materials using shear and extensional flows
Application of Machine Learning (ML) algorithms to model viscoelastic fluid interfaces
Development of a procedure to analyze results and transform them into an algorithm compatible with AI and ML applications

Challenges:

Complex geometry of the flow domain (trachea's first bifurcation)
Complexity of viscoelastic materials and their modeling
Long computation time for numerical simulations and correlation with experiments
Design of a procedure to analyze results and transform them into an algorithm compatible with AI and ML applications

Expected Outcomes:

Determination of mucus rheological properties from image processing and air dynamics inside the trachea
Relation between mucus rheology and air distribution in the trachea's bifurcation
* Potential application to lung dysfunctions and novel technologies in medicine and engineering.