2450533
Project Grant
Overview
Grant Description
Collaborative research: CDS&E: Data-driven next-generation Lagrangian models for swarms of deforming and interacting droplets.
Sprays are formed by breaking down a liquid into tiny droplets, enabling a wide range of practical applications from fuel injection in engines to efficient cooling systems.
It is important to be able to predict how droplets in sprays deform and move while interacting with each other and the surrounding airflow.
However, accurately predicting the behavior of large numbers of droplets in sprays is challenging due to the immense computational power required.
By using advanced computer simulations to create detailed datasets and applying machine learning, this project will improve predictions of spray behavior.
These improvements could help scientists and engineers design more efficient engines, better cooling technologies, and innovative solutions to reduce environmental pollution.
The project also supports education by engaging high school students in science and engineering careers through workshops at engineering summer camps.
The project will employ the BASILISK multiphase flow solver to perform high-fidelity simulations of droplet swarms in gas flows, generating a comprehensive dataset across a wide range of key parameters, including droplet volume fraction, Reynolds, and Weber numbers.
A novel machine-learning model, based on a graph convolutional network, will be developed to capture how individual droplets and their interactions influence deformation and forces.
This model ensures consistency with physical principles, such as maintaining symmetry in motion and orientation.
The project will also use Shapley additive explanations (SHAP) analysis to interpret the machine-learning model, identify the most important factors, and create simpler models for faster predictions.
The resulting models will be made publicly available, enabling researchers to improve spray simulations in applications like combustion and spray cooling.
This award is expected to advance computational fluid dynamics and provide practical benefits for many industrial applications.
This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the foundation's intellectual merit and broader impacts review criteria.
Subawards are not planned for this award.
Sprays are formed by breaking down a liquid into tiny droplets, enabling a wide range of practical applications from fuel injection in engines to efficient cooling systems.
It is important to be able to predict how droplets in sprays deform and move while interacting with each other and the surrounding airflow.
However, accurately predicting the behavior of large numbers of droplets in sprays is challenging due to the immense computational power required.
By using advanced computer simulations to create detailed datasets and applying machine learning, this project will improve predictions of spray behavior.
These improvements could help scientists and engineers design more efficient engines, better cooling technologies, and innovative solutions to reduce environmental pollution.
The project also supports education by engaging high school students in science and engineering careers through workshops at engineering summer camps.
The project will employ the BASILISK multiphase flow solver to perform high-fidelity simulations of droplet swarms in gas flows, generating a comprehensive dataset across a wide range of key parameters, including droplet volume fraction, Reynolds, and Weber numbers.
A novel machine-learning model, based on a graph convolutional network, will be developed to capture how individual droplets and their interactions influence deformation and forces.
This model ensures consistency with physical principles, such as maintaining symmetry in motion and orientation.
The project will also use Shapley additive explanations (SHAP) analysis to interpret the machine-learning model, identify the most important factors, and create simpler models for faster predictions.
The resulting models will be made publicly available, enabling researchers to improve spray simulations in applications like combustion and spray cooling.
This award is expected to advance computational fluid dynamics and provide practical benefits for many industrial applications.
This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the foundation's intellectual merit and broader impacts review criteria.
Subawards are not planned for this award.
Awardee
Funding Goals
THE GOAL OF THIS PROGRAM IS TO SUPPORT RESEARCH PROPOSALS SPECIFIC TO "COMPUTATIONAL AND DATA-ENABLED SCIENCE AND ENGINEERING
Grant Program (CFDA)
Awarding / Funding Agency
Place of Performance
Gainesville,
Florida
32611-1941
United States
Geographic Scope
Single Zip Code
Related Opportunity
University Of Florida was awarded
Project Grant 2450533
worth $150,632
from the Division of Chemical, Bioengineering, Environmental, and Transport System in July 2025 with work to be completed primarily in Gainesville Florida United States.
The grant
has a duration of 3 years and
was awarded through assistance program 47.041 Engineering.
The Project Grant was awarded through grant opportunity Computational and Data-Enabled Science and Engineering.
Status
(Ongoing)
Last Modified 8/21/25
Period of Performance
7/15/25
Start Date
6/30/28
End Date
Funding Split
$150.6K
Federal Obligation
$0.0
Non-Federal Obligation
$150.6K
Total Obligated
Activity Timeline
Additional Detail
Award ID FAIN
2450533
SAI Number
None
Award ID URI
SAI EXEMPT
Awardee Classifications
Public/State Controlled Institution Of Higher Education
Awarding Office
490702 DIVISION OF CHEMICAL BIOENGINEERING
Funding Office
490702 DIVISION OF CHEMICAL BIOENGINEERING
Awardee UEI
NNFQH1JAPEP3
Awardee CAGE
5E687
Performance District
FL-03
Senators
Marco Rubio
Rick Scott
Rick Scott
Modified: 8/21/25