Summary
Multiphase flow (MF) is the simultaneous flow of materials with two or more thermodynamic phases. MF occurs in numerous settings: bioengineering, conventional and nuclear power plants, oil and gas production and transport, pharmaceutical industry, combustion engines, chemical industry, flows inside the human body, biological industry, and process technology, to name a few. Researchers use experimental and theoretical techniques to study MF. Experimental techniques are usually restricted to smaller domains or laboratory scales due to very high costs; in addition, experiments in realistic conditions are very difficult to manage. On the other hand, theory usually requires numerical computation, which is very time-consuming for realistic MF problems.
The objective of this study is to develop a novel machine learning (ML)-based hidden fluid dynamics approach for MF. This innovative method combines ML and fluid dynamics by means of information accessible from photographs of flow visualizations and governing equations. The developed approach will considerably reduce the cost and time required to analyse MF systems. It will unlock many opportunities to improve the design and efficiency of existing and future MF systems, thus lowering design/operational costs.
This will be the first study to develop a ML-based model for MF. The novelty of the proposal is that it will open the opportunity to explore, via a relatively cheap and fast computation method, the vast number of design parameter variations normally needed for the optimisation of MF systems (e.g. efficiency, pressure drop, etc.). This contrasts with expensive and time-consuming experiments and time-inefficient parameter studies involving traditional computational fluid dynamics.
The proposal will be carried out in three stages:
1) To develop a ML-based model for MF.
2) To test the model with several benchmark MF problems.
3) To use the model to improve design and reduce cost in an actual practical environment.
The objective of this study is to develop a novel machine learning (ML)-based hidden fluid dynamics approach for MF. This innovative method combines ML and fluid dynamics by means of information accessible from photographs of flow visualizations and governing equations. The developed approach will considerably reduce the cost and time required to analyse MF systems. It will unlock many opportunities to improve the design and efficiency of existing and future MF systems, thus lowering design/operational costs.
This will be the first study to develop a ML-based model for MF. The novelty of the proposal is that it will open the opportunity to explore, via a relatively cheap and fast computation method, the vast number of design parameter variations normally needed for the optimisation of MF systems (e.g. efficiency, pressure drop, etc.). This contrasts with expensive and time-consuming experiments and time-inefficient parameter studies involving traditional computational fluid dynamics.
The proposal will be carried out in three stages:
1) To develop a ML-based model for MF.
2) To test the model with several benchmark MF problems.
3) To use the model to improve design and reduce cost in an actual practical environment.
Unfold all
/
Fold all
More information & hyperlinks
| Web resources: | https://cordis.europa.eu/project/id/101110330 |
| Start date: | 01-08-2024 |
| End date: | 31-07-2026 |
| Total budget - Public funding: | - 215 534,00 Euro |
Cordis data
Original description
Multiphase flow (MF) is the simultaneous flow of materials with two or more thermodynamic phases. MF occurs in numerous settings: bioengineering, conventional and nuclear power plants, oil and gas production and transport, pharmaceutical industry, combustion engines, chemical industry, flows inside the human body, biological industry, and process technology, to name a few. Researchers use experimental and theoretical techniques to study MF. Experimental techniques are usually restricted to smaller domains or laboratory scales due to very high costs; in addition, experiments in realistic conditions are very difficult to manage. On the other hand, theory usually requires numerical computation, which is very time-consuming for realistic MF problems.The objective of this study is to develop a novel machine learning (ML)-based hidden fluid dynamics approach for MF. This innovative method combines ML and fluid dynamics by means of information accessible from photographs of flow visualizations and governing equations. The developed approach will considerably reduce the cost and time required to analyse MF systems. It will unlock many opportunities to improve the design and efficiency of existing and future MF systems, thus lowering design/operational costs.
This will be the first study to develop a ML-based model for MF. The novelty of the proposal is that it will open the opportunity to explore, via a relatively cheap and fast computation method, the vast number of design parameter variations normally needed for the optimisation of MF systems (e.g. efficiency, pressure drop, etc.). This contrasts with expensive and time-consuming experiments and time-inefficient parameter studies involving traditional computational fluid dynamics.
The proposal will be carried out in three stages:
1) To develop a ML-based model for MF.
2) To test the model with several benchmark MF problems.
3) To use the model to improve design and reduce cost in an actual practical environment.
Status
SIGNEDCall topic
HORIZON-MSCA-2022-PF-01-01Update Date
12-03-2024
Images
No images available.
Geographical location(s)
