Mass Diffusion Apparatus
This YouTube video is a great video showing how gas diffusion occurs through simulation and the effects of temperature.
The following test rig shown on Youtube shows some useful methods to find mass diffusion.
Computational Fluid Dynamics Experimental Apparatus of the Future
I believe that Stafford experiment it can be used to mimic study cases of Ficks Law. This experimental set up can also be used to calculate different diffusion coefficients for different incompressible fluids in addition to the ability to use heat or pressure in the experiment. I can see that this experiment can also be used to model fuel cells at more advanced stages of the development program. Then CFD can be applied to cover the numerical side of the project. Different values used in turbulence models can be extracted such as Eddie dissipation rates or even showing Eddies sizes occurring when using a very viscous fluid with a less viscous one. Studying buoyancy effects are also applicable for the Stafford experiment.
New innovations to the previous experiment can be added, where an inflow (fluid 1 at the top, fluid 2 at the bottom) and outflow is applied to the Stafford experimental set up. Then playing around with the (currents or voltage) frequencies and magnitude this could apply turbulence to the flow. The fluid of concern for the study would be seeded and a high frame rate camera and PIV can be used to capture the occurring flow pattern and velocity fields. The data can be stored on a computer that comes with the experimental setup. This is when a CFD software would be used to model the studied flow and then the researcher is shown the difference between the numerical and experimental results. Pre-programmed turbulence models can be applied for the study of the experiment not forgetting applying different grid types for the study.
Another possible suggestion that can be developed for use with undergraduate students is having a mini wind tunnel where a computer linked to a CFD solver will present on the spot the difference between computational methods and extracted experimental data for standard studied shapes in air tunnels such as spheres, cones and mini car models. The wind tunnel would be provided with some kind of automated Pitot tube to measure the velocity profiles at specific points and then the data is curve fitted for the experimental part and compared with the solved solution for the Navier-Stokes equation.
I can also foresee that this can be applicable on combustion modelling cases and species prediction such as a case studying premixed or non premixed flames through the use of a Bunsen burner, PIV, Camera and flow seeding then showing through the use of a CFD combustion model calculation how well will it perform in predicting the different parameters of interest.
This can also be used for the study of boundary layers, Weir flow studies to calculate and predict friction losses, flows in tubes; where again an automated or a manually operated Pitot tube is used.
MATLAB Simulink package can be used to run the different experiments.
Another experiment that I propose is to use a computer to conduct numerical experiments on it to show how a number of used cores could affect calculation time in relation to the use of different types of mesh, turbulence models, etc...
New innovations to the previous experiment can be added, where an inflow (fluid 1 at the top, fluid 2 at the bottom) and outflow is applied to the Stafford experimental set up. Then playing around with the (currents or voltage) frequencies and magnitude this could apply turbulence to the flow. The fluid of concern for the study would be seeded and a high frame rate camera and PIV can be used to capture the occurring flow pattern and velocity fields. The data can be stored on a computer that comes with the experimental setup. This is when a CFD software would be used to model the studied flow and then the researcher is shown the difference between the numerical and experimental results. Pre-programmed turbulence models can be applied for the study of the experiment not forgetting applying different grid types for the study.
Another possible suggestion that can be developed for use with undergraduate students is having a mini wind tunnel where a computer linked to a CFD solver will present on the spot the difference between computational methods and extracted experimental data for standard studied shapes in air tunnels such as spheres, cones and mini car models. The wind tunnel would be provided with some kind of automated Pitot tube to measure the velocity profiles at specific points and then the data is curve fitted for the experimental part and compared with the solved solution for the Navier-Stokes equation.
I can also foresee that this can be applicable on combustion modelling cases and species prediction such as a case studying premixed or non premixed flames through the use of a Bunsen burner, PIV, Camera and flow seeding then showing through the use of a CFD combustion model calculation how well will it perform in predicting the different parameters of interest.
This can also be used for the study of boundary layers, Weir flow studies to calculate and predict friction losses, flows in tubes; where again an automated or a manually operated Pitot tube is used.
MATLAB Simulink package can be used to run the different experiments.
Another experiment that I propose is to use a computer to conduct numerical experiments on it to show how a number of used cores could affect calculation time in relation to the use of different types of mesh, turbulence models, etc...
Unless otherwise noted, all content on this site is @Copyright by Ahmed Al Makky 2012-2014 - http://cfd2012.com