This is a great video explaining the differences between the turbulence models used in RAS. This also explains why SRH-2D dampens the velocity field when horizontal secondary currents/eddies are experienced. Is there a reason why the k-epsilon turbulence model was not incorporated? This may not be a good model to use for large scale reaches but may be good to have for "smaller" scale type projects.
It is possible to have more details about the example presented as: Sudden Expansion Lab Experiment ? I recreated the model with HEC-RAS 2D, but I don't understand whether the image presented at min 4:00 is a view from above of half the channel or of the entire channel. I modeled with HEC-RAS as in the image, but vortices are also formed on the upper wall, and the speed trend does not correspond to that presented with both conservative and non-conservative turbulence models.
Does aligning the mesh cell faces to be perpendicular and parallel to flow make it easier to calibrate longitudinal and transverse mixing? Longitudinal mixing is described as in line with flow mixing, but what if your mesh cell faces aren't aligned with flow in a structured mesh with a stream crossing the cells diagonally?
This could make a difference, but we suspect the difference would be minor. I have some data sets I could test this with...we'll add it to the to-do list. But for now, I try to align my cells with the channel banks. If it is on a bend you will still get some advection but a straight flume will be just diffusion/dispersion/turbulence.
A question: can we run the turbulence model without a sediment transport simulation.I mean just for a 2D inondation mapping ? or not possible because we need to have solutes to solve the equations for the advection.
Yes, you can run turbulence for a simple hydraulic model. It will affect the velocity distribution across the channel. So if you are interested in detailed velocities it can help. But it is essential in sediment because we are so dependent on detailed veleocities.
Nice!!! Love how both you and Alex explain these complex features in a very easy to understand way. Thanks!
Great explanation regarding HEC-RAS turbulence default values!
This is a great video explaining the differences between the turbulence models used in RAS. This also explains why SRH-2D dampens the velocity field when horizontal secondary currents/eddies are experienced. Is there a reason why the k-epsilon turbulence model was not incorporated? This may not be a good model to use for large scale reaches but may be good to have for "smaller" scale type projects.
Thank you Dr.Gibson
Thanks a bunch! Super informative!
Very good explanation
Fantastic work!
Thank you for sharing
It is possible to have more details about the example presented as: Sudden Expansion Lab Experiment ? I recreated the model with HEC-RAS 2D, but I don't understand whether the image presented at min 4:00 is a view from above of half the channel or of the entire channel. I modeled with HEC-RAS as in the image, but vortices are also formed on the upper wall, and the speed trend does not correspond to that presented with both conservative and non-conservative turbulence models.
We included the Weis data set with the 6.6 release and are going to be posting the full 2D sediment class with videos in the coming year.
Does aligning the mesh cell faces to be perpendicular and parallel to flow make it easier to calibrate longitudinal and transverse mixing? Longitudinal mixing is described as in line with flow mixing, but what if your mesh cell faces aren't aligned with flow in a structured mesh with a stream crossing the cells diagonally?
This could make a difference, but we suspect the difference would be minor. I have some data sets I could test this with...we'll add it to the to-do list. But for now, I try to align my cells with the channel banks. If it is on a bend you will still get some advection but a straight flume will be just diffusion/dispersion/turbulence.
A question: can we run the turbulence model without a sediment transport simulation.I mean just for a 2D inondation mapping ? or not possible because we need to have solutes to solve the equations for the advection.
Yes, you can run turbulence for a simple hydraulic model. It will affect the velocity distribution across the channel. So if you are interested in detailed velocities it can help. But it is essential in sediment because we are so dependent on detailed veleocities.