Sediment Scouring
 

Hello,

I would like to know if there is any limitation in the sediment scouring model to simulate different packed sediments with different species in the same experiment.

I am using one fluid, without air entrainment, and first order solvers because second order solvers stopped the calculation.

I have a ground with two layers (one over the other). However, only the first one is scoured.

Best regads

Simulations can have up to 10 species of sediment (as of v9.4 and later). In your case you would need two.

Thanks Mr Burnham for your reply.

I have two layers with different species (sand with two species and crushed stone with two species). Sand is on top of crushed stone.

As you can see in the pictures, only the first one is scoured. The second one looks like a solid.
If I quit the sand layer, the second layer is scoured without any problem.

Using 8 cores, 100 seconds of simulation are calculated in less than 2 hours (scouring process begin with 5-10 seconds).

Best regards

The sediment physics model has been undergoing changes with pretty much every version since it was introduced: mostly improvements. There are a few possibilities why the scour might stop at a layer change.

The most likely reason is due to the hiding/exposure function: when there is more than 1 species, FLOW-3D calculates a hiding/exposure adjustment to the critical Shields parameter in each cell (and for each species in that cell) based on the percentage of finer and coarser sediments in the cell (smaller sediments are 'hidden' by larger ones, and so are harder to move, and larger sediments are 'exposed' relative to the smaller ones and are easier to move. The method used is that of Egiazaroff (1965), which has an asymptote when the ratio of species diameters is large (like sand and gravel), which causes the Critical Shields Parameter of the smaller particles (sand) to get very large and completely inhibit scouring.

Starting in FLOW-3D v9.4.5, this equation includes the fix of Ashida & Michiue (1972), which corrects the problem and eliminates the asymptote. My guess is that if you are using a version earlier than 9.4.5, you're seeing the Critical Shields Parameter get very high at the sand/gravel interface because of the asymptote in the original Egiazaroff equation. Updating to the current version (10.1) will fix this.

We have 10.0 version
we will try to use fix Critical Shields Numbers, changing in the different time-steps
Thanks for your advices

Dear JBurnham
i am a Flow3D 11.04 Version user. in simulating the sand bed around a bridge pier the sediments are defined as Flow3D 10, but i have a problem in this new version, after about 200s, i see no scouring and the sand bed around the pier look likes a solid.
can you help me?