Multiphase stabilization and mass-momentum consistency updates #1265

wjhorne · 2024-06-10T17:46:56Z

This PR contains all of the multiphase work that was used to successfully run hybrid cases of waves passing over solid bodies. The main changes include

mass-momentum consistency via modifying the mass flux when vof is used. This was found to be a substantial improvement for moving interfaces over long runs.
upwinding stabilization with switching in the momentum equation at interfaces. This was a critical addition to ensure stability for high CFLs and harsh flow events that are encountered when waves pass over bodies.

…nsistency

wjhorne · 2024-06-20T21:00:58Z

@psakievich Do you know anyone that could take a look at this to start getting the ball rolling on getting it in?

psakievich · 2024-06-25T20:30:01Z

I will try to take a look but I'm not sure how timely I can be. @BumseokLee would you also be willing to take a look? @mbkuhn your input would also be helpful. Also @rcknaus if you have some spare time your input is always appreciated. We have a P/T you can charge for your time if needed.

BumseokLee · 2024-06-26T16:36:48Z

Hi Phil,
I am on a business trip until July 6. Also, I don’t have experiences with the multiphase stabilization or mass-momentum consistency in Nalu-Wind. I think it would be better to assign someone else to the work.
Bumseok

psakievich · 2024-06-26T18:13:08Z

@wjhorne will you add some regression tests that utilize the user functions you've added?

marchdf

Great work! A bunch of minor comments.

include/edge_kernels/VOFAdvectionEdgeAlg.h

marchdf · 2024-06-27T14:12:01Z

include/user_functions/DropletVelocityAuxFunction.h

+  virtual ~DropletVelocityAuxFunction() {}
+
+  using AuxFunction::do_evaluate;
+  virtual void do_evaluate(


Kind of unfortunate we call this with 2 verbs ;) Probably not relevant to this PR.

It is strange for sure. There are a few places where patterns like this show up. There is also stuff like EquationSystem.C vs EquationSystems.C which can be confusing when working in the code. Likely worth a pass through the code to fix them down the road.

marchdf · 2024-06-27T14:13:45Z

src/LowMachEquationSystem.C

@@ -165,6 +165,9 @@
 #include <user_functions/KovasznayVelocityAuxFunction.h>
 #include <user_functions/KovasznayPressureAuxFunction.h>

+#include <user_functions/DropletVelocityAuxFunction.h>


Any reason for the ordering here? Want to group the multiphase aux functions together?

This got re-arranged during a conflict. I'll clean it up.

marchdf · 2024-06-27T14:18:30Z

src/VolumeOfFluidEquationSystem.C

@@ -270,12 +281,12 @@ VolumeOfFluidEquationSystem::register_inflow_bc(

  stk::mesh::MetaData& meta_data = realm_.meta_data();

-  // register boundary data; gamma_bc
+  // register boundary data; vof_bc


Should we just remove these comments (and the following)? The code is pretty explicit I think.

I purged a bunch of these comments. We should probably do a pass through the code base to get more of them since this is quite common throughout the code.

marchdf · 2024-06-27T14:20:00Z

src/VolumeOfFluidEquationSystem.C

        ScalarFieldType* density_ = realm_.meta_data().get_field<double>(
          stk::topology::NODE_RANK, "density");
-        std::vector<double> userSpec(1);
+        std::vector<double> userSpec(3);


we could initialize this with an initializer list instead of like this.

I agree and changed this over

marchdf · 2024-06-27T14:23:44Z

src/edge_kernels/MomentumEdgeSolverAlg.C

+      // widths that are ~2 cells thick
+      DblType density_upwinding_factor = 1.0;
+      DblType alphaUpw_w_vof = alphaUpw;
+      DblType om_alphaUpw_w_vof = 1.0 - alphaUpw_w_vof;


We get so many questions about upwinding in nalu-wind and all the settings. It feels like we should document these extra modifications?

Along those lines. I would be in favor of adding a warning or throwing an error if the upwinding is put in a state that you have discovered is unstable. I'm assuming pure central will not work with VOF? Or is you modification below sufficient to ensure that the local upwinding is sufficient? Not knowing is also an acceptable answer, but I would like a clear guide line for the best practice, preferably in the code that is checked at runtime.

The documentation should definitely reflect the changes here. Our general finding is that pure central at the multiphase interface does not work well for VOF when you have high density ratios and strong flow dynamics.

The modifications here respect the users settings except explicitly at the multiphase interface when VOF is turned on. I think, at best, an appropriate warning would be to inform the user that by enabling VOF you have implicitly turned on upwinding at the interface.

marchdf · 2024-06-27T14:25:13Z

src/edge_kernels/MomentumEdgeSolverAlg.C

-                                ? (alphaUpw * uIpL[i] + om_alphaUpw * uiIp)
-                                : (alphaUpw * uIpR[i] + om_alphaUpw * uiIp);
+        const DblType uiUpw =
+          (mdot > 0.0) ? (alphaUpw_w_vof * uIpL[i] + om_alphaUpw_w_vof * uiIp)


the suffix _vof is confusing to me. Can we keep it as it is and just modify the value if vof is on? My thinking is that if I read this code I am thinking that it is vof specific.

Yeah I agree. It seems like the vof changes are to modifying the general upwinding? If there is not a second set of behavior that requires these variables for vof then I think it makes sense to just remove these variables and modify the original as needed when VOF is active.

It ends up being identical to alphaUpw and om_alphaUpw if a non-VOF case is run, or you are not at the interface with VOF. I'll change over the naming.

marchdf · 2024-06-27T14:29:05Z

src/user_functions/FlatDensityAuxFunction.C

+    const double z = coords[2];
+    const double interface_thickness = 0.015;
+
+    fieldPtr[0] = 0.0;


probably don't need this one, Just make the line after than a plain =?

marchdf · 2024-06-27T14:29:41Z

src/user_functions/FlatDensityAuxFunction.C

+    fieldPtr[0] += -0.5 * (std::erf(y / interface_thickness) + 1.0) + 1.0;
+
+    // air-water
+    fieldPtr[0] = 1000.0 * fieldPtr[0] + (1.0 - fieldPtr[0]);


this is confusing? Maybe make the first assignement a variable and then use it here?

I will change this over. There was a lot of fast iteration on this function since it was used for a lot of our method development resulting in the cruft you see. I'll scan through some of the other user functions to make sure they don't have similar issues.

marchdf · 2024-06-27T14:30:20Z

src/user_functions/SloshingTankPressureAuxFunction.C

+    const double z0 = coords[2];
+    const double z1 = domain_height_;
+
+    // These need to come from elsewhere


yeah it should. Before we get into a mess like amr-wind and density.

Constants we should stash during construction. There is more freedom on the API there and we can glob them into a type specific struct.

This user function was added during some of the testing of the pressure behavior of some fully AMR-wind/Nalu-Wind coupled cases. Specifying the pressure is not necessary for this case. I will remove it completely

wjhorne · 2024-07-18T15:42:16Z

@wjhorne will you add some regression tests that utilize the user functions you've added?

We have fully-coupled tests that utilize the user functions. The only exception is the droplet case, which is also covered by VOFDroplet in our regression tests.

wjhorne · 2024-07-18T16:11:22Z

I pushed in changes addressing all of the review here. Let me know how it looks and I'll keep iterating

whorne and others added 5 commits May 21, 2024 12:41

Changes

0c03c7b

Rebase VOF work

7f25584

Multiphase with switched upwinding stabilization and mass-momentum co…

a1b14ad

…nsistency

Remove unused level set data

4f5a078

Merge branch 'master' into multiphase_upwinding

27101d0

psakievich self-requested a review June 25, 2024 20:30

marchdf reviewed Jun 27, 2024

View reviewed changes

psakievich requested a review from itopcuoglu June 28, 2024 18:06

Merge branch 'Exawind:master' into multiphase_upwinding

247d582

Fix user functions and general cleaning

20b88e1

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Multiphase stabilization and mass-momentum consistency updates #1265

Multiphase stabilization and mass-momentum consistency updates #1265

wjhorne commented Jun 10, 2024

wjhorne commented Jun 20, 2024

psakievich commented Jun 25, 2024

BumseokLee commented Jun 26, 2024

psakievich commented Jun 26, 2024

marchdf left a comment

marchdf Jun 27, 2024

wjhorne Jul 18, 2024

marchdf Jun 27, 2024

wjhorne Jul 18, 2024

marchdf Jun 27, 2024

wjhorne Jul 18, 2024

marchdf Jun 27, 2024

wjhorne Jul 18, 2024

marchdf Jun 27, 2024

psakievich Jun 27, 2024

wjhorne Jul 18, 2024

marchdf Jun 27, 2024

psakievich Jun 27, 2024

wjhorne Jul 18, 2024

marchdf Jun 27, 2024

marchdf Jun 27, 2024

wjhorne Jul 18, 2024

marchdf Jun 27, 2024

psakievich Jun 27, 2024

wjhorne Jul 18, 2024

wjhorne commented Jul 18, 2024

wjhorne commented Jul 18, 2024

Multiphase stabilization and mass-momentum consistency updates #1265

Are you sure you want to change the base?

Multiphase stabilization and mass-momentum consistency updates #1265

Conversation

wjhorne commented Jun 10, 2024

wjhorne commented Jun 20, 2024

psakievich commented Jun 25, 2024

BumseokLee commented Jun 26, 2024

psakievich commented Jun 26, 2024

marchdf left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

wjhorne commented Jul 18, 2024

wjhorne commented Jul 18, 2024