Jump to content
GMS, SMS, and WMS User Forum
heimo

RMA2 wetting/drying problems

Recommended Posts

Hi!

I have a beginners question about the wetting/drying in RMA2. I am about to

model a planned artificial lake here in northern Finland and did generate the

mesh for the planned area. To examine the mesh quality I withdrew water from

the mesh (single material) with BQ card to empty (and later fill) the lake.

All worked fine during the withdrawal and the water elevation decreased at

the rate calculated by BQ discharge and surface area information. However,

as I try to fill the lake by changing the BQT value to positive, the

solution sooner or later diverges. I have tried different values for

diffusion (direct assignment and Peclet), roughness coefficient (direct and

automatic assignment), DE card depth limits etc. without success.

To see if the divergence had something to do with bottom topography, I created

a simple geometry with parabolic depth distribution. The same problem however

applies to this geometry as well. During the rewetting, the solution diverges.

Sometimes I manage to run the simulation through but the water level does not

rise as it should. At the perimeter of the wet area the grid cells simply

switch back and forth between wet and dry and the water elevation remains much the same.

The simplified model is 10x10 km (some 6.2 by 6.2 miles) and has maximum

depth of 10 m (30 ft) at the centre. The elements are quadrilateral 25x25 (

400 metre grid spacing, tried also with refined grid of 200 m spacing). I calculate

first 300 hours with BQT value of -8.8E-6 and from 300 to 600 hours with

+8.8E-6. At 300 hours the water elevation has dropped from the initial +0.5 m

to -8.9 m but only rises by approximately one meter (to -7.8m) by the end of the simulation.

I would be grateful if you had time to help me with this problem and tell

which factors I should consider. It must be something simple I'm missing here.

I am ready to send the necessary files of the simplified problem when needed.

Yours, Heimo

  • Like 1

Share this post


Link to post
Share on other sites

You may want to try using marsh porosity. It will keep elements as active even as the water drops below the elevations of the nodes based upon parameters you set. When using marsh porosity, make sure you look at the depths of the elements rather than the active/inactive information.

If you continue to experience problems with wetting and drying, you may want to look into using TUFLOW. TUFLOW handles wetting and drying better than RMA2 or FESWMS.

Share this post


Link to post
Share on other sites

Hi Rusty and thank you for your reply!

I now manage to drain and refill the model domain but am still somewhat lost with this issue. First I must say that unfortunately I'm stuck with RMA2 as I need to later use RMA4 as well (tuflow solution surely can't be used for this?).

Do I understand it correctly that marh porosity when used alone always leaves some water to the element and in order to remove elements on which all nodes will be dry, elemental elimination (DE) must be added? For the lake to be modelled, some 80 % of the surface area will be dry during the minimun water stage so accordingly most elements should be dry then.

Using the previously described parabolic topography I tested combination of DA and DE cards with lowering water elevation. On the borders of the domain water depth falls to approximately 0.1 m but no elements or nodes are (to my understanding) removed (dried/rewetted) although DSET in DE card is 0.25 m. Here are some parts from the screen dump:

After gfgen phase:

GFGEN [binary] information concluded.

$M 1

GCL 1 1175 1173 -1

IC 15.000 0.15200 0.76000E-01

EX 1 2500.00000 2500.00000

EY 1 2500.00000 2500.00000

DE 0.25000 0.50000 2

DA 1 2.0000 0.60000 0.20000E-01 2.0000

HN 1 0.045

HNT 1 0.045

TI 10 10 0.00010000 0.00010000

TR 0 0 0 0 0

-------> INCREMENT IQGEN= 1

=+= KOUNT # OF END-CARDS MARKING END OF TS = 1

PREPARE FOR S.S. PROCESS ... IDNOPT = 1

*** RMA2 CARD INPUT COMPLETE WITH 0 ERRORS ***

Going through the dump file I found out that for all time steps the number of elements/nodes dried or rewetted is zero. Here's a clip from the time of minimum water elevation (when, according to SMS, there are 253 elements and 699 nodes with water depth less than 0.25 metres).

RESULTS AT THE END OF 53 TIME STEPS...

TOTAL TIME = 318.000000 HOURS... ITERATION CYCLE = 2

CONVERGENCE PARAMETERS

DF AVG CHG MAX CHG NODAL_LOCATION

1 0.0000 -0.0020 1175 X-VEL

2 0.0000 -0.0010 1169 Y-VEL

3 0.0001 -0.0038 1175 DEPTH

ACTIVE NODAL STATISTICS FOR THIS ITERATION

NODE XVEL-MAX NODE XVEL-MIN NODE YVEL-MAX NODE YVEL-MIN

1156 0.203 1153 -0.204 1118 0.018 1168 -0.459

NODE DEPTH-MAX NODE DEPTH-MIN NODE ELEV-MAX NODE ELEV-MIN AVE-ELEV NODES ACTIVE/TOTAL

441 2.387 1159
0.107
1 7.429 1175 7.156 7.412 1259 / 1262

REWETM. Re-Set ITRACE to original value

MARSH POROSITY WET/DRY Update Summary, Simulation hr= 318.000 Iteration= 3

Elems dried= 0 Elems rewet= 0

Nodes dried= 0 Nodes rewet= 0

SMS plots velocity vectors to all elements in all time steps which makes me believe all elements are wet. Am I correct in that no elements have been removed from the calculation? Surely I must be able to remove the dry elements. Won't those elements with some 0.1 metres of water otherwise be included in concentration modeling with RMA4.

So briefly, am I still doing something wrongly?

Yours,

Heimo

Share this post


Link to post
Share on other sites

You are right that TUFLOW results are not easy to use with RMA4.

It is possible that all of the elements are included in the calculations for the entire run when using marsh porosity even though some elements should dry. As you mention, this may or may not cause problems with RMA4. The idea is that marsh porosity allows you to reduce the conveyance through area of the mesh that should dry to a level small enough that the overall results are acceptable.

Even though marsh porosity keeps elements active after they would normally go dry, these areas are assigned negative depths. You can use the contour settings to identify which areas would normally be dry. With the depth dataset active, turn on color filled contours set to use a fixed range and set the minimum range to 0.0 and uncheck fill below. This will only contour areas of the mesh that are truly wet.

Good luck.

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...

×
×
  • Create New...