Hi I'am trying to feed west and east boundary with pressure source in a short channel with two piers in the center. The flml mainly follows the manual "8.16 Configuring ocean simulations",and turns on "/mesh_adaptivity/mesh_movement/free_surface/move_whole_mesh" and "adaptive_timestep". It blowed up at 178 secs, and a very large pressure exists at the northwest boundary corner near surface. The adaptive mesh is also sharp around there. But after changing nproc from 2 to 1 and running again. It performs very well. In another case with parallel doesn't get this problem. Does it mean running in parallel may take risk with some meshs? BC: West/East: Pressure Source (use dirichlet read from file) South/North side: no normal flow Bottom: no normal flow and drag Top: free surface Pier:no slip terrain: Thannel:length 55m, width 45m and depth 5m (2 sigma layer) pier: radius 2.1m The geo and flml contents are below. #------------ pier_tsu.geo ------------ // center coordinate xc=0; yc=0; // model center xp1=xc-6.6; yp1=yc; // south pier center 6.6 xp2=xc+6.6; yp2=yc; // north pier center 6.6 rad = 2.1; // pier radius rad = 2.1 d1w=5; // resolution at boumdary 50 d1e=5; d2=2.5; // 5 d3=1; // resolution at pier // boundary wlen=27.5; //27.5 elen=27.5; //27.5 wid=45; // 2345 // 1876 Point(1) = {xc-1*wlen, yc-0.5*wid, 0, d1w}; //sw Point(2) = {xc-1*wlen, yc+0.5*wid, 0, d1w}; //nw Point(3) = {xc-0.8*wlen, yc+0.5*wid, 0, d2};//nnw Point(4) = {xc+0.8*elen, yc+0.5*wid, 0, d2};//nne Point(5) = {xc+1*elen, yc+0.5*wid, 0, d1e}; //ne Point(6) = {xc+1*elen, yc-0.5*wid, 0, d1e}; //sw Point(7) = {xc+0.8*elen, yc-0.5*wid, 0, d2};//ssw Point(8) = {xc-0.8*wlen, yc-0.5*wid, 0, d2};//sse Line(1) = {1, 2}; Line(2) = {2, 3}; Line(3) = {3, 4}; Line(4) = {4, 5}; Line(5) = {5, 6}; Line(6) = {6, 7}; Line(7) = {7, 8}; Line(8) = {8, 1}; // WEST PIER Point(9) = {xp1, yp1, 0, d3}; Point(10) = {xp1-rad, yp1, 0, d3}; Point(11) = {xp1+rad, yp1, 0, d3}; Circle(9) = {10, 9, 11}; Circle(10) = {11, 9, 10}; // EAST PIER Point(12) = {xp2, yp2, 0, d3}; Point(13) = {xp2-rad, yp2, 0, d3}; Point(14) = {xp2+rad, yp2, 0, d3}; Circle(11) = {13, 12, 14}; Circle(12) = {14, 12, 13}; Line Loop(11) = {1,2,3,4,5,6,7,8}; //BOUNDARY Line Loop(12) = {9,10}; // SOUTH PIER Line Loop(13) = {11,12}; // NORTH PIER Plane Surface(14) = {11,12,13}; Physical Line(1) = {1}; //E Physical Line(2) = {2,3,4}; //N Physical Line(3) = {5}; //W Physical Line(4) = {6,7,8}; //S Physical Line(5) = {9,10,11,12}; // WEST EAST PIERS Physical Surface(6) = {14}; #--------- pier_tsu.flml --------- pier_tsu fluids 3 5 2 101 102 1 discontinuous 2 8 8 8 8 101 102 vtk 5 0.0 1.0 870.0 3 3 2 1e-6 1.0 1.1 9.81 0 0 -1 1.0 only first timestep 1.0e-7 0.0 6000 1 def val(X,t): import inputwave_w g=9.81 return inputwave_w.val(X, t)*g def val(X,t): g=9.81 return -0.135*g*min(1,t) 3 def val(X,t): import inputwave_e g=9.81 return inputwave_e.val(X, t)*g def val(X,t): g=9.81 return -0.135*g*min(1,t) 0.0 1 1 0.1 45 1.0e-7 0 10000 0.0 0.0 0.0 101 2 4 102 5 0 0 0 102 2.5e-3 1e-6 1000 2.5e-3

When	Who	Action	New State
2014-03-07	liu chin chi	Give more information	Open
2014-03-07	Matthew Piggott	Answer	Answered
2014-03-08	liu chin chi	Reopen	Open
2014-03-09	liu chin chi	Confirm	Solved
2014-03-09	liu chin chi	Confirm	Solved
2014-03-09	Matthew Piggott	Comment	Solved
2014-03-12	Fangxin Fang	Comment	Solved

Fluidity

History of question #244867