Nektar++
Functions
ExtractCriticalLayerFunctions.cpp File Reference
#include <cstdio>
#include <MultiRegions/ExpList.h>

Go to the source code of this file.

Functions

void Computestreakpositions (MultiRegions::ExpListSharedPtr &streak, Array< OneD, NekDouble > &xc, Array< OneD, NekDouble > &yc, NekDouble cr, NekDouble trans)
 

Function Documentation

◆ Computestreakpositions()

void Computestreakpositions ( MultiRegions::ExpListSharedPtr streak,
Array< OneD, NekDouble > &  xc,
Array< OneD, NekDouble > &  yc,
NekDouble  cr,
NekDouble  trans 
)

Definition at line 42 of file ExtractCriticalLayerFunctions.cpp.

46 {
47  int i;
48  int npts = xc.size();
49 
50  int nq = streak->GetTotPoints();
51  Array<OneD, NekDouble> derstreak(nq);
54  streak->GetCoords(x, y);
55 
56  streak->BwdTrans(streak->GetCoeffs(), streak->UpdatePhys());
57  streak->PhysDeriv(MultiRegions::eY, streak->GetPhys(), derstreak);
58 
59  // set intiial xc to be equispaced over mesh and yc to be zero
60  NekDouble x_max = Vmath::Vmax(nq, x, 1);
61  NekDouble x_min = Vmath::Vmin(nq, x, 1);
62 
63  for (i = 0; i < npts; ++i)
64  {
65  xc[i] = x_min + (x_max - x_min) * i / ((NekDouble)(npts - 1));
66  yc[i] = 0.0;
67  }
68 
69  int elmtid, offset, cnt;
70  NekDouble U, dU;
71  NekDouble F;
72  NekDouble ConvTol = 1e-9;
73  NekDouble CoordTol = 1e-5;
74  int maxiter = 100;
75  Array<OneD, NekDouble> coord(2);
76 
77  // Do Newton iteration on y direction
78  cerr << "[";
79  for (int e = 0; e < npts; e++)
80  {
81  coord[0] = xc[e];
82  coord[1] = yc[e];
83 
84  if (!(e % 10))
85  {
86  cerr << ".";
87  }
88 
89  F = 1000;
90  cnt = 0;
91  while ((abs(F) > ConvTol) && (cnt < maxiter))
92  {
93  elmtid = streak->GetExpIndex(coord, CoordTol);
94  offset = streak->GetPhys_Offset(elmtid);
95 
96  U = streak->GetExp(elmtid)->PhysEvaluate(coord, streak->GetPhys() +
97  offset);
98  dU =
99  streak->GetExp(elmtid)->PhysEvaluate(coord, derstreak + offset);
100 
101  coord[1] = coord[1] - (U - cr) / dU;
102 
103  F = U - cr;
104  cnt++;
105  }
106  ASSERTL0(cnt < maxiter, "Failed to converge Newton iteration");
107 
108  yc[e] = coord[1];
109  }
110  cerr << "]" << endl;
111 
112  if (trans != NekConstants::kNekUnsetDouble)
113  {
114  // output to interface file
115  FILE *fp = fopen("interfacedat.geo", "w");
116 
117  NekDouble y_max = Vmath::Vmax(nq, y, 1);
118  NekDouble y_min = Vmath::Vmin(nq, y, 1);
119 
120  cnt = 1;
121  fprintf(fp, "Point(%d)={%12.10lf,%12.10lf,0,1.0}; \n", cnt++, x_min,
122  y_min);
123  fprintf(fp, "Point(%d)={%12.10lf,%12.10lf,0,1.0}; \n", cnt++, x_max,
124  y_min);
125  fprintf(fp, "Point(%d)={%12.10lf,%12.10lf,0,1.0}; \n", cnt++, x_max,
126  y_max);
127  fprintf(fp, "Point(%d)={%12.10lf,%12.10lf,0,1.0}; \n", cnt++, x_min,
128  y_max);
129 
130  for (i = 0; i < npts; ++i)
131  {
132  fprintf(fp, "Point(%d)={%12.10lf,%12.10lf,0,1.0}; \n", cnt++,
133  xc[i], yc[i]);
134  }
135 
136  fclose(fp);
137 
138  // output to interface_up file as bend of vertical shift and 45 degrees
139  // shift
140  fp = fopen("interfacedat_up.geo", "w");
141 
142  NekDouble nx, ny, norm;
143 
144  fprintf(fp, "Point(%d)={%12.10lf,%12.10lf,0,1.0}; \n", cnt++, xc[0],
145  yc[0] + trans);
146 
147  for (i = 1; i < npts - 1; ++i)
148  {
149  norm = sqrt((xc[i + 1] - xc[i - 1]) * (xc[i + 1] - xc[i - 1]) +
150  (yc[i + 1] - yc[i - 1]) * (yc[i + 1] - yc[i - 1]));
151  nx = (yc[i - 1] - yc[i + 1]) / norm;
152  ny = (xc[i + 1] - xc[i - 1]) / norm;
153 
154  fprintf(fp, "Point(%d)={%12.10lf,%12.10lf,0,1.0}; \n", cnt++,
155  xc[i] + nx * trans, yc[i] + ny * trans);
156  }
157 
158  fprintf(fp, "Point(%d)={%12.10lf,%12.10lf,0,1.0}; \n", cnt++,
159  xc[npts - 1], yc[npts - 1] + trans);
160 
161  // output to interface_up file as bend of vertical shift and 45 degrees
162  // shift
163  fp = fopen("interfacedat_dn.geo", "w");
164 
165  trans = -trans;
166 
167  fprintf(fp, "Point(%d)={%12.10lf,%12.10lf,0,1.0}; \n", cnt++, xc[0],
168  yc[0] + trans);
169 
170  for (i = 1; i < npts - 1; ++i)
171  {
172  norm = sqrt((xc[i + 1] - xc[i - 1]) * (xc[i + 1] - xc[i - 1]) +
173  (yc[i + 1] - yc[i - 1]) * (yc[i + 1] - yc[i - 1]));
174  nx = (yc[i - 1] - yc[i + 1]) / norm;
175  ny = (xc[i + 1] - xc[i - 1]) / norm;
176 
177  fprintf(fp, "Point(%d)={%12.10lf,%12.10lf,0,1.0}; \n", cnt++,
178  xc[i] + nx * trans, yc[i] + ny * trans);
179  }
180 
181  fprintf(fp, "Point(%d)={%12.10lf,%12.10lf,0,1.0}; \n", cnt++,
182  xc[npts - 1], yc[npts - 1] + trans);
183  }
184 }
#define ASSERTL0(condition, msg)
Definition: ErrorUtil.hpp:215
static const NekDouble kNekUnsetDouble
double NekDouble
T Vmin(int n, const T *x, const int incx)
Return the minimum element in x - called vmin to avoid conflict with min.
Definition: Vmath.cpp:1050
T Vmax(int n, const T *x, const int incx)
Return the maximum element in x – called vmax to avoid conflict with max.
Definition: Vmath.cpp:945
scalarT< T > abs(scalarT< T > in)
Definition: scalar.hpp:298
scalarT< T > sqrt(scalarT< T > in)
Definition: scalar.hpp:294

References tinysimd::abs(), ASSERTL0, Nektar::MultiRegions::eY, Nektar::NekConstants::kNekUnsetDouble, tinysimd::sqrt(), Vmath::Vmax(), and Vmath::Vmin().