Welcome, guest | Sign In | My Account | Store | Cart

2D Fluid Simulation using FHP LGCA (Lattice Gas Cellular Automata)

Simulates fluid flow in a circular channel.

It works really slow but I think it can be a lot faster if it modified for NumPy and possibly Py2Exe.

But my main goal was to provide easy to understand code (not performance) anyway.

Python, 221 lines
 ``` 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221``` ```# 2D Fluid Simulation using FHP LGCA (Lattice Gas Cellular Automata) # Simulates fluid flow in a circular channel. # Particles go out from right side and enter back from left. # Reference: # Lattice Gas Cellular Automata and Lattice Boltzmann Models by Wolf-Gladrow # FB - 20140818 import math import random from PIL import Image imgx = 512; imgy = 512 # image size image = Image.new("RGB", (imgx, imgy)) pixels = image.load() # simulation parameters: tilesX = 32 tilesY = 32 n = 8 # coarse graining tile size is n by n timeSteps = 300 nodesX = tilesX * n nodesY = tilesY * n nodes = [[[0 for x in range(nodesX)] for y in range(nodesY)] for z in range(6)] obstacle = [[0 for x in range(nodesX)] for y in range(nodesY)] # insert a square obstacle in the middle for y in range(nodesY / 4): for x in range(nodesX / 4): obstacle[y + nodesY / 2 - nodesY / 8][x + nodesX / 2 - nodesX / 8] = 1 # fill-up with fluid flowing towards right for y in range(1, nodesY - 1): # do not include top/bottom walls for x in range(nodesX): if obstacle[y][x] != 1: nodes[0][y][x] = 1 for t in range(timeSteps): # run the simulation # HANDLE COLLISIONS # collisions at non-boundary nodes for y in range(1, nodesY - 1): # do not include top/bottom walls for x in range(nodesX): if obstacle[y][x] != 1: cell = [nodes[z][y][x] for z in range(6)] numParticles = sum(cell) # only 2 or 3 symmetric particle collisions implemented here if numParticles == 3: if cell[0] == cell[2] and cell[2] == cell[4]: # invert the cell contents for z in range(6): nodes[z][y][x] = 1 - cell[z] elif numParticles == 2: # find the cell of one of the particles p = cell.index(1) # its diametric opposite must occupied as well if p > 2: pass elif cell[p + 3] == 0: pass else: # randomly rotate the particle pair clockwise or # counterclockwise if random.randint(0, 1) == 0: # counterclockwise nodes[0][y][x] = cell[5] nodes[1][y][x] = cell[0] nodes[2][y][x] = cell[1] nodes[3][y][x] = cell[2] nodes[4][y][x] = cell[3] nodes[5][y][x] = cell[4] else: # clockwise nodes[0][y][x] = cell[1] nodes[1][y][x] = cell[2] nodes[2][y][x] = cell[3] nodes[3][y][x] = cell[4] nodes[4][y][x] = cell[5] nodes[5][y][x] = cell[0] # collisions along top/bottom walls (no-slip) for x in range(nodesX): cell = [nodes[z][0][x] for z in range(6)] nodes[0][0][x] = cell[3] nodes[1][0][x] = cell[4] nodes[2][0][x] = cell[5] nodes[3][0][x] = cell[0] nodes[4][0][x] = cell[1] nodes[5][0][x] = cell[2] cell = [nodes[z][nodesY - 1][x] for z in range(6)] nodes[0][nodesY - 1][x] = cell[3] nodes[1][nodesY - 1][x] = cell[4] nodes[2][nodesY - 1][x] = cell[5] nodes[3][nodesY - 1][x] = cell[0] nodes[4][nodesY - 1][x] = cell[1] nodes[5][nodesY - 1][x] = cell[2] # collisions at obstacle points (no-slip) for y in range(nodesY): for x in range(nodesX): if obstacle[y][x] == 1: cell = [nodes[z][y][x] for z in range(6)] nodes[0][y][x] = cell[3] nodes[1][y][x] = cell[4] nodes[2][y][x] = cell[5] nodes[3][y][x] = cell[0] nodes[4][y][x] = cell[1] nodes[5][y][x] = cell[2] # HANDLE MOVEMENTS nodesNew = [[[0 for x in range(nodesX)] for y in range(nodesY)] for z in range(6)] for y in range(nodesY): for x in range(nodesX): cell = [nodes[z][y][x] for z in range(6)] # propagation in the 0-direction neighbor_y = y if x == nodesX - 1: neighbor_x = 0 else: neighbor_x = x + 1 nodesNew[0][neighbor_y][neighbor_x] = cell[0] # propagation in the 1-direction if y != nodesY - 1: neighbor_y = y + 1 if y % 2 == 1: if x == nodesX - 1: neighbor_x = 1 else: neighbor_x = x + 1 else: neighbor_x = x nodesNew[1][neighbor_y][neighbor_x] = cell[1] # propagation in the 2-direction if y != nodesY - 1: neighbor_y = y + 1 if y % 2 == 0: if x == 0: neighbor_x = nodesX - 1 else: neighbor_x = x - 1 else: neighbor_x = x nodesNew[2][neighbor_y][neighbor_x] = cell[2] # propagation in the 3-direction neighbor_y = y if x == 0: neighbor_x = nodesX - 1 else: neighbor_x = x - 1 nodesNew[3][neighbor_y][neighbor_x] = cell[3] # propagation in the 4-direction if y != 0: neighbor_y = y - 1 if y % 2 == 0: if x == 0: neighbor_x = nodesX - 1 else: neighbor_x = x - 1 else: neighbor_x = x nodesNew[4][neighbor_y][neighbor_x] = cell[4] # propagation in the 5-direction if y != 0: neighbor_y = y - 1 if y % 2 == 1: if x == nodesX - 1: neighbor_x = 0 else: neighbor_x = x + 1 else: neighbor_x = x nodesNew[5][neighbor_y][neighbor_x] = cell[5] nodes = nodesNew print '%' + str(100 * t / timeSteps) # show progress # Create an image from the final state # Calculate average velocity vectors for tiles aveVelocityVectorMag = [[0.0 for x in range(tilesX)] for y in range(tilesY)] aveVelocityVectorAng = [[0.0 for x in range(tilesX)] for y in range(tilesY)] pi2 = math.pi * 2.0 dx = [math.cos(i * pi2 / 6.0) for i in range(6)] dy = [math.sin(i * pi2 / 6.0) for i in range(6)] for ty in range(tilesY): for tx in range(tilesX): vx = 0.0 vy = 0.0 for cy in range(n): for cx in range(n): for z in range(6): if nodes[z][ty * n + cy][tx * n + cx] == 1 \ and obstacle[ty * n + cy][tx * n + cx] == 0: vx += dx[z] vy += dy[z] aveVelocityVectorMag[ty][tx] = math.hypot(vx, vy) / n ** 2.0 aveVelocityVectorAng[ty][tx] = (math.atan2(vy, vx) + pi2) % pi2 for ky in range(imgy): iy = nodesY * ky / imgy jy = tilesY * ky / imgy for kx in range(imgx): ix = nodesX * kx / imgx jx = tilesX * kx / imgx if obstacle[iy][ix] == 1: # paint the obstacle(s) red = 0 grn = 0 blu = 255 else: # use vector magnitude and angle for coloring aveVelVecMag = aveVelocityVectorMag[jy][jx] aveVelVecAng = aveVelocityVectorAng[jy][jx] red = int(aveVelVecMag * 255) grn = int(aveVelVecAng / pi2 * 255) blu = 0 pixels[kx, ky] = (red, grn, blu) image.save("FHP_LGCA_2DFluidSim.png", "PNG") ```

#### 2 comments

John Park 4 years, 11 months ago

File "C:/Users/POS/Documents/Python Scripts/2D FLUID SIMULATION USING FHP LGCA.py", line 32, in <module> for y in range(nodesY / 4):

TypeError: 'float' object cannot be interpreted as an integer

I got an error as above. Please advice me!!

FB36 (author) 4 years, 11 months ago

I just downloaded the code and run it no problem. I am guessing you maybe using Python 3.x. I had written this code using Python 2.7.x. They are not fully compatible versions. Also you would need to get and install PIL library for it to work.

 Created by FB36 on Mon, 18 Aug 2014 (MIT)