htrdr-combustion.1.in (19972B)
1 .\" Copyright (C) 2018-2019, 2022-2025 Centre National de la Recherche Scientifique 2 .\" Copyright (C) 2020-2022 Institut Mines Télécom Albi-Carmaux 3 .\" Copyright (C) 2022-2025 Institut Pierre-Simon Laplace 4 .\" Copyright (C) 2022-2025 Institut de Physique du Globe de Paris 5 .\" Copyright (C) 2018-2025 |Méso|Star> (contact@meso-star.com) 6 .\" Copyright (C) 2022-2025 Observatoire de Paris 7 .\" Copyright (C) 2022-2025 Université de Reims Champagne-Ardenne 8 .\" Copyright (C) 2022-2025 Université de Versaille Saint-Quentin 9 .\" Copyright (C) 2018-2019, 2022-2025 Université Paul Sabatier 10 .\" 11 .\" This program is free software: you can redistribute it and/or modify 12 .\" it under the terms of the GNU General Public License as published by 13 .\" the Free Software Foundation, either version 3 of the License, or 14 .\" (at your option) any later version. 15 .\" 16 .\" This program is distributed in the hope that it will be useful, 17 .\" but WITHOUT ANY WARRANTY; without even the implied warranty of 18 .\" MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the 19 .\" GNU General Public License for more details. 20 .\" 21 .\" You should have received a copy of the GNU General Public License 22 .\" along with this program. If not, see <http://www.gnu.org/licenses/>. 23 .Dd January 24, 2024 24 .Dt HTRDR-COMBUSTION 1 25 .Os 26 .Sh NAME 27 .Nm htrdr-combustion 28 .Nd simulate radiative transfer in a combustion medium 29 .Sh SYNOPSIS 30 .Nm 31 .Op Fl fhINsv 32 .Op Fl C Ar persp_camera_opt Ns Op : Ns Ar persp_camera_opt No ... 33 .Op Fl D Ar laser_flux_density 34 .Op Fl d Ar dump_type 35 .Op Fl F Ar rdgfa_opt Ns Op : Ns Ar rdgfa_opt No ... 36 .Op Fl g Ar combustion_chamber_opt Ns Op : Ns Ar combustion_chamber_opt No ... 37 .Op Fl i Ar image_opt Ns Op : Ns Ar image_opt No ... 38 .Op Fl l Ar laser_opt Ns Op : Ns Ar laser_opt No ... 39 .Op Fl O Pa cache 40 .Op Fl o Pa output 41 .Op Fl P Ar ortho_camera_opt Ns Op : Ns Ar ortho_camera_opt No ... 42 .Op Fl R Ar flux_sensor_opt Ns Op : Ns Ar flux_sensor_opt No ... 43 .Op Fl T Ar optical_thickness 44 .Op Fl t Ar threads_count 45 .Op Fl V Ar accel_struct_definition 46 .Op Fl w Ar laser_wavelength 47 .Fl m Pa medium_geometry 48 .Fl p Pa thermo_properties 49 .Fl r Pa refractive_ids 50 .Sh DESCRIPTION 51 The purpose of 52 .Nm 53 is to perform radiative transfer computations in a scene representing a 54 semi-transparent combustion medium enlightened by a laser sheet. 55 The medium may be surrounded by solid boundaries 56 .Pq inner limits of the combustion chamber . 57 The program will currently compute, in the visible at a given frequency, the 58 monochromatic image or the radiative flux density of the combustion medium: 59 collected light comes from the laser, and is scattered by soot aggregates 60 within the flame before eventually reaching the sensor. 61 .Pp 62 Data relating to the gaseous medium are stored on the vertices of an 63 unstructured tetrahedral mesh: pressure, temperature and concentrations 64 of H2O, CO2 and CO are provided for each vertex. 65 In the visible range, these data are useless, since the gas is assumed 66 to be transparent, but they are part of the data expected to anticipate 67 future developments in the longwave domain. 68 .Pp 69 Soot optical properties are computed using the Rayleigh-Debye Gans theory, for 70 Fractal Aggregates (RDG-FA). 71 This requires the knowledge of: soot volumic fraction, soot number 72 density (number of primary particles per aggregate) and primary 73 particles diameter, for each vertex of the tetrahedral mesh. 74 For any position in the volume, these quantities are first interpolated 75 from the values retrieved at the nodes of the current tetrahedron, and 76 are then interpolated for the position of interest. 77 Which then makes possible to compute the absorption and scattering 78 cross-sections of soot aggregates, as well as their scattering function. 79 .Pp 80 The Monte Carlo algorithm that accounts for the visible intensity is inspired 81 from the algorithm used for solar radiation in 82 .Xr htrdr-atmosphere 1 . 83 It was adapted to partially illuminated scenes in order to solve 84 onvergence issues. 85 .Pp 86 In 87 .Nm 88 the spatial unit 1.0 corresponds to one meter while the 89 estimated monochromatic radiances and flux densities are saved in 90 W/sr/m^2 and 91 W/m^2 respectively. 92 The results are written to the output file if the 93 .Fl o 94 option is set and otherwise to standard output. 95 The output image is a list of raw ASCII data formatted using the 96 .Xr htrdr-image 5 97 file format. 98 .Pp 99 .Nm 100 implements mixed parallelism. 101 On a single computer (that is, a node), it uses shared memory 102 parallelism while it relies on Message Passing Interface (MPI) to 103 parallelize calculations between multiple nodes. 104 .Nm 105 can therefore be launched either directly or via a process launcher such 106 as 107 .Xr mpirun 1 108 to distribute the calculation on several computers. 109 .Pp 110 The options are as follows: 111 .Bl -tag -width Ds 112 .It Fl C Ar persp_camera_opt Ns Op : Ns Ar persp_camera_opt No ... 113 Set up a pinhole or thin-lens perspective camera. 114 .Pp 115 The options for a perspective camera are as follows: 116 .Bl -tag -width Ds 117 .It Cm focal-dst= Ns Ar distance 118 Distance to focus on with a thin lens camera, that is, a camera whose 119 .Cm lens-radius 120 is not zero. 121 The default focal distance is 122 @HTRDR_ARGS_DEFAULT_CAMERA_PERSPECTIVE_FOCAL_DST@ meters. 123 .It Cm focal-length= Ns Ar length 124 Focal length of a camera lens. 125 It is another way to control the field of view of a thin lens camera. 126 By default, the field of view is set through the 127 .Cm fov 128 parameter. 129 .It Cm fov= Ns Ar angle 130 Vertical field of view of the camera in 131 ]@HTRDR_ARGS_CAMERA_PERSPECTIVE_FOV_EXCLUSIVE_MIN@, 132 @HTRDR_ARGS_CAMERA_PERSPECTIVE_FOV_EXCLUSIVE_MAX@[ degrees. 133 The default field of view is 134 @HTRDR_ARGS_DEFAULT_CAMERA_PERSPECTIVE_FOV@ degrees. 135 .It Cm lens-radius= Ar radius 136 Radius of the camera lens. 137 A non-zero radius means that the camera is a thin lens camera while a 138 zero radius defines a pinhole camera. 139 The default lens radius is 140 @HTRDR_ARGS_DEFAULT_CAMERA_PERSPECTIVE_LENS_RADIUS@. 141 .It Cm pos= Ns Ar x , Ns Ar y , Ns Ar z 142 Camera position. 143 Default is @HTRDR_ARGS_DEFAULT_CAMERA_POS@. 144 .It Cm tgt= Ns Ar x , Ns Ar y , Ns Ar z 145 Targeted position 146 Default is @HTRDR_ARGS_DEFAULT_CAMERA_TGT@. 147 .It Cm up= Ns Ar x , Ns Ar y , Ns Ar z 148 Upward vector that the top of the camera is pointing towards. 149 Default is @HTRDR_ARGS_DEFAULT_CAMERA_UP@. 150 .El 151 .It Fl D Ar laser_flux_density 152 Laser flux density in W/m^2. Default is 153 @HTRDR_COMBUSTION_ARGS_DEFAULT_LASER_FLUX_DENSITY@. 154 .It Fl d Ar dump type 155 Write the data defined by 156 .Ar dump_type 157 to 158 .Pa output 159 instead of performing a normal calculation. 160 .Pp 161 The 162 .Ar dump_type 163 values supported are as follows: 164 .Bl -tag -width Ds 165 .It Cm laser 166 Write the geometry of the laser sheet in legacy VTK format. 167 .It Cm octree 168 Write the leaves of the space partitioning structures used to speed up 169 radiative transfer calculations. 170 Each leaf stores the minimum and maximum extinction coefficients of the 171 tetrahedra it covers. 172 Data are written in legacy VTK format. 173 .El 174 .It Fl F Ar rdgfa_opt Ns Op : Ns Ar rdgfa_opt No ... 175 RDG-FA phase function parameters. 176 .Pp 177 The parameters of the RDG-FA phase function are as follows: 178 .Bl -tag -width Ds 179 .It Cm dimension= Ns Ar real 180 Fractal dimension. 181 Default is @HTRDR_COMBUSTION_ARGS_DEFAULT_FRACTAL_DIMENSION@. 182 .It Cm prefactor= Ns Ar real 183 Fractal prefactor. 184 Default is @HTRDR_COMBUSTION_ARGS_DEFAULT_FRACTAL_PREFACTOR@. 185 .El 186 .It Fl f 187 Force overwriting of 188 .Pa output 189 file. 190 .It Fl g Ar combustion_chamber_opt Ns Op : Ns Ar combustion_chamber_opt No ... 191 Define the combustion chamber. 192 .Pp 193 Note that the combustion chamber does not prevent the camera from seeing 194 the medium, nor the laser from illuminating the medium, even if either is 195 outside the chamber. 196 The rendering algorithm ensures that they are not occluded by the 197 combustion chamber, like with a two-way mirror. 198 When the laser is out of chamber, its emissive surface is seen as if it 199 were following its interior surface. 200 Likewise, the radiance seen by the camera outside the chamber is the 201 radiance that reaches it as if its image plane were following the 202 surface of the geometry. 203 .Pp 204 The combustion chamber options are as follows: 205 .Bl -tag -width Ds 206 .It Cm mats= Ns Ar materials 207 Combustion chamber materials saved in 208 .Xr htrdr-materials 5 209 format. 210 .It Cm obj= Ns Ar mesh 211 Combustion chamber geometry saved in 212 .Xr htrdr-obj 5 213 format. 214 .Pp 215 In accordance with the 216 .Xr htrdr-obj 5 217 file format, the mesh interface must be defined as a fine interface, 218 i.e. it must be composed of 3 elements separated by the 219 .Li \&: 220 character. 221 By convention, 222 .Nm 223 expects the external environment to be called "air" and the medium 224 inside the combustion chamber to be called "chamber". 225 No assumptions are made about the name of the surface material, except 226 that it must refer to a valid material. 227 .El 228 .It Fl h 229 Display short help and exit. 230 .It Fl I 231 Use an isotropic phase function rather than the RDG-FA phase function 232 .Pq option Fl F . 233 .It Fl i Ar image_opt Ns Op : Ns Ar image_opt No ... 234 Configure sensor image. 235 .Pp 236 The image options are as follows: 237 .Bl -tag -width Ds 238 .It Cm def= Ns Ar width Ns x Ns Ar height 239 Image definition. 240 Default is 241 @HTRDR_ARGS_DEFAULT_IMG_WIDTH@x@HTRDR_ARGS_DEFAULT_IMG_HEIGHT@. 242 .It Cm spp= Ns Ar samples_per_pixel 243 Number of samples to solve the Monte Carlo estimation of each pixel. 244 Default is @HTRDR_ARGS_DEFAULT_IMG_SPP@. 245 .El 246 .It Fl l Ar laser_opt Ns Op : Ns Ar laser_opt No ... 247 Laser emission surface. 248 .Pp 249 The laser options are as follows: 250 .Bl -tag -width Ds 251 .It Cm pos= Ns Ar x , Ns Ar y , Ns Ar z 252 Center of the laser emission surface. 253 Default is @HTRDR_ARGS_DEFAULT_RECTANGLE_POS@. 254 .It Cm tgt= Ns Ar x , Ns Ar y , Ns Ar z 255 Targeted position, i.e.\& 256 .Cm tgt No - Cm pos 257 is the normal of the laser surface. 258 Default is @HTRDR_ARGS_DEFAULT_RECTANGLE_TGT@. 259 .It Cm up= Ns Ar x , Ns Ar y , Ns Ar z 260 Upward vector that the top of thr laser is pointer towards. 261 Default is @HTRDR_ARGS_DEFAULT_RECTANGLE_UP@. 262 .It Cm sz= Ns Ar width , Ns Ar height 263 Size of laser surface in meters. 264 Default is @HTRDR_ARGS_DEFAULT_RECTANGLE_SZ@ 265 .El 266 .It Fl m Ar medium_geometry 267 Tetrahedra of the combustion medium saved in 268 .Xr smsh 5 269 format. 270 .It Fl N 271 This speeds up runtime performance by calculating normals once and for 272 all rather than re-evaluating them every time a tetrahedron is queried 273 at a given position. 274 In return, the memory space used to store normals increases the memory 275 footprint. 276 .It Fl O Pa cache 277 File where acceleration structure is stored/loaded. 278 If the 279 .Pa cache 280 file does not exist, it is created and filled with the acceleration 281 structure constructed from the medium geometry 282 .Pq option Fl m , 283 termodynamic properties 284 .Pq option Fl p 285 and refractive indices 286 .Pq option Fl r 287 input files. 288 This cached data can then be reused in subsequent executions, provided 289 that the input files supplied to the command are the same as those used 290 to set up the cache, thus considerably speeding up the pre-processing 291 stage. 292 .Pp 293 If 294 .Pa cache 295 contains data generated from input files that are not those submitted on 296 the command line, an error is notified and execution is aborted, thus 297 avoiding the use of bad cached data. 298 .It Fl o Pa output 299 Output file. 300 If not defined, data is written to standard output. 301 .It Fl P Ar ortho_camera_opt Ns Op : Ns Ar ortho_camera_opt No ... 302 Set up an orthographic camera. 303 .Pp 304 The options for an orthographic camera are as follows: 305 .Bl -tag -width Ds 306 .It Cm height= Ns Ar lenght 307 Image plane height. 308 Its width is calculated from this length and the image ratio 309 to guarantee square pixels 310 .Pq see Fl i No option . 311 .It Cm pos= Ns Ar x , Ns Ar y , Ns Ar z 312 Camera position. 313 Default is @HTRDR_ARGS_DEFAULT_CAMERA_POS@. 314 .It Cm tgt= Ns Ar x , Ns Ar y , Ns Ar z 315 Targeted position. 316 Default is @HTRDR_ARGS_DEFAULT_CAMERA_TGT@. 317 .It Cm up= Ns Ar x , Ns Ar y , Ns Ar z 318 Upward vector that the top of the camera is pointing towards. 319 Default is @HTRDR_ARGS_DEFAULT_CAMERA_UP@. 320 .El 321 .It Fl p Ar thermo_properties 322 Thermodynamic properties of the combustion medium saved in 323 .Xr atrtp 5 324 format. 325 .It Fl R Ar flux_sensor_opt Ns Op : Ns Ar flux_sensor_opt No ... 326 Set up a flux sensor. 327 .Pp 328 The flux sensor options are as follow: 329 .Bl -tag -width Ds 330 .It Cm pos= Ns Ar x , Ns Ar y , Ns Ar z 331 Sensor center. 332 Default is @HTRDR_ARGS_DEFAULT_RECTANGLE_POS@. 333 .It Cm tgt= Ns Ar x , Ns Ar y , Ns Ar z 334 Targeted position. 335 Default is @HTRDR_ARGS_DEFAULT_RECTANGLE_TGT@. 336 .It Cm up= Ns Ar x , Ns Ar y , Ns Ar z 337 Upward vector that the top of the sensor is pointing towards. 338 Default is @HTRDR_ARGS_DEFAULT_RECTANGLE_UP@. 339 .It Cm sz= Ns Ar width , Ns Ar height 340 Sensor size in meters. 341 Default is @HTRDR_ARGS_DEFAULT_RECTANGLE_SZ@. 342 .El 343 .It Fl r Ar refrative_ids 344 Refrative indices of the combustion medium as a function of wavelength, 345 saved in 346 .Xr atrri 5 347 format. 348 .It Fl s 349 Use Single Instruction Multiple Data (SIMD) instruction sets if available. 350 This should speed up calculation time. 351 .It Fl T Ar optical_thickness 352 Optical thickness used as threshold criterion for building the acceleration 353 structure. 354 Default is @HTRDR_COMBUSTION_ARGS_DEFAULT_OPTICAL_THICKNESS_THRESHOLD@. 355 .It Fl t 356 Advice on the number of threads to use. 357 By default, 358 .Nm 359 uses many threads as processor cores. 360 .It Fl V Ar accel_struct_definition 361 Definition of the discrete field storing the upper limit of the 362 radiative coefficients from which the acceleration structure is built. 363 The extent of the grid corresponds to the axis aligned bounding box of 364 the combustion medium. 365 .Pp 366 The definition can be established as follows: 367 .Bl -tag -width Ds 368 .It Ar x , Ns Ar y , Ns Ar z 369 Define grid definition on X, Y and Z axes. 370 .It Ar expected_definition 371 Provide an expected definition of the grid along its longest axis. 372 The definition along the two remaining axes is then calculated 373 internally to tend towards cubic cells. 374 This is the default behavior with an expected definition set to 375 @HTRDR_COMBUSTION_ARGS_DEFAULT_GRID_DEFINITION_HINT@. 376 .El 377 .It Fl v 378 Make 379 .Nm 380 verbose. 381 .It Fl w Ar laser_wavelength 382 Laser wavelength in nanometers, which is also the wavelength at which 383 calculations are performed. 384 Default is @HTRDR_COMBUSTION_ARGS_DEFAULT_WAVELENGTH@. 385 .El 386 .Sh OUTPUT IMAGE 387 Images calculated by 388 .Nm 389 are saved in 390 .Xr htrdr-image 5 391 format. 392 This section describes the nature and arrangement of image data 393 depending on the type of calculation performed. 394 .Ss Shortwave monochromatic image 395 For a monochromatic image rendering, the expected value and the standard 396 deviation of the pixel radiance (in W/sr/m^2) are saved on the first and 397 the second components. 398 All other components are unused excepted the seventh and eighth 399 components that store the estimate of the radiative path computation 400 time in microseconds and its standard error. 401 .Ss Shortwave flux density map 402 A flux density map 403 .Pq option Fl R 404 store on its first and second component 405 the expected value and the standard error of the pixel radiative flux 406 density 407 .Pq in W/m^2 . 408 All other components are unused excepted the seventh and eighth 409 components that store the estimate of the radiative path computation 410 time 411 .Pq in microseconds 412 and its standard error. 413 .Sh EXIT STATUS 414 .Ex -std 415 .Sh EXAMPLES 416 Make 417 .Nm 418 verbose 419 .Pq option Fl v 420 and render an image of a combustion medium whose tetrahedral mesh is 421 stored in 422 .Pa tetra.smsh 423 and whose associated thermodynamic properties are recorded in 424 .Pa thermprops.atrtp . 425 Refractive indices are listed in 426 .Pa refract_ids.atrri . 427 The laser's surface emission center is positioned at the origin and its 428 direction aligned with the Y axis 429 .Pq option Fl l . 430 The calculated image resolution is 431 .Ar 800 No by Ar 600 432 pixels 433 .Pq option Fl i 434 and the monochromatic radiance 435 of each pixel is estimated at 436 .Ar 500 437 nanometers 438 .Pq option Fl w 439 with 440 .Ar 64 441 Monte Carlo realisations 442 .Pq option Fl i . 443 The resulting image is written to 444 .Pa output 445 unless the file already exists, in which case an error is notified, the 446 program stops and the output file remains unchanged: 447 .Bd -literal -offset Ds 448 htrdr-combustion -v \\ 449 -m tetra.smsh \\ 450 -p thermprops.atrtp \\ 451 -r refract_ids.atrri \\ 452 -l pos=0,0,0:tgt=0,1,0:up=0,0,1:sz=0.001,0.2 \\ 453 -w 500 \\ 454 -C pos=0.6,0,0.1:tgt=0.5,0,0.1:up=0,0,1:fov=30 \\ 455 -i def=800x600:spp=64 \\ 456 -o output 457 .Ed 458 .Pp 459 Add a combustion chamber to the previous example 460 .Pq option Fl g : 461 its mesh is defined in 462 .Pa chamber.obj 463 while its materials are listed in 464 .Pa materials.mtls . 465 Save the acceleration structure in 466 .Pa octree.cache 467 or reuse it if it has already been filled in a previous run with 468 compatible input data. 469 Set the finest resolution of this acceleration structure to 470 .Ar 1000 471 voxels along the main extension of the medium 472 .Pq option Fl V 473 and use an optical 474 thickness criterion of 475 .Ar 5 476 to build it 477 .Pq option Fl T . 478 Use the 479 .Fl f 480 option to force 481 overwriting of the 482 .Pa output 483 file if it exists, and use the 484 .Fl s 485 option to accelerate rendering with available SIMD instruction sets: 486 .Bd -literal -offset Ds 487 htrdr-combustion -v \\ 488 -m tetra.smsh \\ 489 -p thermprops.atrtp \\ 490 -r refract_ids.atrri \\ 491 -g obj=chamber.obj:mats=materials.mtls \\ 492 -l pos=0,0,0:tgt=0,1,0:up=0,0,1:sz=0.001,0.2 \\ 493 -w 500 \\ 494 -C pos=0.6,0,0.1:tgt=0.5,0,0.1:up=0,0,1:fov=30 \\ 495 -i def=800x600:spp=64 \\ 496 -O octree.cache \\ 497 -V 1000 \\ 498 -T 5 \\ 499 -fo output \\ 500 -s 501 .Ed 502 .Pp 503 Calculate a flux density map 504 .Pq option Fl R . 505 The sensor on which the flux density is calculated is a square with 506 sides measuring 507 .Ar 0.05 508 meters. 509 Its center is placed at the origin and points towards the Z axis. 510 The flux density map has a resolution of 511 .Ar 500 No by Ar 500 512 pixels 513 .Pq option Fl i . 514 The flux density per pixel is estimated with 515 .Ar 64 516 realisations; the flux density for the entire sensor is therefore 517 calculated with 16 million realizations (500*500*64): 518 .Bd -literal -offset Ds 519 htrdr-combustion -v \\ 520 -m tetra.smsh \\ 521 -p thermprops.atrtp \\ 522 -r refract_ids.atrri \\ 523 -l pos=0,0,0:tgt=0,1,0:up=0,0,1:sz=0.001,0.2 \\ 524 -w 500 \\ 525 -R pos=0,0,0:tgt=0,0,1:up=0,1,0:sz=0.05,0.05 \\ 526 -i def=500x500:spp=64 \\ 527 -O octree.cache \\ 528 -V 1000 \\ 529 -T 5 \\ 530 -fo map.txt 531 -s 532 .Ed 533 Write a representation of the acceleratrion structure in 534 .Pa accel_struct.vtk 535 .Pq option Fl d : 536 .Bd -literal -offset Ds 537 htrdr-combustion -v \\ 538 -m tetra.smsh \\ 539 -p thermprops.atrtp \\ 540 -r refract_ids.atrri \\ 541 -O octree.cache \\ 542 -d octree \\ 543 -o accel_struct.vtk 544 .Ed 545 .Sh SEE ALSO 546 .Xr htrdr-atmosphere 1 , 547 .Xr atrri 5 , 548 .Xr atrtp 5 , 549 .Xr htrdr-image 5 , 550 .Xr htrdr-materials 5 , 551 .Xr htrdr-obj 5 , 552 .Xr smsh 5 553 .Rs 554 .%A Morgan Sans 555 .%A Mouna El\ Hafi 556 .%A Vincent Eymet 557 .%A Vincent Forest 558 .%A Richard Fournier 559 .%A Najda Villefranque 560 .%T Null-collision meshless Monte Carlo - A new reverse Monte Carlo \ 561 algorithm designed for laser-source emission in absorbing/scattering \ 562 inhomogeneous media 563 .%J Journal of Quantitative Spectroscopy and Radiative Transfer 564 .%V 271 565 .%D 2021 566 .%U https://doi.org/10.1016/j.jqsrt.2021.107725 567 .Re 568 .Rs 569 .%A Jérôme Yon 570 .%A Fengshan Liu 571 .%A Alexandre Bescond 572 .%A Chloé Caumont-Prim 573 .%A Claude Rozé 574 .%A François-Xavier Ouf 575 .%A Alexis Coppalle 576 .%T Effects of multiple scattering on radiative properties of soot \ 577 fractal aggregates 578 .%J Journal of Quantitative Spectroscopy and Radiative Transfer 579 .%V 133 580 .%P 374-381 581 .%D 2014 582 .%U https://doi.org/10.1016/j.jqsrt.2013.08.022 583 .Re 584 .Sh STANDARDS 585 .Rs 586 .%A OpenMP Architecture Review Board 587 .%D March 2002 588 .%T OpenMP C and C++ Application Interface 589 .%O version 2.0 590 .Re 591 .Pp 592 .Rs 593 .%A Message Passing Interface Forum 594 .%D July 1997 595 .%T MPI-2: Extensions to The Message-Passing Interface 596 .Re 597 .Sh HISTORY 598 .Nm 599 has been developed as part of 600 .Li ANR-18-CE05-0015 601 Astoria project.