meso-web

Sources of the |Méso|Star> website
git clone git://git.meso-star.fr/meso-web.git
Log | Files | Refs | README | LICENSE

validation.md (6516B)


      1 # Validation
      2 
      3 This webpage provides the references to validations of the Stardis code
      4 and the theoretical framework it is based on.
      5 
      6 1.  We first present [validations against analytical model](#analytical),
      7     which are directly provided in Stardis as non-regression tests.
      8 
      9 2.  We then [refer to scientific papers](#papers) in which Monte Carlo
     10     algorithms are compared to deterministic solvers on non trivial
     11     configurations.
     12 
     13 3.  Finally, we present [a validation test case of SYRTHES](#SYRTHES),
     14     the thermal code developed by *Électricité de France*, which
     15     provides both a finite element solver and a Monte Carlo solver
     16     powered by Stardis.
     17 
     18 <span id="analytical"/>
     19 
     20 ## Comparison against analytical results
     21 
     22 Stardis provides comparisons against analytical solutions.
     23 These non-regression tests are available in the `src/` directory of
     24 [Stardis Solver](https://gitlab.com/meso-star/stardis-solver/-/tree/master/src)
     25 (denoted by `test_*`).
     26 Note that the tests are performed on the direct Monte Carlo simulation
     27 and the propagator (path-replay with different conditions) when
     28 possible.
     29 For each test, the scene geometry and
     30 physical parameters are described in the header.
     31 Among these, we note the following tests:
     32 
     33 - [test\_sdis\_conducto\_radiative.c](https://gitlab.com/meso-star/stardis-solver/-/blob/master/src/test_sdis_conducto_radiative.c)
     34   validates the steady resolution of the coupled conduction and
     35   radiative transfer in a solid surrounded by two different fluids
     36   (left/right faces).
     37 
     38 - [test\_sdis\_convection\_non\_uniform.c](https://gitlab.com/meso-star/stardis-solver/-/blob/master/src/test_sdis_convection_non_uniform.c)
     39   validates the transient resolution of the convection for a fluid
     40   inside a cube with faces of different known temperatures.
     41 
     42 - [test\_sdis\_transient.c](https://gitlab.com/meso-star/stardis-solver/-/blob/master/src/test_sdis_transcient.c)
     43   validates the transient resolution of conduction in nested cubes.
     44 
     45 - [test\_sdis\_solve\_boundary.c](https://gitlab.com/meso-star/stardis-solver/-/blob/master/src/test_sdis_solve_boundary.c)
     46   validates the steady computation of the boundary temperature on a
     47   solid cube interfaced with a fluid with known temperature.
     48 
     49 - [test\_random\_walk\_robustness.c](https://gitlab.com/meso-star/stardis-solver/-/blob/master/src/test_sdis_solid_random_walk_robustness.c)
     50   validates the random walk in a solid with / without a source term in
     51   complex geometry.
     52 
     53 <span id="papers"/>
     54 
     55 ## Cross-comparison against deterministic solvers
     56 
     57 Stardis is also validated against usual deterministic codes, on more
     58 complex geometries where no analytical solution exists.
     59 We list here the academic papers which include such validations and
     60 provide a description of the configuration and mention the code used for
     61 comparison.
     62 
     63 1.  [Penazzi et al.](https://doi.org/10.1016/j.cpc.2023.108911),
     64     "Path integral formulations leading to propagator evaluation for
     65     coupled linear physics in large geometric models", Computer Physics
     66     Communications 2024, appendix C.
     67     - Validation against [COMSOL](https://www.comsol.fr/)
     68     - Solid with fluid cavities
     69     - Coupled conduction, convection (perfectly mixed cavity) and
     70       radiation; homogeneous coefficients
     71     - Stationary state
     72     - Validation of the propagator
     73 
     74 2.  [Ibarrart et al.](https://hal.science/hal-03818899v2),
     75     "Advection, diffusion and linear transport in a single path-sampling
     76     Monte-Carlo algorithm: getting insensitive to geometrical
     77     refinement", Preprint 2022, figures F.9 and F.10.
     78     - Validation against [COMSOL](https://www.comsol.fr/) or
     79       [ANSYS Fluent](https://www.ansys.com/products/fluids/ansys-fluent)
     80     - Poiseille duct or Kelvin cells
     81     - Coupled conduction, convection (with advection) and radiative
     82       transfer; homogeneous coefficients
     83     - Stationary state
     84 
     85 3.  [Caliot et al.](https://hal.science/hal-02096305v1),
     86     "Combined conductive-radiative heat transfer analysis in complex geometry
     87     using the Monte Carlo method", Eurotherm 2018, figures 6 to 9.
     88     - Validation againsta
     89       [ANSYS Fluent](https://www.ansys.com/products/fluids/ansys-fluent)
     90     - Kelvin cells
     91     - Coupled conduction and radiative transfer
     92     - Stationary state
     93 
     94 4.  [Retailleau et al](https://hal.science/hal-04059892),
     95     "Résolution d’un problème de transferts thermiques couplés en géométrie
     96     urbaine par la méthode Monte Carlo", in SFT 2023, figure 4.
     97     - Validation against finite differences
     98     - Slab with Robin conditions
     99     - Coupled conduction, convection (perfectly mixed cell) and
    100       radiative transfer
    101     - Un-stationary state
    102 
    103 <span id="SYRTHES"/>
    104 
    105 ## Stardis in SYRTHES
    106 
    107 Stardis is used in the
    108 [SYRTHES](https://www.edf.fr/en/the-edf-group/inventing-the-future-of-energy/r-d-global-expertise/our-offers/simulation-softwares/syrthes)
    109 code of the French electric company *Électricité de France*.
    110 Both deterministic and stochastic resolutions can therefore be compared
    111 on the exact same CAD input.
    112 Here we provide the validation on one stationary test case of
    113 conduction inside a [square](#figure1).
    114 Both the finite elements and the Monte Carlo (using Stardis) resolutions
    115 are [compared](#figure2).
    116 
    117 <span id="figure1"/>
    118 
    119 <figure style="text-align: center">
    120   <img src="images/geometry.svg" alt="geometry"
    121    style="width: 50%; display: inline;">
    122   <a href="images/temperature.png">
    123     <img src="images/temperature.png" alt="temperature"
    124      style="width: 45%; display: inline;">
    125   </a>
    126 </figure>
    127 
    128 > The left figure describes the configuration of the test case.
    129 > The system to be simulated is a solid square with one edge having a
    130 > known temperature.
    131 > Another edge has a convective exchange with a fluid whose temperature
    132 > is also known.
    133 > The two other edges are adiabatic.
    134 > The right image illustrates the temperature field corresponding to
    135 > this configuration at steady state.
    136 
    137 <span id="figure2"/>
    138 
    139 [![Y Profile](images/TprofY.png)](images/TprofY.png)
    140 [![X Profile](images/TprofX.png)](images/TprofX.png)
    141 
    142 > Validation of the Finite element solver and the Monte Carlo solver
    143 > (i.e. Stardis) of SYRTHES against the analytical solution of the test
    144 > case [presented above](#figure1).
    145 > Both curves are computed at steady state at probe positions varying
    146 > along the X axis (Top) or the Y axis (Bottom).
    147 
    148 The version of SYRTHES used for this validation is still on development
    149 and available on-demand.
    150 Please [contact us](mailto:contact@meso-star.com) to obtain this
    151 version.