Physics-aware meshing techniques for generating grids for objects

Patent metadata

Field	Value
Publication number	US-2025131160-A1
Application number	US-202318491263-A
Country	US
Kind code	A1
Filing date	Oct 20, 2023
Priority date	Nov 24, 2022
Publication date	Apr 24, 2025
Grant date	—

Abstract

Computational Fluid Dynamic, Finite Element Methods (FEM), and other engineering modelling and simulation requires meshing (numerical grid generation) of the geometry. The meshing is done by meshing software and the CFD expert need to provide meshing parameters. Existing approaches involve a lot of human intervention which leads to additional computational time and may be prone to errors. Present disclosure provides system and method that construct meshed geometry based a first mesh generated using meshing parameters wherein the meshed geometry is compared with original geometry of an object under consideration to estimate the surface size. Further, the first mesh is simulated in an iterative manner to obtain parameters such as domain size, refinement zone, and layer parameters of the object under consideration until these parameters reach an associated threshold and a mesh independent grid is obtained for the object based on the above-mentioned parameters.

First claim

What is claimed is:

1 . A processor implemented method, comprising:
receiving, via one or more hardware processors, an input comprising a first geometry of an object in a predefined format; and iteratively performing:
constructing, via the one or more hardware processors, a second geometry for the object based on a first mesh generated using a first set of meshing parameters comprised in a meshing dictionary;
performing, via the one or more hardware processors, a comparison of the first geometry and the second geometry;
estimating, via the one or more hardware processors, one or more surface sizes based on the comparison;
calculating, via the one or more hardware processors, one or more layer parameters on the object based one or more y-plus values calculated from a simulation of the first mesh;
calculating, via the one or more hardware processors, one or more refinement zones associated with the object based on one or more absolute errors of one or more field variables calculated by taking an absolute difference of (i) a value of a first set of field variables on the first mesh and (ii) a value of a second set of field variables of a second mesh being mapped onto the first mesh, wherein the value of the first set of field variables on the first mesh and the value of the second set of field variables on the second mesh are obtained by simulating the first mesh and the second mesh; and wherein the first mesh and the second mesh are generated using the first set of meshing parameters and a second set of meshing parameters respectively; and
calculating, via the one or more hardware processors, a domain size for the object based on the simulation of the first mesh,
until a respective objective associated with the one or more surface sizes, the domain size, the one or more layer parameters, and the one or more refinement zones reach an associated threshold to obtain a mesh independent grid.

2 . The processor implemented method of claim 1 , wherein the step of performing the comparison of the first geometry and the second geometry comprises performing a comparison of a set of point clouds of the first geometry and an associated set of point clouds of the second geometry.

3 . The processor implemented method of claim 1 , wherein the step of calculating, via the one or more hardware processors, the domain size for the object based on the simulation of the first mesh comprises:
iteratively performing: incrementing a domain size by adding one or more layers on a plurality of directions of the object based on the first mesh; simulating the first mesh based on the incremented domain size; and calculating a force based on the simulation of the first mesh, until a difference in the force of a current iteration and a previous iteration reaches a force threshold.

4 . The processor implemented method of claim 3 , wherein number of the plurality of directions is based on the object being placed on a first region and in a second region.

5 . The processor implemented method of claim 1 , wherein the one or more layer parameters comprise number of layers to be added on the object, a size of a first layer amongst the one or more layers, and an associated expansion ratio.

6 . The processor implemented method of claim 5 , wherein number of layers to be added on the object is calculated based on the simulation of the first mesh, comprises:
calculating the one or more y-plus values over a wall type of the object based on the simulation of the first mesh; calculating a minimum size of the first mesh at a wall of the object based on the one or more y-plus values, one or more target y-plus values and a current cell size in the first mesh; and calculating the number of layers to be added on the object based on the minimum size of the first mesh with a pre-defined expansion ratio.

7 . The processor implemented method of claim 1 , wherein the one or more refinement zones are calculated by:
calculating a first mesh size based on a pre-computed background mesh size; calculating a second mesh size based on the first mesh size; creating the first mesh and the second mesh using the first mesh size and the second mesh size, respectively; simulating physics on the first mesh and the second mesh, respectively; calculating the one or more absolute errors of the one or more field variables calculated by taking an absolute difference of (i) a value of the first set of field variables obtained from the simulation of the first mesh and (ii) a value of the second set of field variables obtained from the simulation of the second mesh being mapped onto the first mesh; performing a comparison of the one or more absolute errors with an error threshold; and calculating the one or more refinement zones in the second mesh based on the comparison by enclosing one or more error regions with an iso surface of an error equal to the error threshold.

8 . A system, comprising:
a memory storing instructions; one or more communication interfaces; and one or more hardware processors coupled to the memory via the one or more communication interfaces, wherein the one or more hardware processors are configured by the instructions to: receive an input comprising a first geometry of an object in a predefined format; and iteratively perform:
constructing a second geometry for the object based on a first mesh generated using a first set of meshing parameters comprised in a meshing dictionary;
performing a comparison of the first geometry and the second geometry;
estimating one or more surface sizes based on the comparison;
calculating one or more layer parameters on the object based one or more y-plus values calculated from a simulation of the first mesh;
calculating one or more refinement zones associated with the object based on one or more absolute errors of one or more field variables calculated by taking an absolute difference of (i) a value of a first set of field variables on the first mesh and (ii) a value of a second set of field variables of a second mesh being mapped onto the first mesh, wherein the value of the first set of field variables on the first mesh and the value of the second set of field variables on the second mesh are obtained by simulating the first mesh and the second mesh, and wherein the first mesh and the second mesh are generated using the first set of meshing parameters and a second set of meshing parameters respectively; and
calculating a domain size for the object based on the simulation of the first mesh,
until a respective objective associated with the one or more surface sizes, the domain size, the one or more layer parameters, and the one or more refinement zones reach an associated threshold to obtain a mesh independent grid.

9 . The system of claim 8 , wherein the step of performing the comparison of the first geometry and the second geometry comprises performing a comparison of a set of point clouds of the first geometry and an associated set of point clouds of the second geometry.

10 . The system of claim 8 , wherein the step of calculating the domain size for the object based on the simulation of the first mesh comprises:
iteratively performing: incrementing a domain size by adding one or more layers on a plurality of directions of the object based on the first mesh; simulating the first mesh based on the incremented domain size; and calculating a force based on the simulation of the first mesh, until a difference in the force of a current iteration and a previous iteration reaches a force threshold.

11 . The system of claim 10 , wherein number of the plurality of directions is based on the object being placed on a first region and in a second region.

Assignees

• Tata Consultancy Services Ltd

Inventors

• Ahmad Dilshad
• Kulkarni Hrishikesh Nilkanth
• Maurya Mithilesh Kumar

Classifications

• G06T17/20

  Finite element generation, e.g. wire-frame surface description, {tesselation} · CPC title

• G06F2111/20

  Configuration CAD, e.g. designing by assembling or positioning modules selected from libraries of predesigned modules · CPC title

• G06F2119/14

  Force analysis or force optimisation, e.g. static or dynamic forces · CPC title

• G06F30/23 Primary

  using finite element methods [FEM] or finite difference methods [FDM] · CPC title