In order to compute the solution for a given simulation case the simulation domain needs to be discretized. It essentially means to split up one large problem into multiple smaller mathematical problems. Solving the entire continuous domain at once is not possible or reasonable whereas solving multiple small pieces of it becomes feasible.
This discretization process is the base for methods such as the Finite Difference Method (FDM), Finite Volume Method (FVM) and Finite Element Method (FEM), whose purpose is to take equations in the continuous form and generate a system of algebraic difference equations. This process generates a set of discrete points and cells, which, as a whole, cover the entire simulation domain.
SimScale strives to make the meshing process as simple and user-friendly as possible. In the end, the user should only have to decide on the trade-off between the fineness of the mesh (and thereby the accuracy of the result) and the required computation time (as larger mesh requires more computational resources to be solved).
SimScale offers various meshing capabilities to create computational grids. Several algorithms are available to create three-dimensional tetrahedral and hexahedral meshes on the platform. Due to robustness and general applicability of the algorithms, both automated and manual versions of these meshing algorithms are provided. Details for each operation can be found below.
Based on the simulation type, a default mesh setup will be created, which in most cases is a good starting point to begin with. Some mesh types are applicable for certain analysis types only and therefore will not be available for others.
All meshes within a project can be shared across all simulations in the project, given that the mesh type is compatible with the chosen analysis type. E.g. a tet-dominant mesh, that has been created in a static analysis can also be assigned to a heat transfer analysis in the same project. Making changes to the mesh in one simulation will also update it in all other simulations where it is assigned.
Open the mesh select menu to create a new mesh with default settings or by duplicating/importing mesh settings from another mesh.
Before a simulation run can be started, the assigned mesh needs to be generated.
While meshes are created within a simulation, all meshes, independent of where they were first created, can be referenced in, assigned to, and edited from any other simulation setup. Copying a simulation does not also copy meshes, but simply references the original mesh as an independent object.
Most of the mesh settings depend on the chosen mesh algorithm. Current meshing strategies available on the platform are:
As mentioned above, the available meshing algorithms depend on the chosen analysis type, but also on the CAD domain file format. E.g. FEA simulations are limited to the Tet-dominant mesher, while most CFD cases can be solved with any of the given meshing types. CAD files, uploaded in the STL file format can currently only be meshed using the parametric mode of the hex-dominant mesher.
The setting for the number of processors defines the size of the cloud computing instance that will be used to compute the mesh. The chosen setting defines the number of computing cores (CPUs) of the machine. However, an instance with more CPUs also comes with more memory, which in many cases is the limiting factor in meshing. The default setting is usually a good starting point.
Besides creating a mesh within SimScale, users have the possibility to use meshes created with others tools by uploading them directly to the platform. Find out more about mesh upload here.
This offering is not approved or endorsed by OpenCFD Limited, producer and distributor of the OpenFOAM software and owner of the OPENFOAM® and OpenCFD® trade marks. OPENFOAM® is a registered trade mark of OpenCFD Limited, producer and distributor of the OpenFOAM software.