Altair nanoFluidX 2019 Release Notes

nanoFluidX is a software to simulate single- and multi-phase flows based on the Lagrangian Particle Method "Smoothed Particle Hydrodynamics" (SPH) and is developed and maintained by Altair Engineering.

The purpose of this code is to simulate complex flows that are difficult or infeasible to handle using classical CFD approaches such as Finite Volume Methods. The main advantages of SPH are the conservation of mass and momentum (linear and angular) in the absence of physical dissipation e.g. due to viscous shear effects and the exact advection of the particles. As a consequence, SPH is very powerful in dealing with free-surface flows as well as multi-phase flows with complex physical phenomena such as surface-tension or interfacial transport processes.

Highlights

  • SPH weakly compressible code.
  • Unique numerical solutions bringing unprecedented accuracy among industrial SPH codes.
  • Free-surface and high-deformation fluid flows.
  • Meshless approach, leading to simplified pre-processing.
  • Multi-phase flows.
  • Aeration-viscosity models.
  • Temperature equation and viscosity-temperature dependence models.
  • Surface tension models.
  • Adhesion models for fluid-solid contacts.
  • Defined motion of solid bodies.
  • Passive rigid body motion (as in fluid-rigid body momentum exchange).
  • Created and optimized for use on clusters of Graphical Processing Units (GPUs), making it extremely fast.

New Features

Optional use of low-dissipation 1st order Riemann solver
Greatly improves accuracy and the representation of the pressure field.
Note: This feature is considered experimental when used in multiphase mode.
Moving Least Squares correction for gradients
Further improves accuracy.
New optional particle position correction scheme
APD (Artificial Particle Displacement)
nanoFluidXcompanion (nFXc)
A post-processing tool for converting SPH particle data to a mesh-based data for easier treatment and visualization in ParaView.
Aeration-viscosity models
Account for the change of viscosity as a function of the air entrapment. Has beneficial influence on the torque prediction, especially in cases where torque comes mostly from shear stress.
Primitive adhesion model
Adhesion models which depend on an adjustable numerical parameter. Though it improves the capability of capturing film formations, the model has to be tuned to the experimental results.
Monitoring probes
Observe the flow at a specified location for any physical property.
Circular inlets
Improved motions
Improved impose regions (momentum sources)
Fluid Contact Time (FCT)
Show contact time history between solid elements and a chosen fluid, which is useful for qualitative analysis of oil supply and lubrication.
Operational modes
Allow for a “single command setup”, depending on what is the required level of speed/accuracy ratio.

Enhancements

Supported Platforms
All Unix-based OS with GCC newer than 4.4.7 and GLIBC 2.12 (RHEL 6.x and 7.x and compatible Scientific Linux, CentOS, Ubuntu 14.04 and 16.04, OpenSUSE 13.2, and so on.)
NVIDIA CUDA 8.0 and OpenMPI 2.1.5 – shipped with the binary.

Current Limitations

  • Surface tension coefficient for single phase flows requires tuning, as no universal single phase surface tension model exists.
  • Adhesion coefficient needs to be calibrated against experimental results or visually estimated.
  • Large negative pressures cannot be accurately handled by the current weakly compressible formulation without total particle volume preservation.

Resolved Issues

  • Improved robustness of the restart (disappearing phases at rank borders are addressed).
  • Planetary motion – rotation around y axis is now fixed.
  • Fixed particle overlapping check, so that it shows consistent numbers between phases.