### What is it
Appealing method for simulation of fluid flows
- A competitive numerical tool for simulating fluid flows over a wide range of complex physical problems
- LBMs were initially derived from **lattice gas cellular automata (LGCA).** The basic idea of LGCA is to *simulate the macroscopic behavior of a fluid flow* by implementing an extremely simplified model of the microscopic interactions between particles.
- LBMs were developed, starting from LGCA, in the attempt to overcome their major drawbacks:
- statistical noise,
- increasing complexity of the collision operator (for three dimensional problems),
- high viscosity (due to small number of collisions)
- Nowadays, LBM has consolidated into a powerful alternative to more classical computational fluid dynamics models based on the *discretization of the Navier-Stokes equations of continuum mechanics.*
### Challenges & New ideas
Applicability of the traditional Lattice Boltzman method to **uniform, regaular lattices** is restricted by the** intrinsic coupling of momentum and space discretization**. This is often disadvantageous in practice.
**Available off-lattice Boltzmann algorithms** have *stability problems* which are to be *handled at the expense of additional computational cost.*
> A **general characteristic-based algorithm** for off-lattice Boltzmann simulations that preserves all appealing properties of the standard Lattice Boltzmann method while **extending the method to unstructured grids.**
>
> Both, **finite-element and finite-difference implementations** of the algorithms are exemplified.
Applications:
1. Particulate Fluid Simulation
2. Multi-component flows
3. Radiative Transport
4. Thermal Fluid Flows
5. Turbulence Simulations
6. Flows in Complex Geometries
---
Magnus Performance Results: MLUPps (**M**ega **L**attice **UP**dates per **p**rocessing unit and **s**econd) as a function of cores