Skip to content

Getting Started

This user guide presents the third-generation DNS code developed at the University of Magdeburg, building upon decades of experience in direct numerical simulations (DNS) of turbulent reactive flows. Earlier developments from our group include the PARCOMB family 12 and the π³ code 34.

Originally, the name DINOSOARS — now shortened to DINO — stood for:

DIrect Numerical, high-Order Simulation and On-the-fly Analysis of Reacting flows and Sprays

Background

DNS of gaseous reactive flows has achieved remarkable success over the past three decades (e.g., 56789101112). Nevertheless, it remains an active research field, as numerous challenges persist due to the extreme computational demands involved (see, e.g., 13141516). Beyond purely gaseous flames, reacting sprays have also been extensively investigated via DNS (e.g., 1718).

Our group has specialized in DNS studies that incorporate detailed kinetic models. As discussed in the first DINO paper 19, considering complex fuels such as n-heptane or ethylene greatly increases the computational cost compared to non-reacting or single-step-chemistry simulations. Furthermore, achieving physically consistent results requires accurate models for thermodynamic and molecular transport properties, such as viscosity and diffusion coefficients. Consequently, DINO simulations are designed to run on massively parallel supercomputers.

Code Philosophy and Design

To meet the challenges of simulating realistic reacting and two-phase flows, DINO integrates several key numerical and algorithmic strategies to ensure accuracy, efficiency, and scalability.

  1. Low-Mach-number formulation
    Most practical flows are either incompressible or occur at low Mach numbers. Using a low-Mach-number solver significantly reduces computational cost compared to a fully compressible approach 20.

  2. Semi-implicit time integration
    Removing acoustic wave constraints allows larger timesteps, but for reacting flows, the timestep is still limited by chemical kinetics. To overcome this, DINO employs a semi-implicit, high-order time integration scheme that ensures both stability and accuracy.

  3. Efficient parallelization and Poisson solver
    DINO is parallelized using the open-source library 2DECOMP&FFT 21, which provides efficient domain decomposition and communication routines. A major bottleneck in incompressible and low-Mach-number solvers lies in solving the Poisson equation that couples pressure and velocity fields. To address this, DINO implements an FFT-based solver capable of handling non-periodic boundary conditions, ensuring scalability on modern high-performance computing systems.

Immersed Boundary Capabilities

With increasing demand for simulations involving complex geometries while maintaining low numerical dissipation,DINO integrates a Direct Boundary Immersed Boundary Method (DB-IBM). This method allows the representation of semi-complex geometries on a fixed Cartesian grid, extending DNS applicability to a wider range of practical and academic configurations.

Next Steps

The following sections provide:
- Instructions for downloading the code from GitLab
- A description of all code dependencies and external libraries required to build and run DINO - A step-by-step installation procedure

  1. Thévenin D., Behrendt F., Maas U., Przywara B., Warnatz J. Development of a parallel direct simulation code to investigate reactive flows. Comput. Fluids, 25(5):485–496, 1996. ↩

  2. Hilbert R., Thévenin D. Influence of differential diffusion on maximum flame temperature in turbulent non-premixed hydrogen/air flames. Combust. Flame, 138:175–187, 2004. ↩

  3. Thévenin D., Gicquel O., de Charentenay J., Hilbert R., Veynante D. Two- versus three-dimensional direct simulations of turbulent methane flame kernels using realistic chemistry. Proc. Combust. Inst., 29:2031–2039, 2003. ↩

  4. Thévenin D. Three-dimensional direct simulations and structure of expanding turbulent methane flames. Proc. Combust. Inst., 30:629–637, 2005. ↩

  5. Oran E.S., Boris J.P. Numerical Simulation of Reactive Flows. Elsevier, 1987. ↩

  6. Reynolds W.C. The potential and limitations of direct and large eddy simulation. Lecture Notes in Physics, Berlin, Heidelberg, Springer, 357:313–343, 1990. ↩

  7. Poinsot T., Veynante D., Candel S. Quenching processes and premixed turbulent combustion diagrams. J. Fluid Mech., 228:561–606, 1991. ↩

  8. Baum M., Poinsot T., Haworth D.C., Darabiha N. Direct numerical simulation of H\(\_2\)/O\(\_2\)/N\(\_2\) flames with complex chemistry in two-dimensional turbulent flows. J. Fluid Mech., 281:1–32, 1994. ↩

  9. Poinsot T., Candel S., Trouvé A. Applications of direct numerical simulation to premixed turbulent combustion. Prog. Energy Combust. Sci., 21:531–576, 1996. ↩

  10. Vervisch L., Veynante D. Direct numerical simulation of non-premixed turbulent flames. Annu. Rev. Fluid Mech., 30:655–691, 1998. ↩

  11. Thévenin D., Behrendt F., Maas U., Warnatz J. Simulation of reacting flows with a portable parallel code using dynamic load-balancing, in: high-performance computing and networking. Lecture Notes in Computer Science, Berlin, Heidelberg, Springer, 919:378–383, 1995. ↩

  12. Thévenin D., Behrendt F., Maas U., Warnatz J. Parallel simulation of reacting flows using detailed chemistry. Lecture Notes in Computer Science, Berlin, Heidelberg, Springer, 796:125–130, 1994. ↩

  13. Hawkes E.R., Sankaran R., Sutherland J.C., Chen J.H. Scalar mixing in direct numerical simulations of temporally evolving plane jet flames with skeletal CO/H\(\_2\) kinetics. Proc. Combust. Inst., 32:1633–1640, 2007. ↩

  14. Coussement A., Gicquel O., Fiorina B., Degrez G., Darabiha N. Multicomponent real gas 3-D-NSCBC for direct numerical simulation of reactive compressible viscous flows. J. Comput. Phys., 245:259–280, 2013. ↩

  15. Bansal G., Mascarenhas A., Chen J.H. Direct numerical simulations of autoignition in stratified dimethyl-ether (DME)/air turbulent mixtures. Combust. Flame, 162:688–702, 2015. ↩

  16. Bhagatwala A., Luo Z., Shen H., Sutton J.A., Lu T., Chen J.H. Numerical and experimental investigation of turbulent DME jet flames. Proc. Combust. Inst., 35:1157–1166, 2015. ↩

  17. Duret B., Reveillon J., Menard M., Demoulin F.X. Improving primary atomization modeling through DNS of two-phase flows. Int. J. Multiphase Flow, 55:130–137, 2013. ↩

  18. Kitano T., Nishio J., Kurose R., Komori S. Effects of ambient pressure, gas temperature and combustion reaction on droplet evaporation. Combust. Flame, 161:551–564, 2014. ↩

  19. A. Abdelsamie, G. Fru, T. Oster, F. Dietzsch, G. Janiga, and D. Thévenin. Towards direct numerical simulations of low-mach number turbulent reacting and two-phase flows using immersed boundaries. Computers & Fluids, 131:123 – 141, 2016. ↩

  20. de Charentenay J., Thévenin D., Zamuner B. Comparison of direct numerical simulations of turbulent flames using compressible or low-Mach number formulations. Int. J. Numer. Meth. Fluids, 39:497–515, 2002. ↩

  21. Li N., Laizet S. 2DECOMP&FFT - a highly scalable 2D decomposition library and FFT interface. In Cray User's Group 2010 conference. 2010. ↩