Passive scalar interface in a spatially evolving mixing layer (A. Attili and D. Denker)

Quartz nozzle sampling (D. Felsmann)

Dissipation element analysis of a planar diffusion flame (D. Denker)

Turbulent/non-turbulent interface in a temporally evolving jet (D. Denker)

Dissipation elements crossing a flame front (D. Denker and B. Hentschel)

Particle laden flow (E. Varea)

Turbulent flame surface in non-premixed methane jet flame (D. Denker)

DNS of primary break up (M. Bode)

Diffusion flame in a slot Bunsen burner (S. Kruse)

Various quantities in spatially evolving jet diffusion flame (D. Denker)


Turbulence is omnipresent in everyday life, as well as in engineering applications. It can be observed in the mixing of milk and coffee at breakfast or in the jets of exhausts exiting car mufflers on a cold day. One identifies these turbulent flows as unsteady and chaotic with eddies of various sizes. Mixing in turbulent flows is greatly enhanced, which enables the operation of highly efficient combustion engines and sophisticated chemical reactors in the first place. Therefore, a better understanding of the behavior and underlying mechanisms of turbulent flows is of great practical importance which is not limited to combustion applications.

At the ITV, turbulent flows are investigated by means of direct numerical simulations (DNS). These are large scale simulations where no models for the turbulence are employed. Therefore, these simulations are also referred to as "numerical experiments". Since all turbulent scales need to be resolved in DNS, these massive simulations are run in a heavily parallelized fashion on Europe's fasted supercomputers including JUWELS, SuperMUC and Hazel Hen.

To clearly separate between effects of turbulence and combustion, as well as to achieve higher turbulence intensities, some DNS are performed without chemical reactions. The configurations of these DNS are highly idealized and include flow types such as temporally and spatially evolving jets and mixing layers and isotropic turbulence.

Figure: Scalar dissipation rate in DNS of isotropic turbulence.

Turbulent/non-turbulent interface

Free shear flows, such as jets, mixing layers and wakes possess two distinctive flow regions, i.e. an outer laminar flow region and an inner, fully turbulent part. These regions are separated by an interface layer called the turbulent/non-turbulent interface (TNTi). Across the thin TNTi, flow properties such as the vorticity or any transported scalar changes drastically.    

Since the properties of turbulence are important for a wide variety of engineering fields, the prediction of the location and behavior of the TNTi is of great interest. As the TNTi is very thin, large eddy simulations (LES) are unable to resolve it and the inherent steep gradients of various flow properties in the vicinity of the TNTI. Additionally, in RANS simulations of the k - ε type there is a further difficulty because both, the kinetic energy k and its dissipation ε tend to zero as one moves towards the laminar region. Since the eddy viscosity is proportional to k²/ε, it is ill-defined close to the TNTi.

The TNTi is investigated at the ITV using data of DNS containing up to 19 billion grid points and taylor based Reynolds numbers of up to 300.

Video: TNTi in the DNS of a planar spatially evolving jet

Dissipation element analysis

Dissipation element (DE) analysis is a physically motivated space filing compartmentalization of turbulent scalar fields. The scalar fields are subdivided into regions of monotonic scalar behavior, i.e. regions where diffusive fluxes are directed to the same extremal points of the respective scalar fields. The DEs provide non-arbitrary measure of the local turbulent scales and divide the scalar fields into smaller sub units which are more amenable for detailed investigation.   

One of the characteristics of the length scale statistics obtained with DEs is its invariance towards changes in Reynolds number and underlying scalar. This holds true even for highly anisotropic flow configurations where traditional methods of obtaining statistics, such as structure functions, fail to reproduce scalings.  

DE analysis is applied to scalar fields of reacting and non-reacting DNS with up to 45 billion grid points and Jet Reynolds numbers of up to 22400. These turbulent fields include the mixture faction, passive scalar, temperature and turbulent kinetic energy. 

Figure: Dissipation element in the mixture fraction field of a methane jet flame.