Home | english  | Impressum | Datenschutz | Sitemap | KIT


The DITEFA facility (see Figure 1) is a small Gallium Indium Tin (GaInSn) liquid metal loop (~30 liters). GaInSn has the characteristic of being a non-toxic liquid metal at room temperature. The facility was designed upon a modular manufacturing concept which allows to adapt future liquid metal experiments to the facility with ease.

The first experiment built into the DITEFA facility will be a benchmark test case for computational fluid dynamics (CFD) code validation, namely, a 3D vertical backward facing step (BFS). This experiment is an intermediate step with the aim of integrating a BFS-experiment into the KASOLA-facility.

A BFS experiment tests CFD codes against flow separation and its reattachment due to a sudden cross-section expansion of the duct (see Figure 2). We call this a “3D”-BFS because of the low ratio (1:2) between the step´s height (“h” in Figure 2) and the span-wise (“into” the screen) length relatively to the main flow´s direction. This low aspect ratio will make the influence on the secondary motions (generated by the effect of the sidewalls’ corners) significant.

In order also to test a CFD code in the case of fluid flow coupled buoyancy effects, it is then of interest to study the fluid flow for the case when the reattachment zone of the BFS is heated (“heating plate” in Figure 2). Considering the very particular thermal-hydraulic properties of liquid metals (Pr<<1), this buoyancy effect cannot be neglected, even for very high Reynolds numbers.

The objective of the 3D vertical BFS is to provide the scientific community with reliable and accurate experimental data on velocity profiles, temperature profiles, mean average reattachment point and turbulent heat flux. On one hand the data will be used for CFD code validation and, in the other, very interesting insight into the physical phenomena of fluid flow and convective heat transfer regime transition will be gained as well. 

The data will be measured with a set of permanent magnet probes (PM-probes). These probes are capable of detecting very low mean-average velocities locally. When thermocouples are used instead of regular copper electrodes, local temperature fluctuations can also be measured. Special efforts in the further development of these probes will be made in order to be able to measure velocity fluctuations accurately and to be able to calculate the turbulent heat flux. This still remains a challenge due to the very low signal voltage (nanovolts).

               Figure 1: DITEFA Test Facility                   Figure 2: Sketch of backward facing step experiment

For more information about the project, please click on the following links:

Project A1: Thermo-hydraulic flow in a sudden expansion

Project B1: Transition between free, mixed and forced convection


Open positions within the DITEFA-project: