29.08.2018 Views

atw 2018-09v3

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

<strong>atw</strong> Vol. 63 (<strong>2018</strong>) | Issue 8/9 ı August/September<br />

470<br />

AMNT <strong>2018</strong> | YOUNG SCIENTISTS' WORKSHOP<br />

| | Fig. 3.<br />

Comparison of simulations with empirical contact angle model and the laboratory experiments by Becker<br />

Technologies [5] (in false color representation) with mass flow rate ṁ = 11 g/s (ṁ =12 g/s for inclination<br />

of 10°), three different inclinations (left 2°, middle 10°, right 20°) and without aerosol loading.<br />

given water velocity parallel to the<br />

surface such that a specified mass flow<br />

rate is achieved. The flat plate is<br />

bounded by vertical sidewalls and<br />

has an inclination angle α. Material<br />

properties of water and air are used.<br />

For snapshots of the resulting flow<br />

fields see Figure 3.<br />

The simulations are conducted<br />

with the empirical contact angle<br />

­model and the filtered initial<br />

­randomized contact angle field [6].<br />

The contact angle is specified in the<br />

boundary conditions of the water field<br />

and is taken into account to calculate<br />

the curvature of the water-air interface.<br />

The contact angle has a huge<br />

impact on the formation of rivulets<br />

and their stability as shown in previous<br />

studies [7]. The empirical contact<br />

angle model accounts for the<br />

wetted history and therefore enforces<br />

a spatially and temporally stable<br />

­rivulet flow.<br />

4 Simplified geometry<br />

This study also considers a simplified<br />

geometry with dimensions of 6 cm<br />

x 5 cm, 60° inclination and different<br />

water loadings. As a first step the<br />

­simplified geometry, for which additional<br />

benchmark data from CFD<br />

simulations and experiments are<br />

available, is used for the parameter<br />

variation to save computational effort<br />

and time. Later the findings are<br />

transferred to the larger laboratory<br />

geometry. Also the experimental<br />

data can be used to investigate the<br />

empirical contact angle model [6] in<br />

another scenario than the laboratory<br />

geometry where it was developed. For<br />

the simplified geometry Singh et al.<br />

[8] provide results of CFD simulations,<br />

as do Hoffmann [9] and Iso et.<br />

al [10]. Experiments are conducted by<br />

Ausner [11]. All of the latter use the<br />

identical geometry, but different inlet<br />

conditions (overflow weir and feed<br />

tube) and various simulation tools<br />

(Singh and Iso Fluent, Hoffmann<br />

CFX). In the present study simulations<br />

with constant contact angle and with<br />

empirical contact angle model are<br />

performed. The results for different<br />

Weber numbers are evaluated and<br />

compared to the results of the studies<br />

mentioned above for validation. Five<br />

different Weber numbers (We = 0.02,<br />

We = 0.24, We = 0.47, We = 0.76 and<br />

We = 1.10) are investigated, which<br />

correspond to an increasing water<br />

mass flow rate:<br />

We =<br />

with liquid density ρ l , inclination<br />

angle α, volumetric mass flow rate Q,<br />

surface tension σ, plate width W<br />

and viscosity μ. As the water load<br />

­increases, the flow pattern changes<br />

from a thin rivulet to a more pronounced<br />

rivulet to a fully wetting<br />

­water film (see Figure 4).<br />

The influence of the side walls<br />

is also clearly visible and was also<br />

observed by Hoffmann [9] and Ausner<br />

[11]. With a constant contact angle of<br />

70° (which is the value frequently<br />

quoted in the literature for the material<br />

combination water on steel) the<br />

percentage of wetted area in the<br />

present CFD calculations and in the<br />

calculations of Hoffmann and Iso<br />

tends to be underestimated, whereas<br />

the similar setup of Singh yields, for<br />

an unknown reason, larger values<br />

of wetted area. In Figure 5 the<br />

measurements are shown as blue<br />

triangles, the results of Hoffmann, Iso<br />

and Singh in purple, yellow and green,<br />

respectively, and the current calculations<br />

with the different contact angles<br />

in red, gray and black. The simulations<br />

with constant contact angle and<br />

empirical contact angle model with<br />

70° for a dry surface and 50° for a wet<br />

one still slightly underestimate the<br />

wetted surface. With the empirical<br />

contact angle model 30°/70° the<br />

results are very well within the variation<br />

of the experiments.<br />

5 Wash-off model<br />

The particle wash-off consists of a<br />

two-stage process. First the sedimented<br />

particles on the plate floor are<br />

| | Fig. 4.<br />

Comparison of simulations with empirical contact angle model with<br />

θ dry = 70° and θ wet = 30° for different Weber numbers We. The water height<br />

is indicated by color.<br />

| | Fig. 5.<br />

Normalized wetted surface A wn for different Weber numbers. Blue triangles<br />

indicate the experimental results; results of CFD simulations are displayed<br />

with differently colored lines.<br />

AMNT <strong>2018</strong> | Young Scientists' Workshop<br />

Development and Validation of a CFD Wash-Off Model for Fission Products on Containment Walls ı Katharina Amend and Markus Klein

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!