28 June 2021 – 02 July 2021

Simulation parameters:

AbstractBuffered chemical polishing (BPC) is the reference surface polishing adopted for ESS and MYRRHA SRF spoke resonators at IJCLab. This chemicaltreatment, in addition to improving the RF performance, fits into the frequency adjustment strategy of the jacketed cavity during its preparationphase. In the framework of the collaboration with Fermilab for PIP-II project, IJCLab has developed a new setup to perform rotational BCP. Theimplementation of a rotation during chemical etching improves significantly the homogeneity and quality of surface polishing. In this paper, wepresent the numerical analysis based on a fluid dynamics model. The goal is to estimate the acid flow characteristics inside the cavity, determine theinfluence of several parameters as mass flow rate and rotation speed and propose the best configuration for the new experimental setup.

CONCLUSION

Jules Demercastel-Soulier†, P. Duchesne, D. Longuevergne, G. Olry, T. Pépin-Donat, F. Rabehasy, D. Reynet, S. Roset, L.M. Vogt, Université Paris-Saclay, CNRS/IN2P3, IJCLab, Paris, France

Figure 6 : Flow rate of 5L/min and rotation at 0,1 rad/sCavity cooled with forced water flow

Figure 5 : Flow rate of 5L/min without rotationCavity cooled with forced water flow

Figure 4 : Flow rate of 5L/min without rotation, Cavity cooled with air

0,0

20

2

0,0

26

2

0,0

10

2

0,0

10

4

0,0

10

1

0,0

08

8

0,0

10

8

0,0

15

9

0,000

0,005

0,010

0,015

0,020

0,025

0,030

0,0E+00

1,0E-01

2,0E-01

3,0E-01

4,0E-01

5,0E-01

6,0E-01

7,0E-01

8,0E-01

9,0E-01

WithoutrotationFlowrate5L/min

withoutrotation

Flow rate17L/min

Flow rate: 5L/min

Rotation :0,1 rad/s

z

Flow rate: 5L/min

Rotation :0,12 rad/s

z axis

Flow rate: 4L/min

Rotation :0,1 rad/s

Flow rate: 5L/min

Rotation :0,1rad/s y

axis

Flow rate: 5L/min

Rotation :0,1rad/s yaxis only

1 inlet

Flow rate: 5L/min

Rotation :0,1 rad/s

y axishorizontal

Sta

nda

rd d

evia

tion

Etc

hin

g r

ate

in µ

m/m

in

Mean Minimum Maximum Standard deviation

Inlet

Outlet

In conclusion, the etching rate of a cavity depends on the rotating speed and the mass flow inlet. To improve homogeneity of etching, it isimportant to have a good balance between both of them. From our simulation, a flow rate of around 5 L/min for both inlets which is around 0,125kg/s and a rotating speed of around 1 rad/s correspond to an optimal configuration. Moreover, these parameters values are easy to achieve withthe new rotational BCP setup. Finally, too much flow rate or higher rotational speed will increase the etching rate discrepancies between thedifferent regions of the cavity. To go further, we will try new CFD calculations (transient and multiphase analyses) with different cavity fillingfactors and also estimate the overall RF frequency shifts from the calculated local etching rates.

Fluid properties (acid)

Density 1530 kg/m3

Specific heat 2330 j/kg.K

Thermal conductivity 0,236 W/m.K

Viscosity 0,022 kg/m.s

Solid Properties (Niobium)

Density 8570 kg/m3

Specific heat 265 j/kg.K

Thermal conductivity 53,7W/m.K

Heat Flux 1200 W/m²

Model k-epsilon realizable

Spatial dizcretization Second Order Upwind

Pressure-velocity Coupled

INTRODUCTION

Figure 1 : BCP setup currently used at IJClab(Myrrha cavity on the left, ESS cavity on the right)

Figure 2: Overview of the rotational BCP tooling

Figure 3 : fluid domain used in the simulation

Figure 9 : Etch rate in µm/min Flow rate of 5L/min and rotation at 0,1 rad/s

Figure 8 : Etch rate in µm/min Flow rate of 5L/min without rotation

Improvement of Chemical Etching Capabilities (BCP) for SRF Spoke Resonators at IJCLab

Buffered chemical polishing is the chemical treatment performed at IJCLab on theSpoke resonators equipped with their helium vessel (Fig. 1). The total amount ofmaterial removed from the cavity surface by BCP is around 200µm in order toeliminate the damaged layer created during the manufacturing. Today, the cavity isstatic during the treatment. The average removal rate measured on ESS andMYRRHA cavities by weighing is between 0,25 and 0,6 µm/min. The chemicaltreatments induce a frequency variation. For the Myrrha single spoke cavity, theaverage frequency sensitivity is +1kHz/µm. A new setup to perform rotational BCPwas developed at IJCLab for SSR2 cavities for PIP-II project. The implementation of arotation during chemical etching improves significantly the homogeneity and qualityof surface polishing. A gear mechanism makes the cavity rotate slowly back andforth with an amplitude of +/-180° (Fig. 2).

CFD SIMULATIONSThe simulation were done using ANSYS Fluent. The fullgeometry of the fluid domain and the Niobium walls areconsidered. A double mass flow inlet at 6°C and outletare defined.

Thermal analysisThe reaction between acid and niobium is an exothermic reaction. When the cavity remains fixed, wecan see that the air cooling at 20°C is less efficient than a water cooling system at 15° C and themaximum cavity walls temperature can achieve 47°C (320K) (Fig. 4 and Fig. 5). With a constant rotationof the cavity (0,1 rad/s), the cavity wall temperature is nearly uniform, around 15°C (Fig. 6). Rotatingwalls drag the fluid thanks to its high viscosity. This kinetic energy helps cooling down the cavity becausethe fluid is not stationary and the exothermic energy of the reaction is dissipated as the fluid is moving.

Etching rate analysisFor the study of etching rate, we used the following equation giving the etching rate versusthe turbulent kinetic energy (TKE) described in [1]:

Etch rate = 3.1585 x TKE0.186 µm/min

Without rotation, the etching rate on the SSR2 cavity is estimated by simulation around0.25 µm/min. Compared to a static treatment, the rotation improves the homogeneity andthe efficiency of the surface polishing: less standard deviation and higher average etchingrate (Fig. 8, 9 and 10). An optimum was found around 5L/min for the flow rate and arotating velocity of 0.1 rad/s. The average etching rate is then 0.42 µm/min. Using the Yaxis as rotating axis gives very similar results in terms of etching rate.

Figure 10 : Histogram of etching rate and standard deviation