Mean surface roughness (Ra) vs. impact angle distribution --> latter clearly more relevant for rough surfaces --> action for next year's tasks to take this into account
Dynamic changes in the impact angles due to plasma exposure: analyses of samples before and after the experiments --> Calls for Fourier transform of surface morphology patterns
Dynamic changes in roughness call for comparison also with data from linear machines --> next year's task
3
Effect of surface roughness on erosion of fusion relevant materials (JSI)
Overview of the project of samples with controlled roughness (produced by ENEA) for lab studies
Sputter yields agree with SDTrimSP-3D predictions generally except for the highest roughnesses
Discussion:
Christian: using AFM as input - how many figures taken? Only one randomly selected --> should one take rather a statistical average over larger set of images (like 10)?
Udo: deconvolution of the response function from AFM? This may mean that the angular dependence is not fully caught --> correction factor may be necessary for better match with experiments and SDTrimSP
Andreas: even small depressions on the surface impact the sputtering yield via deposition; production of atomistically smooth surfaces however very challenging
Smooth surface wrt. collision cascade --> not only length scale but also energy plays a role (typically: < 1 nm)
Roughness can be inspected on different length scales - all these provide different overviews of the studied surfaces
Different QCMs used for sputtering experiments --> angular dependence of sputtering yield
Roughest samples result in "collimation effect"
Data from nano-columnar W surfaces deviates largely from SDTrimSP predictions
Discussion
Sebastijan: PWIE request for off-normal operation (resulting from e.g. melting), possibility for detailed surface characterization of W monoblocks & supplemented by sputtering studies of "simulated samples" (resembling ones resulting from off-normal events)
How about measuring the velocity distribution of sputtered particles? Maybe in connection with a reference experiment --> combination of a task for fundamental and practical understanding (is our model sufficient in predicting what would happen during the 2nd discharge of ITER or so); Udo thinks no major deviations would occur (based on MD) but interesting to probe it anyway
Udo: how robust is QCM for rough samples? lateral inhomogeneity for deposited material may occur; in the end the question is on making a compromise
5
Role of roughness in sputtering process of W by GyM He plasma (ENEA)
Doppler shift used as an indicator for roughness (HRS diagnostics in PSI-2)
Roughness before/after plasma operation --> requires specific set of samples --> make use of already existing samples (ENEA, OEAW,...) --> Oleksandr provides us with PSI-2 boundary conditions
Discussion
Udo: Removal of Al oxides may have affected some of the observations? Clearly also Al sputtered --> further investigations needed to prove the "coolness" of the method.
7
Discussion
8
Study on nano-columnar tungsten under Ar irradiation (VTT)
Impact angle good parameter for instantaneous sputter yield estimates but does not take into account fluence --> model extended to impact angle distributions
Also local atomistic structure should be taken into account --> Crystal-SDTrimSP developed (inclusion of lattice structures) --> explains existing experimental data well (factor of 2-8 dependence between varying crystallographic directions)
Discussion:
Different values in different crystallographic directions may impact experiments as well --> SP B link
However, dependence on crystallinity not that strong following multiple erosion-deposition steps --> factor of 2-8 dependence relevant only in the start-up phase; in a tokamak the structure around the OSP may not have been preserved following exposure to high fluences (in a magnetic environment)
Surface roughness and morphology & their implementation in ERO2.0 --> comparison between ERO results and samples (pyramidal structures, different roughnesses) from GyM exposures
Input
SEM and AFM data of all samples & weight-loss measurements
statistical analyses of W thickness variations after exposure
Obtained up to now
implementation of surface mesh
studies on the dependence of mesh resolution and surface-evolution model
SOLPS-ITER plasma background for GyM cases
Discussion:
Klaus: is triangular-surface model appropriate for modelling surface evolution? smoothing may become an issue; in modelling time step may need to be reduced to diminish smoothing?
Sebastijan: evolution should be also assessed based on very-high fluence exposures in MAGNUM-PSI --> calls for comparisons
12
Discussion
Re-deposited layers not yet addressed --> calls for actions, both experiments and modelling
Which model to be used for modelling re-deposits? either explaining experimental data or building "artificial models" into MD
Issue with tokamak surfaces in the finest scale --> use surrogate models, mean surface angle etc. --> data for understanding global migration picture in tokamaks
Which parameters are the most important to be included in the models (that would have the largest impact on the results)? Trends more important than exact details? Roughness decreasing during erosion but also cases where roughness may increase --> all these points to be addressed in the future