ISOLDE Coaxial Sampling Probe Prepared by Gennaro Bozza

ISOLDE Coaxial Sampling Probe Prepared by: Gennaro Bozza 06 -06 -2012 01/03/2021 G. Bozza, E. Da Riva 1

1 st Attempt: sharp elbow As a first attempt, Enrico and I tried to realize the coaxial geometry by just lengthening along the stack axis the 45° cut pipe. Unfortunately this attempt didn’t work as the existing mesh doesn’t suit the new coaxial pipe. As a result, the velocity pattern provided is not physical. Indeed, the velocity direction at the inlet is positive, just as the flow were exiting instead of entering the pipe. 01/03/2021 G. Bozza, E. Da Riva 2

1 st Attempt: sharp elbow Arrows are clearly flowing in the opposite direction than the main stream. 01/03/2021 G. Bozza, E. Da Riva 3

1 st Attempt: sharp elbow We tried to adjust the mesh in the following two ways: • The difference in cells length along the mainstream direction was too sharp at the sampling inlet when swapping from inside to outside the sampling pipe. Therefore, we added a grading to make it gradual. Unfortunately this didn’t solve the problem. • We added a refinement box at the sampling inlet, right where the solution appeared to be wrong. The result is not available yet 01/03/2021 G. Bozza, E. Da Riva 4

2 nd Attempt: elbow with a curvature ratio = 1 In order to definitively solve the problem, we realized a new geometry and a new mesh, in which: • the elbow is not sharp but has a curvature ratio of 1 • the mesh has a layer not only at the inner wall of the sampling inlet, but also at the outer wall. As a result, the velocity patterns appear to be excellent, and very close to a 2 D check-simulation Enrico did for the purpose. 01/03/2021 G. Bozza, E. Da Riva 5

2 nd Attempt: elbow with a curvature ratio = 1 01/03/2021 G. Bozza, E. Da Riva 6

2 nd Attempt: elbow with a curvature ratio = 1 All the arrows are entering the pipe 01/03/2021 G. Bozza, E. Da Riva 7

2 nd Attempt: elbow with a curvature ratio = 1 Being the 2 nd Attempt fine, the particle loss was evaluated. Here’s a graph with the comparison of sampling efficiencies between Coaxial and 45° Cut inlet probe design. 01/03/2021 G. Bozza, E. Da Riva 8

ICEM Attempt At the same time, some simulation with the new ANSYS mesher ICEM are being run. ICEM is a powerful mesher, but tough to use for new users, being in fact not intuitive as it operates through the new concept of “blocking” which should, at least according to the developers, enhance meshing capabilities. ICEM allows to directly mesh CAD geometry imported by, for example, CATIA, in the. stp format which preserves, unlike the. stl format, all the geometry features. 01/03/2021 G. Bozza, E. Da Riva 9

ICEM Attempt Mesh near the sampling pipe 01/03/2021 G. Bozza, E. Da Riva 10

ICEM Attempt Mesh near the sampling pipe inlet 01/03/2021 G. Bozza, E. Da Riva 11

ICEM Attempt Mesh at the sampling outlet Mesh at the stack outlet 01/03/2021 G. Bozza, E. Da Riva 12

ICEM Attempt Mesh of the sampling pipe 01/03/2021 G. Bozza, E. Da Riva 13

ICEM Attempt • This mesh has a good quality (as confirmed by the quality-check algorithms of ICEM) • This mesh was correctly exported into FLUENT • This simulation is being run and results will be available within a few days. 01/03/2021 G. Bozza, E. Da Riva 14