Ansys - Old Geometry - Cathode 1 Ansys - New Geometry - Cathode - - PowerPoint PPT Presentation

Ansys - Old Geometry - Cathode

1

Ansys - New Geometry - Cathode

• lamella (PCB and copper strips) geometry is closer to realitiy
• additional gas volumes

2

Ansys - New Geometry - GEM unit cell

• Kapton layer with gas holes
• complete unit cell
• expansion in x-direction → later: in Garfield++ periodicity in z-direction

3

Ansys - Full geometry

• pink: gas
• Cathode area with

lamella (cyan)

• Drift area
• Partial GEM foil

(about +/- 2 mm from bottom of active volume) and gas volume at left and right

• Induction area

corrected gap size compared to geometry showed before

4

Ansys – Partial GEM Foil - Detail

5

Ansys Geometry - Meshing - Node Minimization

• about 441.500 of 512.00 nodes used
• Challenge: geometry with complete GEM Foil or 3 lamella with partial

GEM exceeds free license node limitation

• Node

minimization for different volumes via ESIZE

• no mid-node and

angle warnings

• less than 1% shape

warnings

6

Potential distribution - new geometry with GEM

• diffx = 90V
• diffy = 400V
• diff(cathode-GEMO) = 100V
• Diff(GEMO-GEMU) = 300V
• Add voltage at bottom of induction

area!!!

17.4

7

Potential distribution GEM

• +/- 2 mm from bottom of active volume
• 300 V between GEM top and bottom
• without cathode

16.4

8

Potential distribution without GEM

• diffx = 400V, diffy = 90V, diff(anode) = 100V
• voltage applied on same copper stripes (in new geometry on inner stripes)

→ slope of lines with same potential differs in active volume of both geometries

20.4

• ld geometry

new geometry without GEM

9

Resulting contours of electric field (y-component)

• diffx = 400V, diffy = 90V, diff(anode) = 100V
• voltage applied on same copper stripes (in new geometry on inner stripes)

→ contours of the electric field show differences

20.4

• ld geometry

new geometry without GEM

10

Efficiency of electron detection

• diffy = 400V, diff(anode) = 100V, diffx = changed

→ about same maximum detection efficiency reached → but maximum moved from about diffx_max=88V (old) to diffx_max=115V (new) → gas volumes have influence on potential distribution and also on electron distribution

21.4

• ld geometry

new geometry without GEM

11

TO DOs

install new Ansys Simulation and get it to work get GEM geometry to work (convergence problem) add voltage at bottom of induction area material properties take closer look at Garfield++ simulation (distribution of detected electrons, influence of GEM, correlation of y- position of created electrons and detected charge?, ...)

12

Ansys 2020 R1 – finally working

• adapt my geometry script
• adapt script for running

complete simulation including ansys2020 and Garfield++

• 1st: compare simulation

results from ansys15 and ansys2020 to check if it’s working correctly

12.5

13

Potential distribution: Ansys 15 vs. Ansys 2020 R1

12.5

14

• diffx = 400V, diffy = 90V, diff(anode) = 100V
• voltage applied on same copper stripes (in new geometry on inner stripes)

→ same potential distribution

Efficiency of electron detection: Ansys 15 vs. Ansys 2020 R1

13.5

15

• diffy = 400V, diff(anode) = 100V, diffx = changed

→ same detection efficiency for same voltage combinations → no differences: adapted scripts seem to work correctly

Continuing: Efficiency of electron detection

13.5

16

• diffy = 400V, diff(anode) = 100V, diffx = changed

→ about same maximum detection efficiency reached → but maximum moved from about diffx_max=88V (old) to diffx_max=117V (new) → gas volumes have influence on potential distribution and also on electron distribution

• ld geometry

new geometry without GEM (ansys2020)

Task: Add complete GEM Foil

12.5

18

• Potential distribution for diffx = 90V, diffy = 400V, diff(cathode-GEMO) = 100V,

Diff(GEMO-GEMU) = 300V

• complete GEM: areas with 0V

partial GEM complete GEM

Closer look at GEM Foil

12.5/13.5

19

• add partial GEM: (13+x+3) mm
• ∆U(top_GEMO)=100V; ∆U(GEMO_GEMU)=300V; ∆U(GEMU_bottom)=100V

Closer look at GEM Foil

12.5/13.5

20

• add partial GEM: (13+x+3) mm

Closer look at GEM Foil

12.5/13.5

21

• add partial GEM: (13+x+4) mm
• ∆U(top_GEMO)=100V; ∆U(GEMO_GEMU)=300V; ∆U(GEMU_bottom)=100V

→ same voltage → 1 mm more at the right compared to previous geometry → but different potential distribuation

Closer look at GEM Foil

12.5/13.5

22

• add partial GEM: (13+x+4) mm

→ volume with 0V → Electric field vectors “stay” in their volume

Field in GEM

08.5

23

Update: Closer look at GEM Foil

14.5

24

• whole GEM volume is now embedded in one gas volume including drift and induction area
• ∆U(top_GEMO)=100V; ∆U(GEMO_GEMU)=300V; ∆U(GEMU_bottom)=100V
• left: 5 GEM unit cells
• right: extend to 121 GEM unit cells

Electric field vector

14.5

25

Add cathode

20.5

26

• with about 5mm (121 unit cells) gas volume at right → VGLUE error ↯ → use

4mm at right

• also without cathode the number of GEM unit cells is limited → with 141 unit cell:

VGLUE error (maybe VSBV doesn’t work) ↯

• Potential distribution for diffx = 90V, diffy = 400V, ∆U(cathode_GEMO)=100V,

∆U(GEMO_GEMU)=300V, ∆U(GEMU_bottom)=100V and electric field vectors

Add GEM step by step

19.5

27

GEM metal top and Kapton layer

• nly GEM metal top

Add GEM step by step

20.5

28

complete GEM: potential distribution

diffx = 90V, diffy = 400V, ∆U(cathode_GEMO)=100V, ∆U(GEMO_GEMU)=300V, ∆U(GEMU_bottom)=100V → no electrons can be detected ↯

Contours of the electric field in (y-component)

Add GEM step by step

20.5

29

diffx = 90V, diffy = 400V, ∆U(cathode_GEMO)=100V, ∆U(GEMO_GEMU)=300V, ∆U(GEMU_bottom)=100V → seems like electric field in GEM is not homogenous ?↯

Contours of the electric field in (y-component) corresponding vector plot