MONTECARLO CALCULATIONS OF PRESSURE PROFILES FOR THE ESRF DIPOLECROTCH AREA

1

MONTECARLO CALCULATIONS OF PRESSURE PROFILES FOR
THE ESRF DIPOLE/CROTCH AREA R. Kersevan Vacuum
Group, Tech.Serv.Division, ESRF, Grenoble

ALBA-MaxLab Workshop on Vacuum Systems for
Synchrotron Radiation Light Sources, Barcelona,
12-13 Sept 2005
2

• Motivation
• Among many other things, during the summer 2005
  shutdown (Aug 2005), the vacuum group has carried
  out
• Vertical re-alignment (-2 mm vertical
  displacement, no venting) of all 64 crotch
  absorbers in the machine, following the failure
  (water leak) of one of them1,2
• First time in years that such a global
  intervention was carried out
• One crotch has been removed and replaced with a
  brand new one (cell vented)
• Synchrotron radiation-induced dynamic pressure
  rise seen in all 32 cells
• MC simulations done in order to get an estimate
  of the outgassing yield of each crotch
• ? Work in progress ?

(1) Post Mortem Analysis of Cell-15s Crotch
Absorber Water Leak Failure, R. Kersevan, this
workshop (2) Identifying Vacuum Problems and
the Recovery from Failures, D. Schmied, this
workshop
3
ESRF STANDARD CELL IP/NEG DISTRIBUTIONAND
LOCATION OF VACUUM GAUGES
PEN6
PEN3RGA
PEN4
PEN1
PEN5
PEN7 RGA
CELL LENGTH 844 / 32 26.4 m
5000 mm
Ion pumps (Varian) ID120 l/s CV4/CV11 120
l/s CV3/CV10/CV15 45 l/s dipoles 60
l/s crotches 400 l/s
NEG pumps (SAES) crotches GP500 elsewhere
GP200
4

Global Vacuum Conditioning after Re-Start Beam
lifetime lower than usual (20 - 25 hours less)
at beginning of USM

End of machine-dedicated time X-ray beam to users
5

Local Vacuum Conditioning after Re-Start Cell
6 replacement of crotch-2 re-alignment of
crotch-1 (-2 mm V) new CV3 chamber (precedes
dipole-1) partially pre-conditioned NEG-coated,
extruded Al CV5073 (not shown)

PEN4 CV4 PEN5 CROTCH-1 PEN6 CV11 PEN7
CROTCH-2
6

CROTCH-2 AND CV13 GEOMETRY Total flux _at_200ma
5.82E18 ph/s (46mrad)
TOP-VIEW
to FE absorber
X-rays from preceding dipole (430W)
X-rays from this dipole (7200W)

PEN7 location
lt --------------------------- SIDE-VIEW /
CUT-OUTS --------------------------- gt
7
DIPOLE-2 (CV12) GEOMETRYradius of curvature
23.366 m
Pumping-port geometry simulated separately

8

400 l/s SIP GP500 NEG simulated total pumping
speeds of 500, 750, 1000 l/s
pressure test facet 320x20mm2
PEN7

crotch absorber
Closer view of the Molflow simulation model
total number of points 339 total number of
facets 131 computation time 15 hours
9

400 l/s GP500
GP200 Pumping Port
60 l/s SIP Pumping Port
PEN7 (star symbols)
10

ESRF DIPOLE PUMPING PORT GEOMETRY MODELED AND
SIMULATED WITH MOLFLOW

320 mm
B
CPPORT
280 mm
A
C
beam chamber molecules are generated here
D
CPPORT
CPPORT
sticking coefficient s1.0 ? 913 l/s s0.0656 ?
60 l/s
11

The effective pumping speed is calculated via
Seff 1 / ltPgt

CPPORT
A
B
pressure test facet
sticky surface (pump)
12

CPPORT
CPPORT
The effective pumping speed is calculated via
Seff 1 / ltPgt Statistical error proportional to
1/SQRT(molecules generated)
13
Estimation of Gas Load Q

• Photon Flux F
• 46 mrad of dipole radiation correspond to
  5.82E18 ph/s/200mA (ephgt10 eV)
• Outgassing load Q Fkh 0.236h mbarl/s
• with k4.05E-20 mbarl/mol (_at_20C)
  hSR-induced outgassing yield (mol/ph)
• Substituting into the graph seen before,
• assuming the case
• 1000, 60, 200 l/s,
• and comparing with
• Pmeas4.0E-7 mbar
• (_at_ D0.1 Ahr ? 5.5E22 ph/m),
• one gets
• h 3.2E-4 mol/ph

14
Effect of Switching On/Off the IP

• switching off the 400 l/s IP and looking at the
  effect on pressure at PEN7. Pressure increases
  1.7x (1.4E-9 ? 2.4E-9)

Ratio of PEN7 for the two cases (1000,60,200) and
(500,60,200) is 2.25 / 1.38 1.63
15
CONCLUSIONS

• An example of use of the montecarlo code Molflow
  has been shown. The code has been used to
  evaluate
• the 3D geometry of ESRF dipole-crotch absorber
  vacuum chambers
• the conductance of its two side pumping-ports
• an estimate of the outgassing yield of the
  crotch absorber (obtained by comparing
  calculations with measured pressures)
• a reasonable agreement of calculated vs
  measured pressure rise when the 400 l/s IP is
  switched off
• These preliminary results are encouraging, more
  to come soon
• The contribution to this presentation from
  I.Parat, is acknowledged