- Fluka
studies - Luis
Showed detailed results for peak temperature rises from Fluka study
using full Cu spoiler; Al+graphite and Cu+graphite "option 2";
and summarised all spoiler material/geometry configurations
investigated. Conclusion is that Ti-allow+graphite option 2
is most robust against damage, full Cu is most easily damaged,
Al+graphite composite would not melt but may fracture
(temperature for this was not known). Luis had also written a
draft Eurotev memo on his fluka studies so far, which had been
circulated to several people for comments.
Luis had also simulated the conditions for Single
pulse damage in copper by Marc Ross et al.. He estimated area of
melted region and compared with that found in Marc's paper,
found good agreement. This gives confidence in simulation for
Cu. Noted a systematic difference (Fluka prediction
~100micron2 larger than observed, but "hole" size in
paper is where material vaporised, so expect larger area of
melt, plausible this is consistent with simulation. Very
promising!
Adriana (outside of meeting) was also to write a similar note
summarising her studies in Geant4.
- ANSYS
update - Justin (for George)
Used energy deposition
from 250 GeV beam with 10x10x500 micron mesh from Luis, and
considered effect of this within a small model (to keep
calculation time sensible when considering long time periods)
of 1000x100x30000 microns. Larger mesh was used to avoid
warnings about aspect ratios (Justin commented this
may not be a real problem, warnings are when any
aspect ratio>5). Studied thermal rise initially for 1, 2
bunches, linear rise observed. Plan for future to have
variable mesh size, smallest in beam region. Temperture rises
now found to be compatible with Luis, previously had been much
smaller due to larger mesh sizes considered.
In ANSYS,
have three options for volume elements: heat transfer, stress,
or "coupled field" (combination of previous two - more
flexible, but less precise than either separately). With
these coupled field elements, found temperature rise ~10k
lower than the 231K using thermal elements, and ~17k less for
two bunch heating, and same mesh cell gave max. rise in both
cases.
Showed time dependence of the von Mises stress of
the mesh cell which had largest temperature rise after a
single bunch. Noted also general rise in mean stress from
~2.108 to ~4.108 modulated by
oscillatory behaviour (with period ~100ps). Max. stress found
(~740MPa) below ultimate tensile stress and yield stress
(880MPa), but only room temp. data...
Future work: To study max. stress at any point in the mesh as function
of time, rather than of a fixed mesh cell, and reduce time
steps to get increased precision on oscillatory behaviour
(could be an artefact?).
- Damage proposal - discussion
Frank commented on beam size at ESA - the 20 micron spot size
which had been considered assumes zero dispersion. With best
information on dispersion from Jan. 2006 run, more likely that
spot size would be closer to 100 microns - more measurements
need in Apr/May run to confirm this. If so, ESA would be less
useful for damage tests. Note that earlier test by Marc Ross
et al had spot size ~7 microns.
Roger J. had photo of
jig which could hold samples at TTF2; samples would be
20.0+/-0.1mm diameter, max. thickness 4+-0.1mm. For Ti-alloy, this
would limit samples to ~0.1X0, i.e. very similar to
coupon tests with 1.4mm thick Cu at SLAC.
Discussion about what could be learned from beam test and
extremely tentative schedule:
- 3-Mar: 2-3 page proposal written
- ~9-13 Mar (at LCWS): Roger J. to take this for
discussion with e.g. Nick Walker/Marc Ross
- ~22-Mar: follow up by email/set up phone discussion with
additioanl interested people eg Keller, Seidel not at LCWS.
- Apr-May-Jun, organisation/setting up at DESY?
- Jul - test at DESY??
