ALARM2ESS!
BEWARE OF TOF SINGLE XTAL DIFF!
MY KEY TECHNICAL INPUT TO ISIS :
10 YEARS ALSO UNJUSTLY UNDERRATED
IN COMPARISON WITH OTHER STAFF!
I
- ISIS SX (Magnetic) DIFF DVT :
QUOTING MATTHIAS HIMSELF:
"OH BOTHER! TEN YEARS LATER...."
NOTHING CHANGES OR IT'S WORSE !
SX
TOF = FORGET ABOUT SMALL
MOLECULE (& MAG) PRECISE BRAGG :
SCATTERING : PASCAL's
SCARY PLOT WILL
MAKE YOU SCREAM FOREVER !
|
3 years On SXD 2005-2008 A. Daoud-Aladine + Matthias Guttman |
7 years On WISH 2012-2019 P. Manuel + F. Orlandi |
|
Software Dilema: Work with what exisist SXD2001 IDL or rebuilt all from Scratch? ("Dirty prog" Block buster & not modular code almost impossible to be dvt in collaboration with third party)
|
Software : Work with Mantid C++ (in
dvt) |
|
Conclusion : Using SXD2001 + Lots of pearl/python scripting to retrive equivlent hkl mesured in different setting => Exctinction = BIG PB : 100% correlated to mag moment in simplest ONE param AF! |
Conclusion : Same : Beckker Coppens seems to be insufficient, primary/secondary are known to be highly correlated .... (not discussed..) |
|
Correlation with Absorption ? No Discussion ! |
Same... No Discussion neither about exponents different to Extinction law lambda dependance = idicating the effect on intensites of "hkl set" dependant corrections, far more complex than what's modelled! |
| Strategy :
Generate/test single crystal MCstas module containing a simulation of
extinction process refinable by SXD2001 (visualisation done) +
ultimately compare simulated/real intensities (not achieved): Huge interest from P. Willemdrup (MCStast): but abandonned after move on HRPD (Summer 2008) This meant: My single crystal expertise = early Warnings original management plans are far too ambitious given difficulties, DENIED! |
? |
|
Refinements : though not in Ideal way & with ergonomy forcing integration of all bank, not enabling access to single reflexions... Propagation Vectors are In SXD2001 => Though not in optimised way (loosing to
much time!) Peaks can be integrated to Test real Mag Str. refinement
with FullprofDone against comparison with ILL
eq. * |
Refinements Wait for Vicky?= Politic view => Test on Crystal structure NOT on a mag struture => making of the talk title a mere lie!!!! |
|
Coverage, expt planner: |
Coverage, expt planner : Other big issue |
|
Is Matthias (in my time, well settled in his « niche »**) working & collaborating with them, or is he delegating that to Sylvia Capelli...? |
Gonio ready BEFORE data treatement stuck@no more than 30 % (an optimistic figure) of dvt compared to SXD2001 capabilities*** !!! |
*eg on TbMnO3 : implying Juan Rodriguez-C insulted by the INS group through me (& Loreynne Pinsard, bringing in ISIS a Orsay University crystal!) in the developement of SXD/magnetic diffraction!
** I analysed and compared the single crystal output of ILL and ISIS during my end in ISIS (winter 2014): Bragg scattering is restricted to few studies where Data redudancy is large enough (large cell) = after Chick Wilson, SXD has been "saved" by Matthias program enabling studies of Diffuse Scattering
=> APPART FROM THIS COLLABORATION
I HAD ONLY 2 MORE PUBLICATION ON SXD
OVER 3 YEARS (=> my move on HRPD), ONE,
& ANOTHER SAVED BY HRPD!
WHILST P. RADAELLI & CHAPON
BUILD THEIR SCIENCE EXPERTISE/FAME
BY APPLYING & DOING ALL THEIR SX
MULTIFFEROIC WORK@ILL! =>
to get structure right,
or High profile
Polarised
Work
!
*** Extending the CRYSFML SXTAL Laue Suite
in FORTRAN90 to add the Time dimension can
Solve most issues in HALF THE TIME
AND 10 TIMES LESS ACADEMIC
= FAR MORE CHEAPER!
RESSOURCES
II - HRPD UPGRADE COMPARISON :
FROM
HARD WORK &
COMPETENT
=>
MADE UP INCOMPETENT WORK!!!
|
~ 2 years Sept 2013 – Eviction March 2015 A. Daoud-Aladine Master directing Steve Hull PostDoc Jordi... |
~ 2 years 2018/19 A. Gibbs+D. Fortes+HRPD-X W.Group ( Who ???) |
|
MCstat workshop Sept 2013 & |
|
|
Several month of investigations… to get the table used 1st design & base concept + managenent aknowledged essential bechmark of newly provided moderator files |
The « New » Concept = Slide 25 : |
|
CONCEPT I (Proper MC simulation) Tough Software & Hardware choices : To cunningly explore the parameter space of the beamline requiring the most lenghty simulation of all ISIS instrument (100 components)… => python Scripting of an « Automate » generating
Si MCstas pattern piped to automated refinment all peaks &
outputting refined params as function of d-space** |
CONCEPT II (Vague analytic idea...) Blender pictures making all nice see slide (I wonder : made by L. Chapon ??? he's a specialist !!!) to mask that , a "out of the bleu" reuse! « Old Historical idea » : Logarithmic spiral ! = idea of time focusing ! BUT, |
|
Conclusion : due to elliptic guide + moderator finite size : beam divergence => L2 & Dphi have NO effect, resol controlled by L1 & Dtheta only : Two years of honest hard work well justifyging the expensive Upgrade |
Conclusion : given our conclusion : TIME FOCUSING WILL BE INEFFECTIVE !!! => & the design got complicated uselessly ! Two years wasting time/effort => certainly MC would prove as only, a COSMETIC improvment! |
|
Extra*** : had popular Seq ref scripts almost ready to Refine also data live & retrieve any varible parameter for plot : a user dream for data planning & optimize stats counting time ! |
|
* that lead me get a TEAM BONUS in 2014 signed by
Langdridge, who refused to insit with Mike Glazer as suggested by Kevin
Knight, I should be on his Relaxor Ferroelectric Nature paper,
Allegedly spyed with a pair of binoculars (!!!) by Dickon
Champion his Neighbor at a time! Sean Langdrige is who in the end, absurdly
signed my dissmissal me 6 month after signing that Reward as my
division Head...!!! I actually used & scripted with python Fullprof
Profile fitting input/output BECAUSE mantid fitting engine had been bugged for 4 years during which change of version were making scripts NOT BACKWARD COMATIBLE = LYING ON PRISTINE MANTID SPECs!
** It is the script Steve Hull asked me to teach
Silvia Imberti to use which made me conclude she's dangerously
incompetent! she was incapable of converting in polar coordinates (my script convention) the geom of her instrument design givin detectors locations in cartesian coordinate
: i told her that's High School math: she never came back to me: it was
almost like if implicitely, I had to do that BASIC calculation for
her!!!
*** That extra dvt, had certainly more "pratctical value" to the user community than the upgrade : you can ask, e.g Zsoltz Gersci was greatly impressed, Jeremy Turcaud, published thanks this, a Phys Rev B within 1 month after dtta collection, a record, Adroja, on site was very kind about using it etc...
LAST : GUIDE UPGRADE (2007) INDUCED BEAM DIVERGENCE
=> AFFECTED MORE THAN RESOLUTION,INTENSYTIES HAVE
SYSTEMATIC ERRORS IN NORMALISATION BY USING IN PRACTICE
VANADIUMs SHAPES MISMATCHING THE SAMPLE SHAPE
THIS IS THE MAIN HRPD ISSUE (INCORRCT INTENSITY)
*
CORRELATING ABS PARAMs & MAKING THEM
BANK DEPENDANT => HARD TO GET CONSISTENT
TEMPERATUR FACTORS : EFFECTs& CORRELATION
THAT HAVE NEVER BEEN STUDIED IN DETAILs
*
Laurent Chapon insted I correct instead
effects of divergence profile on Resolution
but this was less CRITICAL : besides, this
was anoter 2 years of useless Mantid work
IN GENERAL:
ISIS (technocratic) Philosophy,
IS SIMPLY NOT THE RIGHT ONE =
IT IS NOT ADAPTED AT ALL TO NEUTRON
SCATERRING SPECIFIC LOW BRILLANCE :
FUTURE REQUIRE FAIR REFLEXION
& COMPARISON OF CW vs TOF WORK
HIGHLIGHTING COMPLEMENTARITY
& QUESTIONNING METHODOLOGIES :
CONCLUSION : for SX diffraction, TOF is but Ok
to map Reciprocal space, but will rarely beat &
n dit will remain essentiallly qualitative, wrt to
CW Bragg-Xtallo requirement in Bragg diffration!
III - ISIS INSTRUMENT DVT
& MANGANITES SCIENCE :
THE INS INTERFERENCES =>
LEAD CRY TO DENY MY KEY
INPUT TO THE CRY GROUP
AND A SERVICE GREATLY
APPRECIATED BY ALL USERs
!
* overrating ISIS single xtal/INS*
TECHNICAL ISSUE UNTIL 2008:
FROM START PIGGY IN MIDLE!
COLLABORATE? OR COMPETE?
WITH MATTHIAS DVT OF SXD2001?
CONSIDRING WELl HOW PROTECTIVE...
=> HOW LITTLE KNOWLEDGE TRANSFER
ISIS = MANAGENENT OF EGOs CAN INDUCE!
MY "DVT" = A TOUGH BENCHMARK vs ILL!
PROVING ILL SUPERIOR : METHODOLIGICALLY!
=> UNREWARDED AS PERSONAL FAILURES! =>
CONSOLIDATING THE INS ZP SCIENCE A PRIORIs
DURING 4 YEARS = ALSO DIVERTED ON MULTIFERROICS
UNTIL I GET ILL-PROOF OF ZP SYMMETRY MARCH 2013:
||
V
MANAGENMENT STRATEGICALLY ASKING ME THEN!
TO START ON HRPD THE TITAN WORK (sept 2013):
ADDING NOW 2 INS MANAGEMENT TOTAL CTRL:
THE TOUGH HRPD UPGRADE : MEANT TO KILL ZP!
THE TIME I NEEED TO FINISH ILL D9 WORK
ON THE HALF DOPED PCMO XTAL =>
* underrating all my powder work efforts meant*
*2 bring better Science to Me & Users!...*
I COLLECTED WISH DATA TO STUDY
THE (x,T) MAGNETIC PHASE
DIAGRAM OF THE CMR-PCMO
HOPPING MY HRPD SCRIPTS
WILL HELP ME AS PROVEN
FIRST, WITH MY USERS =>
I CAN REFINE ALL! = ALL
ALONE! ATOMATE TYPIC
PHD's = HARD WORK =
& ACHIEVE IT WITH
NO RESSOURCES!
2 BACK UP MY 2nd
ZP VIEW ON PCMO
= TOWARDS CMR
& BEAT THE INS
CENSORED ZPO
AT 1/2 DOPING
PROOF
@ILL
!
During my INS plagiarism/falsification started mid 2014
P. Radaelli send me N. Perks (Following ISIS love harassment Logic!)
Explicitely asking besides pretending she's interested in SXD data,
Which I had collected on NaMn7O12 in my SXD time (before 2008),
she may like to have access at my collaborator's (A. Prodi) Lab X-ray data!!!
= Radealli pertinently knowing SXD data analysis he witness Me Fight with,
as my GRP Leader, would be far too cumbersone = litteraly sueing
A. Gauzzi's (Prodi's supervisor), who was visiting in Oxford at the time!!!!
=> If Science is a race : we were the First
to overtake P. Radaelli, in that Matter where
P. Radaelli in Fact Falsifies ISIS' INS narrative
regarding magnetism/Charge order in this material!
IV- TWO REMARKs 2 ISIS,
ON AREAs WHERE ILL IS EXPERT
Laurent Chapon himself,
saying one day that@ESS
SXTAL diff would be better done
by monochromatising the long pulse &
import instruments from CW sources
to do 4 Circle SXTAL diff
I am not joking!!!
For Polarisation Analysis & TOF,
I'm not an expert, but from memory (WISH test) ,
ISIS really struggles, polarisation efficiencies
are really at the limit of reliability, and certainly,
all issues about its wavelenght dependance
are far from being understood
V- LAST BUT NOT LEAST:
IN SAFETY AS WELL!!!
2 POIDS = 2 MESUREs!
HRPD (old TS1) vs WISH (New shiny, TS2)
Both experiencend a similar safety issue:
more seriously considered on WISH
only beacause HRPD is sort of "hidden
far away" in its own little blockhouse,
& WISH became an TS2 flagship
It was a spillage of active sample
But plans to secure HRPD
(Dating from at least 2013
& Maybe as early as 2011!
can't remember when...)
got neglected totally
during 2 years =
not started
When I
Left
!
WELL : SAFE...
BYE BYE GUYs!
TIME TO DANSE!