Tests on IN14

[Andrew Wildes]

Hi Gents,

Just a little e-mail to let you know what we're planning here at the ILL.

I was in Risoe the other week and spoke to Kim concerning some tests we're
planning on IN14.  The question is:  is there any benefit in putting a
supermirror 'trompet' between monochromator and sample?  We've been given
some m=2 supermirror by Ian Anderson and Peter Hoghoj, and have made a test
piece of equipment to install it.  The dimensions of the entrance of the
'trompet' are fixed, but the horizontal distance of the exit may be changed
to create a converging guide.

The tests will be backed up by Monte-Carlo simulations using both Jan
Saroun's RESTRAX code and your McSTAS code.  We've been doing a lot of
simulations of IN14 recently as we've been considering changing the current
nickel guide for a supermirror guide, and consequently the instrument is
well characterised from the technical drawings and the current flux of the
instrument has been successfully reproduced.  This should be an excellent
test of the two Monte-Carlo packages, and will hopefully lend confidence in
the simulations we've already done.  Myself and Jan will be concentrating
on the RESTRAX simulations, Emmanuel Fahri will be doing the simulations in
McSTAS.

The tests are scheduled to begin on Thursday next week.  I've attached at
the bottom the tests we're hoping to do and the simulations we'll be
carrying out.  I'll certainly keep you posted as to how things progress
here, otherwise, if you have any comments, questions or queries please let
me know!

All the best,

Andrew

-------------------------

Hi Gents,

First of all, the information for the second collimator in the
configuration file should read something like:
2nd collimator (distance,length,hor1,hor2,ver1,ver2,ro[m-1], gh, gv [Ni nat.]):
90.     82.5   5.    ?.     5. 5.    0.  2.  2.  1.     1.

i.e. it is an m=2 supermirror, the initial horizontal distance is 5 cm, the
final can be adjusted between 5cm and 0.245 cm

The tests I propose we do are as a function of ki from 1.1<ki<2.6, the
values of ki will be defined after each test.  All tests will be done with
a vertically curved monochromator.

i)  Measurement of flux (i.e. number of neutrons in a 1x1 sample area) and
intensity (i.e. number of neutrons in a 4x4 sample area) as a function of
ki.  ki=1.1,1.2,1.3,1.4,1.5,1.6,1.7,1.8,1.9,2,2.1,2.2,2.3,2.4,2.5,2.6
The test will use the calibrated monotor at the sample position.  Due to
considerations of space we can't use a higher order filter, and
consequently 2*ki and even 3*ki should also be simulated.  This is done in
RESTRAX by halving the d-spacing of the monochromator and doubling the
value of ki, I think the same should be done in McSTAS.

ii) Shape of the beam at the sample position as a function of ki.
ki=1.1,1.4,1.7,2,2.3,2.6
This will be done with an image plate at the sample position, and may
require an attenuator (the same image plate on LADI is saturated in the
main beam in 5sec).

iii) Measurement of delta(k) at the sample position as a function of ki.
ki=1.1,1.4,1.7,2,2.3,2.6
This will be done with a series of pinholes at the sample position and the
image plate positioned at some distance behind.  I have the idea of taking
5 pinhole positions in a pattern like the 5 on a die. That is, if we define
the central pinhole at (0,0) and all dimensions in centimeters, there will
be four pinholes at (2,2) (2,-2) (-2,-2) (-2,2).  To save time I can't see
any reason why we couldn't make a mask like this and take all 5 photos at
once.  Is there any problems with this?  If a cadmium mask is used, will
that cloud the film?

iv)  Measurement of delta(E).
ki=1.1,1.2,1.3,1.4,1.5,1.6,1.7,1.8,1.9,2,2.1,2.2,2.3,2.4,2.5,2.6
This will be done with a vanadium sample, and with two configurations of
secondary spectrometer.  The first will be a flat graphite analyser and 40'
collimation between sample and analyser, analyser and detector, the second
with no collimation and a curved analyser.  The reason for doing the two
configurations is important for the final test:

v)  Measurement of the resolution function.
ki=1.1,1.2,1.3,1.4,1.5,1.6,1.7,1.8,1.9,2,2.1,2.2,2.3,2.4,2.5,2.6
This will be done with the above configurations.  Measurement of a vanadium
sample will give the projection of the resolution functiononto the E=0
plane.  A Si sample will also be measured, a QQ grid scan will give a slice
through the resolution function at E=0, and a GU scan will give an
indication of the vertical divergence.

The above should be repeated for final horizontal widths of the supermirror
of 5cm, 4cm, 3cm, 2cm, 1cm and 0.5cm.

Now, I know what you're saying - there's no way all these tests can be
completed in 4 days.  However, I think it's worthwhile to do the
simulations at least, we can then choose the most promising and focus on
them for the actual measurements.

Emmanuel and I should confer this afternoon before launching our respective
simulations to be sure we're doing exactly the same thing, and hopefully be
early next week we'll have a more refined gameplan.

Any questions, comments, etc.?

Cheers,

Andrew

====================================================
Andrew Wildes               ph:   +33 (0)4 76 20 70 37 (office)
Institut Laue-Langevin          +33 (0)4 76 20 73 31 (IN14)
BP 156                           fax:  +33 (0)4 76 48 39 06
38042 Grenoble Cedex 9
France
====================================================