Schnelle dynamische Radiographie an einem thermischen Hochfluss-Neutronenstrahl. Fast dynamic radiography at a high-flux thermal neutron beam

56 Pages·5.191 MB·German

by  Gildemeister A.E.

Fakult¨at fu¨r Physik und
Astronomie
Ruprecht-Karls-Universit¨at Heidelberg
Diplomarbeit
im Studiengang Physik
vorgelegt von
Arnd E. Gildemeister
aus Bad Schwalbach
2003

Fast dynamic radiography at
a high-flux thermal neutron
beam
Schnelle dynamische
Radiographie an einem
thermischen
Hochfluss-Neutronenstrahl
Die Diplomarbeit wurde von
Arnd E. Gildemeister
ausgefu¨hrt am
Physikalischen Institut Heidelberg
unter der Betreuung von
Herrn Privatdozent Dr. Hartmut Abele.

Fast dynamic radiography at a high-flux ther-
mal neutron beam
Wereportonfirstdynamicradiographieswithmillisecondtimeresolution
measured at the new neutron radiography and tomography facility Neu-
trograph at the Institut Laue-Langevin (ILL) in Grenoble (France). We
visualized the fuel jet from a diesel direct injection nozzle and recorded
dynamic radiographies of a running combustion engine and an externally
drivencarengine. Furthermeasurementsincludedtwo-phaseflowofpen-
tane in a thin heated steel pipe, water penetration in building materials,
and water transport into a lecithin lamellar phase. We put Neutrograph
into operation at the ILL beamline H9. It provides the high thermal flux
of2.9 109 n cm−2 s−1 andthesufficientlylowdivergenceof6mradthat
×
permit fast dynamic radiography with sub-millimeter resolution. The
properties of Neutrograph are discussed, especially those of the detector.
Schnelle dynamische Radiographie an einem
thermischen Hochfluss-Neutronenstrahl
Wir berichten u¨ber erste dynamische Radiographien mit einer Zeitauf-
l¨osung im Millisekunden-Bereich, aufgenommen an der neuen Neutronen
Radiographie- und Tomographieanlage Neutrograph am Institut Laue-
Langevin(ILL)inGrenoble(Frankreich). WirkonntendenSpru¨stoßeiner
Diesel-Direkteinspritzdu¨se sichtbar machen und haben dynamische Ra-
diographien eines laufenden Verbrennungsmotors und eines geschleppten
Automotorsaufgenommen. WeiterhinhabenwirunteranderemdieZwei-
phasenstr¨omung von Pentan in einem du¨nnen erhitzten Stahlr¨ohrchen,
das Eindringen von Wasser in Baumaterialien, sowie den Wassertrans-
port in eine laminare Phase aus Lezithin untersucht. Neutrograph haben
wir am StrahlH9des ILLinBetriebgenommen. Dieserbietet denhohen
thermischen Fluß von 2.9 109 n cm−2 s−1 und die hinreichend niedrige
×
Divergenz von 6 mrad, die schnelle dynamische Radiographie mit einer
sub-Millimeter Au߬osung erlauben. Die Eigenschaften von Neutrograph
und insbesondere des Detektors werden dargestellt.
Contents
1 Introduction 4
2 Basics of Neutron Radiography 5
2.1 Measuring Principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.2 Tomography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.3 Comparison with other Methods . . . . . . . . . . . . . . . . . . . . . 6
3 Beam and Detector 8
3.1 The Beam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.2 The Detector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.2.1 The Scintillator . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.2.2 The Camera . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
3.2.3 The Detector Housing . . . . . . . . . . . . . . . . . . . . . . . 12
3.3 The Casemate. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
3.3.1 Radiation Protection . . . . . . . . . . . . . . . . . . . . . . . . 13
3.3.2 Safety . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
4 Error Analysis 17
4.1 Systematic Errors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
4.1.1 Scintillator Degradation . . . . . . . . . . . . . . . . . . . . . . 17
4.1.2 Intensity Fluctuations . . . . . . . . . . . . . . . . . . . . . . . 20
4.2 Statistical Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
4.3 Contrast . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
4.4 Spatial Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
4.5 Time Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
4.6 Image Intensifiers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
4.7 Data Treatment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
5 Measurements 31
5.1 Combustion Engines . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
5.1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
5.1.2 Injection Nozzle . . . . . . . . . . . . . . . . . . . . . . . . . . 32
5.1.3 Running Engine . . . . . . . . . . . . . . . . . . . . . . . . . . 32
5.1.4 Car Engine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
5.2 Further Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
5.2.1 Two-Phase Flow . . . . . . . . . . . . . . . . . . . . . . . . . . 40
5.2.2 Building Materials . . . . . . . . . . . . . . . . . . . . . . . . . 40
5.2.3 Myelin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2
CONTENTS 3
6 Outlook 44
A Derivations 45
Bibliography 47
Acknowledgements 50
Declaration 51
Instructions for the CD 52
Chapter 1
Introduction
We put the new neutron radiography brightness by up to 28% within 90 min,
andtomographyfacilityNeutrographinto and systematic beam intensity fluctua-
full operation in 2003. It is located at tions of up to 1.5% in the range of sec-
the beamline H9 of the Institut Laue- onds. Statisticalerrorsareessentiallyde-
Langevin’s research reactor in Grenoble terminedbythenumberofneutronsused
(France). With its unprecedented ther- forameasurementandcanbelowerthan
mal flux of 2.9 109 n cm−2 s−1 at the 0.1%. The spatial resolution is typically
×
moderate divergence of 6 mrad (L/D between 0.2mmand0.6mm. Time reso-
≈
150) it opens a wide range of new oppor- lutiondependsonthetypeofexperiment;
tunities in neutron radiography. it can be as low as 100 µs.
The high flux allows the study of dy- Ourmeasurementsfocussedonthevi-
namic processes in the millisecond range sualization of fuel inside combustion en-
and even below if stroboscopic methods gines. The fuel jet of an injection noz-
can be used. Applications include the vi- zle could be visualized. Dynamic radio-
sualizationofthegasolinedistributionin- graphies of a running model-sized com-
side combustion engines, two-phase flow bustion engine were recorded but vibra-
in fuel cells, and water transport. tionssofarpreventedthevisualizationof
Neutronradiographymeasurestheat- the fuel. First measurements with a car
tenuation of a neutron beam due to ab- engine were done. We also did measure-
sorption and scattering in the sample. It mentstogetherwithagrowingnumberof
yields information on the nuclear compo- externalusers. Theseincludedtwo-phase
sition of the sample and in this respect it flow in a small steel tube, water pen-
differs greatly from the more common X- etration in building materials, and wa-
ray radiography, which is sensitive to the ter transport into a biochemical sample.
atomic composition. This is discussed in Measurements are described in Chapter
Chapter 2. 5.
We record radiographies with a de-
tector that consists mainly of a neutron
scintillator andaChargeCoupledDevice
(CCD) camera. It was optimized to fully
exploit the high flux. The properties of
the beam and the detector are discussed
in Chapter 3.
Chapter 4 provides a detailed error
analysis. Principal sources of systematic
errorsarethedegradationofthescintilla-
torsinthehighflux,whichcanreducethe
4
Chapter 2
Basics of Neutron
Radiography
2.1 Measuring Principle
neutrons. Weassume a parallel beam ge-
ometry; the small divergence limits the
Neutron radiography images samples by spatial resolution (cf. Section 4.4).
measuring their attenuation of a neutron Onecalculatestheattenuationfroma
beam with a two-dimensional position- measurement of I with the sample in the
sensitive detector that determines the beam and an “open beam” measurement
transmitted neutron flux in a plane per- of I without the sample. As the beam is
0
pendicular to the beam. not homogeneous, I is a function of the
0
Asneutronscarrynoelectricalcharge position in the image.
they interact predominantly with the nu- The attenuation coefficient µ is given
clei as they pass through matter. The by
short-ranged strong interaction leads to µ=σ ρNA (2.3)
both absorption and scattering of the t M
neutrons. Electromagnetic interaction with the material’s total cross section σ
t
withtheneutron’smagneticmomentonly for neutrons, its density ρ, Avogadro’s
plays a minor role in radiography. The number N = 6.022 1023mol−1, and the
A
beam attenuation at a given position in molar mass M. For samples consisting
the image is described well by of several materials this must again be
rewritten as
I =I e−µ∆x (2.1)
0
ρ (~x)N
i A
µ(~x)= σ . (2.4)
whereI andI aretheincidentandtrans- t,i M
0 i
i
mitted beam intensities, ∆x is the thick- X
ness of the sample and µ is the attenua- The total cross section is
tion coefficient, a material property that
σ =σ +k σ (2.5)
will be discussed in more detail below. t s n a
For samples consisting of more than one
with the scattering cross section σ ,
material µ will be a function of space s
the absorption cross section σ , and
and 2.1 must be rewritten as a
the neutron-energy dependent factor k .
n
Scatteringoccursbothcoherently,includ-
I =I exp µ(~x)ds , (2.2)
0 − ing Bragg scattering, small angle scat-
(cid:18) ZS (cid:19) tering at grain boundaries, and thermal
the path S being a straight line through diffuse scattering for light elements, and
thesampleinthedirectionoftheincident incoherently. For neutron wavelengths
5
6 CHAPTER 2. BASICS OF NEUTRON RADIOGRAPHY
greater than the Bragg cutoff the total Neutrons penetrate matter more eas-
cross section sharply drops for many ma- ily than charged particles as the strong
terials. force is very short-ranged compared
At thermal neutron energies and be- to the long-range Coulomb interaction.
low, the absorption cross section of most While it takes 45 cm of aluminum to at-
materials shows a reciprocal dependence tenuateabeamofthermalneutronsto1%
on the neutron velocity. As absorption the range of 10 MeV electrons is 2.2 cm
cross sections found in tables are usually [LB90] and that of 10 MeV protons only
validforthermalneutronswithavelocity 630 µm [Bal96].
of v =2200 ms−1 the factor Photons are also uncharged and X-
0
rays penetrate matter well. Interac-
v
0
k = (2.6) tion occurs mainly with the atoms’ elec-
n
v
tron shell [Sie55] and the cross sections
has to be used for neutrons with a ve- roughlyfollow apowerlawinthenuclear
locityv [Hug57]. Thisapproximativefor- charge Z. This is very different for neu-
mulaisvalidformostmaterials. Anoted trons,wherethecrosssectiondependson
exceptionistheresonanceinGadolinium. the nuclear structure and even two iso-
topes of the same element may have to-
tallydifferentcrosssections(cf. Fig.2.1).
2.2 Tomography
Thismeansthateventhickerlayersof
metals may be relatively transparent to
Computer tomography allows a measure- neutrons while being opaque for X-rays
ment of the attenuation in three dimen- and that neutrons may be sensitive to
sions. To do so, radiographies of the small differences in nuclear mass where
sample are recorded under different an- X-rays cannot resolve the difference. For
gles by rotating the sample in the beam. example, neutrons are highly sensitive to
These radiographies are two-dimensional hydrogen, which is practically invisible
projections of the three-dimensional at- forX-rays. Table2.1showscrosssections
tenuation of the sample. The inverse for the interaction of neutrons with some
Radon transform allows to reconstruct isotopes.
the three-dimensional structure from the The major drawback of neutron ra-
projections[Rad17]. Various methods diography is the low availability of high
have been developed for the computa- fluxsources. Atthemomentonlynuclear
tionally intense reconstruction. Details reactors and spallation sources are capa-
on computer tomography, its artifacts ble of producing the flux necessary for
and limitations have been described else- dynamic radiography. Portable sources,
where, e.g. [Kak88]. Specifics on neu- based on spontaneous fission in 252Cf,
trontomographycanbefoundin[Sch99], cannotusuallyprovidethenecessaryflux.
[Sch01], and [Fer03]. Caremustbeusedtoavoidstrongactiva-
tion of samples through neutron capture,
especially for a high flux.
2.3 Comparison with
other Methods
Electrons, protons, or X-rays are also
used for radiography. The best choice
of radiation depends both on the sample
and practical considerations. Neutrons
areoftenwell-suitedforthicksamplesbut
their availability is low.