Scholarly article on topic 'The Scope of the Imaging Instrument Project ODIN at ESS'

The Scope of the Imaging Instrument Project ODIN at ESS Academic research paper on "Physical sciences"

Share paper
Academic journal
Physics Procedia
OECD Field of science
{"neutron imaging" / "energy resolved neutron imaging" / "time-of-flight techniques" / "neutron scattering in imaging" / instrumentation}

Abstract of research paper on Physical sciences, author of scientific article — Markus Strobl

Abstract In late 2019 the European Spallation Source (ESS) is supposed to produce first neutrons. The concept of this future world leading neutron source is based on long pulses in contrast to the nowadays leading short pulse neutron sources. The high flux as well as the time structure of the source hold strong potential for a new generation of imaging instruments. Among the first approved instruments to be constructed simultaneous with the source is the Optical and Diffraction Imaging with Neutrons instrument ODIN. The basic concept, scope and expected performance for this instrument, for which it has been chosen for construction, shall be introduced here.

Academic research paper on topic "The Scope of the Imaging Instrument Project ODIN at ESS"


Available online at


Physics Procedia 69 (2015) 18-26

10 World Conference on Neutron Radiography 5-10 October 2014

The Scope of the Imaging Instrument Project ODIN at ESS

Markus Strobl*

European Spallation Source ESS AB, Tunavägen 24, SE-221 00 Lund, Sweden


In late 2019 the European Spallation Source (ESS) is supposed to produce first neutrons. The concept of this future world leading neutron source is based on long pulses in contrast to the nowadays leading short pulse neutron sources. The high flux as well as the time structure of the source hold strong potential for a new generation of imaging instruments. Among the first approved instruments to be constructed simultaneous with the source is the Optical and Diffraction Imaging with Neutrons instrument ODIN. The basic concept, scope and expected performance for this instrument, for which it has been chosen for construction, shall be introduced here.

© 2015TheAuthors.Publishedby ElsevierB.V.This is an open access article under the CC BY-NC-ND license


Selection and peer-review under responsibility of Paul Scherrer Institut

Keywords: neutron imaging; energy resolved neutron imaging; time-of-flight techniques; neutron scattering in imaging; instrumentation

1. Introduction

The advent of wavelength resolved imaging methods and their potentials on the one hand, and the development of ever more powerful pulsed spallation neutron sources on the other hand have made the combination of imaging with pulsed sources very favorable and promising [Strobl (2009) and Strobl et al. (2011a)]. Accordingly, a number of instrumentation projects at pulsed neutron sources are on the way, namely RADEN at JPARC [Kiyanagi et al. (2011)], which has meanwhile entered commissioning, IMAT at ISIS [Kockelmann et al. (2015)] which is in the final stage of construction, VENUS at SNS which reportedly has received funding for construction, but also an imaging instrument project at the pulsed reactor source IBR-2 [Lukin et al. (2015)]. On top of that the instrument project Optical and Diffraction Imaging with Neutrons (ODIN) [Strobl (2013)] has been approved as one out of

* Corresponding author. Tel.: +46 721 79 20 68. E-mail address:

1875-3892 © 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license


Selection and peer-review under responsibility of Paul Scherrer Institut

doi: 10.1016/j.phpro.2015.07.002

three first instruments for construction at the European Spallation Source (ESS) in Lund in Sweden in 2013. The project has since entered its planning and preliminary engineering design phase and is scheduled to for operation,

1.e. to enter commissioning, with the source by the end of 2019. ODIN will hence be the first imaging instrument at a pulsed spallation source to be planned from the very beginning. The ESS will be a novel kind of source, as in contrast to other pulsed spallation sources it will produce long neutron pulses with a burst time t of 2.86 ms, at a frequency f of only 14 Hz and a projected power of 5 MW. This will allow to produce not only the highest peak brightness for cold neutrons (Fig. 1), but a time averaged flux at least like that at the highest flux continuous, i.e. reactor sources [Peggs et al. (2013)]. This unique combination of high flux and specific time structure will not only enable a suite of world leading instruments, but has a significant impact on the scope of an imaging instrument to be built at ESS, namely ODIN. In the following paragraphs this impact and the corresponding basic concept of ODIN as well as its scope will be discussed. Furthermore it will be described what is currently envisaged and shall be set at the next major project review in 2015. However, the ESS is a project carried out by 17 European partner countries, and accordingly ODIN is a project to be realized by a number of partners, providing expertise and inkind contributions. Accordingly also the specification of ODIN and key components has and is to a large extend carried out by partners, in particular involving HZB, TUM, PSI, Copenhagen University, DTU and TUD. This article will refer to, but will not discuss these preliminary specifications in detail, as this is and will be done elsewhere. Many specifications are not final and subject to significant change due to still changing boundary conditions like moderator designs and geometries, which are, however, close to being finalized. Hence, only key principal design features, but no final specifications can be communicated yet.

2. Wavelength resolution and time-of-flight (ToF) principles

The key strength of pulsed sources lies in the utilization of time-of-flight approaches, from which most scattering methods can profit due to the fact that e.g. the modulus of the scattering vector q is defined as

where 0 and X are the scattering angle and the neutron wavelength, respectively. This implies that to probe q in elastic scattering applications the wavelength has to be well defined as well as the angles 0 in which the scattered neutrons are measured. As q is the Fourier space equivalent of a real space dimension, measuring scattering as a function of q allows investigating the scattering structures. In the case of imaging, i.e. when in most cases only the transmitted beam is measured, either beam deviations to small angles can be resolved (dark-field) or scattering angles can be identified for specific cases like Bragg edges in the transmission spectrum occurring at 0=90deg. However, the wavelength has to be defined. At continuous sources this is achieved by monochromatisation or by chopping the beam for a ToF approach, through which the wavelength at any given time at the detector is well defined. In both cases a large fraction of neutrons from the source, differing wavelengths and neutrons not arriving within the chopper opening time, respectively, cannot be utilized. Hence the use of neutrons is highly inefficient. In contrast, at a pulsed source and exploiting ToF the source replaces the initial chopper of a ToF approach at a continuous source, and hence in principle all generated neutrons can efficiently be used in a scattering experiment, as their wavelength is well defined by their flight time from the source to the detector. However, that also implies that a ToF approach is only efficient, when a measurement on a defined wavelength band is required or useful, but not for a measurement which only requires a single wavelength to be detected. In the latter case only the time averaged flux, from which the wavelength can be chosen is decisive. How well defined a wavelength is at the detector in a ToF measurement can easily be derived from the ToF equation

q = —sind ,

X _ AZ2L and consequently

m lToF

3A dToF r hr


where h is the Planck constant, m the mass of the neutron and LToF the distance corresponding to the ToF, i.e. from pulse source to detector. The last transformation includes t, the pulse burst time as the principal uncertainty of the ToF. While for a certain wavelength LToF defines the ToF, the time window of the burst time, the source pulse

width, can in principle be compared with the width of the pinhole in the real space collimation L/D. In analogy to the real space divergence resolution (local divergence in a point in the sample) the wavelength resolution can be increased by shortening the burst time t (time window width) and by increasing the distance LToF. With the principal difference that the spatial parameters are 2-dimensional and the ToF parameters apply only in one dimension, the corresponding trade off in available flux is obviously the same: decreasing the time window of the neutron pulse burst time, e.g. by a pulse shaping chopper, decreases the flux correspondingly. Increasing the length LToF decreases the flux per defined time bin AToF due to the spread of the pulse with (flight) time. Adapting the time binning of a measurement to the conditions provided by the burst time implies the same efficiency optimization as the adaptation of the pixel size to blur induced due to spatial collimation conditions. Correspondingly, an adaption of resolution is only efficiently done when adapting both, the intrinsic resolution given by the geometry of the setup and the detection resolution (binning) accordingly. For example, binning the detector pixels to an efficient size of 200 micrometer, but not adapting the pinhole and measuring with a blur of just 50 micrometer (full width at half maximum) means to waste nearly an order of magnitude in available flux and hence efficiency. Correspondingly measuring with t<AToF, where AToF is the time binning of the detection, means to give away a factor AToF/t in flux and efficiency for the corresponding measurement. Consequently only the potential to adapt the wavelength resolution to the required resolution of a measurement can guarantee to measure with high efficiency for a broad variety of applications. This is possible up to now only at continuous sources, where the pulse width and frequency can be tailored to the requirements of a measurement. At a short pulse source the pulse width t and the length of the instrument LTOF define a fixed wavelength resolution for a hence very specialized instrument.

Fig. 1 Source pulse brightness of ESS as compared to other sources (depicted for 4 A)(left), the corresponding wavelength resolution and bandwidth at instrument lengths for 14 Hz and 2.86 ms pulse width at ESS (mid) and a representation of the option of pulse shaping at ESS for

tuning the wavelength resolution for any instrument length.

3. ESS source conditions with respect to neutron imaging

The minimum resolution at which a pulsed source can provide full efficiency, i.e. where the peak flux defines its performance can be approximated by the duty cycle c=t/T, where T=1/f is the repetition period [Mezei (2007)]. For short pulse sources values range between 2% and 0.1%, depending on the moderator. For the ESS this value is 4%. However, because the resolution is wavelength dependent and the wavelength bandwidth due to the low repetition frequency of ESS can be rather broad, a mean resolution of that order is a quite crude approximation (compare section 5 below). Nevertheless, it underlines, that the ESS as a long pulse source, in contrast to short pulse sources is very well suited also for very moderate resolutions of around 10% and is hence particularly suited also for e.g. efficient SANS and reflectometry instruments, which operate at such resolutions. Calculating on the other hand the instrument length required to achieve sub percent resolutions like required e.g. for high resolution powder diffraction or stress/strain measurements yields unrealistic lengths. E.g. 1% (0.5%, 0.33%) resolution at 2 A requires an instrument length of 600 m (1200 m, 1800 m). However, on the other hand the long pulse nature of the source allows to apply pulse shaping choppers and hence the flexibility otherwise only available at continuous sources. The high peak flux over the long pulse time thereby still guarantees for highest brightness and efficiency, as can be seen in a schematic representation for the cold spectral region in Fig. 1. Correspondingly the main strengths of the ESS long pulse source convey (i) a world leading time averaged flux, (ii) an outstanding peak

brightness in particular in the cold neutron domain as well as (iii) the efficiency for ToF methods of a pulsed source combined with (iv) a flexibility comparable to that of a continuous source.

4. Method and requirements

Not only neutron sources, but also neutron imaging has made significant progress with major impact. Recent developments and novel methods imply novel requirements for neutron imaging instruments, in particular also in the field of wavelength dispersive measurements. As a consequence neutron imaging instruments have developed a potential to significantly profit from pulsed sources and especially from the strengths of ESS [Strobl (2009) and Strobl et al. (2011a)]. It is therefore worth to examine such methods, their potential and requirements in detail.

(a) "white" beam imaging

The conventional method of attenuation contrast imaging utilizes a broad spectral range and aims at the highest achievable flux conditions for large and homogeneous fields-of-view (FoV). In turn it provides highest spatial, and in case of kinetic studies also highest time resolution. However, concerning kinetic studies the time structure of the beam at a pulsed source has to be taken into account carefully. With respect to the 14Hz frequency of the ESS the best time resolution with a white beam that can be achieved is coinciding with the period T of approximately 70 ms or multiples of it, in order to guarantee the same spectrum for every individual measurement. Faster measurements are only possible if the specific spectral range/wavelength do not play an important role or can easily be corrected for. Exceptions are repetitive motions probed with an adapted stroboscopic approach, were the phase of the process and the wavelength distribution of individual images are sorted carefully and accumulated during post-processing accordingly. As white beam attenuation contrast imaging is the classical approach, the field of applications is very broad and well documented [Banhart (2008), Kardjilov et al. (2011)] and it can be expected to be the dominant method for a variety of applications still in the near future. Traditionally there are instruments that utilize either a thermal or a cold spectrum, depending on the utilized moderator.. The moderator concept of the ESS, however, will allow to view a cold and a thermal moderator at the same time and hence to work either with a combined and very broad spectrum, or to select a certain spectral range. Such approach can aid correct quantification e.g. by choosing a spectral range beyond the last Bragg edge of a specific sample, avoiding Bragg scattering and beam hardening effects [Treimer et al. (2006)]. The brightness of ESS will provide a world-leading performance at least competitive to those at the brightest continuous source instruments for all of these cases. Efficient transfer of the moderator brilliance is required over a large divergence range and bandwidth to create a large and homogeneous field of view. The requirement for ODIN has been set to at least 25 x 25 cm2 for the FoV.

(b) wavelength dispersive/resolved Bragg edge imaging

The most prominent imaging applications involving wavelength resolution and in particular ToF are certainly studies of crystalline features of condensed matter samples based on the distinct Bragg edges in the attenuation spectrum due to coherent elastic Bragg scattering at the lattice planes of such materials (Fig. 2). Thus neutron imaging is entering the domain of neutron diffraction, and first corresponding and pioneering measurements have been reported from ToF diffractometers at ISIS [Santisteban et al. (2001)]. However, as many of the early examples already reveal, just as much as the variety of specialized diffraction instruments, the focus of such studies can be manifold. First approaches focused on strain mapping, which can in principle be performed with image resolution by measuring the shift of Bragg edges due to lattice distortions. However, as only the transmission direction of distortions can be probed that way, certain limitations apply. Additionally, such measurements require highest wavelength resolution, i.e. better than about 0.5%, like usually only available at high resolution and engineering diffractometers. When on the other hand concentrating on the investigation of crystalline phases and local variations and distributions of these [Santisteban et al. (2002), A. Steuwer et al. (2005), Salvemini et al. (2012), Woracek et al. (2014)] as well as corresponding phase transitions [Makowska et al. (2015)], resolutions can be relaxed to around 1% and according flux gains can be utilized, potentially allowing for in-situ studies of such phenomena. In this range also texture phenomena can be resolved in qualitative studies [Woracek et al. (2014)]. Fitting of significantly wide Bragg edge spectra also hold the potential to retrieve microstructural information and to perform a kind of Rietveld refinement [Vogel (2000) and Sato at al. (2010)]. Further relaxing the resolution decreases the sensitivity to such phenomena, but can in some cases still provide

relevant results. Hence, in this regime resolution and flux have to be tailored carefully in order to achieve the required Bragg-edge and time resolution as required for a specific study and corresponding flexibility can be key to achieve the desired results. At continuous sources, the ToF approach enables best tailoring of wavelength band and resolution, which in particular at short pulse sources are fixed by the instrument length, to the needs of a specific measurement for optimum efficiency. When e.g. concentrating a study on a single Bragg edge, the wavelength range can be limited and the repetition increased, increasing the flux delivered in this bandwidth. At a spallation source the minimum bandwidth is fixed in terms of repetition rate and instrument length, which is hence a key choice of instrumentation (Fig. 1). With the first dedicated ToF imaging instruments available in the near future, the development and potential of these approaches is expected to be further explored and exploited, and hence the full scope of applications is still to grow significantly. Currently most applications focus on engineering materials and components as well as on cultural heritage applications.

(c) diffraction complemented imaging

Investigations of crystalline characteristics in imaging can be complemented with angular resolved diffraction data [Grazzi et al. (2014)]. Already in forward direction ToF dependent diffraction spots can be measured in addition to the (partial) extinction in transmission when the detector is significantly larger than the sample. However, efficiency can be increased with additional detectors perpendicular to the beam or even in backscattering direction. Such studies in principle allow for the resolution of individual crystal grains, potentially full grain maps of poly crystalline samples, including orientation information and even individual strains in such grains. Corresponding approaches exist with x-rays and with a quasi Laue approach for neutrons in a pilot study [Peetermans et al. (2014)]. A ToF approach can also here add efficiency and potential.

Fig. 2 Overview of modern neutron imaging capabilities. Examples from left to right referring to [Santisteban et al. (2001)], [Woracek et al. (2014)], [Strobl et al. (2012a)], Manke et al. (2010)], [Hilger et al. (2010)] and.[Kardjilov et al. 2008] as well as conventional attenuation

contrast (white beam) images from PSI and TUM.

(d) dark-field imaging

Dark-field contrast imaging in contrast to the above is utilizing small and ultra-small angle neutron scattering ((U)SANS) signals in order to visualize the distribution of microstructures beyond the direct spatial resolution limits of neutron imaging, i.e. in a range of a few micrometer to some tens of nanometers. After the introduction of such imaging signal [Strobl et al. (2004)] the advent of grating interferometers has initialized the breakthrough of small angle scatter imaging as dark-field imaging with neutrons [Strobl et al. (2008)]. Though grating interferometers require a relaxed monochromaticity with a resolution of about 10%, recent studies show that they can be operated with monochromatic neutrons within a range of a few Angstrom, and it has been shown theoretically and with corresponding x-ray data, that wavelength dispersive measurements can yield quantitative (U)SANS information [Strobl (2014)]. Hence, the method can qualify as a 2D spatially resolved (U)SANS technique and even holds the potential for quantitative SANS-tomography. Besides the grating interferometers measuring the dark-field signal by a cosine spatial modulation of the beam, a novel technique for according modulation utilizing spin-echo [Bouwman et al. (2011)] has been introduced recently. With this technique, which is analogue to grating interferometry [Strobl 2014)], quantitative SANS results could already be demonstrated [Strobl et al. (2012a)] and an extension to spatial resolved measurements is straightforward. The specific advantage

of this SEMSANS approach is that it can be tuned remotely and utilized straightforwardly in ToF in order to support quantification and to extend the resolvable range. However, some technical optimisations for implementation at an imaging beamline are still to be considered. As a spin-echo method it also requires a polarized beam, and there is hence some synergy in instrumentation for polarized neutron imaging. Applications of such technics are expected to span a range from biology and soft matter to engineering as they should allow for detecting small angle scattering and corresponding local structural information in real systems, in contrast to SANS instruments, which integrate the signal of the illuminated area.

(e) polarized neutron imaging

Polarized neutron imaging is a relatively young method as well and utilizes approaches to keep track of the local spin rotation of a beam transmitting magnetic structures and samples. It can hence provide spatially resolved information on the magnetization in bulk samples, which is not amenable to other techniques [Kardjilov et al.

(2008)]. In order to keep track of the spin, the beam has to be monochromatic, but it has been argued [Strobl

(2009), Strobl et al. (2009b)] and demonstrated shortly after [Shinohara et al. (2011), Tremsin et al. (2011)], that a ToF approach can provide additional information that enables quantification beyond the modulus of spin rotation. Additionally, methods have been proposed to allow 3D vector field reconstructions of magnetic fields [Strobl et al. (2009b)] in particular through ToF methods [Strobl (2009)]. While basic approaches are already applied successfully, more sophisticated methods increasing the scope of the method are still under development. However, applications are promising, as neutrons are an outstanding probe for investigating magnetism in bulk samples, like demonstrated extensively in hard matter diffraction studies in past decades. The required wavelength resolution depends on the magnetic field strengths and structures to be investigated, and hence, like in the case of Bragg edge studies, tailoring the resolution with efficiently to specific applications can be key to allow corresponding studies. Resolutions required are in the range of 1% and better over a wide range of neutrons, which however also depends on the performance of polarization and spin analyzer devices that can be utilized.

Together, these methods, which are judged to profit most from the ESS source in imaging [Strobl (2009) and Strobl et al. (2011a)] define the highest level requirements for the instrument. Other methods in the focus of developing imaging facilities at pulsed sources, and most prominently fast neutron resonance absorption imaging, is not under consideration at ODIN, as the long pulse does not allow to resolve such resonances.

5. Basic scope and instrumentation concept of ODIN

Most of the described potential and according solutions have already been pointed out in Strobl (2009). However, to realize such potential requires novel concepts in instrumentation and hence the finally proposed and approved solutions will be introduced and discussed on a conceptional level of detail. The first decision to be taken for the future instrument was the choice of the moderator and hence available spectrum. The original baseline of the ESS as published in the Technical Design Report (TDR) 2013 [Peggs et al. (2013)] includes two equal moderators each consisting of a cold para-hydrogen moderator with thermal moderator wings (H2O), so that each instrument at ESS can either view an about 12 x 12 cm2 cold moderator or 12 x 12 cm2 thermal wing. However, this specific design also allows bi-spectral extraction [Mezei and Russina (2002)], by directly viewing a thermal wing moderator and viewing the cold moderator via supermirrors in the beam extraction system within the target monolith. ODIN was decided to take advantage of the option to be operated in the thermal and cold range, unlike other existing instruments where an a priori choice has to be taken. This extraction has been designed for the described baseline moderator and provided promising simulation results concerning its performance [Zendler et al. (2013)]. The detailed specification of this component has been performed with the Helmholtz Zentrum Berlin (HZB), a partner laboratory of ESS during the design update phase. However, meanwhile the moderator designs at ESS have been updated, which requires a re-optimisation of this component, which is currently on the way for cold and thermal moderator surfaces of 6 x 6 cm2. The correct choice of the instrument length requires to carefully consider the most relaxed wavelength resolution requirement as well as wavelength band and bandwidth requirements. With the fixed pulse length of the source of 2.86 ms the wavelength resolution at a certain moderator to detector distance is wavelength dependent. Fig. 1 provides a rough idea of the wavelength resolutions and

bandwidths available using the full pulse width at ESS for realistic instrument lengths. Baring in mind, that wavelengths down to about 2 Angstrom should be available for studies with the most moderate wavelength resolution of 10% (compare Section 2), then the length of the instrument has to be chosen with about 60 m. Due to the ESS source frequency of 14 Hz the wavelength bandwidth for ToF measurements at such distance is about 4.5 A (e.g. 1 - 5.5 A or 1.5 - 6 A etc.). This bandwidth was judged to be a good compromise between e.g. dark-field contrast measurements and strain measurements, requiring a significantly wider and most efficient with a significantly narrower range, respectively. On the other hand many Bragg edge studies require at least initially the coverage of a bandwidth of such width in order to analyse and identify phases present in the sample. In addition only at higher resolutions (app. 1% and better), which require pulse shaping choppers, like will be discussed later, a longer distance and hence shorter wavelength band could increase the available flux for specific studies. It has to be noted also, that at such length the wavelength resolution at 3 A (4 A) is already 6,66 % (5%). However, obviously the "natural", i.e. full pulse resolution at a length of 60 m does not satisfy the requirements of many foreseen and important applications such as in Bragg edge imaging or polarized neutron imaging. Therefore the concept of pulse shaping choppers has been proposed for such an instrument early on [Strobl (2009)]. At a long pulse source, however, such approach requires wavelength frame multiplication (WFM) [Mezei (2002)], because only a limited wavelength band can pass through a single opening of a pulse shaping chopper, which can only be positioned outside the target monolith, i.e. about 6.5 m from the moderator. Realizing the WFM pulse shaping with an optical blind double chopper set-up, does not only allow choosing one constant resolution for the desired wavelength band, which obviously optimizes the efficiency for such resolution, but to tune the resolution with a free choice over a certain range. The specification and optimization of this kind of system has been described and demonstrated in [Strobl et al. (2013)] and has meanwhile been realized at an ESS test beamline starting commissioning at HZB in early 2015. A corresponding adaptation will allow tuning the wavelength resolution at ODIN between 1% and about 0.3%, in addition to the "natural" resolution (Fig. 1 for 60 m). It has been specified in collaboration with TUM and will be reported elsewhere in detail. An additional feature of this sophisticated chopper system conveying 8 disc choppers, will be, that the WFM mode will also be extendable to a one-fold pulse suppression. Therefore in particular for resolutions down to around 1% also a doubled wavelength band will be accessible simultaneously by utilizing only every second source pulse. The chosen length of the instrument implies the use of a neutron guide system [Strobl (2009)]. The chopper system foreseen requires on the other hand some boundary conditions for the guide system to fulfil. These boundary conditions convey a limitation of the cross section of the guide, which has been limited to about a maximum of 10 x10 cm2, and in particular to utilize an eye-of-the-needle concept at least in horizontal direction in the position of the pulse shaping choppers, in order to allow for tuning to high resolutions. Hence, it was proposed to consider this focal position at about 6.7 m as a virtual source, a virtual pinhole in terms of an imaging instrument, and to then transport neutrons by a first diverging and then converging guide into the pinhole of the final imaging geometry at 50m. The consequently proposed guide solution for the baseline moderator of ESS has been characterized and optimized in collaboration with HZB and is described in detail in [Hilger et al. (2014)]. The current change of the final moderator geometry, however, requires re-optimisation of the guide, which is currently undertaken in collaboration with PSI and TUM. Guide simulations have been done for a straight geometry as well as for different approaches to loose line of sight in order to suppress the prompt pulse background. In the first case a T0, i.e. a prompt pulse suppression chopper is foreseen in the beamline at about 9 m [Hilger et al. (2014)]. The feasibility of such T0 chopper and other implications concerning shielding requirements, maintenance and costs as well as background considerations have not yet allowed to make a final decision on whether to loose line of sight, with corresponding effects on the beam homogeneity, or not. Required data, simulations and a decision are expected within the next months. The imaging end station, the so called cave, will start at the pinhole position at about 50m and will extend about 14m in length, allowing for variable pinhole to detector distances up to about 14m and hence enable a maximum FoV of > 25 x 25 cm2 [Hilger et al. (2014)]. The collimation ratio can be chosen by the pinhole to detector distance and a variable pinhole size. The potential of having a continuously tuneable pinhole size instead of a drum-like pinhole exchanger with a few fixed choices will depend on considerations concerning prompt pulse contamination. In addition, positions for diffuser, filters and at least one ToF monitor detector are foreseen around the pinhole position. The cave has to be planned spacious in order to host all the required add-on installations like gratings, polarizer analyser devices etc.

It is expected to have at least 3 m space from the beam axis to each side wall. This shall also allow for an upgradeability with diffraction detector banks, in case this is found promising later on and with respect to experience with IMAT at ISIS [Kockelmann et al. (2015)]. Some of the corresponding imaging applications might however also be covered with the help of the Beamline for European Engineering Research (BEER) also approved at ESS. Similar considerations as for the width of the instrumental cave apply for the height and no final decision on roof openings and a potential in-cave crane has been made yet. However, the cave design depends strongly on the beamport decision, which is entangled with moderator and space considerations as well as potential neighbouring beamlines and has not been made yet. The considerations concerning the best suited detector suite for ODIN is the dedicated topic of another paper in this issue [Morgano et al. (2015)]. The required software tools to take full advantage of measurements at ODIN are the content of another work package developed in close collaboration with ODIN partner institutes.

6. Conclusion and Discussion

On the basis of the above conceptual ideas and scope of the instrument together with the expected source and moderator parameters a first estimate of the expected performance of ODIN can be deduced, which are supported by initial simulation results (Fig. 3). Obviously, these have to distinguish between different wavelength resolution modes of ODIN. For the comparison the performance at corresponding resolutions is approximated for different existing and upcoming instruments in Europe, based on optimum flux for resolution trades, i.e. assuming high efficiency for e.g. monochromator devices. For the IMAT instrument at the short pulse source of ISIS the available flux for different resolutions is considered constant, as the resolution cannot be changed. Furthermore it is assumed that the whole wavelength range covered by ODIN is relevant for specific measurements. ODIN gain factors continuous source instruments reduce correspondingly, if that is not the case. Results are normalized to the calculated ODIN performance. The chosen resolutions are representative with respect to different imaging methods to be utilized at ODIN, like discussed above. 100% wavelength resolution denotes the white beam imaging case. The results show the obvious trend that ODIN performs comparable to an instrument at a high flux reactor as long as wavelength resolution is not required, i.e. in the white beam case.


1.00E+00 -

100 10 1 0.3 100 10 1 0.3

wavelength resolution [%] wavelength resolution [%]

Fig. 3 ODIN performance estimate compared to other European imaging instruments at PSI, TUM and ISIS.

This is exactly where continuous source instruments at advanced sources are expected to excel over short pulse source instruments. Towards higher wavelength resolution continuous source instruments fall short behind ODIN performance, however, short pulse source instruments show the opposite trend and gain. These results coincide very well with first predictions in 2009 [Strobl (2009)] as well as with the general ESS source gain factor of about 25, due to time average flux estimates (competitive to ILL) and the time structure with the corresponding ratio (T/t = 25). From the shown comparison it can be expected that the instrument RADEN at the advanced short pulse source of JPARC might surpass or be competitive with the performance of ODIN at least for highest resolution. Simulations show, that it will be decisive for the exact outcome, how small the eye-of-the-needle at about 6.7m from the moderator for the ODIN neutron guide system will have to be, and the final optimization is still ongoing. Of comparable outstanding importance are the available detector performances for ToF imaging.


Banhart, J., 2011, Advanced Tomographic Methods in Materials Research and Engineering, Oxford University Press, Oxford, UK, 2008 Barzagli, E., Grazzi, F., Salvemini, F., Scherillo, A., Sato, H., Shinohara, T., Kamiyama, T., Kiyanagi, Y., Tremsin, A. and Zoppi, M., 2014.

Wavelength resolved neutron transmission analysis to identify single crystal particles in historical metallurgy. Eur. Phys. J. Plus 129, 158 Bouwman, W.G., Duif, C.P., Plomp, J., Wiedenmann, A., Roland Gaehler, R., 2011. Combined SANS-SESANS, from 1 nm to 0.1 mm in one instrument.,Physica B 406, 2357

Hilger, A., Kardjilov, N., Kandemir, T., Manke, I., Banhart, J., Penumadu, D., Manescu, A., Strobl, M., 2010. Revealing micro-structural

inhomogeneities with dark-field neutron imaging. J. Appl. Phys. 107, 036101 Hilger, A., Kardjilov, N., Manke, I., Zendler, C., Lieutenant, K., Habicht, K., Banhart, J.,Strobl, M., 2015. Neutron guide optimisation for a

time-of-flight neutron imaging instrument at the European Spallation Source. Optics Express, 23, 1, 301-311 Kardjilov, N., Manke, I., Strobl, M., Hilger, H., Treimer, W., Meissner, M., Krist, T., Banhart, J., 2008. Three-dimensional imaging of

magnetic fields with polarized neutrons. Nature Physics, 4 Kardjilov, N., Manke, I., Hilger, A., Strobl, M. and Banhart J., 2011. Neutron imaging in materials science. Materials Today 14, 6, 248 - 256 Kiyanagi, Y., et al. 2011. Design studyoftheimagingbeamlineatJ-PARC MLF, ERNIS. Nucl. Instr. Meth. A 651, 16 - 20 Kockelmann, W., et al. this issue (2015)

Lukin E.V., Kozlenko, D.P., Kichanov, S.E., Rutkauskas, A.V., Bokuchava, G.D., Savenko, B.N., 2015. First attempts on energy-selective

neutron imaging at IBR-2. Phys. Procedia, this issue (2015) Makowska, M.G., Strobl, M., Lauridsen, E.M., Frandsen, H.L., Tremsin, A.S., Karjilov, N., Manke, I., Kelleher, J.F. and T.-Kuhn, L., 2015.

Effect of stress on NiO reduction in solid oxide fuel cells - a new application of energy-resolved neutron imaging. J. Appl. Cryst. Manke, I., Kardjilov, N., Schäfer, R., Hilger, A., Strobl, M., Dawson, M., Grünzweig, C., Behr, G., Hentschel, M., David, C., Kupsch, A., Lange, A., Banhart J., 2010. Three-dimensionl imaging of magnetic domains. Nature Commun. 1, 125, D0I:10.1038/ncomms1125 Mezei, F., Russina, M., 2002. Patent Application of 23.1.2002, No. 10203591.1, Germany

Mezei, F. 2002. Instrumentation concepts: advances by innovation and building on experience, in: The ESS Project, vol. II, New Science and

Technology for the 21st Century, ESS Council, Juelich, Germany, 2002 (Chapter 3). Mezei, M., 2007. New perspectives from new generations of neutron sources. C. R. Phys. 8 (2007) 909 Morgano, M., Lehmann, E., Strobl, M., 2015. Detectors requirements for the ODIN beamline at ESS. Phys. Proceedia, this issue Peetermans, S., King, A., Ludwig, W., Reischig, P., Lehmann, E.H., 2014. Cold neutron diffraction contrast tomography of polycrystalline

material. Analyst 139(22):5765-71 Peggs, S., 2013. ESS Technical Design Report TDR. Salvemini, F., et al., 2012. Quantitative characterization of Japanese ancient swords through Energy-Resolved Neutron Imaging. J. Anal. At. Spectrom., 27, 1347-1354

Santisteban, J.R., Edwards, L., Steuwer, A., Withers, P.J., 2001. Time-of-flight Neutron transmission diffraction. J. Appl. Cryst.. 34, 289-297 Santisteban, J. R., Edwards, L., Fitzpatrick, M. E., Steuwer, A. and Withers, P. J., 2002. Engineering applications of Bragg edge neutron

transmission. Appl. Phys. A 74, 1433-1436 Sato, H., Takada, O., Satoh, S., Kamiyama, T., Kiyanagi, Y., 2010. Development of material evaluation method by using a pulsed neutron transmission with pixel type detectors. Nucl. Instr. And Methods A 623, 1, 597-599

Shinohara, T. et al., 2011. Quantitative Magnetic Field Imaging by Polarized Pulsed Neutrons at J-PARC. Nucl. Instr. And Methods A 651, 121 Steuwer, A., Withers, P. J.; Santisteban, J. R. and Edwards, L., 2005. Using pulsed neutron transmission for crystalline phase imaging and

analysis. J. of Appl. Phys. 97, 074903 Strobl, M., Treimer, W., Hilger A., 2004. Small angle scattering signals for (neutron) computerized tomography. Appl. Phys. Lett., 85, 3, 488 Strobl, M., Grünzweig, C., Hilger, A., Manke, I., Kardjilov, N., David, C., Pfeiffer F., 2008. Neutron dark-field tomography. Phys. Rev. Lett. 101,123902

Strobl, M., 2009. Future prospects of imaging at spallation neutron sources. Nucl. Instr. And Meth. A 604, 646 - 652.

Strobl, M., Kardjilov, N., Hilger, A., Jericha, E., Badurek, G., Manke, I., 2009b. Imaging with polarized neutrons. Phys. B 404, 2611-2614

Strobl, M., Kardjilov, N., Hilger, A., Penumadu, D., Manke, I., 2011a. Advanced neutron imaging methods with a potential to benefit from

pulsed sources. Nucl. Instr. and Meth. A 651, 57 - 61 Strobl, M., Tremsin, A.S., Hilger, A., Wieder, F., Kardjilov, N., Manke, I., Bouwman, W.G., Plomp, J., 2012a. TOF-SEMSANS - Time-offlight spin-echo modulated small-angle neutron scattering. J. Appl. Phys. 112, 014503 Strobl, M., 2013.

Strobl, M., Bulat, M., Habicht, K., 2013. The wavelength frame multiplication chopper system for an ESS test- beamline and corresponding

implications for ESS instruments. Nucl. Instr. and Meth. A 705, 74-84 Strobl, M., General solution for quantitative dark-field contrast imaging with grating interferometers. Scientific Reports 4, 7243 (2014) Treimer, W., Strobl, M., Kardjilov, N., Hilger, A., Manke, I., 2006. A wavelength tunable device for neutron radiography and tomography. Appl. Phys. Lett. 89, 203504

Tremsin, A. S., Strobl, M., Kinane, C.J., Boag, S., Dalgliesh, R.M., Kelleher, J., 2011. Magnetic imaging using a time of flight neutron beam. Highlights Report, ISIS Annual Review 2011

Vogl, S., 2000. A Rietveld-Approach for the Analysis of Neutron Time-Of-Flight Transmission Data. Ph.D. Thesis, Christian Albrecht University, Kiel, 2000.

Woracek, R., Penumadu, D., Kardjilov, N., Hilger, A., Boin, M., Banhart, J. and Manke, I., 2014. 3D Mapping of Crystallographic Phase

Distribution using Energy-Selective Neutron Tomography. Advanced Materials 26, 24, 4069-4073 Zendler, C., Lieutenant, K., Nekrassov, D., Cussen, L.D., Strobl, M., 2013. Bi-spectral beam extraction in combination with a focusing feeder. Nucl. Instr. and Meth. A 704, 68-75