Reviving and Extending the Neutron Imaging Capabilities at the Penn State Breazeale Reactor

We describe the current situation of the neutron imaging technology, based on known “user facilities” and projects at prominent neutron sources world-wide. Although this method has become highly accepted, there is a great potential for further methodical and technical progress. Continued access to most suitable beam ports and future neutron sources are keystones for the future of neutron imaging. Promising new methods and prominent new applications are stimulating this process. Introduction Neutron Imaging is today a well-established technique for scientific and technical applications. It is very complementary to similar X-ray methods and can often be used symbiotically [1]. The layout of a dedicated imaging beam line should be based on state-of-the-art technologies and the valuable experience of facility operators. The future of this technique will depend on the continuous access to best suitable beam ports at present and future useful neutron sources. This paper starts with a “generic neutron imaging facility”, will reflect the situation on neutron sources and their developments, have a look onto new facility projects and upgrades of existing ones, describe options for best facility utilization and highlight the methodical progress and other new options in neutron imaging. The generic neutron imaging facility As shown in Fig. 1, a generic neutron imaging (NI) facility consists of four major components: the neutron source, including moderation media and filters, the beam forming equipment (collimation), the sample environment and the neutron imaging detector. Modern NI stations [2] have become quite complex and sophisticated systems if all modern trends in this technology should be involved. Depending on complexity level and desired performance, the investments required are in the range from some ten thousands to some ten millions of Euros [3]. The technical level of a NI installation depends on the lab strategy, framed by the funding, the major applications and the demands of the user community [4]. The available facilities can be categorized roughly into four classes: 1. Operational user labs, open for an international access by scientific and industrial partners 2. Operational in-house and test facilities, mainly used by own researchers for domestic projects Neutron Radiography WCNR-11 Materials Research Forum LLC Materials Research Proceedings 15 (2020) 3-10 https://doi.org/10.21741/9781644900574-1 4 3. On-going new NI installation projects or facility upgrade activities 4. Projects under considerations for potential new or upgrade facility installations Fig. 1: Layout of a generic neutron imaging facility with the major components and supplementary features The determining conditions underpinning the most advanced NI systems are: • Well collimated (high L/D-ratio) neutron beam • Beam size adequate to the sample dimensions • High neutron beam intensity • Narrow energy band (thermal or cold), well-known spectral conditions, necessary for quantification • Low background from gamma rays or fast neutrons in the primary beam • Flat beam profile • No interference from back-scattered neutrons (and process gammas) In general, a high intensity neutron beam is required to achieve the highest temporal, spatial or spectral resolution in a reasonable acquisition time. In addition, many more sophisticated techniques are possible in reasonable acquisition time when the intensity is suitable. Neutron source development From all options for the generation of neutrons, until now the research reactors remain the most common, flexible, powerful and even cost-efficient sources of neutrons. They are also by far the majority of the sources where NI stations are presently located and utilized. There exist a few intensive spallation neutron sources as well as other accelerator based neutron sources with lower output of well collimated beams of moderated neutrons. One notes separately that isotopic neutron sources have by far no chance to compete in NI performance when compared to the above mentioned accelerator or reactor based facilities. Neutron Radiography WCNR-11 Materials Research Forum LLC Materials Research Proceedings 15 (2020) 3-10 https://doi.org/10.21741/9781644900574-1 5 The development of the NI technology has to be seen in a global context. In order to develop, apply and utilize modern techniques, the access to well-suited beam ports is necessary. As a general trend, the number of appropriate neutron sources is decreasing in developed countries, but more are being installed in developing countries. On the other hand, the most advanced NI facilities are still situated in a few labs in developed countries. Therefore, the knowledge transfer towards the newly implemented facilities is essential for the progress in the field and broader access for usage of the technique. Fig. 2 describes well the world-wide situation of operational research reactors as summarized in the IAEA Research Reactor data base [5]. From the currently running research reactors with a suitable power above 100 kW, 75 facilities declare to perform “neutron radiography”. Since no detailed specification is given in the data base for many facilities, it is difficult to estimate on which technological level these installations are. Fig. 2: Number of newly commissioned research reactors per year in an inverse time scale: more than 140 have an age of more than 40 years. Some famous sources are highlighted (ILL is indicated twice since NI started just now – 50 years after its startup); SINQ and SNS are not reactors, but spallation neutron sources – shown for comparison. The data are taken from the IAEA RR data base, see: https://nucleus.iaea.org/rrdb. The show examples are by far not complete, but indicate some milestones for the imaging community. A more pragmatic way for data achievement about NI facilities has been made by the “International Society for Neutron Radiography (ISNR)” survey as published on their homepage [6]. A list of “user facilities” can be found in [7], but an update will be given in the appendix. Only few research reactor installations are expected to come to operation in the next years, while aged reactors will continue to shut down for different reasons. On the other hand, all spallation sources (ISIS, JPARC, SNS, ESS) involve NI as a key technology in dedicated projects, while SINQ already operates a few different stations [8] since many years with great success. Neutron Radiography WCNR-11 Materials Research Forum LLC Materials Research Proceedings 15 (2020) 3-10 https://doi.org/10.21741/9781644900574-1 6 The situation within Europe is illustrated by Fig. 3 and the number of NI facilities (running and projected) is added. A similar analysis is not available for the rest of the world in the same quality. Given the fact that new reactor based sources are not presently being built or planned in the Western world, there are initiatives to evaluate and design accelerator based neutron sources with specific performance, e.g. “high brilliance” [9] customized for specific applications. Also in such cases, NI installations are or could be foreseen from the beginning. New installations and upgrades of NI facilities Most of the prominent and powerful neutron sources have been “taken” by the neutron scattering community, by irradiation experiments for nuclear technologies, including isotope production and silicon doping. Therefore, only a few most suitable beam ports remained available and used for NI facilities in the past. The situation has changed slightly after the development, at the end of last century, of digital imaging detector systems with superior performance compared to film based methods. It was possible to make very competitive installations ready in Japan, Europe and America, and more recently in Australia. Based on that progress, the family of “user facilities” has been established which have a similar operational approach now like neutron scattering instruments. Fig. 3: Neutron sources in Europe (data base: [6]); the numbers indicate existing (red) and planned NI facilities New sources tried to follow this trend already at the planning stage, including CARR in China, and planned NI facilities in reactors in Argentina, Jordan and Brazil. Other countries, operating a research reactor for a long time, intend now to rebuild beam ports into NI stations or intend to make a major upgrade, e.g. South Africa and Indonesia. The level of performance these new setups will have depends on the financial situation, the involved Neutron Radiography WCNR-11 Materials Research Forum LLC Materials Research Proceedings 15 (2020) 3-10 https://doi.org/10.21741/9781644900574-1 7 (qualified) manpower and the particular source conditions. Tables 1 to 3 summarize the particular activities were the categories 2-4 as mentioned above are taken. Table 1: In-house and test facilities, mainly for own projects Table 2: Running NI installation projects and upgrade activities Table 3: Potential options and intentions for installations


Introduction
The Breazeale reactor operated by the Radiation Science and Engineering Center (RSEC) of the Pennsylvania State University, is a 1-MW TRIGA type reactor. It has been successfully utilized in the past for neutron imaging and its imaging beam line, which has been operational until very recently (mid 2018), has been upgraded on few occasions. The layout of this old imaging beam line is shown in Fig. 1. It was an ASTM E 544 Category 1 facility with a tangential collimator. It featured a steady neutron flux of 1.7*10 7 n/cm 2 /s at full power and has an L/D collimation ratio of around 150 at the sample position. Further details on the beam line and on past imaging activities, can be found in [1] and [2].
A drawback of the old moderator and beam port design was that the beam ports, except for the imaging beam line #4 on Fig. 1, were not axially aligned with the core center therefore could not feature the highest possible flux. This was related to a change of the fuel type of PSBR from high-enriched uranium to low-enriched TRIGA fuel. Furthermore, the imaging beam line #4 had a pretty high gamma contribution. This was mainly attributed to the thermal neutron capture reaction by hydrogen in pool water which mainly takes place at the sides of the old coremoderator assembly (see Fig. 1) as pointed out in [3]. The high gamma background was not only disturbing for certain imaging applications (see below) but also for some other neutron irradiation techniques sporadically performed at the beam line.
Due to these deficiencies and in a broader context to enhance neutron science capabilities around the reactor, a conceptual design of a new core moderator assembly and new set of beam ports have been prepared using extensive Monte Carlo and thermalhydraulic simulations [3]. This has resulted in a crescent-shaped core-moderator assembly filled with D2O and a number of radial beam ports (see Fig. 2 left). The main advantages of this new design are: larger number and variety of beamlines aligned axially with the core, higher fluxes than in the existing beam line(s), improved n/γ ratios due to the reduced 1 H(n, γ)H 2 reaction by the new moderator geometry and improved Cd ratios. The new thermal beam line NBP4 will be exclusively dedicated to neutron imaging. Some of the design parameters of this beam line are as follows: the thermal neutron flux is around 7*10 7 n/cm 2 /s and the fast flux (> 2MeV) is about 6*10 6 n/cm 2 /s. These are unfiltered flux estimates. Depending on the type and thickness of the applied in-beam filters (Bi, sapphire) these values can decrease up to a factor five for thermal and a factor twenty for fast flux. Regarding flexible beam filtering and spectral shaping see the text below. The expected L/D collimation ratio is around 130 at the sample position (1 m away from the outer wall of the biological shielding). One of the new cold beam lines will also be partly used for neutron imaging. Furthermore, the exploratory free beam NBP2 could also be occasionally used for imaging related studies.  Imaging technique developments at PSBR General thermal neutron imaging. Not long before the reactor refurbishment, we have started extensive developments to enhance the imaging capabilities and activities at PSBR. First, efficient state-of-the-art digital, camera-based detectors have been introduced. One such detector for general thermal neutron imaging purposes have been obtained from NeutronOptics [4] featuring a 200x250 mm 2 field-of-view (FOV) and a CCD camera with a 1 inch Sony ICX694ALG EXview HAD CCD II chip with 2750x2200 pixels. Dark current is strongly reduced (0.002 e/pix/s @-10 °C) by thermo-electric cooling at -35C. The camera has 16 bit digital output and a high, ~75% quantum efficiency (QE). The camera is coupled with a highresolution f/1.4 Fujinon C-mount lens. A photo of the camera box and the CCD is shown in Fig.  3.  Some example neutron images of several test objects are shown in Fig. 4 taken by the above detector in combination with a 400 µm thick, high efficiency LiF/ZnS(Cu) screen from Scintacor [5] on the detector box to illustrate imaging quality at the beamline using the above detector. The high detection efficiency of the screen enabled to take reasonable quality radiographs at only 10% reactor power and with exposure as low as 5s (see Fig. 4 left). The image shows visually observable noise but the overall quality is still reasonable (SNR=22.2) only the fine bore holes in the aluminum block go undetected due to the noise as is illustrated compared to Fig. 4 on the right, which is obtained for 60s exposure and with very good quality (SNR=62.1). The arrows indicated the bore holes that are from left to right 3.5, 2, 1.5, 1 mm in diameter in a 20 mm thick aluminum block and the rightmost, 1.5 mm, is filled with a steel bar. The test objects in Fig. 4 represent different levels of structural complexity (from a simple aluminum step wedge to a PCB board of a PC network card and a highly structured coral) and cover a broad variety of material compositions (organics, metals, minerals etc.). Note that due to the thick screen the spatial resolution, is estimated based on the edge spread function of the upper part of the aluminum step wedge, is only around 500 um, however the primary purpose of the screen was to use it for bright flash imaging (see below). Bright flash neutron imaging. TRIGA reactors, due to their special fuel composition, allow operation in pulse mode [9]. In pulse operation one of the control rods, in case of the Penn State reactor the transient rod, is rapidly ejected to a certain extent from the core with the help of a pneumatic system making the reactor prompt supercritical and engaging it on a power excursion. The power excursion is then mitigated by the strong prompt negative feedback from fuel temperature on reactivity brought about by the fuel composition (uranium mixed intimately with the zirconium hydride moderator) and the reactor power settles down to practically zero again. Peak power values can reach almost 1 GW [6]. For the definition of reactivity in $ and for details on prompt supercritical reactor physics the reader is referred to [9].
This paves the way towards high-speed neutron radiography for fast transient processes. Such processes require high acquisition rate potentially in the kilo frame per second (kfps) range to capture them with minimal motion blur. However, due the short exposure times a high neutron flux is needed to obtain images with acceptable signal-to-noise ratio and statistics.
We have recently demonstrated up to 4 kfps bright flash radiography of a two-phase bubbly flow in a simple bubbler made of aluminum using the old imaging beam line. Some illustrative results are shown in Fig. 5. For this purpose, we have replaced the CCD camera with a legacy high-speed CMOS camera with special, high brightness optics. We report on the details of our bright flash imaging developments elsewhere.
Fast neutron imaging, multi-spectral imaging. Besides thermal neutron imaging we plan to develop fast neutron or multi-spectral imaging methods. This is will be implemented by designing the new thermal imaging beam line in the most flexible way. It will feature (re)moveable in-beam filters to enable multi spectrum imaging ranging from the standard thermal to epithermal and fast energy ranges. Typically, sapphire filters are applied in thermal beam lines to reduce both the gamma and mainly the fast neutron contribution. We plan to have a removable sapphire filter seated deep in the beam port not so far away from the new moderator tank.
We have recently tried to perform fast neutron imaging using the old beam line. The test object was a massive Li-fueled power source shown in Fig.6. The detector was equipped with a 3.8 mm thick PP/ZnS:Cu fast neutron imaging screen from RC Tritec AG, Switzerland [7]. The thermal content of the beam has been removed by applying a 1-cm thick Mirrobor (a borated rubber matt [8]) in-beam filter to avoid unnecessary activation of the sample and the detector. Even exposure times as high as 12 minutes delivered suboptimal image quality (see Fig.6), though the upper oxidized area can be clearly distinguished from the lower elemental Li parts. The low image quality was likely due to the insufficient fast flux in the old tangential beam line, that was looking into the middle of the D 2 O tank. Furthermore, a permanently placed sapphire  As the new beam line leaves the biological shield and enters the new beam cave, we plan to place a rotary beam shutter as is shown on the right of Fig. 2. The rotary beam shutter will have several apertures enabling an easy and fast shutting down of the beam and a flexible in-beam filtering. One aperture will be open for normal beam operation and it will also give access to the removable sapphire filter seated deeper inside the beam port towards the reactor core. This aperture could also house a removable Mirrobor filter for fast neutron imaging. Another aperture will shut the beam down, while a third one will have a few mm thick Cd plate filter to enable epithermal neutron imaging. Switching easily from a thermal to an epithermal beam could have a lot of interesting applications for dual-spectrum or differential imaging of samples that contains component(s) with distinctly different attenuation coefficient e.g. strong resonance(s) in the epithermal range. Those components could be quantified by differential imaging. Epithermal imaging is also promising alternative for samples that are slightly too attenuating from thermal neutrons but not that much that only fast neutrons could be used to examine them.

Summary
The Penn State Breazeale Reactor has just recently undergone a major refurbishment. A new core-moderator assembly has been designed and built enabling a lot of new capabilities or more optimal conditions for neutron beam science. One of the major topics that is being extensively and dynamically developed is neutron imaging. New imaging beam lines are going to be established with thermal and cold spectra. A major effort is also ongoing on imaging detector and imaging methodology development. These include besides general state-of-the-art thermal neutron imaging high-speed, bright flash neutron radiography, fast neutron imaging and dual/multi-spectrum imaging such as thermal/epithermal differential imaging. First results and the corresponding developments have been reported here.