Evaluation of Low Dose Silicon Carbide Temperature Monitors

The Nuclear Science User Facilities (NSUF) is the U.S. Department of Energy Office of Nuclear Energy’s only designated nuclear energy user facility. Its mission is to provide nuclear energy researchers access to world-class capabilities and to facilitate the advancement of nuclear science and technology. This mission is supported by providing access to state-of-the-art experimental irradiation testing, postirradiation examination facilities, and high-performance computing capabilities, as well as technical and scientific assistance for the design and execution of projects. As part of an NSUF project, low dose silicon carbide monitors were irradiated in the Belgian Reactor 2 and were then evaluated both at the SCK•CEN and at the Idaho National Laboratory (INL) High Temperature Test Laboratory to determine their peak temperature achieved during irradiation. The technical significance of this work was that the monitors were irradiated to a dose that was significantly less than recommended in published literature. This article will discuss the evaluation process, the irradiation test, and the performance of the low dose silicon carbide temperature monitors.


I. INTRODUCTION
S INCE the early 1960s, SiC has been used as a postirradiation temperature monitor. Several researchers have observed that neutron irradiation-induced lattice expansion of SiC anneals out when the postirradiation annealing temperature exceeds the peak irradiation temperature.
Irradiation temperature is determined by measuring a property change after isochronal annealing or during a continuously monitored annealing process [1]- [8]. There are many properties that may be measured, including electrical resistivity, density, thermal diffusivity, and lattice spacing. In general, electrical resistivity is accepted as a robust measurement technique resulting in accuracies within 20 • C [9].
Twelve silicon carbide (SiC) temperature monitors were irradiated in the Belgian Reactor 2 (BR2) as part of a Nuclear Science Users Facilities (NSUF) Project and were delivered to the High Temperature Test Lab (HTTL) for evaluation Manuscript   to determine their peak temperature achieved during irradiation. Each monitor had a sister monitor exposed to identical irradiation test conditions. Monitors with the "A" designation (six in total) were evaluated using an electrical resistivity method [10]. Sister monitors with the "B" designation are to be evaluated using a new method [11]. The evaluation of the "B" temperature monitors will not be described in this report but will be described in a subsequent report once the evaluations have been completed. Table I provides identification for each "A" monitor with its dose and an expected peak irradiation temperature based on thermal analysis. The quality of material used to manufacture the SiC temperature monitor has a major impact on the radiation-induced swelling, and thus, the ensuing peak irradiation temperature evaluation. Temperature monitors were fabricated from material meeting the Rohm Haas specification SC003. This material was produced via a chemical vapor deposition (CVD) process with high purity (99.9995%) and a density equal to theoretical. No voids or microcracks were allowed. Independent verification was not performed. Due to its polycrystalline β-cubic structure, the SiC material characteristics are isotropic. Using this characteristic, the SiC monitors were manufactured to exceed a resistivity >10 m. SiC monitors used in the experiment were manufactured as cylinders with a 1-mm diameter and a 12.5-mm length (Fig. 1).
This report discusses the laboratory evaluation method, the irradiation capsule design, thermal analysis, dose calculations, evaluation of the SiC monitors, and subsequent computed tomography (CT) evaluations.

II. METHOD
HTTL uses resistivity measurements to infer peak irradiation temperature [12]- [14]. SiC monitors may be evaluated for peak irradiation temperatures ranging from 200 • C to 800 • C U.S. Government work not protected by U.S. copyright.  with a recommended dose ranging from 1 to 8 dpa [10]. For this evaluation, the temperature criterion was met, but it is significant to note SiC monitors with the M1 designation received a dose that was half the minimum recommended dose. Fig. 2 depicts the equipment at the HTTL used to evaluate the SiC monitors. The SiC monitors are heated in the annealing furnace using isochronal temperature steps. Annealing temperatures are recorded using a National Institute of Standards and Technology (NIST) traceable thermocouple inserted into an alumina tube in the furnace. After each isochronal annealing, the specimens are placed in a resistance measurement fixture located in the constant temperature chamber (maintained at 40 • C) for a minimum of 30 min. This resistance measurement fixture is shown in Fig. 3. Copper spring-loaded rods hold the SiC monitor in place. The rods have a conical recess machined into the end to secure the SiC monitor in place. Electrical contacts are held in place with screws. A comprehensive description of the equipment and processes used may be found in [1] and [10].

III. IRRADIATION CAPSULE
The reactor exposure was performed using the Basket for Material Irradiation (BAMI) rig of the BR2, using standard noninstrumented capsules. The capsule consists of an external aluminum body, which is hermetically sealed in a He atmosphere at two bar pressure using plastic deformation. For providing a required temperature under radiation, a dedicated holder with the samples is inserted into the capsule. In this experiment, this holder included three sections made as thick-wall stainless steel tubes separated by ceramic spacers. The vertical cross section of the capsule is schematically shown in Fig. 4(a), and a photograph of the holders and spacers is shown in Fig. 4(b). Each 15-mm-long stainlesssteel section had four symmetrically located 1.1-mm diameter channels, in which the SiC monitors were inserted. The 0.1 mm difference in the channel and the sample diameter were made to allow easy retrieval of the SiC samples after irradiation.
The 5-mm-thick spacers were made of low thermal conductivity (k = 1.3 W/mK) Alumo-silicate Aremcolox 502-1100 ceramic with a density of 2.70 g/cm 3 . This material can sustain long-term temperatures up to 1200 • C (melting at ∼1600 • C). Impurities include Fe 2 O 3 , MnO, MgO, CaO, TiO 2 , K 2 O, Na 2 O, P 2 O 5 , C, and S with the total amount less than 3% by mass. The spacers provide thermal insulation between different sections, helps to center the plugs on the body axis, and keep the SiC samples in the channels.

IV. THERMAL ANALYSIS
The capsule has no active heating element; the temperature of the samples during irradiation is defined by the thermal balance between the radiation heat generation in the internal components and heat transfer through the He gas gap and the aluminum body to the reactor cooling water. For a given radiation heating level, the inner and outer diameters of the stainless steel holder can be adjusted to obtain a desirable irradiation temperature.
The following phenomena were considered to define the temperature distribution: 1) heat generation inside the holder, spacers, and in the capsule walls;  (2). The holders are separated with ceramic disks (1), which are held in place with springs (4 and 5).
(b) Photograph of the three holders and the three ceramic spacers stacked before irradiation.
2) heat exchange through the different materials inside the capsule, including gas gaps; 3) heat exchange on the capsule surface. The volumetric heat generation is dominated by the gamma heating and can be found using a simple approximation where Q [W/g] is gamma heat generation in reference material (aluminum), which is a standard material for BR2 gammaheating evaluation, and ρ is the material density of a component. Heat generation in stainless steel holders is slightly higher. This difference was neglected because it is comparable with the uncertainty of the computational model. In a closed capsule filled with He gas and small gaps, there is no forced convection, and the influence of the natural convection is small. The maximum temperature considered was 400 • C. For this temperature, the radiative heat transfer was negligible when compared to the conduction. Therefore, heat exchange occurs via a heat conduction mechanism.
For the irradiation, the capsule was placed inside a standard BR2 driver fuel element, which was cooled with a water flow at ∼10 m/s. It is known that the cooling of the fuel elements occurs in a mono-phase (no boiling) regime. This means that the heat exchange conditions were well defined, and the temperature of the external surface of the capsule was maintained between 48 • C and 52 • C throughout the irradiation. A 50 • C temperature was used for the capsule design. This conclusion is consistent with the thermal balance calculations below. The design parameters are summarized in Table II.
Compared to steel and aluminum, He gas has a relatively low thermal conductivity. As a result, the holder-body gap thickness has a big influence on the temperature of the holder. This made it possible to obtain three different temperatures in one capsule by fabricating the sections with different He gap thicknesses. The gap between the spacers and the capsule body was minimized to prevent excessive He leakage and to provide accurate centering of the plugs but was made large enough to allow for capsule dismantling.
Due to the low thermal conductivity of the spacers and the geometry on the BAMI rig, heat flux in the axial direction is much lower when compared to heat flux in the radial direction. Therefore, the temperature distribution calculation problem is effectively reduced to an axially symmetric 2-D problem. This allowed the use of the analytical approach for the preliminary capsule design.
For the preliminary design, the system of the heat transfer equations was solved analytically by where i is the layer number, n = 4, T i is the temperature distribution for i th layer. This layer structure corresponds to the horizontal cross section of the capsule, i.e., the layers 1 and 3 are He, the layer 2 is stainless steel AISI 304, and layer 4 is Al 95.5 Si 0.5 alloy, see Table II for details. The following boundary conditions were applied.
−k n dT n dr where T ∞ is the coolant temperature. The thermal expansion of the capsule components was taken into account by defining the dimensions at the design temperature. The analytical model was verified using ANSYS numerical simulations. The difference was less than 10 • C. The numerical simulations were used for part fabrication. Calculated peak irradiation temperatures were presented in Table I.

V. DOSE CALCULATION
Two identical capsules (M1 and M2) were prepared. The capsule M1 was exposed in the BR2 reactor during two cycles and the capsule M2 during the first cycle only. The irradiation was performed by placing the capsules inside BR2 driver fuel elements with a high burn-up located in the same B120 channel for both cycles. The capsules were dismounted simultaneously, following a two-month cooling period after the end of the second cycle. The channel selection was made based on the precycle neutronic Monte-Carlo computations. The estimated irradiation conditions are given in Table III. The nominal reactor power during the two cycles was the same. The gamma-heating levels in the irradiation channel during the two cycles were nearly identical, which means that the temperatures of the samples were also nearly the same. The difference in the neutron fluxes of ∼10% is related to a lower fuel burn-up in the second cycle.
Displacements per atom (DPA) calculations were completed using flux outputs from MCNP [15] and displacement crosssections from SPECTER [16]. SPECTER is a computational tool developed at Argonne National Laboratory to assist in material damage calculations. The DPA rate is calculated using with σ d being the 41 elemental displacement cross-sections from SPECTER and φ being the neutron flux. Both variables are energy-dependent, and 100 energy groups from E m and E M , 0-20 MeV, respectively, were used for the calculation. The displacement cross-sections are calculated with where E a is the available energy that is found in the SPECTER manual or can be calculated using NJOY [17], E d is the displacement threshold energy, which is also found in the SPECTER manual. Using the INL Advanced Test Reactor (ATR) MCNP neutron flux results, the experimental positions with the most similar conditions to those in the BR2 experiment were found [18]. The thermal and fast fluxes of E < 0.5 eV and E > 0.1 MeV are much higher in the BR2 reactor than in the ATR. However, the ratio of fast to thermal flux can be used to compare the two differing neutron fluxes. After finding the most similar fast to the thermal ratio in the ATR experimental position, the ratio of the BR2 fast flux to the ATR fast flux was found and used as a multiplier for the DPA rate. The DPA rates of silicon and carbon in ATR are 36 days/DPA and 60 days/DPA, respectively. For calculating the DPA rate of the compound, a weighted average is taken based on the atom fraction of silicon and carbon in SiC. This has been shown to be a reasonable method of approximating compound material displacement cross-sections when compared to tools like SPECOMP [19] that calculate compound displacement cross-sections. Last, the time for each experiment is multiplied by the DPA rate, and the actual DPA is found. At 21 days, the DPA is approximately 0.5 and 1.0 at 49 days.
A second irradiation DPA calculation was also performed using MCNP. The M1 and M2 capsules were located at the mid-plane of the BR2 core. For evaluating the irradiation conditions, F4 tally cards were used, which calculate the neutron flux averaged over a cell in neutrons/cm 2 s. This calculated value is normalized to obtain the actual neutron flux in the region of interest. The F4 tally normalized result was used for calculating the flux spectrum and the DPA. Fig. 5 presents the neutron flux spectra on the capsules. The DPA is approximately 0.5 and 1.1 at 21 and 49 days, respectively. This updated calculated DPA is in close agreement with the first calculation. The doses for each SiC monitor were presented in Table I.

VI. SIC TEMPERATURE MONITOR EVALUATIONS
This section discusses the evaluation of the SiC monitors and presents the results. This work was conducted in accordance with an approved evaluation plan [10].
An ohmic response curve was generated for each monitor prior to heating. Monitor BR2 M1-High-A exhibited a  typical ohmic response and is displayed in Fig. 6. These data were used to check for linearity and to select a target voltage (with the corresponding current) that would result in minimal heating of the SiC monitor during resistance testing while remaining within the range of the test instrumentation. For this evaluation, the voltage ranged from 16 to 20 V.
Electrical resistivity is used by HTTL to infer the peak irradiation temperature [1]. Figs. 7-12 present the resistivity data taken at each isochronal annealing temperature for each SiC temperature monitor. The peak irradiation temperature, using an electrical resistivity technique, can be taken as the point where the resistivity begins and consistently remains above the error band. The error band bounds the data and is represented by the dotted lines. For this evaluation, the error band was established as the ±2σ value based on a sample size of the first five data points taken below 150 • C. Table IV shows the results for the evaluation. The calculated versus measured peak irradiation temperatures had good agreement with published data [9]. This result is significant, considering that half of the monitors received doses that were much less than 1 dpa.   As evident from the presented data (Figs. 7-12), all of the monitors responded well with the exception of the BR2 M2-Low-A (see Fig. 11). This temperature monitor received the lowest dose (0.5 dpa) and was exposed to the  lowest temperature (255 • C). Also, the error band was much larger than expected. There are several factors that may be considered as to why BR2 M2-Low-A did not respond to the isochronal heating. Further analysis, such as microscopy, CT, and material analysis, could be used to determine why M2-Low-A did not respond. CT was performed and is discussed below.

VII. CT EVALUATION
In an effort to uncover the cause of the indeterminate response of monitor BR2 M2-Low-A, 3-D micro-focus CT  was performed on this monitor and on BR2 M1-LOW-A for comparison. The scan energy was 50 kV at a resolution of 6 μm. The purpose of performing CT was to determine if there were any defects (voids, cracks, porosity, etc.) or foreign material in the BR2 M2-LOW-A monitor that would singularize it from the BR2-M1-LOW-A monitor. Figs. 13 and 14 show CT slices taken from each scan. After evaluating the scans, no apparent anomalous indications were found in the BR2 M2-Low-A monitor. It is interesting to note that highdensity indications were found on or near the surface of the BR2 M1-Low-A monitor, the monitor that performed well.
Both monitors should have been identical in composition. As evident in the scan data, high-density material was found on or near the surface of the BR2 M1-Low-A monitor. Visual inspection of the monitor did not find material on the surface; however, it is believed that the material is platinum, which was transferred from a resistivity measurement taken with a different resistivity apparatus than the one used in this evaluation. Independent resistivity measurements were taken after this evaluation ended but before the CT scan. 3-D micro-focus CT did not uncover a plausible cause for the indeterminate response of the BR2 M2-Low-A monitor.

VIII. CONCLUSION
SiC temperature monitors were irradiated in BR2 as part of an NSUF project and were evaluated at the HTTL to determine their peak temperatures achieved during irradiation. Dedicated irradiation capsules were designed and fabricated to allow irradiation at specified temperatures ranging from 240 • C to 380 • C. After the irradiation, the peak irradiation temperature of each monitor was evaluated using the resistance measurement method. This method recommends a minimum dose of 1 dpa. These monitors received doses ranging from 0.5 to 1.1 dpa. Deviations between the calculated temperature and the evaluated temperature were within or near published limits. A significant finding from this evaluation is that it is possible to evaluate SiC temperature monitors at dose levels much less than 1 dpa. SiC monitors were successfully evaluated that were irradiated to 0.5 dpa with temperatures ranging from 240 • C to 380 • C.