Experimental Studies on the Thermal Hydraulics of a Fuel Column for a Gas-Cooled Micro Reactor (GMR)

Abstract

A thermal-hydraulic test facility is designed to explore the thermal-hydraulic characteristics inside a fuel assembly under normal operating conditions, thereby providing data for validating computer codes for a novel gas-cooled micro reactor (GMR). The primary loop supplies helium at a prototypic temperature and pressure to the test section containing a full-size fuel assembly. The experimental procedure and the test conditions were elaborated. Pre-test simulations using the COMSOL Multiphysics 5.0 software yield detailed 3D distributions of the temperature and flow fields inside the test section, which were employed to guide the positioning of thermocouples. The maximum temperature and its locus, the pressure drop of the coolant through the test section, and the helium temperature at the outlet duct were determined. The simulation indicates that the “mixer” component can effectively enhance the mixing of helium in the rear plenum and reduce the outlet helium temperature. The measured data of preliminary tests at the facility agree well with the predicted values, which proves the accuracy and reliability of the thermocouples. An unheated section at the end of the heating rods leads to a relatively large deviation of the results on the last measuring plane.

1. Introduction

The high-temperature gas-cooled reactor (HTGR) is one of six promising candidates for the Gen-IV nuclear energy system [1]. Attracted by its innovative features, great R&D efforts have been made worldwide since the 1960s. Several test reactors and demonstration plants have been constructed, e.g., PBMR [2], GT-MHR [3], MHTGR [4], HTTR [5], HTR-10 [6], HTR-PM [7], etc. The gas-cooled micro reactor (GMR) presented in this paper is one of the conceptual designs evolving from the state-of-the-art HTGR technology, aiming at providing a versatile, reliable, and portable power source [8]. In general, the GMR is a thermal-spectrum, graphite-moderated, helium-cooled reactor with a block-type core. A passive core cooling system (PCCS) is adopted, which removes the residual heat of the reactor in the event of accident scenarios so as to maintain its integrity.

Thermal-hydraulic analysis and accident assessment are essential for the design and licensing of a novel reactor. Because of the unique features of the HTGR, several thermal-hydraulic characteristics are of great concern and needed to be addressed, including the convective heat transfer of helium through the stacked fuel spheres (for pebble bed type) or a coolant hole (for block type), effective thermal conductivity of the pebble bed, radial and axial heat transfer within the core, radiative heat transfer, potential natural circulation of helium within the reactor during an accident, flow resistance as helium travels through the active core, flow distribution within the rector, fraction of bypass/leakage flow, natural circulation and heat removal capacity of a passive core cooling system, and transient response of the reactor during accident scenarios, etc. In addition, the capability and reliability of computer codes for performing these analyses should also be fully validated. Experimental studies are one of the important methodologies to explore the above issues and, thus, support the development of the HTGR. In the context of the R&D of the HTGR, several experimental campaigns concerning the thermal-hydraulic phenomena have been conducted. The SANA facility was built to investigate the transport of heat through a pebble bed at atmospheric pressure [9]. The test facility features a pebble-bed core made of graphite or aluminum oxide granules, which is heated by a central and three radial heating elements. The datasets collected from this experiment have been employed to validate computer codes, such as MGT-3D and ANSYS CFX 14.0. The NACOK tests (phase I and II) were completed to address the air ingress accident for a prismatic block-type HTGR [10]. The natural convection of air driven by the chimney effect and the chemical reaction (i.e., graphite corrosion) were mainly concerned. The test section is a vertical channel with a square cross-section of 30 × 30 cm, containing three layers of stacked graphite blocks at the bottom. This experiment has served as a benchmark to verify several computer codes, including TINTE, DIREKT, and SPECTRA [11]. In support of the development of the General Atomics MHTGR, the High Temperature Test Facility (HTTF) was constructed at Oregon State University, which is an integral system test facility with a 1/4-scale in both height and diameter to its reference design [12]. The primary loop of the facility comprises a reactor, a steam generator, and a gas circulator. A reactor cavity cooling system (RCCS) is also adopted as a reliable heat sink to ensure a well-defined and adjustable boundary condition for the vessel wall. The measuring datasets obtained from a steady state experiment, a pressurized conduction cooldown (PCC) transient, and a depressurized conduction cooldown (DCC) transient were utilized to validate the computer code models, including RELAP5-3D [13], ASYST4.1 [14]. The Helium Engineering Demonstration Loop (HENDEL) is a large-scale test facility built by JAERI to perform demonstration tests for the HTTR, aiming at verifying the performance and integrity of high-temperature components such as fuel elements, core support structures, and the hot gas duct [15]. The HENDEL facility contains two test sections: a fuel stack test section (T₁) and an in-core structure test section (T₂). The T₁ test section consists of a single-channel test rig T_1-s and a multi-channel test rig T_1-M to study heat transfer and fluid dynamic characteristics inside a coolant channel and a fuel block, respectively [16,17]. The T₂ test section simulates the full-scale core bottom structure of the HTTR [18].

To investigate the thermal-hydraulic characteristics of the GMR, an experimental project has been initiated. Several test facilities are planned to be constructed to perform a series of experiments spanning a wide range of scales from a single coolant channel to the full-size reactor. Specifically, this paper focuses on a test facility simulating a single fuel column of the same dimensions as the prototype under normal operation conditions. The physical phenomena mainly concerned by this experiment are fluid flow and heat transfer within a fuel column under the prototypical conditions, including: the maximum temperature and its location, temperature profile of a fuel block, pressure drop through a fuel assembly, etc. Steady-state proof tests will be performed under the prototypic conditions at this facility. The collected measuring dataset specific to the GMR is expected to be of high value for quantitatively validating self-developed computer codes/models.

This paper proposes a preliminary design of the test facility consisting of a primary loop and a test section; the corresponding experimental methodology to carry out the tests is also presented. In parallel with the experiment, pre-test analyses were carried out by adopting a generic computational fluid dynamics (CFD) software COMSOL Multiphysics 5.0 to verify and assist in the apparatus’s design. Details of the CFD numerical model were elaborated. Simulation results of the steady-state test corresponding to normal operation conditions were presented. The CFD model developed in this paper is also ready for future validation by comparing against the experimental data once it is available.

2. Overview of the GMR Core

The active core contains an array of prismatic fuel assemblies (FAs) stacked in a cylindrical arrangement and surrounded by the graphite reflector. Each assembly is a column piled up by four fuel elements. The lateral reflector contains twelve holes around the active core for control rods’ insertion.

A standard fuel element (see Figure 1) is composed of a hexagonal graphite block (serving as the moderator), fuel holes for fuel compacts, and coolant channels. The fission heat released from the fuel compacts is transferred to the graphite block by means of conduction and is finally absorbed by the coolant via forced convection. The helium temperatures at the inlet and outlet of the core are 445 °C and 750 °C, respectively.

3. Experimental Apparatus

3.1. Primary Loop

The primary loop is capable of circulating and supplying helium to the test section at a temperature and pressure identical to the GMR prototype. The schematic flowchart of the circulation loop is illustrated in Figure 2. Initially, the primary loop is prefilled with pure helium sourced from compressed helium bottles until the loop pressure reaches the desired value. Then, the centrifugal fan is activated to circulate helium within the loop. A portion of the helium exiting the fan flows into the bypass branch, with its flow rate controlled by a regulating valve (V2), whereas the remainder flows towards the test section and is further subdivided into two branches. One branch of the helium, designated as “main flow”, passes through the pre-heater and is heated up to a desired temperature before entering the test section. Another branch, designated as “protective flow”, enters the test section directly without being heated, thereby maintaining a relatively low temperature. The main flow subsequently passes through the active core of the test section, while the protective flow travels through the gap between the barrel and the pressure vessel. The purposes of the protective flow are twofold: (1) to protect the internal metallic structures of the test section, especially the pressure vessel; (2) to mix with the main flow helium and reduce its temperature at the outlet duct, thus protecting the downstream pipes. The proportion of these two helium flows can be finely controlled by adjusting the openings of the regulating valves on each branch (V3 and V4). The helium exiting the test section is further cooled down after passing through a water-cooled heat exchanger that extracts heat to an external sink. Finally, it returns to the inlet of the centrifugal fan to complete a circulation cycle in the loop. Additionally, a pressurizer is connected to the primary loop to compensate for the fluctuation of fluid volume and, thus, stabilize the loop pressure. A safety valve is also connected to the loop, which can be instantaneously triggered to open once the loop overpressure occurs. The data acquisition system of the loop includes several temperature gauges, pressure transmitters, differential pressure transducers, and flow meters installed in the loop to acquire the thermophysical status of the coolant.

3.2. Test Section

As illustrated in Figure 3, the test section adopts a mock-up fuel column stacked horizontally by four fuel elements with the same dimensions as the reference GMR prototype. Electrically heated rods are inserted in the holes drilled in the hexagonal graphite block. Each heating rod can generate the same power as a fuel rod of the GMR during normal operation. Moreover, the input power of each heating rod is designed to be controlled individually to produce a required power profile in the circumferential direction.

As previously described, the helium of the main flow passes through the active core and is heated up to 750 °C by the fuel block. On the other hand, the protective flow helium of relatively lower temperature travels through the annular channel and ultimately mixes with the main flow in the rear plenum of the test section. Specifically, a spoiler is installed around the periphery of the barrel, which is a spiral baffle intended to redirect and rotate the protective flow in order to enhance convective heat transfer. Additionally, a component called “mixer” is installed in the rear plenum, whose purpose is to redirect and facilitate the mixing of the main flow and the protective flow (see Figure 4). The mixer is a truncated-conical grid with perforated lateral surfaces. One end of a large radius is open and encircled by a ring plate to alter the flow path of the protective flow, whereas the other end of a small radius is enclosed by a plate to block and then force helium to penetrate through the holes on the conical surface after mixing.

To monitor the detailed temperature distribution of the fuel column, K-type thermocouples (TCs) are embedded in the test section and located at five individual measuring planes (P1~P5) along the axial direction (see Figure 5). P1 and P5 correspond to the top and bottom surfaces of the fuel column; P2~P4 are located exactly at the interfaces of two adjacent fuel elements, which are easily accessible by the lead wires of the TCs. Figure 5 also depicts the layout of TCs on a representative measuring plane (P1 or P5). These TCs are categorized into three groups to measure the temperatures of (i) the surface of heating rods, (ii) the graphite block, and (iii) the coolant in the channels, respectively. Representative measuring positions are carefully selected to install TCs, such that TCs are placed evenly, and the radial temperature profile can be well-captured to a large extent by a limited number of TCs. Each TC, whose outer diameter is 1 mm and is sheathed with an Inconel tube, is attached to the surface through a groove and fixed using cement.

Other key thermodynamic variables are also monitored, including the pressure drop of helium flow through the test section, the flow rates of main flow and protective flow, and the helium temperature after mixing in the rear plenum, etc.

3.3. Test Procedure and Conditions

All tests are carried out under steady-state conditions in this experiment. A test run normally begins with vacuuming and prefilling the primary loop with pure helium up to the desired pressure. After activating the centrifugal fan, the valve V2 of the bypass line is adjusted to achieve an approximate bypass flow rate close to the required value. The openings of the valves V3, V4, and V2 (if necessary) are then carefully manipulated until the desired flow rates of both the main flow and protective flow are obtained. Prior to applying power to any equipment with heating elements, it is essential to ensure that a forced flow of cooling water has already been supplied to the cooling heat exchanger. Then, the pre-heater is turned on and the power is increased in a stepwise manner, ensuring that the temperature increase rate is limited to be less than 100 °C per hour. Once the helium temperature at the inlet of the test section has reached the prescribed value, the input power of the heating rods is gradually increased. The data acquisition system starts to record and store data once the steady state of the test section has been achieved.

A benchmark test corresponding to the normal operation of the GMR, referred to as the “base case”, will be first carried out with parameters identical to the prototypic condition as listed in

Table 1. Experimental conditions of the base case.

Parameter	Value
Loop operational pressure (gauge)	2.9 MPa
Core inlet temperature	445 °C
Annular channel inlet temperature	75 °C
Mass flow rate of helium through the core	0.08 kg/s
Input power density of the heating rods	2.4 kWth/m3

. In addition, parametric studies will also be conducted with variations in the experimental conditions based on the “base case”, such as coolant flow rate, inlet temperature, input power, etc.

4. Pre-Test Simulation

4.1. Numerical Model

The pre-test analyses of the experiment primarily focusing on the test section were performed. The thermal-hydraulic behaviors in this work, with respect to the single-phase helium flow and conjugate heat transfer between fluid and solid, are addressed by means of the computational fluid dynamics (CFD) methodology. The COMSOL Multiphysics 5.0 software was employed in this simulation, which features built-in physics interfaces that make it easy to implement multi-physics simulations [19].

The mathematical representations of the thermal-hydraulic behaviors are the governing equations summarized below, describing conservation of mass, momentum, and energy, respectively [20].

Continuity equation:

$$\begin{array}{c}{\displaystyle \frac{\partial \rho }{\partial t}}+{\displaystyle \frac{\partial \left(\rho \mathit{u}\right)}{\partial x_i}}=0 \end{array}$$

(1)

Navier–Stokes equation:

$$\begin{array}{c}{\displaystyle \frac{\partial \left(\rho \mathit{u}\right)}{\partial t}}+\nabla ·\left(\rho \mathit{u}⨂\mathit{u}\right)=-\nabla p+\nabla ·\tau +\rho F\end{array}$$

(2)

where the viscous stress tenor, $\mathit{\tau }$, is expressed as $\mu \left(\nabla \mathit{u}+{\left(\nabla \mathit{u}\right)}^T\right)-{\displaystyle \frac{2}{3}}\mu \left(\nabla ·\mathit{u}\right)\mathit{I}$.

Energy equation:

$$\begin{array}{c}{\displaystyle \frac{\partial \left(\rho c_{p}T\right)}{\partial t}}+\nabla ·\left(\rho c_{p}T\mathit{u}\right)=\nabla ·\left(k\nabla T\right)+\Phi +S\end{array}$$

(3)

To model the turbulent effect, Reynolds averaging decomposition is applied to the above equations by decomposing flow variables into mean (time-averaged) and fluctuating components, yielding the following equations (in component form):

Mean continuity equation:

$$\begin{array}{c}{\displaystyle \frac{\partial \rho }{\partial t}}+{\displaystyle \frac{\partial \left(\rho {\overline{u}}_i\right)}{\partial x_i}}=0\end{array}$$

(4)

Mean momentum equation (RANS):

$$\begin{array}{c}{\displaystyle \frac{\partial \left(\rho {\overline{u}}_i\right)}{\partial t}}+{\displaystyle \frac{\partial \left(\rho {\overline{u}}_i{\overline{u}}_j\right)}{\partial x_j}}=-{\displaystyle \frac{\partial \overline{p}}{\partial x_i}}+{\displaystyle \frac{\partial }{\partial x_j}}\left[\mu \left({\displaystyle \frac{\partial {\overline{u}}_i}{\partial x_j}}+{\displaystyle \frac{\partial {\overline{u}}_j}{\partial x_i}}-{\displaystyle \frac{2}{3}}{\delta }_{ij}{\displaystyle \frac{\partial {\overline{u}}_k}{\partial x_k}}\right)\right]\\ +{\displaystyle \frac{\partial }{\partial x_j}}\left(-\rho \overline{u_i^{\prime }{u}_j^{\prime }}\right)+\rho f_i\#\end{array}$$

(5)

Mean energy equation:

$$\begin{array}{c}{\displaystyle \frac{\partial \left(\rho c_p\overline{T}\right)}{\partial t}}+{\displaystyle \frac{\partial \left(\rho c_p\overline{T}{\overline{u}}_i\right)}{\partial x_i}}={\displaystyle \frac{\partial }{\partial x_i}}\left(k{\displaystyle \frac{\partial \overline{T}}{\partial x_i}}-\rho c_p\overline{u_i^{\prime }{T}^{\prime }}\right)+\overline{\mathsf{\Phi }}+S\end{array}$$

(6)

Specifically, to enclose the Reynolds stress terms in the RANS (Equation (5)), the Boussinesq hypothesis is adopted as follow:

$$\begin{array}{c}-\rho \overline{u_i^{\prime }{u}_j^{\prime }}={\mu }_t\left({\displaystyle \frac{\partial {\overline{u}}_i}{\partial x_j}}+{\displaystyle \frac{\partial {\overline{u}}_j}{\partial x_i}}\right)-{\displaystyle \frac{2}{3}}\rho k{\delta }_{ij}\end{array}$$

(7)

Similarly, the turbulence-related terms in Equation (6) are approximated as follows:

$$\begin{array}{c}\rho c_p\overline{u_i^{\prime }{T}^{\prime }}={\displaystyle \frac{{\mu }_t{C}_p}{{Pr}_t}}{\displaystyle \frac{\partial \overline{T}}{\partial x_i}}\end{array}$$

(8)

$$\begin{array}{c}\overline{\mathsf{\Phi }}={\mu }_t\left({\displaystyle \frac{\partial {\overline{u}}_i}{\partial x_j}}+{\displaystyle \frac{\partial {\overline{u}}_j}{\partial x_i}}\right){\displaystyle \frac{\partial {\overline{u}}_i}{\partial x_j}}\end{array}$$

(9)

To solve Equations (7)–(9), a turbulent model is required to be supplemented, such as k-ε, k-ω, SST, etc.

Figure 6 depicts the computational domains of the simulation. Most of the details of the test section were modeled with relatively high fidelity, including the active core, the pressure vessel, and most internal structures/components. The active core consisted of a full-size fuel assembly and a lateral insulation layer. In line with the prototype, the fuel assembly contains a hexagonal graphite block, coolant channels, and holes for fuel compacts. Both the heated and unheated segments of heating rods were modeled. To simulate the helium flow at the inlet of the active core more accurately, especially concerning the flow distribution and the flow resistance caused by the heating rods obstacle, the front chamber was modeled. The outlet duct was also included in order to capture the mixing process of the main flow and protective flow. The internal structures of great interest were also considered in this model, including the spiral spoiler (Figure 6c) and the mixer in the rear plenums (Figure 6d).

Figure 7 shows the mesh of the model. For the region of the active core, including the fuel column and side insulation, an unstructured 2D face mesh was first created using triangular elements on the radial surface. This planar mesh was subsequently swept uniformly in the axial direction to generate a 3D mesh. The remaining domains were filled with unstructured tetrahedral elements. Boundary layers for fluid flow in the coolant channels of fuel assemblies were finally inserted with the first layer thickness carefully adjusted to satisfy the requirement of the dimensionless distance y+ value for the applied turbulent model. A mesh sensitivity study was also conducted, primarily by varying the elements’ sizes in the graphite block and the coolant channels. Two global quantities were selected to monitor, including the maximum temperature of the heating rods and the average coolant temperature at the outlet. The study result is illustrated in Figure 8, indicating that a refined mesh with 2.8 million elements is adequate to produce mesh-independent results.

Based on the conditions given in Table 1, it is estimated that the Reynolds numbers are 2.3 × 10⁴ for the main flow and 1.5 × 10⁴ for the protective flow, implying that the flow patterns of both flows are in the turbulent mode. As an exploratory study, the standard k-ε turbulent model is employed in the current simulation, which is widely applied in the area of engineering design, renowned for its relatively low computational cost and acceptable accuracy. The transport equations for k and ε are:

Turbulent kinetic energy (k) equation:

$$\begin{array}{c}{\displaystyle \frac{\partial \left(\rho k\right)}{\partial t}}+{\displaystyle \frac{\partial \left(\rho k{\overline{u}}_i\right)}{\partial x_i}}={\displaystyle \frac{\partial }{\partial x_j}}\left[\left(\mu +{\displaystyle \frac{{\mu }_t}{{\sigma }_k}}\right){\displaystyle \frac{\partial k}{\partial x_j}}\right]+P_k-\rho \epsilon \end{array}$$

(10)

where P_k is the production of k due to the mean velocity gradient:

$$\begin{array}{c}{P}_k={\mu }_t\left({\displaystyle \frac{\partial {\overline{u}}_i}{\partial x_j}}+{\displaystyle \frac{\partial {\overline{u}}_j}{\partial x_i}}\right){\displaystyle \frac{\partial {\overline{u}}_i}{\partial x_j}}\end{array}$$

(11)

Dissipation rate (ε) equation:

$$\begin{array}{c}{\displaystyle \frac{\partial \left(\rho \epsilon \right)}{\partial t}}+{\displaystyle \frac{\partial \left(\rho \epsilon {\overline{u}}_i\right)}{\partial x_i}}={\displaystyle \frac{\partial }{\partial x_j}}\left[\left(\mu +{\displaystyle \frac{{\mu }_t}{{\sigma }_{\epsilon }}}\right){\displaystyle \frac{\partial \epsilon }{\partial x_j}}\right]+C_{\epsilon 1}{\displaystyle \frac{\epsilon }{k}}{P}_k-C_{\epsilon 2}\rho {\displaystyle \frac{{\epsilon }^2}{k}}\end{array}$$

(12)

The turbulent viscosity (${\mu }_t$) is correlated in the form of

$$\begin{array}{c}{\mu }_t=\rho C_{\mu }{\displaystyle \frac{k^2}{\epsilon }}\end{array}$$

(13)

The values of the empirical constants in the above equations of the standard k-ε turbulent model are listed in

Table 2. Constants of the standard k-ε turbulent model.

Constant	Value	Description
$C_{\mu }$	0.09	$\mathrm{Coefficient} \mathrm{relating} {\mu }_t$ to k and ε
$C_{\epsilon 1}$	1.44	Coefficient scaling ε production
$C_{\epsilon 2}$	1.92	Coefficient scaling ε destruction
${\sigma }_k$	1.0	Turbulent diffusivity for k
${\sigma }_{\epsilon }$	1.3	Turbulent diffusivity for ε

.

It should be pointed out that extensive validation exercises are to be carried out in future work to justify the applicability of the turbulent model applied in this work. The thickness of the first layer mesh was adjusted such that the dimensionless distance y+ fell within the range of 30~300, as depicted in Figure 7.

The thermal heat produced by the electrically heated rods is modeled as a volumetric heat source. A uniform power density is employed in this simulation.

The specific boundary conditions applied in this model are summarized in

Table 3. Boundary conditions of the model.

Boundary Location	Type
	Fluid Flow	Heat Transfer
Inlet of the main flow	Mass flow rate	Temperature (=445 °C)
Inlet of the protective flow	Mass flow rate	Temperature (=75 °C)
Outlet at the outlet duct	Outflow	Zero temperature gradient
Outlet of the annular channel	Outflow	Zero temperature gradient
Outer wall of the pressure vessel	None	Convective (h = 5 W/(m2-K), ambient temperature = 30 °C)

. Default boundary types were automatically assigned to the rest boundaries if no boundary type was specified, i.e., a non-slip wall for fluid flow and a thermal adiabatic wall for heat transfer. Specifically, instead of explicitly modeling the air flow and the corresponding natural convection around the pressure vessel of the test section, a convective boundary type is adopted to reduce the computational cost, yet it can still obtain a relatively good accuracy. The ambient environment temperature is set as 30 °C. The effective convective heat flux coefficient is specified as 5 W/(m²-K), evaluated using the Churchill–Chu empirical correlation [21].

The materials of components of the test section and their corresponding thermo-physical properties employed in this simulation are listed in

Table 4. Materials and thermo-physical properties of components of the test section.

Component	Material	Density (kg/m3)	Conductivity (W/(m-K))	Specific Heat Capacity (J/(kg-K))
Hexagonal block	Graphite	1900	129	710
Heating rod	MgO	3400	58	960
Side insulation	Aluminum silicate	150	0.05	840
Pressure vessel and internal structures	Stainless steel	7800	15	460

.

The COMSOL Multiphysics interfaces of a specific physical field involved in this simulation include (1) Single phase turbulent flow, and (2) Heat transfer in solids and fluids. The conjugate heat transfer is implemented by adding a multi-physics node “Non-isothermal Flow” to enable the data exchange between the relevant variables of the aforementioned two physical interfaces, e.g., temperature, flow velocity, fluid properties, etc.

A segregated solver was utilized to solve the system of equations for each physical interface after discretization in a step-by-step manner. The PARDISO algorithm is used as the direct solver for each step. The convergence of the solution was judged based on the criteria that the residuals for all physical variables were less than 10⁻⁵ and that the mass and energy differences between the inlet and outlet were indiscernible.

4.2. Simulation Results

The “base case”, whose conditions are defined by Table 1 is dealt with in this section. Figure 9 shows the simulated temperature distributions within the test section. The simulation results indicate that the maximum temperature of heating rods is 887 °C. The hottest spot is located radially at the center of the innermost heating rods and axially in the vicinity of the outlet of the fuel column (but not on the very end surface). To more accurately capture the peak temperature, more TCs are suggested to be placed close to this area when refining the instrumentation design scheme. The average temperature of helium at the outlet of the fuel column is 746.4 °C, which is slightly lower than the expected 750 °C, yet is still acceptable. This discrepancy is primarily attributed to the heat loss through the insulation layer surrounding the fuel column, which is ultimately dissipated by the protective flow. Due to the presence of the mixer in the rear plenum of the pressure vessel, the high-temperature main flow helium and the cooler protective flow can be stirred and blended thoroughly, resulting in a more uniform and lower helium temperature at the outlet duct. The average temperature of helium after mixing the main flow with the protective flow was 194.4 °C, which will not cause thermal damage to the downstream piping.

The calculated flow field distributions of both the main flow and the protective flow helium are shown in Figure 10. The maximum velocity of helium flow in the fuel column is 33.8 m/s; the average velocities at the outlet of the fuel column and the outlet duct are 24.9 m/s and 1.95 m/s, respectively. It can be seen that when the main flow helium enters the front chamber, although it needs to flow around the non-heated section of the heating rods and a 90° elbow, the overall flow field is still relatively uniform and smooth; no evident local eddy is observed. The combined effects of the spiral baffle and the mixer are significant as they can effectively alter the flow path and enhance the mixing of the main flow and the protective flow. The pressure drops of helium flow are 132.6 Pa from the inlet of the main flow to the inlet of the fuel column and 4.18 kPa throughout the fuel column, respectively. Consequently, the total pressure loss of the whole test section is summed up to be 4.313 kPa; the pressure head delivered by the centrifugal fan in the primary loop is, therefore, required to be sufficient to compensate for this pressure loss.

5. Preliminary Test

Currently, the overall construction of the experimental installation has been completed. Prior to commencing experiments with the facility, preliminary tests were carried out to check the reliability of the heating rods and the accuracy of the instrumentation. The experimental conditions of one preliminary test are listed in

Table 5. Experimental conditions of a preliminary test.

Parameter	Value
Loop operational pressure (gauge)	0.8 MPa
Core inlet temperature	262 °C
Annular channel inlet temperature	51 °C
Mass flow rate of helium through the core	0.067 kg/s
Input power of the heating rods	38 kW

.

Figure 11 shows the comparison between the measured data (solid symbols) and the calculated values (solid lines) regarding the axial temperature profiles at representative measuring points. Generally, the temperature distributions of the graphite block, the helium gas, and the heating rod surface agree well with the values predicted by the CFD model developed in this work. Specifically, the maximum deviation of the helium gas temperature is within 0.9%, indicating satisfactory accuracy of the measurements (see Figure 11b). However, as illustrated in Figure 11a, a notable deviation between the measured and predicted graphite block temperature is observed, which appears radially on the periphery and axially on the last measuring plane close to the helium outlet, resulting in a large discrepancy of 33 °C. This measured temperature is unreasonable as its value (352 °C) is even lower than the helium temperature (363 °C) in the neighboring coolant channel. A possible explanation is that a short segment of the heating rod near the end adjacent to the helium outlet does not generate heat as expected, as indicated in Figure 11c by a significant decrease in the measured temperature on the heating rod surface compared with the calculated value. Nevertheless, the temperatures of the graphite block and the heating rod surface obtained by other TCs are still proven to be credible, with the largest difference between the measured and calculated values being less than 8 °C.

6. Concluding Remarks

A test facility is designed to study flow and heat transfer within a fuel column under the prototypic conditions during normal operation, thus providing high-quality datasets to quantify the accuracy of computer codes and models in support of the development of the GMR. Experimental methodology at this facility was presented, including the design of the primary loop and the test section, the test procedure, and the test conditions.

Pre-test CFD analyses for the steady-state “base case” condition produced detailed 3D distributions of the temperature and velocity fields inside the test section. The peak temperature is 887 °C, which is located radially at the center of the innermost heating rods and axially in the vicinity of the outlet of the fuel column. The simulation shows that the “mixer” component can enhance helium mixing in the rear plenum, thus effectively reducing its temperature and protecting downstream pipelines. The heat loss through the side insulation is acceptable. The pressure drop of the coolant through the test section is about 4.313 kPa, which should be overcome by the centrifugal fan.

The predicted axial temperature profiles match well with the measured data of preliminary tests at this facility, thereby verifying the accuracy and reliability of the thermocouples. The deviation between the measured and calculated results on the last measuring plane is relatively large due to an unheated section at the end of the heating rods.