2.1 Design and structure of the irradiation device.

Es Salam is a multipurpose heavy water research reactor (MHWRR), cooled and moderated using heavy water and a graphite as the reflector. It is a tank type with a nominal power of 15 MW. Its maximum thermal neutron flux reaches 2.4x10¹⁴ n/cm²·s at the central irradiation channel position. The availability of a flat neutron flux and low nuclear heating in the heavy water and graphite reflector satisfy the requirements of several types of irradiation. The rather large core of the Es Salam enables a high slowdown of fast neutrons. Moreover, the thermal to fast neutron flux ratio is higher farther from the core.

The design of the reactor has been optimized to satisfy radioisotopes and doped silicon productions requirements, to supply high flux neutron beams for researchers. The reactor has quite a pure thermal-neutron spectrum and a large space available for arranging many vertical and horizontal channels. There are more than 40 dry vertical irradiation channels with 3 different diameters disposed in different regions, 2 pneumatic irradiation devices that are dedicated to neutron activation analysis and delay neutron counting, 5 horizontal beam radial channels, and a thermal column [15].

The irradiation device prototype, as illustrated in (Fig. 1-a), is designed to ensure the irradiation of cylindrical silicon ingots with diameters of 2 to 3 in; it was adapted to the vertical experimental channels situated either in the heavy water reflector region or the graphite reflector region of the Es Salam research reactor core. The device consisted of two concentric aluminum thin tubes, the inner one, coupled to a rotating driven system, and the outer one is motionless and used to ensure the circulation of the coolant and as a thermal barrier against the reactor core. Two silicon ingots with a length of 200 mm are introduced in the aluminum containers and placed remotely on the bottom of the rotating tube using a top cabinet cantilever.

Fig. 1Full-screen View

(Color online) NTD-Si irradiation device: a) mechanical structure of an NTD-Si device, b) flow path of coolant at the bottom of an NTD-Si irradiation device.

Graphite blocks can be used to place silicon ingots in a suitable vertical position and avoid any possible neutron flux depression. An annular space of 3/4 of circle, located opposite to the reactor core, is envisaged between the rotating tube and the fixed tube to locate the gammas attenuation materials or neutron absorber materials for axial neutron flux flatting. A neutron flux measuring system is connected to the device for on line integrated thermal neutron flux measurement. The device is accommodated to ensure remote loading/unloading and, eventually, position change of the silicon containers using the cantilever of the reactor top cabinet.

The rotating driven system of the irradiation device is composed of a 0.75 KW asynchronous motor which has a maximum rotating velocity of 996 rpm, a gear mechanism, variable speed drives ATP-16U18N4, and a tachometer. This system provides a precise rotating speed in the range of 0 to 10 rpm.

The neutron measuring system is comprised of two Rhodium-103 self-powered neutron detectors (SPND) with a sensitivity 5.6x10^-21 A/cm²·s and a 4695 A Studsvik amplification module, which has a nominal amplification factor of 0.5x10⁷ V/A by channel.

A control command system, based on the client server architecture with TCP/IP communication liaison, provided through a Labview exploitation program was developed to remotely command and supervise the principal parameters of the silicon doping operation.

2.2 Mechanical performances tests of the prototype

In order to address the challenge of safety for the Es Salam reactor, as in real use, a prototype of the irradiation device at a reduced scale was fabricated with all necessary auxiliary systems. The main goal of this process is to enable the performance of a series of out of pile qualification tests in normal operation conditions. One of the tests verifies the operation and identifies its mechanical performances.

Unfortunately, the presence of singularities; the cross section change of the coolant passage, which involves a variation of coolant flow velocity; and the direction changes of coolant inside the irradiation rig generate pressure drops and could disturb the normal circulation of the coolant. Considering the interaction of fluid structure and the mechanical forces caused by the rotation driven system, all those phenomena and events could threaten the integrity of the NTD-Si irradiation device and its vicinity.

To quantify the impact of those factors on the conceptual design of the NTD-Si irradiation device, experimentation tests were conducted on the prototype during its operation.

This qualification was done in order to ensure that the design is reliable from points of view of operability, mechanical, and hydraulic behaviors. It is also a way of checking whether the irradiation device operation parameters generate strains and stresses above the limitation values. Furthermore, it is a helpful step in enhancing the design and bringing technical solutions in case of malfunctions.

As illustrated in Fig. 2a, the prototype was fixed on a rigid carbon steel platform in the similar assembly conditions as that on the reactor core. It is connected to an instrumented compressed air coolant circuit composed mainly of a three piston compressor, which delivers a maximum flow rate of 60 m³/h and a maximum pressure of 7 bar, a tank with a volume of 1 m³ equipped with safety valve and battery of filters (vapors oil and dust filters), a flow meter, two pressure gauges for the measurement of the inlet and outlet pressures of coolant inside the irradiation device, an anemometer, a pressure reducing valve to regulate the pressure at the entry of the irradiation device, and a set of check and throttle valves.

Fig. 2Full-screen View

(Color online) a) Experimental setup of the NTD-Si prototype, b) acceleration vs. coolant velocity.

A vibration system was used to measure the vibrations at various positions of the prototype under operation conditions of pressure, which varied from 3 to 7 bar, and the coolant velocity of compressed air at the exit part of the irradiation device was varied from 2.5 to 35 m/s. Based from practical knowledge, about 13 points of measurement were chosen. The points were chosen based on the connecting points between parts, fixation points, and maximum displacement, which can occur on the tube. Among those points, seven were along the tube in order to describe the deformation.

This vibration measurement system is comprised of sensors (accelerometers and force transducer), amplifiers and hammer (Brüel & Kjær) [16, 19], a laptop with a PCMCIA card, and a home software (DAQ700) for signal treatment and vibration analysis. For the test method, the choice was done on the impact hammer instead of the electrodynamics exciter. In order to reduce errors, measurements were done several times from 5 to 10 times, taking the average values [16-17, 19].

The vibrational inspection of the prototype behavior under different operation conditions (pressure, flow rate of compressed air coolant and velocity of rotating tube of the irradiation rig) reveals that the operation is normal, the noise level is weak and lower than 50 dB, small mechanical frictions are present, the motion transmission system is adequate and reliable, and the circulation of compressed air inside the irradiation device is appropriate.

Vibration measurements, as illustrated in Fig. 2b, show that there is a direct correlation between acceleration and compressed air flow rate; however, a slight range of stability is observed at high coolant velocity up to 23 m/s, where an acceleration about 1 m/s² is obtained. This maximum acceleration is quite weak, and the equivalent generated forces do not produce harmful mechanical stresses, remaining less than the allowable stress for the irradiation device material (Aluminum 6061, σ_adm = 87 MPa [18]).

In parallel, a numerical model was elaborated and validated using the obtained experimental data. The model is extended to simulate the irradiation device at a real scale. The results of the simulation at static and dynamic mode show a stress of about 2 MPa [18] which is less than the limits as shown previously. These results prove the robustness of the model in spite of the severe conditions applied.

3.1 Heat source and cooling process

Nuclear heat generation in a non-fissile bulk and high density material in or near the reactor core comes mainly from the energy deposited by the collision and interaction of primary gammas (fission gammas and fission product decay gammas that constitute 2/3 of the total gammas released in the core) and from, at a less degree, secondary gammas (resonance, thermal absorption, and inelastic scattering) [20-22]. According to the type and the power of the reactor and irradiation position in the core, gamma heat on silicon varies in the range of 0.02 to 0.5 W/g. Previous studies [1, 23] found that heat densities of 0.02 W/g and 0.25 W/g are generated, respectively, in 3 in diameter silicon irradiated in the graphite reflector region and in 4 in diameter silicon irradiated the in heavy water reflector region of a DR3 research reactor, considering that DR3 is 10 MW tank type reactor cooled and moderated with heavy water.

Koziel and Pytel [24] found that a heat density of 0.25 W/g is generated in 2 in diameter silicon irradiated in the light water reflector region of the EWA research reactor (EWA is a 10 MW type tank reactor cooled and moderated with light water). They also mentioned that a heat density of 0.5 W/g is generated in 3 in diameter silicon irradiated in the graphite reflector region of the MARIA research reactor (MARIA is a 30 MW power pool reactor cooled with light water and moderated with light water and beryllium).

Kim et al. [25] experimentally measured the nuclear heat on aluminum samples around the experimental channels of the HANARO research reactor through the calorimetric method. They obtained a nuclear heat density of 0.143 W/g at the power of 8 MW and prospected a nuclear heat density of 0.473 W/g at the full power of 30 MW (HANARO is a 30 MW pool type reactor cooled and moderated with light water).

Under gamma and neutron radiations, all structural materials of the NTD facility become sources of heat. In the present paper, only the heat generated from the two silicon ingots were considered in the thermal hydraulic analysis. We assume, for both irradiation positions, a heat density of 0.25 W/g as heat source generated in the 2 inch silicon ingot irradiated inside the NTD-Si irradiation device prototype. The maximum temperatures in the vertical experimental channel situated in graphite reflector region and in the vertical experimental channel located in heavy water reflector region of the Es Salam core were 393.15 and 343.15 K, respectively

Compressed air, in an open cycle mode, is considered as a coolant that removes nuclear heat from the structural material of the irradiation device and silicon ingots. This coolant is selected for its low void coefficient, because the heat source is weak, and to facilitate loading and unloading operations.

The argon-41 concentration (half-life of 109.2 min) generated by the activation and irradiation of air through the nuclear reaction ⁴⁰Ar (n,γ) ⁴¹Ar, mainly contributes to the total activity of air. Trace Ar-40 of 0.933% by air volume activates with a thermal capture cross of 0.7 barn and emits a gamma ray of 1,294 KeV. This phenomenon was considered in the design to keep the dose rate below the limits. The use of light or heavy water as coolant requires a complicated design, an important investment cost (irradiation device, thermal-hydraulic loops, loading/unloading, and transfer system), and can cause neutron flux distortion. Filtered compressed air with a temperature of 393.15 K and a pressure of 3 bar was introduced at the top part of the irradiation device. It circulates between the inner wall of the isolation tube and the outer wall of the rotating tube until the hemispherical part where it emerges through the four circular holes, towards the annular section constituted between the inner wall of the rotating tube and the outer walls of the silicon containers. Thus, it carries out heat transfer by the forced convection phenomena between the air coolant and silicon ingots (Fig. 1-b).

Three-dimensional temperature distributions of the two irradiated silicon ingots, as well as the flow path of coolant through the singular points of the irradiation rig (holes, hemispherical part, direction change, etc.), were calculated for three flow rates, 10, 20 and 30 m³/h, and in both irradiation sites. Table 1 summarizes the irradiation conditions and the thermal hydraulic parameters considered.

3.2 Governing equations

To study the characteristics of cooling effect on the target by varying the inlet air mass flow, the governing equations of continuity, momentum, and energy are expressed as follows [26-27]:

where S_m is the mass added to the continuous phase from the dispersed second phase and any user-defined sources.

where $p$ is the static pressure, $\stackrel{=}{\tau }$ is the stress tensor, and $\rho \overrightarrow{g}$ and $\overrightarrow{F}$ are the gravitational body force and external body forces, respectively.

where $k_{\text{eff}}$ is the effective conductivity. The first two terms on the right-hand side of Eq. (3) represent energy transfer due to conduction, and viscous dissipation, respectively. $S_{\text{h}}$ is the volumetric heat source.

3.3 Turbulence model

Two-equation turbulence models allow the determination of both turbulent length and time scale by solving two separate transport equations. The standard k-ε model in ANSYS Fluent falls within this class of models and has become the workhorse of practical engineering flow calculations since it was proposed by Launder and Spalding.

The turbulence kinetic energy, k, and its rate of dissipation, ε, are obtained from the following transport equations [26-27]:

and

where $G_{\text{k}}$ represents the generation of turbulence kinetic energy due to the mean velocity gradients, calculated as described in the modeling turbulent production in the k-ε models.  $G_{\text{b}}$ is the generation of turbulence kinetic energy due to buoyancy, calculated as described in effects of buoyancy on turbulence in the k-ε models. $Y_{\text{M}}$ represents the contribution of the fluctuating dilatation in compressible turbulence to the overall dissipation rate, calculated as described in effects of compressibility on turbulence in the k-ε models.  $C_{1\epsilon }$, $C_{2\epsilon },C_{3\epsilon },$ and C_μ  are constants. ${\sigma }_k$ and ${\sigma }_{\epsilon }$ are the turbulent Prandtl numbers for k and ε, respectively. $S_{\text{k}}$ and $S_{\epsilon }$ are user-defined source terms.

The turbulent (or eddy) viscosity ${\mu }_{\text{t}}$ is computed by combining k and ε as follows:

3.4 Numerical modeling

Computational fluid dynamic method (CFD), by means of ANSYS Fluent version 14.5, is utilized to simulate the thermal hydraulic behavior of silicon ingots inside the irradiation device prototype. This is justified by the fact that this method is useful and efficient for describing the local behavior of fluid flow and heat transfer. It also gives a detailed visualization of flow and temperature distribution at any point of the studied domain. The CFD codes solve the transports equations (conservation of mass, momentum, and energy [26]), which are discrete with the finite volume method (FVM), governing these physical phenomena. The k-ε standard turbulence flow model is considered for the analysis of heat transfer and the flow streams of compressed air in different regions of the simulation domain. The problem is treated as a steady state mono-phase turbulent flow. The structured meshing type adopted is dominant for the geometry model. In order to compare the dependency of the calculated physical parameters (as example temperatures) on grid size, several densities of grid were tested to obtain the appropriate one which ensures minimum uncertainties (1,208,906 nodes and 1,973,701 elements). A difference of 1% between two consecutive results of calculated physical parameters (as example temperatures) is considered as the selection criterion of the adopted grid [27].