Cooling analysis of nuclear reactors fuel element

Udayraj (M.tech (Thermal sciences)) NIT Calicut udayrraj@gmail.com

Abstract:
In the present work, finite difference solutions is obtained for steady state temperature distribution in a nuclear fuel element that is maintained at constant temperature at one end, and the exposed portion of the element is cooled by convective heat transfer. The finite difference scheme used to solve the problem utilizes an implicit formulation of the governing equation. Results are presented for a wide range of convection, heat generating parameters and length to diameter ratio for a specific fluid liquid sodium. It is found that the radial temperature distribution and length to diameter ratio in the fuel element are significant. The fluent analysis is also done. The obtained results are in good agreement with the results obtained by numerical analysis. Keywords: Heat diffusion equation, Uniform heat generation, Thomas Algorithm, Line Method, Z-sweep, Consistent.

INTRODUCTION:
During normal operation of a nuclear reactor, the internal heat generated in the fuel element must be dissipated and carried away by a stream of coolant passing over it. If the heat generated is not removed fast enough by the coolant, the fuel element may heat up so much that eventually a part of core may melt. Hence the problem is the one of heat removal without excessively high temperature inside fuel elements or on their surfaces. Knowledge of temperature distribution in

the fuel element is needed in order to predict its performance, in particular the highest fuel element temperature and rate of heat removal. This necessitates a analysis of temperature distribution within the fuel element. Sparrow and Chyu [1] carried out a conjugate heat transfer analysis for a vertical plate fin washed by laminar forced convection boundary layer flow. They assumed the heat conduction in the fin to be one-dimensional. Huang and Chen [2] have studied a vertical thin circular pin fin in forced convection flow. They have considered one-dimensional heat conduction in the longitudinal direction. Velusamy and Garg [3] have obtained the heat transfer characteristics for a vertical cylinder fin washed by a combined forced and free convective flow. An accurate solution of the coupled forced convection-conduction problem for a horizontal flat plate has been given by Pozzi and Lupo [4]. They analyzed the entire thermofluid-dynamic field by means of two expansions in terms of coupling parameter. Yu et al. [5] proposed a very effective solution method to solve the conjugate problems of forced convection in incompressible laminar boundary layer flow with constant properties and heat conduction in a solid wall. Two-dimensional heat conduction in the solid for the analysis was not considered in any of the above investigations. In case of nuclear fuel element, radial heat conduction also need to be taken into account in addition to the axial heat conduction. Lawrence Agbezuge [6] have studied the transient heat transfer in a partially cooled cylinder rod. A portion of the rod is immersed in a coolant reservoir that is maintained at constant temperature, and exposed portion of the rod is cooled by convection heat transfer. Because the thermal conductivity of the rod is temperature dependent, the governing partial differential equation is nonlinear. The present work deals with the analyses of the two-dimensional steady-state heat conduction in a heat generating nuclear fuel element. This fuel element is cooled by a upstream flow of a liquid sodium coolant. The objective of this investigation is to study the effect of various parameters on temperature profiles and radial heat conduction. The governing partial differential equation is converted into finite differential equation using second order accurate implicit scheme. The obtained finite difference equation is solved by line method.

2. MATHEMATICAL FORMULATION:
A heat generating cylindrical fuel element of a nuclear reactor cooled by a upward flowing stream of liquid sodium is considered as shown in fig.1.The lower end of the fuel rod is assumed to be maintained at the free stream temperature while the upper end is insulated. The heat transfer mechanism involves the following two phenomena: (a) Conduction inside the fuel rod , and (b) Convection from the fuel element surface to the flowing liquid sodium.

Fig.1. Physical model of the problem

Temperature distribution within the fuel element is governed by the general heat diffusion equation. General form of the governing equation is given as:  

Where, T(r, , z, t) = temperature variation in cylindrical coordinate system (r, ,z) k = thermal conductivity of fuel element, = density of fuel element, = specific heat of fuel material, t = time, and q = heat generation rate within fuel element. 2.1. ASSUMPTIONS: The following important assumptions are made in the present analysis: (1) Heat generation is uniform throughout the fuel element, (2) The thermal conductivity of fuel element material does not change with temperature, (3) Fuel element material is homogeneous and isotropic, (4) Steady state problem, and (5) Axial and rotational symmetry.

Because of rotational symmetry, variation with respect to the

-coordinate is ignored.

Because thermal conductivity does not change with temperature, the governing equation in cylindrical coordinates is 

 For steady state, 

2.2. BOUNDARY CONDITIONS: The lower end of the fuel rod is assumed to be maintained at the free stream temperature i.e., ambient temperature T.  For convective heat flux at r = a,   The upper end is insulated at z = b,      

Symmetry about the fuel element centerline enables only half of the cylindrical element needs to be taken as the domain for obtaining the two-dimensional temperature distribution inside the fuel element. Also the temperature gradient along the element centerline is zero. Hence, at r =0,    

The governing equation pertaining to the problem is solved using a finite difference technique. The two-dimensional heat conduction equation is elliptic in nature and is solved using line method.

3. FINITE DIFFERENCE SOLUTION:

In order to represent the governing partial differential equations into finite difference form, a two-dimensional rectangular mesh is superimposed on the computational domain. Indices ( i , j) are used to indicate position in the ( r , z ) directions, respectively. The choice of the coordinates is such that (0,0) represents the leading edge both in terms of the indices and the spatial coordinates. A small change by increases i and j by 1. r and z in the r- and z-directions respectively,

3.1. DISCRETIZATION: Central differencing is employed in both r- and z-directions while discretizing heat conduction equation as given below:  Let,   

Then, discretized governing equation becomes,  The above discretization is second accurate in both the radial and axial directions.

3.2. SOLUTION PROCEDURE: The two-dimensional heat conduction equation (eq. (6)) is solved iteratively by the line method. Each iteration consists of one sweep in z-direction. The procedure starts with an assumed temperature distribution over the fuel element and computation is carried out iteratively. The iterations are continued till the temperature distributions over the element obtained in two successive iterations satisfy the convergence criterion of the order of 10-5. Since Eq. (6) forms a penta-diagonal matrix system, during the z-sweep a set of equations forming tridiagonal matrix system is solved using Thomas algorithm for each fixed j. The terms corresponding to the nodes of (i,j-1) and (i,j+1) are transferred to the right hand side and these are assumed from the previous iteration. The resulting tridiagonal matrixes are shown below: (1) For i=1, 2 imax; j=1

(2) For j=2 to jmax-1 : i=1 to imax

      



(3) For j= jmax : i=1 to imax        



For the solution of the problem a constant grid size of 0.01 is chosen in both the axial and radial directions. The flowchart of solution procedure is shown in fig. 2.

START

GOVERNING EQUATION (HEAT DIFFUSION EQUATION FOR STEADY STATE AND UNIFORM VOLUMETRIC HEAT GENERATION)

FINITE DIFFERENCE EQUATION (CENTER DIFFERENCING IN BOTH DIRECTIONS)

APPLYING BOUNDARY CONDITIONS AND OBTAIN TWO TRIDIAGONAL MATRICES

TRIDIAGONAL MATRICES ARE SOLVED BY APPLYING THOMAS ALGORITHM AND ITERATIVELY BY LINE METHOD.EACH ITERATION CONSISTS OF ONE SWEEP IN Z-DIRECTION.

SOLUTION CONVERGED

NO

YES RESULTS FIG.2. Flowchart of solution procedure.

4. RESULTS AND DISCUSSION:

The computer code developed has been used to analyze the conduction problem in a heat generating nuclear fuel element. The results are presented for a range of convection parameter, heat generation parameter. Special emphasis is given to the effect of these parameters on the radial temperature distribution in the fuel element. 4.1. TEMPERATURE DISTRIBUTION WITHIN THE ELEMENT: Temperature contour are plotted for q = 2MW/m3 , h=1000 W/m2.K , k = 20 W/m.K , Kg/m .
3

= 11500

Fig.3. Temperature contour within nuclear fuel element.

It is noticed that the fuel element temperature at the center increases along the axial direction. The maximum temperature, 1472 K, obtained at the topmost surface which is within the allowable limit.

Fig.4. Temperature at the surface of the fuel element.

4.2. RADIAL TEMPERATURE DISTRIBUTION: Fig.5. presents the radial temperature distribution in the element at four different axial locations.

Fig.5. Radial temperature distribution in the element at different Z-locations

It is noticed that element temperature at the center increases along the axial direction. However, increase in temperature is very much reduced after Z = 0.15 m. This trend reveals important information because temperature in the element must be below certain allowable limit. It is also clear from the figure that radial temperature distribution becomes steeper along the axial direction. This is due to the fact that the rate of heat flux from the surface of the element near the leading edge is more than that away from it. Since the internal heat generation is assumed to be uniform, lesser heat flux from the surface results in energy storage giving rise to steeper temperature gradient.

4.3. EFFECT OF CONVECTION PARAMETER: Effect of convection parameter, h, on the radial temperature distribution in the element is shown in Fig.6.

Fig.6. Effect of convection parameter on the temperature distribution in the element at Z=0.4m.

It is seen that increase in h, results in lower temperature distribution in the element. However, this trend becomes insignificant when h value is increased from 500 W/m2K to 1000 W/m2.K. The maximum temperature at the center has similar trend. The above observed trend can be attributed to the increased energy dissipation to the flowing liquid sodium at higher values of h. Obviously there will be a limit to which energy dissipation can be increased by increasing the value of h as internal heat generation is kept constant.

Temperature contour for different values of convective parameter, h, are also shown in fig.7. and fig.8.

Fig.7. Temperature contour for q=2 MW/m3 and h= 500 W/m2.K.

Fig.8. Temperature contour for q=2 MW/m3 and h=200 W/m2.K.

4.4. EFFECT OF HEAT GENERATION PARAMETER: Fig.9. illustrates the effect of heat generation parameter, q, on the radial temperature distribution in the cylinder at Z=0.4 m. It is quite evident from these figures that increase in q results in temperature distribution with higher values. The radial temperature gradient is also significantly increased. However, it is interesting to note that the maximum operating temperature in the fuel element limits the extent to which heat generation can be increased. Evidently the above

observation is in good agreement of the fact that all the extra energy generated within the fuel element cannot be dissipated keeping h value constant.

Fig.9. Effect of heat generation parameter on the radial temperature distribution in the element at Z=0.4 m and h=1000 W/m2K.

4.5. EFFECT OF LENGTH TO DIAMETER RATIO: Fig.10 shows the effect of length to diameter ratio L/D on the radial temperature distribution in the element at Z=0.2 m. It is seen that higher value of L/D ratio results is radial temperature distribution with higher values. Even though increase in temperature distribution values are not very significant, increase in L/D ratio will result into higher radial heat flux which is desirable in case of fuel element of nuclear reactor.

Fig.10. Effect of L/D ratio on the radial temperature distribution in the element at Z= 0.2 m for q= 2 MW/m3 and h=1000 W/m.K.

4.6. EFFECT OF CONVECTION PARAMETER ON AXIAL TEMPERATURE DISTRIBUTION: Fig.11. shows the effect of convection parameter, h, on the axial surface temperature distribution. It is observed that axial surface temperature increases sharply up to a distance very close to the leading edge and thereafter remains more or less constant. Increase in h value results in relatively lower value of the surface temperature. In addition the region of variable surface temperature becomes closer to the leading edge. The above observation is due to the fact that higher h values

result in higher energy dissipation to the adjacent sodium fluid resulting in lowering of surface temperature since internal heat generation is kept constant.

Fig.11. Effect of convection parameter on the axial surface temperature distribution at R=0.1 m and q =2 MW/m3.

5. SOFTWARE ANALYSIS:
Modeling software GAMBIT is used to model the nuclear fuel element same as the above problem. Then the modeled element is analyzed using analysis software FLUENT. The energy equation is solved using in two-dimensional case with double precision.

5.1. GAMBIT MODELING:

GAMBIT is used to model our two-dimensional fuel element. The dimensions of the fuel element are same as the case of numerical solution. The element is meshed using triangular mesh with interval size 0.02 and interval count 1. The discrete and continuum boundary condition are given as the case of our mathematical model in numerical analysis.

Fig.12. Meshed model of fuel element in GAMBIT.

5.2. FLUENT ANALYSIS:

After importing this designed model from GAMBIT to FLUENT we have to perform grid check operation. Solving for energy equation we applied all the boundary conditions and numerical values of variables same as the case of numerical analysis. The obtained results are shown in figs. below.

Fig.13. Scaled residuals at q=2000000W/m3, h=1000W/m2.k, k=20W/m.k.

Fig.14. Grid representation in FLUENT.

Fig.15. and Fig.16. represents the solution or temperature distribution obtained by using FLUENT and by using numerical analysis, respectively. The obtained results are same and hence we obtain the same temperature distribution in both of our analysis. This concludes that our computer codes are right. The maximum temperature obtained by both analysis are same.

Fig.15. Temperature contour obtained in FLUENT analysis for q=2MW/m3 , h=1000 W/m2.K , k=20 W/m.K.

Fig.16. Temperature contour obtained in numerical analysis for q=2MW/m3 , h=1000 W/m2.K , k=20 W/m.K.

Fig.17. Temperature contour obtained in numerical analysis for q=1MW/m3 , h=1000 W/m2.K , k=20 W/m.K.

6. CONCLUSION:
This paper presents the consideration of radial heat conduction along with internal heat generation in the nuclear fuel element. Although the results obtained have qualitative similarity with previous studies, this numerical work for nuclear fuel element has demonstrated that consideration of radial heat conduction and internal heat generation presents more significant and realistic results. In addition to this the length to diameter ratio L/D is a very important significance in case of the nuclear fuel element. Higher L/D ratio increases the radial heat conduction from the element which is desirable in case of nuclear fuel element. Apart from these, the effect of heat convection parameter is also obtained in axial temperature distribution. This indicates the significance of high convection parameter, which is mandatory in case of nuclear reactor safe operations. The fluent analysis is done for the same problem and the obtained results are same as that for numerical analysis. This validates our computer codes generated for the analysis.

7. REFERENCES

[1] A.V. Luikov, V.A. Aleksashenko, A.A. Aleksashenko, Analytical methods of solution of conjugate problems in convective heat transfer, Int. J. Heat Mass Transfer 14 (1971) 1047-1056. [2] A.V. Luikov, Conjugate convective heat transfer problems, Int. J. Heat Mass Transfer 17 (1974) 257-265. [3] R. Karvinen, Some new results for conjugated heat transfer in a flat plate, Int. J. Heat Mass Transfer 21 (1978) 1261-1264. [4] E.M. Sparrow, M.K. Chyu, Conjugate forced convection-conduction analysis of heat transfer in a plate fin, ASME J. Heat Transfer 104 (1982) 204-206. [5] M.J. Huang, C.K. Chen, Vertical circular pin with conjugated forced convection-conduction Flow, J. Heat Transfer 106 (1984) 658-661. [6] K. Velusamy, V.K. Garg, Heat transfer from a cylindrical fin in combined free-forced flow, Int. J. Heat Fluid Flow 9 (1988) 233-240. [7] A. Pozzi, M. Lupo, The coupling of conduction with forced convection over a flat plate, Int. J. Heat Mass Transfer 32 (1989) 1207-1214. [8] Ramis, M. K. and Jilani, G. (2007), "Heat transfer Characteristics of a Rectangular Fuel Element Cooled in the Surrounding Medium", Energy Heat and Mass Transfer, Vo1.29, pp. 271288. [9] Ramis M. K. , Jilani G, Jahangeer S,(2007), "Conjugate Conduction Forced Convection Heat Transfer Analysis of a Rectangular Nuclear Fuel Element with Non-uniform Volumetric Energy Generation", Int. J. Heat and Mass Transfer. [10] G.Jilani, S.Jayaraj, M. Adeel Ahmad (2002), Conjugate forced convection-conduction heat transfer analysis of a heat generating vertical cylinder, Int. J. Heat and Mass Transfer. [11] Lawrence Agbezuge (2009), Transient Heat Transfer in a Partially Cooled Cylindrical Rod, Int. J. of Heat Transfer.