'Modeling object cooling - comparison (CHT, HT fluid, HT solid)' simulation project by psosnowski2


#1

I created a new simulation project called 'Modeling object cooling - comparison (CHT, HT fluid, HT solid)':


More of my public projects can be found here.


#2

Analysis scope

The aim of this study was to compare different numerical models and analysis approaches used to simulate heat transfer and cooling of an object.

Simulation setup and run

Studied object was a 0.3 m diameter steel cylinder with a spherical top cup. Total height of the object was 0.65 m. It was placed in the center of an air box 3.0 x 2.0 x 1.5 m size. Inlet and outlet faces were added (refer to Drawing 1). The overview of the geometry is presented in Fig.1.

In order to demonstrate differences between thermal heating models, six simulations were performed.

  1. First analysis considered natural convection, with closed inlet and outlet faces. Two approaches to heat introduction were compared:
    a. 1000 W introduced through bottom face othe object
    b. 1000 W generated within the volume of the object
    In both cases heat was removed through top and side walls of the box which were kept at 293 K. This excluded in/out channels and bottom wall that were selected to be adiabatic.

  2. Second approach studied effects of forced convection with 1m/s inlet of 293 K. All air domain walls were considered adiabatic. Again, two means of heat introduction were studied:
    a. 1000 W introduced through bottom face of the object
    b. 1000 W generated within the volume of the object

  3. The third analysis aimed to investigate the flow patterns within the air domain. For this analysis the object was modeled only by a boundary condition that introduced 1000 W through its surface. Two setups were investigated:
    a. Natural convection, with box side walls and top wall kept at 293 K (similar to case 1).
    b. Forced convection study, where an inlet of 1 m/s pushed air at 293 K, while all walls were considered adiabatic (refer to case 2).

Simulations in Case 1 and Case 2 used Conjugated Heat Transfer model (fluid+solid), while Case 3 simulated used Convective Heat Transfer (fluid only). Case 1 and Case 2 shared the same mesh, while Case 3 used one of the same quality (refer to Fig.2). Meshes consisted of around 1.7 mln elements.

All analysis were carried out for 1000 iterations using 32 core machines. Overall runtimes of all simulations ranged from 30-50 minutes. Residual plots of all the runs indicate convergence to a solution (refer to Fig.3). At the same time it is important to note that in all cases the residuals reached a “plato”, indicating that only the best solution for a given mesh quality was obtained. This means that there is room for improvement of the results quality given that a finer mesh would be used.

Result comparison and discussion

Analysis of natural convection and forced convection deserve separate comparison due to different nature of the physical phenomena that are present in the system.

Natural convection

In this case the heat is introduced to the system through the object and is transported with the convective motion of air to the external box walls. Fig.4 presents comparison of vertical cross-slices through the flow domain while Fig.5 demonstrates the horizontal slice. It is clearly observable that temperature stratification develops in the domain for each case. Introduction of the heat uniformly through the surface as in Case 3.a results in a reverse temperature pattern on the surface of the object in comparison to CHT setup of heating from below (Case 1.a). Heat generation within the whole body (Case 1.b) produces an uniform temperature distribution on the object surface. For comparison of object’s surface temperature refer to Fig.6. Significant difference in the way temperature is introduced to the air domain result in generating a more stratified (Case 1.a and Case 1.b) or uniform (Case 3.a) temperature distribution.

Forced convection

Forced convection setup investigates the way heat is removed by blowing it away with a air flux at fixed temperature. Cases 2.a, 2.b and 3.b will be compared in this section. It is expected that due to the presence of forced flow effects of temperature stratification will be negligible. At the same time a heat-wake should develop behind the body as the flow drags the heat towards the outlet. Fig.7 visualizes temperature profiles at vertical slices across the domain and temperature plot on a line connecting two opposite corners of the air box. As expected, stratification effects are not dominant in this case. Fig.8 demonstrates horizontal slice of the domain at 0.6m. We can clearly observe the heat wake developed behind the object. It clearly visible that CHT simulations provide similar results regardless of the heat source method, while convective case results in different temperature profile. The similarity of CHT cases can be explained by high thermal conductivity of object’s material (steel). Regardless whether heat is introduced through the bottom face (Case 2.a) or is generated inside the volume (Case 2.b), it is easily transported across the whole body and results in a similar, more uniform temperature range. In contrast, convective case assumes uniform heat-flux distribution on the surface of the body. In result, the effect of blowing the heat away from the object generates big temperature differences at the surface (see Fig.9).

Results mapping

Description of simulation cases is summarized in Table.1
Case 1.a - CHT natural convection, heating from below S3 R1
Case 1.b - CHT natural convection, volume heat generation S4 R1
Case 2.a - CHT forced convection, heating from below S1 R1
Case 2.b - CHT forced convection, volume heat generation S2 R1
Case 3.a - Fluid heat transfer - natural convection S6 R4
Case 3.a - Fluid heat transfer - forced convection S5 R2
Table.1 - Mapping of simulation data to project cases

Fig.1 - Geometry overview

Drawing 1 - Air domain dimensions

Fig.2 - Mesh quality comparison. Top: two-region mesh for Case 1 & 2, Bottom: single region mesh for Case 3

Case 1.a

Case 1.b

Case 2.a

Case 2.b

Case 3.a

Case 3.b

Fig.3 - Residual plots

Fig.4 - Natural convection case - vertical temperature slice comparison. From left: Case 1.a heating from below, Case 1.b volumetric heat generation, Case 3.a surface heat generation

Fig.5 - Natural convection case - horizontal temperature slice at Z=0.6m comparison. From left: Case 1.a heating from below, Case 1.b volumetric heat generation, Case 3.a surface heat generation

Fig.6 - Natural convection case - comparison of temperature on the surface of the object. From left: Case 1.a heating from below, Case 1.b volumetric heat generation, Case 3.a surface heat generation

Fig.7 - Forced convection case - vertical temperature slice comparison. From left: Case 2.a heating from below, Case 2.b volumetric heat generation, Case 3.b surface heat generation

Fig.8 - Forced convection case - horizontal temperature slice at Z=0.6m comparison. From left: Case 2.a heating from below, Case 2.b volumetric heat generation, Case 3.b surface heat generation

Fig.9 - Forced convection case - comparison of temperature on the surface of the object. From left: Case 2.a heating from below, Case 2.b volumetric heat generation, Case 3.b surface heat generation