0
Research Papers

# Heat Transfer and Pressure Investigation of Dimple Impingement

[+] Author and Article Information
K. Kanokjaruvijit

Department of Mechanical Engineering, Naresuan University, Muang, Pitsanulok, 65000 Thailandkoonlaya@gmail.com

R. F. Martinez-Botas

Department of Mechanical Engineering, Imperial College London, London, UKrmbotas@imperial.ac.uk

J. Turbomach 130(1), 011003 (Dec 14, 2007) (11 pages) doi:10.1115/1.2220048 History: Received October 01, 2004; Revised February 01, 2005; Published December 14, 2007

## Abstract

Heat transfer and pressure results of an inline array of round jets impinging on a staggered array of dimples are reported with the consideration of various geometric and parametric effects; results are normalized against flat plate data. The heat transfer was measured by using transient wideband liquid crystal method. The geometrical configurations considered were crossflow (or spent-air exit) scheme, dimple geometries, and impinging positions. Three crossflow schemes were tested such as one-way, two-way, and free exits. These led to the idea of the coupling effects of impingement and channel flow depending on which one dominated. Hemispherical and cusped elliptical dimple shapes with the same wetted area were considered and found that both dimples showed the similarity in heat transfer results. Impinging positions on dimples and on flat portions adjacent to dimples were examined. Throughout the study, the pitch of the nozzle holes was kept constant at four jet diameters. The investigated parameters were Reynolds number $(ReDj)$ ranged from 5000 to 11,500, jet-to-plate spacing $(H∕Dj)$ varied from 1 to 12 jet diameters, dimple depths $(d∕Dd)$ of 0.15, 0.25, and 0.29, and dimple curvature $(Dj∕Dd)$ of 0.25, 0.50, and 1.15. The shallow dimples $(d∕Dd=0.15)$ improved heat transfer significantly by 70% at $H∕Dj=2$ compared to that of the flat surface, while this value was 30% for the deep ones $(d∕Dd=0.25)$. The improvement also occurred to the moderate and high $Dj∕Dd$. The total pressure was a function of $ReDj$ and $H∕Dj$ when $H∕Dj<2$, but it was independent of the target plate geometry. The levels of the total pressure loss of the dimpled plates werenot different from those of the flat surface under the same setup conditions. Wall static pressure was measured by using static taps located across each plate. $ReDj$ and $H∕Dj$ affected the level of the static pressure while the dimple depth influenced the stagnation peaks, and the crossflow scheme affected the shape of the peaks.

## Figures

Figure 1

Experimental apparatus

Figure 2

Crossflow schemes

Figure 3

Dimple geometries: (a) Hemispherical dimples and (b) Cusped elliptical dimples

Figure 4

Impinging positions

Figure 5

Static pressure tapping on target plate

Figure 6

Comparisons of overall average results of flat plate to literature—consideration of crossflow scheme effect. (a) Minimum crossflow. (b) Intermediate crossflow. (c) Maximum crossflow.

Figure 7

Comparisons of overall average results of flat plate to literature—results of maximum crossflow scheme at various ReDj and H∕Dj (corresponding to Phase 2, Table 1◆ represents data obtained in the present work,—(11-12) (a) ReDj=11,500. (b) ReDj=8000. (c) ReDj=5000.

Figure 10

Streamwise average Nusselt number distribution of different dimple geometries compared to that of the flat surface H∕Dj=4, ReDj=11,500, maximum crossflow, impinging on dimples (Phase 1 in Table 1)

Figure 11

Normalized average Nusselt numbers of both dimple geometries for maximum crossflow scheme, ReDj=11,500 (Phase 1 in Table 1)

Figure 12

Streamwise average Nusselt numbers of dimples at H∕Dj=2 and ReDj=11,500 (Phase 2, Table 1). (a) d∕Dd=0.25. (b) d∕Dd=0.15.

Figure 13

Normalized overall average Nusselt numbers of dimples with different depths at ReDj=11,500 (Phase 2, Table 1)

Figure 14

Normalized streamwise average Nusselt numbers of hemispherical dimples, H∕Dj=8 (Phase 1, Table 1(5))

Figure 15

Effect of Reynolds number on hemispherical dimples, overall average Nusselt numbers at various H∕Dj values (Phase 1, Table 1(5))

Figure 16

Streamwise average Nusselt number of hemispherical dimples, maximum crossflow scheme, ReDj=11,500 (Phase 1, Table 1(4))

Figure 17

Normalized overall average Nusselt numbers of hemispherical dimples, maximum crossflow scheme, Re=11,500 (Phase 1, Table 1)

Figure 18

Cross section along inline dimples (underneath jets) of Nusselt number distribution at ReDj=11,500, H∕Dj=4 (Phase 2, Table 1(5))

Figure 19

Overall average Nusselt numbers of dimpled plates of different dimple depths at H∕Dj=4 (Phase 2, Table 2 (5))

Figure 20

Streamwise average Nusselt number at H∕Dj=2, ReDj=11,500, impinging on dimples (see Table 1)

Figure 21

Contour plots of different Dj∕Dd. (a) Dj∕Dd=0.25 (Phase 3, Table 1). Note that the pitch of the jet plate has become 8Dj. (b) Dj∕Dd=0.50 (Phase 2, Table 1). (c) Dj∕Dd=1.15 (Phase 4, Table 1). Note that the white filled circles represent dimple areas, which are not taken into account, and the filled rectangles are the inline dimples where the jets are impinging.

Figure 22

Effect of crossflow scheme, ReDj=11,500, impinging on dimples

Figure 23

Total pressure for different setups, impinging on dimples

Figure 24

Effect of impinging positions

Figure 25

Effect of dimple geometry

Figure 26

Dimpled plate with imprinted diameter, Dj=40mm. Projected view of dimpled plates d∕Dd=0.15 and 0.25

Figure 27

Crossflow scheme effect on flat plate at ReDj=11,500, H∕Dj=2, streamwise direction

Figure 28

Surface static pressure measurements across the plate in streamwise direction along all rows at H∕Dj=4, ReDj=8000, d∕Dd=0.25, impinging on dimples

Figure 29

Surface static pressure along inline dimples at different Reynolds numbers, maximum crossflow, H∕Dj=2, d∕Dd=0.15

Figure 30

Surface static pressure along inline dimples at different jet-to-plate spacings, maximum crossflow, ReDj=11,500, d∕Dd=0.15

Figure 31

Surface static pressure along inline dimples at different dimple depths, maximum crossflow, H∕Dj=2, ReDj=11,500

Figure 32

Surface static pressure along inline dimples at different dimple curvature (Dj∕Dd), maximum crossflow, H∕Dj=2

Figure 9

Normalized overall average Nusselt numbers of hemispherical dimples at different crossflow schemes, impinging on dimples, ReDj=11,500 (Phase 1 in Table 1)

Figure 8

Streamwise average Nusselt numbers for hemispherical dimples at H∕Dj=2, impinging on dimples (Phase 1 in Table 1(14))

## Discussions

Some tools below are only available to our subscribers or users with an online account.

### Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related Proceedings Articles
Related eBook Content
Topic Collections