0
Research Papers

# Effects of a Realistically Rough Surface on Vane Heat Transfer Including the Influence of Turbulence Condition and Reynolds Number

[+] Author and Article Information
E. L. Erickson

Third Wave Systems, 7900 West 78th Street, Suite 300, Minneapolis, MN 55439ethan.erickson@thirdwavesys.com

F. E. Ames

Department of Mechanical Engineering, University of North Dakota, Grand Forks, ND 58202forrestames@mail.und.edu

J. P. Bons

Department of Aerospace Engineering, Ohio State University, 2300 West Case Road, Columbus, OH 43017bons.2@osu.edu

J. Turbomach 134(2), 021013 (Jun 27, 2011) (8 pages) doi:10.1115/1.4003026 History: Received July 08, 2010; Revised July 09, 2010; Published June 27, 2011; Online June 27, 2011

## Abstract

Heat transfer distributions are experimentally acquired and reported for a vane with both a smooth and a realistically rough surface. Surface heat transfer is investigated over a range of turbulence levels (low (0.7%), grid (8.5%), aerocombustor (13.5%), and aerocombustor with decay (9.5%)) and a range of chord Reynolds numbers ($ReC=500,000$, 1,000,000, and 2,000,000). The realistically rough surface distribution was generated by Brigham Young University’s accelerated deposition facility. The surface is intended to represent a TBC surface that has accumulated 7500 h of operation with particulate deposition due to a mainstream concentration of 0.02 ppmw. The realistically rough surface was scaled by 11 times for consistency with the vane geometry and cast using a high thermal conductivity epoxy $(k=2.1 W/m/K)$ to comply with the vane geometry. The surface was applied over the foil heater covering the vane pressure surface and about 10% of the suction surface. The $958×573$ roughness array generated by Brigham Young on a $9.5×5.7 mm2$ region was averaged to a $320×191$ array for fabrication. The calculated surface roughness parameters of this scaled and averaged array included the maximum roughness, $Rt=1.99 mm$, the average roughness, $Ra=0.25 mm$, and the average forward facing angle, $αf=3.974 deg$. The peak to valley roughness, Rz, was determined to be 0.784 mm. The sand grain roughness of the surface $(kS=0.466 mm)$ was estimated using a correlation offered by Bons (2005, “A Critical Assessment of Reynolds Analogy for Turbine Flows,” ASME J. Turbomach., 127, pp. 472–485). Based on estimates of skin friction coefficient using a turbulence correlation with the vane chord Reynolds numbers representative values for the surface’s roughness Reynolds number are 23, 43, and 80 for the three exit condition Reynolds numbers tested. Smooth vane heat transfer distributions exhibited significant laminar region augmentation with the elevated turbulence levels. Turbulence also caused early transition on the pressure surface for the higher Reynolds numbers. The rough surface had no significant effect on heat transfer in the laminar regions but caused early transition on the pressure surface in every case.

<>

## Figures

Figure 1

Large scale linear cascade test section

Figure 2

Large scale low speed wind tunnel facility

Figure 3

Smooth vane surface pressure distribution

Figure 4

Digital photograph of heat transfer vane with high thermal conductivity tiles installed

Figure 5

Comparison of surface pressure distribution for smooth and rough vanes for ReC=500,000, 1,000,000, and 2,000,000

Figure 6

Smooth vane heat transfer distributions compared with boundary layer code predictions, ReC=500,000, low, grid, and aerocombustor turbulence

Figure 7

Smooth vane heat transfer distributions compared with boundary layer code predictions, ReC=1,000,000, low, grid, and aerocombustor turbulence

Figure 8

Smooth vane heat transfer distributions compared with boundary layer code predictions, ReC=2,000,000, low, grid, and aerocombustor turbulence

Figure 9

Rough vane heat transfer compared with smooth vane distributions, ReC=500,000, low, grid, and aerocombustor turbulence with and without spool

Figure 10

Rough vane heat transfer compared with smooth vane distributions, ReC=1,000,000, low, grid, and aerocombustor turbulence with and without spool

Figure 11

Rough vane heat transfer compared with smooth vane distributions, ReC=2,000,000, low, grid, and aerocombustor turbulence with and without spool

Figure 12

Vane pressure surface roughness Reynolds number distributions, with varying turbulence condition and Reynolds number

Figure 13

Vane surface acceleration parameter distributions, ReC=500,000, 1,000,000, and 2,000,000

## 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