Short pin fins are often used to increase the heat transfer to the coolant in the trailing edge of a turbine blade. Due primarily to limits of casting technology, it is not possible to manufacture pins of optimum length for heat transfer purposes in the trailing edge region. In many cases the pins are so short that they actually decrease the total heat transfer surface area compared to a plain wall. A heat transfer data base for these short pins is not available in the literature. Heat transfer coefficients on pin and endwall surfaces were measured for several staggered arrays of short pin fins. The measured Nusselt numbers when plotted versus Reynolds numbers were found to fall on a single curve for all surfaces tested. The heat transfer coefficients for the short pin fins (length to diameter ratios of 1/2 and 2) were found to be about a factor of two lower than data from the literature for longer pin arrays (length to diameter ratios of about 8).

Recently, several experiments concerning heat transfer from short pin fins have been conducted with the results indicating lower heat transfer from short pin fins than from longer pin fins found in tube banks and other similar configurations. Assessments of the effect of the number of pin rows and row geometry have also been made. It was felt that there was a need to determine the relative contribution of pin length to diameter ratio and pin row geometry on the heat transfer. Array-averaged heat transfer coefficients on pin and endwall surfaces were measured for two configurations of staggered arrays of short pin fins (length to diameter ratio of 4). One configuration contained eight streamwise rows of pins, while the other contained only four rows. Results showed that both the eight-row and the four-row configurations for an Lp/D of 4 exhibit higher heat transfer than in similar tests on shorter pin fins (Lp/D of 1/2 and 2). It was also found that for this Lp/D ratio the array-averaged heat transfer was slightly higher with eight rows of staggered pins than with only four rows.

An experiment was conducted to measure the heat transfer from a heated cylinder, three diameters in length, in crossflow in an array of identical circular cylinders. Both in-line and staggered array patterns were studied. The cylinders were spaced on a 2.67 × 2.67 pitch. The Reynolds number range, based on cylinder diameter and average unobstructed channel velocity, was from 5000 to 125,000. Turbulence intensity profiles were measured for each case at a point one-half space upstream of the row containing the heated cylinder. The basis of comparison for all the heat transfer data was the single row with the heated cylinder. For the in-line cases, the addition of a single row of cylinders upstream of the row containing the heated cylinder increased the heat transfer by an average of 50 percent above the base case. Adding up to five more rows caused no increase or decrease in heat transfer. On the other hand, adding rows in the staggered array cases resulted in average increases in heat transfer of 21, 64, 58, 46, and 46 percent for one to five upstream rows, respectively.