Experimental investigation of attachment line transition on a large swept cylinder
Transition of the attachment fine boundary layer was investigated using a large swept cylinder. Results for natural transition and transition tripping with two-dimensional trip wires were simila to those obtained by Poll using a similar, but smaller, model. ]Lower displacement thickness Reynolds numbers but larger trip sizes, than for the flat-plate boundary layer, were required for transition. The investigation of transition tripping was then extended to involve three-dimensional trips. The attachment line boundary layer was less susceptible to three-dimensional trips than to two-dimensional trips but upper and lower bounds of attachment line Reynolds number for transition were identical. It was also found that the roughness Reynolds numbers for fully effective three-dimensional trips were similar for the attachment line and flat-plate boundary layers. Another common feature was the more abrupt upstream movement of the transition front with increasing Reynolds number for three-dimensional trips than for two-dimensional trips. Turbulence spreading downstream of a three-dimensional trip was also examined and, as in the flat-plate boundary layer, was found to be heavily dependent on Reynolds number (varying from 3° at low Reynolds number to a value approaching 10° as Reynolds number exceeded the value for natural transition), but was also dependent on either the trip size or the initial conditions at which the trip first introduced turbulent spots. The effects of higher levels of freestrearn turbulence were then investigated for both two-dimensional and three-dimensional trips. With a small increase in freestrea turbulence the conditions for transition with twodimensional trips were affected far more than those for three-dimensional trips, for which only the transition completion conditions were affected signfficantly, resulting in a reduced extent ofthe transition region. Larger levels of turbulence appeared to have similar effects on the two trip types. Restrictions in model length and windspeed for the higher turbulence tests prevented an accurate investigation of the effects of turbulence,on the upper and lower bounds for transition tripping and on the influence of spanwise distance at higher levels of turbulence. Finally, the interaction between two trips positioned on the attachment line was examined. The effect of the second trip on the transition Reynolds number was found to a function of the streamwise separation distance between the two trips.