Shawn,
Those certainly are two interesting points indeed.
If we think about the erosion on the Bell 212 blades first. To start let me ask a question...If flow separation causes erosion to the surface of rotor blades wouldn't we expected to see rather a lot of erosion near the root - i.e. in the reversed flow region? I don't really know the answer to your original question I can only offer an educated guess. As we all know the erosion on the leading edge of rotor blades near the tip is essentially impact damage, from debris thrown around by the rotor wake, insects and other bits and bobs. It is most noticeable near the tips since this is where the velocities are greatest. I would argue that the erosion that you see on the trailing edge is likely to be related to the entrainment of that same debris etc, by the role-up of the tip vortex. I can't say whether or not the flow there will be separated, but it will certainly be highly three-dimensional.
If you are interested in the boundary layer behaviour on the Huey series then Tanner and Yaggy did a fascinating study on this subject using a UH-1B in the hover. The reference for this is:
Experimental boundary layer study on hovering rotors, Journal of the American Helicopter Society, Vol.11(3) pp. 22-37, 1966.
One of the figures in the report clearly shows the 'affected' area of the upper surface of the blade, over which the vortex lies, if this is approximately the same region in which you have noticed the erosion then we have validated my argument. The report also shows the extent of the laminar and turbulent regions on the rotor, which is essentially what this thread has been all about.
So the short answer is that I think the erosion on the aft of the blade in the tip area is likely to be the result of debris entrainment into the tip vortex flow field, which may or may not involve local flow separations.
The point about the CH47D blades is similarly interesting, but I feel that saying that surface roughness helps delay separation is way too simplistic to be particularly helpful. I have never seen any convincing evidence that the general rule is that separation location is delayed with increasing surface roughness levels for a fully turbulent flow. However, as we have discussed above and as you will see in the paper I have highlighted - even on relatively large helicopters operating at high tip Reynolds numbers, large regions of laminar flow exist on both the upper and lower surfaces of the main rotor. The report cited, illustrates a laminar region of approximately 10% chord on the upper surface and 40% chord for the lower surface for a severely eroded blade in steady hover. Under these conditions then surface roughness can make a significant difference. However, the roughness must exist ahead of the transition location, in order to trigger laminar-to-turbulent boundary layer transition prior to the location at which it would naturally occur. If this is done (on the upper surface only) then yes, leading edge laminar separation can be eliminated. However, I find it quite hard to believe that it can be demonstrated that surface roughness on a region of the blade that will be turbulent anyway will have a significant effect on the separation characteristics. All that I can see this doing is increasing the viscous drag.
So in this case I feel quite strongly that you should take that explanation with a pinch of salt. I think it is a classic case of simplifying things to the point that they don't actually make sense any more!
With regards your question about trying turbulator tape on rotors, the answer is no. The reason is that only in the last few years have we been able to predict boundary layer transition accurately on helicopter rotors in the hover. We still can't do it for forward flight! If you can't predict the transition behaviour then you can't simulate the transitional flow. If you can't simulate the transitional flow then you won't predict the bubble behaviour correctly. And if you can't predict the bubble behaviour correctly how will you know where to put the tape to avoid the laminar separation that causes the bubble? Remembering that because of the vast array of aerodynamic conditions any balde section will be subjected to these transitional features will be moving all over the place and if you put the tape in the wrong place it will make things worse not better!
I hope this helps and keeps this interesting discussion going...
All the best,
CRAN
