Re: bottom only tape application
j...@oxaero.com
Yes, it was very exciting to see improved performance on the vario all across the high speed range. It felt like I could zip the sky and never come down.
Since I am sending this to the group, I need to say that this (http://deturbulator.org/Progress-01292011-LowerLET1.asp) was just one measurement in the present configuration and sink-rate measurements are susceptible to corruption from large scale vertical air movements such as shear waves. It usually takes at least three flights to average out the scatter for accurate results. But, these deviations are much larger than normal scatter deviations that are typically around 20 fpm. So, the broad outlines of this result are not likely to change with further flights.
The exact locations of the rear edge of the tapes on the lower surface can be determined from the pictures and tabulated top-surface measurements posted at http://www.deturbulator.org/LET-Yourself.asp. You can run a tape measure from the TE over the top surface to the top edge of the tape and then go around the LE 2 inches, the tape width. For the outer panel, you must be sure to measure from the TE to the lower of the two tapes. These measurements are given for three span stations: root, inboard end of aileron, where the outer panel begins, and the tip. The inner panel, root to aileron, is a linear transition zone from the root to the outer airfoil, so a straight line tape edge between the measured points works. Just stretch a line. The outer panel has one airfoil shape, so a stretched line works there too. This will give you a starting point from which you can make adjustments to improve on my results.
The thing to keep on mind when adjusting the position is that the rear edge of the tape is in a region where the pressure gradient is negative and extremely steep. It's a region where the pressure is recovering from the direct impact of the stagnation point. And, since the stagnation point moves around the LE with changes in speed/AOA, the rear edge of the lower-surface tape sees a stronger gradient at slow speeds and a weaker gradient at high speeds.
Here is my understanding of what is going on. I'm no aerodynamicist, so my terminology is unconventional and my thinking may be wrong, but there are so many supporting indications that I feel confident. In any case, there must be some explanation for the consistency of so many measurements over the years by me and by Dick Johnson in 2006 that are in agreement with mine. So, here goes.
Think of the pressure gradient near the leading edge as a force that accelerates the flow. Near the LE, you have much more pressure behind and much less ahead of an element of air in the boundary flow. At the rear edge of the tape, the skin friction is broken, freeing the flow for unrestrained acceleration. A low pressure bubble in the dip behind the rear edge of the tape initiates rolling turbules that fill the space between the detached flow and the surface. This can be seen in images at http://www.deturbulator.org/LET-Yourself.asp and also in a video clip that can be reached from the link "Navier-Stokes Simulation of Rear-Facing Step" at http://www.deturbulator.org/index.asp.
Now, you must suspend disbelief and accept that this condition of a stabilized, (essentially) laminar, marginally detached flow can extend all the way back to the normal reattachment point (about .6 chord on Workmann airfoils) with the abnormally accelerated flow, because of it's increased momentum, resisting the influence of the pressure profile more than it normally would. And, as a result, the flow can be made to follow a cambered trajectory that flattens or eliminates the bubble and reattaches gently at a grazing angle. (This condition is clear in numerous oil-flow images at http://www.deturbulator.org/Progress-OilFlows.asp where the oil is not wiped away at the reattachment point as it normally would be. In addition, there is at least one peer reviewed paper showing theoretically this sort of thing with a plasma actuator step instead of a passive step: http://www.deturbulator.org/AIAA-2011-733-896.pdf)
I said the flow "can be made" to follow a desired trajectory because a clean break with the surface at the rear edge of the tape does not produce the desired condition. What a coincidence that would be! In fact, I believe that is the cause of the extremely narrow performance notch (deficit) that always appears in my LE-tape-only measurements. The notch is so narrow that a change of one or two kts makes the difference between a deficit and a very large improvement. This corresponds to an AOA change of just .2 to .4 degrees. I think the narrowness of the notch tells us two things. First, the notch is not caused by tripping the flow. And, second, the notch is a one-surface only phenomenon since we would not expect the precise condition for a notch to occur on both surfaces simultaneously.
Ok, so there needs to be some form of control to achieve the desired detached flow trajectory. The source of this control can be seen in the afore mentioned Navier-Stokes images and video clip. Instead of breaking cleanly, the flow can be seen bouncing off the surface behind the rear-facing step. This local resistance behind the step pulls the downstream flow toward the surface. The degree of contact/resistance is controllable by changing the airspeed/AOA slightly. I think this is why performance improves greatly on each side of the notch, because the amount of local contact behind the step is being adjusted by changes in airspeed to values that shape the detached flow camber for a good, low loss, reattachment. At the precise speed of the notch we get too much of a good thing. The flow trajectory straightens too much, the normal transition bubble becomes larger and reattachment is even steeper and more violent than normal, hence the performance loss.
When everything is right, the energy loss per unit time (power loss) of the glider - rate of loss of altitude-potential energy (i.e., sink rate), is reduced in the following ways:
1. Reduced area of skin friction drag from attached laminar flow at the front of the wing. Since the area if attached laminar flow is greater on modern critical laminar wings, I wonder if they have more to gain from this efficiency.
2. Flattened transition bubble for reduced form drag.
3. More gentle reattachment resulting in a delayed growth of the attached turbulent flow so that the trailing edge wake is thinner, carrying off energy at a reduced rate.
4. On the top surface, there are clear indications that the lift coefficient increases. This appears to be due to the detached flow effectively increasing the camber of the airfoil. The indications include:
a. Abnormal, very large nose dip with one or two kts reduced speed (should increase AOA) and corresponding performance increase (http://www.deturbulator.org/20100402-PerformanceEvent.asp).
b. Reduced wing bending with only the inner panel of wing modified. (http://www.deturbulator.org/Progress-02262005.asp)
c. Cruising with nose lower than usual. When on tow at 70 kts with the glider aligned vertically so that the tug wheels are on the horizon, daylight can be seen between the tug wheels and the instrument panel cover. Normally, the panel cover obscures the belly of the tug.
d. On tow at 70 kts with the trim set to the most forward notch, normally, releasing the stick will allow the glider to nose up sharply. But with an upper surface tape edge the rise is much more gradual. And, with turbulator panels installed, there is almost no tendency for the nose to rise at all.
Compared to conventional turbulator tapes, this approach has no penalty, only additional efficiencies.
That's what I think, and I await the real aerodynamicists to do the hard work of finding out what is actually happening. With the Johnson flight test in 2006 and my own numerous performance measurements, parallel flights, oil-flow images and drag probe measurements, the professionals have ample cause to take up the work.
I suspect that my assumption that the forward edge of the lower-surface-only tape would not be a problem is not correct when it is above the stagnation point. This is one possible explanation for the dramatic performance loss at positive AOAs in my 1/29/2011 measurements. So using a narrower tape that puts the forward edge at least 10 degrees blow the nose of the wing may solve the problem. I'd smoothing is needed, perhaps a hard wax will do the job.
Thanks, Tom, for the kind remarks.
Jim Hendrix
On Feb 1, 2011, at 9:18 PM, Tom <tpho...@gmail.com> wrote:
> Hi Jim Hendrix,
>
> This test is pretty exciting especially the improvement over the broad
> speed range above 60KIAS. The idea of just using narrow width tape as .
> 50" or >75" also seems like it would allow much easier tape placement
> at the trip line desired. Have you described the exact location for
> this trip line for the underside as you proceed from root to tip? Can
> that be described by an angle in relation to the chord line through
> the wing profile for example? I would think that would make it fairly
> easy to determine points along the leading edge length with a jig. Do
> I understand correctly that you think the leading edge of the narrower
> tape idea just wouldn't have much negative impact? Might it even help?
> If it needed smoothing at the leading edge, what did you have in mind?
>
> Nice work Jim! Good to be coming upon a smile from time to time with
> all the efforts you have put into this! That must have been quite
> pleasant moment for you.
>
> Following with great interest;
>
> Tom Hopper
>