Hi Graham,

Sounds all good so far!

I suggest considering the turbulators to be composed of two sections. An inner section which spans most of the wing, and an outer section that will cover perhaps the outer 8 to 10 inches of tip. I'll have to run a few more numbers to give you the break location between these sections.

The outer turbulator should likely be zig-zag tape. But the inner turbulator should likely be a single layer of narrow striping tape, in a fairly straight line.

The reason for this difference is how far back from the turbulator we want the air to trip to turbulent. Out near the tips, we want it to go ahead and trip a few percent behind the turbulator. So the turbulator needs to be a little aggressive. But on the inner wing, we want the flow to trip quite a ways behind the turbulator. We want to keep the bubble intact just make it a little shorter. So we aren't really trying to trip it as much as to just add a little energy to the surface flow, in the hope that it will now trip a little sooner than it otherwise would. That part is going to be a little tricky and require testing.

So I would start with the outer turbulation as that is the most straight-forward. I'd do one side only, and fly to see what differences show up. However, beware and don't give it a full power launch! It may well do a roll...

Things to look for are a change in yaw or roll trim, possible change in pitch trim, change in behavior in turbulent air, change in roll response to ailerons, change in behavior in very tight turns. I don't think the dead air difference will be measurable, unless the handling is improved enough to make a difference. Anyway some of the differences should show up as asymmetry in response for left and right. Testing needs to be done when flying at non-optimal speeds. Flying a little too fast is where most of the difference should show up.

Once the tips are worked out (if we're lucky, first attempt), then duplicate it on the other wing. Then start on the inner turbulator which is where most of the potential performance boost is located - if it works! That one is trickier to figure out the thickness required. Too thin and it will do essentially nothing. If the difference is detectable then likely it is thick enough for the intended purpose.

I'll get you some numbers for locations.

Gerald

Ok here are the promised pics.

Similar to my Zone Ver2, this plane has a fat wing with a root chord of 190mm. Wing design followed Gerald's procedure for these foils with a little tweaking here and there. XFLR5 predicted performance for the turbulated wing is impressive. It remains to be seen whether the plane will live up to the predictions.

Wing area is 2247 cm2 and flying weight is 292gm. Would have been about 280gm if I didnt have to deal with the short nose of the Raptr pod.

At the present time, I am trying to spend as much time as possible with the plane to get to know it before I start experimenting with turbulators. Performance without turbulators is roughly the same as the Zone Ver2 plane with maybe a little less penetration speed. However, that may be due to setup which still needs lots of work. I have not yet had any good weather to spend time in serious trimming sessions - too much wind and turbulence and the next week or so looks to be unflyable because of high winds.

Graham

For the "60" airfoil, the bubble begins 55% back at an alpha of 1.5 degrees. So we need to turbulate perhaps 52% back from the leading edge. Since the desired trip point is rather far back at 85%, the action of the turbulator needs to be weak.

For the "50" airfoil, the bubble begins 46% back at an alpha of 1.7 degrees. So we need to turbulate perhaps 43% back from the leading edge. Since the desired trip point is rather far back at 85%, the action of the turbulator needs to be weak.

For the "40" airfoil, the bubble begins 42% back at an alpha of 1.9 degrees. So we need to turbulate perhaps 40% back from the leading edge. Since the desired trip point is rather far back at 85%, the action of the turbulator needs to be weak.

For the "30" airfoil, the bubble begins 32% back at an alpha of 2 degrees. So, we need to turbulate perhaps at 30% back from the leading edge. Since the desired trip point is rather far back at 78% or thereabout, the turbulator needs to be weak in action.

For the "20" airfoil, the bubble begins at 32% back at an alpha of 2 degrees. Same suggestion as previous foil.

Based on all this, I'd start with a full span turbulator on one wing only. Make the turbulator a single thickness of narrow striping (or other convenient) tape, with the back at 52% chord at the root and 30% back near the tip. I'd make the tape a straight line, not zig-zag. We don't want to trip the flow immediately. We want to persuade it to trip a little easier.

Testing should consist of flying at optimal airspeeds in a straight line, where one hopes to find that there is little difference between the wings (turbulator is inside the bubble at the front where surface speed is essentially zero), and flying a couple meters per second faster, where one hopes to find the turbulated wing (turbulator is now exposed) has less drag so the plane turns to the other side.

If the straight line flight is compromised but the higher speed flight behaves as desired, then the turbulator needs to be thinner (or possibly farther back - results likely are sensitive to accuracy of XFLR5 predictions).

If the higher speed flight does not show a turn tendency away from the turbulated wing, the turbulator needs to be thicker.

If thickening reaches the point where straight line flight is compromised and higher speed flight is not improved, then this approach, at least with these foils, is a failure.

This approach should be sufficient to determine whether the approach has merit. If it does, then one can go back and do a more detailed checkup of turbulation for different regions of the wing. That would include location, thickness, and method.

Air turbulence conditions will likely affect the results. It should be interesting.

Gerald

I have fitted a turbulator strip to one wing. The tape I used is similar to hinge tape, but has a textured surface. 2 layers gives me 0.3mm, which is about what Martin Hepperle recommends. I am not sure about the width of this tape and suspect that it should be a bit narrower. With the textured surface, it may be a bit too aggressive in effect.

Conditions were not ideal this evening. Wind was 10-12mph but at least it was fairly smooth. I used the wind to slope a small 2m high dune. The one advantage of this type of flying is that you can virtually hover the plane right in front of you and watch it closely. The main problem is that air speed is difficult to judge as the plane is virtually standing still. You can't fly too slow or you get pushed over the back. I am pretty sure that most of my flying was above optimum speed.

The good news is that there was a consistent and marked tendency to turn towards the unturbulated wing. I am also pretty sure that this tendency was reduced when I tried to slow it down. When flown much too fast, as I was forced to do some of the time, the difference seemed less.

The main problem with these tests today is that the small dune meant that I could not fly in a straight line for very long.

I really want calmer weather for more definitive tests

Graham

PS
After looking at the pics for a while, I am not sure whether the placement of the strip is correct near the tip. I was in a hurry to get it on and go fly so I placed it in a straight line from root to the last panel break and then continued in a straight line to the tip. It should maybe be curved back a bit as the tip is approached. Please comment.

Your planes are looking great Graham!

Well done!

The turbulator strip seems too wide to me. At that width, it will behave as much as an alteration of the shape (camber, thickness, etc) of the foil as anything else. What you want is a narrow strip, and likely not as thick. The classical thickness of a turbulator is intended to trip the flow pretty much immediately. That is not what we want to do here. We just want to energize the flow a little bit but NOT trip it. We are just encouraging it to trip a little sooner than it otherwise would do, left on its own. So I'd suggest starting with one layer of 1/8" striping tape or something similar.

As to location... At your last panel break (start of the tip panel, or wherever you have the "30" airfoil), set the turbulator's back side to be 30% back from the leading edge. At the root, set the turbulator's back side to be 52% back from the leading edge. Connect with a straight line, and run it out to the tip. Nothing fancy needed, just a thin narrow straight strip of tape.

Interesting results anyway. But some of it could be due to the size (width and thickness) altering the shape of the airfoil in addition to turbulation. That is what using a narrow strip will sort out.

Gerald

Thanks J

Gerald - As usual, I did not read carefully enough what you said. I have changed the tape. This is about 0.1mm thick. Now you can see clearly where the tape is.

If the weather Gods are kind, will do some more tests tomorrow. Otherwise, as soon as the wind abates.

Graham

Graham, I would love to see some of your building techniques, That is one beautiful bird.

I do not normally do a lateral balance on my planes but realised that these tests really require the plane to be properly balanced laterally. I found the plane required about 3gm on the left wing, which is the one I have been placing the turbulator on. This is not good. So I did a carefull lateral balance and shifted the CG slightly forward. The plane was a little too responsive to turbulence, which was potentially masking the effects I was looking for. It is now at about 8% static margin.

This morning we had a steady wind of 7 to 8m/s (measured). I have come to realise that if the wind is in the right direction where it comes off the sea (so has low turbulence) and I am able to slope a small dune that this is an ideal test situation as I can hover the plane close to me and observe effects close-up.

I set up 2 flight modes on the Tx, both of which use 5 deg of camber. The first trims for an air speed of about 6-7m/s which is about optimum. The second trims for an air speed of about 1 to 2m/s faster. This way I could hover the plane in front of me and then hit the switch and move away slowly without having to push on the elevator and risk adding tiny aileron input as I am pushing down elevator.

I first flew with the wing clean to make sure that the trim was correct. I then tried 3 different turbulators, the red striping shown in a previous post, a single layer of the wider textured tape and a double layer of the textured tape.

In all configurations I could not detect any difference in the behaviour of the plane. Obviously, the previous differences that I saw were due to lateral imbalance.

In all, I flew for about 1.5 hours and really tried to find some evidence of the effect of the turbulators, but there was nothing to see. Conditions were pretty smooth so even small effects should have been visible.

This has come as a surprise to me as I was really expecting to see SOME differences between the wings and I am not sure how to carry this forward.

Graham

5 degrees of camber sounds rather high to me for those flight speeds, unless you are carrying lots of ballast, FWIW. 2 degrees is probably more appropriate.

Using the narrow striping, add layers until a difference is observed. At some point, it will be different. Whether the difference, when first observed, is positive or negative, remains to be seen!

If the XFLR5 analysis is off a little, or the actual wing is off a little, then the bubble location may be farther forward. If it is a little farther forward, then this will mask the turbulator rendering it essentially invisible. Flying progressively faster should unmask the turbulator. The ideal location, thickness, and type of turbulator would result in improvements when the plane is flown a little too fast, but no observable difference when the plane is flown optimally. The analysis I did for this wing was at an Ncrit of 9, which should correspond to decently smooth air. It is not at all unreasonable that things may change in turbulent air. I'll have to run a bunch more numbers to see what the pattern might be.

If one has to go too much too fast to get the first observable difference, then the turbulator probably needs to be moved a little bit towards the leading edge. Although it could be that the turbulator is just not quite thick enough. The boundary layer gets thinner for the forward part of the wing as airspeed goes up.

If one has to be in air that is a little turbulent to see an observable difference, then the turbulator probably needs to be a little thicker or a little more aggressive (zig-zag).

You are flying in conditions that are already probably somewhat turbulent (low altitude wind tends to be that way). If the air flow starts out as turbulent, then it may be that attempting to add additional turbulation is not of benefit. I have not run any numbers for this wing at higher turbulence numbers. I have done so for other wings. By memory, generally as turbulence level goes up (Ncrit) there is an increase in overall drag at optimal conditions, but a decrease in drag when flown a little fast (relative to smoother air results). The latter is essentially what we are trying to achieve with these turbulators...

Gerald

For the "60" airfoil in air corresponding to Ncrit=7 (a little more turbulent than the smooth air I did the analysis with), the bubble begins about 66% back at an alpha of 1.5 degrees. This is well behind the smooth air position. The optimal turbulator position would then be somewhere closer to 60% back and would need to be a little thicker (boundary layer is getting thicker farther back).

Turbulence level makes a big difference...

Gerald

Yesterday I did do some 70 to 80% effort launches with the thin red turbulator strip fitted. There was no noticible effect on the upward trajectory during launch so the red strip is probably too thin.

I take your point about turbulence in low level winds so maybe I really should wait for calm conditions for further tests.

Just for reference, after lateral balancing and CG shift, the plane is now flying at 301gm with a wing area of 2247cm2 for a wing loading of 13.4gm/dm2 which is quite close to the example wing at 12.9gm/dm2.

I will do the next testing in calm weather, adding strips on top until I can see a difference.

Looks like we may have some calm weather middle of next week.

More to come

Graham

PS
Is there any value in doing tests with a full forced transition at the foil design location

If you mean doing an aggressive turbulator back towards the rear, not really. It would be in the separation bubble it is intended to prevent!

Just going thicker or going to zig-zag at the location I specified should be sufficient. Though your observation earlier about curving the turbulator back in the last panel is probably valid, if for no other reason than to keep it from adding extra drag.

It is ok to fly in turbulent conditions. You can reduce the effect somewhat by simply flying higher and not trying to maintain a position on the slope. For testing purposes, it really doesn't matter if you are flying in lift or not.

I would expect there could be differences in calm vs. turbulent air, and it is all useful information!

If differences don't show up in launch (roll, yaw, possibly pitch) then the turbulator isn't going to do anything at any speed.

I'd suggest thickening up the turbulator a fair bit and seeing what happens. Perhaps go thicker than you think you need, just to force it to do something. Then back off as much as you can, progressively, until it stops working. Then add back a layer...

Are you checking the lateral balance after adding the turbulator? Just curious.

Gerald

Yes - I now try to keep lateral balance correct after every change.

I like your suggestion of going very thick on the turbulator and then backing off. At least, hopefully I will see some effects sooner. I think that when I thicken it up, I will do some launch tests first.

I realise that these tests can be done in turbulent air. Its just that I dont think I am good enough to always know whether it is the turbulator or turbulence that has caused some effect that I am observing.

Its just a little beach dune so the lift did not go high up. I quite like watching for effects with the plane stationary about 20 ft away from me. Getting carried away and not thinking enough. I should have just launched it higher.

I need some launch practice anyway so starting with launch tests should be good.

Graham

On another track, - there is something I am not getting here.

The polar data that we are working with for the foils, was generated was generated by Xfoil, which assumes that the bubble is present. If we manage to prevent the bubble, then does that not invalidate our analysis.

Graham