X-foil turbolator predictions

[Peytr]

Tried the PRofili forum and Mr Drela but no luck so far.

I'm working on small and micro gliders most of the time and like to know how I should value the turbolator predictions made by X-foil. I tend to fly micro gliders and have to encounter the very low end of the Re nrs scale.

I ask because the results of the simulations look very similar on all profiles. It seems the program assumes a constant result of the use of tubolators on all foils. Can anybody comment on this?

[green66]

Hi Peytr,

This may not address your specific problem since I use Profili, which incorporates XFoil, but is more user-friendly. Anyway, here's my approach to determine the optimum chord location and thickness of turbulators.

The goal is to attach a turbulator as far back as possible on the wing without going behind the transition point. In that regard, I typically mount the turbulator 5-10% ahead of the calculated optimum location for conservatism.

Again, this method uses Profili, which I can highly recommend for anyone designing or modifying their own wings (it's only $10 US), does more than XFoil alone and is much easier to use.

PROCEDURE
Turbulator Position
1) Determine the anticipated Reynolds Number (Re) for the various chords of your wing. This is the Re value to use in the following steps.

2) In Profili, run a Type 4 analysis, where all parameters except turbulator location are the same (see the screen shot in my next post for clarification). This is easily accomplished by entering all parameters for the 1st plot only, then select menu item "Copy from Polar..." to set all the other plots the same, then manually change only the turbulator chord position for each plot. After the first analysis, you can use this feature to change the airfoil, Re, etc for all plots in one shot. . . very convenient :-)

I usually set the chord locations from 20% to 80% in 10% increments. Remember to compare a non-turbulated plot to avoid turbulation if it actually worsens things.

3) From the Profili plots you'll be able to see at what chord position the drag polars start to break down, indicating the optimal chord position. Again, I typically put the turbulator 5 to 10% of chord ahead of the predicted optimum position, i.e. if calculated optimum position is 50%, then put turbulator at 40-45% of chord.

4) After the first analysis, it's easy to try other parameters by changing the first plot only then, as above, using the "Copy from Polar..." feature. It might be overkill, but you can also refine the analysis by specifying a narrower range of chord positions, e.g. 40 to 60% in 5% increments, etc.

5) Repeat above for remaining wing chords.

Turbulator Thickness

I use the chart below taken from this informative page Turbulation (http://www.mh-aerotools.de/airfoils/turbulat.htm) at Martin Hepperle's site to determine turbulator thickness.

Note that the recommended turbulator height shown on the x-axis is not the actual turbulator thickness - The x-axis values must be multiplied by the chord to obtain the proper turbulator thickness.

I'm not sure of the basis for the chart, but it's the only (understandable) information that I've found on the subject. I'm sure it's based on something rational, probably calculated boundary layer thickness.

Hope this helps.

[green66]

Here is the Profili Type 4 input screen - Each plot differs only by turbulator location, and includes a non-turbulated airfoil (Polar N.1).

[green66]

Here are the calculated polars for the various turbulator locations. The analysis predicts a significant benefit from turbulation compared to the non-turbulated airfoil.

I arbitrarily chose the HN-1027 airfoil because it's fairly thin and at Re = 40,000 - Figured this might be similar to your small-glider situation.

For this example, it appears that the magenta colored curve provides the best result, suggesting a turbulator positioned at 60% of chord. Again, for conservatism, I would go with 50-55% of chord.

[green66]

From the Hepperle chart, the recommended turbulator thickness for this example (turbulator at 50%, Re = 40k) would be about .0021 to .0025 X chord, i.e. for a 4" chord use a turbulator about .008" to .010" thick.

IMO, the easiest way to make turbulators is to stack several layers of thin tape of known thickness, then cut out straight strips, or the zigzag type with pinking shears.

[Peytr]

@green66

Thanks for your replies.

I forgot to mention I use Profili. The chart I didn't know, so that's a help.

Basically I work the way you do, what I'd like to know is how the tubolator simulations relate to reality. The plain polar calculations are quite good, because there are enough windtunnel polars available to compare with the ones generated by Profili. Windtunnel generated polars including turbolator applications are very scarce however, so I can't get a comparison to judge how real the predictions are. To a certain extend I don't really trust the X-foil predictions involving turbolators. The basic foil predictions however, are excellent, imo.

[green66]

....what I'd like to know is how the tubolator simulations relate to reality Here are the only practical ways, i.e. no wind tunnels and elaborate instrumentation required, that I can suggest to get physical proof of the effectiveness of turbulators, and you may have already tried them. I haven't actually tried these approaches, so don't know if they're meaningful in practice or at small scale.

First trim the unturbulated glider for stable straight flight. Apply a turbulator to one wing only and see if the model tends to bank or turn (away from the turbulated wing). Then remove the turbulator and apply it to the other wing and check for a tendency to turn the other way, which will confirm that the turbulator provides a beneficial effect.

Finally, again with a turbulator on one wing only, move the turbulator slightly aft until no turning tendency is observed, indicating that the turbulator is behind the point of transition and needs to be moved forward.

Then compare actual results to the best chord location predicted by Profili. Ditto for turbulator thickness from the Hepperle chart.

Offhand, the only way I can think of to quantify the actual reduction in profile drag is to compare L/D by comparing distances glided from a known height in "dead calm" conditions. Provided the model isn't flying in ground effect, I believe the actual reduction in profile drag will be twice the improvement in glide distance, since total drag = profile drag (mostly) + induced drag, and profile drag = induced drag when flying at the best glide (max L/D) speed, i.e. a 5% improvement in glide distance indicates a 10% reduction of profile drag. This will require taking the average of many glide measurements to be statistically valid.

L/D can also be compared by flying a glider equipped with a data logging altimeter such as the Alti2 (http://www.lomcovak.cz/a2/a2.html). It's tiny, but still might not be practical if your glider is really small.

Another approach I've read about and would like to try involves kerosene, maybe other volatile liquids also. Not sure if it works at very low Re, and only practical if your model can be wiped clean, i.e. covered with plastic film. Basically a film of dyed kerosene is applied to the wing, which will supposedly reveal the location of transition by leaving a brighter-colored film behind the transition point because of faster drying occurring in the turbulent boundary layer.

I've also heard about colored mineral oil being used similarly, but I think this method is based on getting a physical "imprint" in the film to show the transition point, based on a pressure difference between laminar and turbulent flows. Based on that, I suspect that this approach might be useful only at higher Re where pressure differences may be greater. An everyday analogy would be a car having a steep rear window, where the line between the dirty and clean portions reveals the transition point.

[green66]

The plain polar calculations are quite good, because there are enough windtunnel polars available to compare with the ones generated by Profili. Windtunnel generated polars including turbolator applications are very scarce however, so I can't get a comparison to judge how real the predictions are. To a certain extend I don't really trust the X-foil predictions involving turbolators. Hi Peytr,

Measured data may be better when considering a single airfoil, however for comparing different airfoils or airfoil modifications a pure analysis may actually be preferred instead of wind tunnel data because the analytical approach eliminates the errors and uncertainties inherent to any kind of testing where results are sensitive to several variables, especially when data is collected from different facilities by different people.

Analytical methods will provide absolute values that may differ from reality, but comparisons and relative effects between several airfoils can often be more accurately predicted because of the reasons above - The total effect of instrumentation inaccuracies, different measurement methods and protocols, variances of test specimen profile and surface quality, quality of the tunnel flow, etc. can significantly influence results.

With Profili / XFoil or other analytical method, all of these outside influences inherent to real testing are eliminated, therefore a reliable (IMO) comparison between airfoils can be made where the difference in polars is based solely on the airfoils' influence on the local momenta and separation of a pure 2D flow, plus viscous effects.

[Peytr]

Thanks for all yout input.

First trim the unturbulated glider for stable straight flight. Apply a turbulator to one wing only and see if the model tends to bank or turn (away from the turbulated wing). Then remove the turbulator and apply it to the other wing and check for a tendency to turn the other way, which will confirm that the turbulator provides a beneficial effect.

This is so simple I should have thought of this myself! :D
Will definately give it a try. Thanks again.

[Antonsoarer]

Thats interesting I hadn't managed to get any results using turbulators in Profili, I use type 4 polars a lot, will have another go.