NCrit setting in XFoil for DLG sailplanes

[kcaldwel]

A question about turbulators came up in the DLG group. XFLR5 is showing 25% drag reductions for Dr. Drela's AG45 airfoil with an upper surface turbulator at 70%, but only with the NCrit set to 11 or above.

NCrit = 11 is supposed to represent the turbulence level in a clean wind tunnel, and 13 is supposed to be the turbulence level around the wing of a sailplane.

Real world testing does not seem to show any change with a turbulator on the AG series airfoils. Dr. Drela has indicated in the past, that he doesn't think a turbulator would benefit the AG 45 family of airfoils.

When I now re-run some of my airfoil design work at various NCrit values, I am seeing fairly big changes in the relative performance of airfoils. This has me concerned that the results I have will not be representative in the real world. The results I get at NCrit =9 are quite different than I get with it set to 13. Profili Pro 2 states that 13 should represent sailplane flight, and 9 would be a standard wind tunnel level of turbulence.

My question is, what would be a representative value of NCrit for a DLG? Most discus launch flying is done pretty low, 200' and below, so maybe there is a higher turbulence level in this boundary layer than you would normally expect for a sailplane?

The turbulator testing so far would indicate that either the real world NCrit is closer to 9 than 13, or that Xfoil is giving funny results for the turbulators on the AG45.

Thanks,

Kevin

[Neil Stainton]

Sorry, I can't answer your question, but when flight testing most models (including my AVA) I see a measurable improvement in min sink on windless days after sunset, when the air becomes truly still. This suggests to me that atmospheric turbulence is reducing performance, so I cant see that adding a turbulator will improve things.

BTW does turbulence increase with lower NCrit numbers?

Neil.

[kcaldwel]

Neil,

Lower NCrit numbers simulate more air mass turbulence in XFoil. Wind tunnels are generally more turbulent than a glider flying in free air, so you would use lower NCrit numbers to simulate wind tunnel results.

I think you are seeing other effects than a lowering of atmospheric turbulence that improves your min sink. At least from XFoil, a higher level of simulated turbulence gives better low Re performance with model airfoils, and with DLG airfoils in particular. You should get worse min sink rate performance if the air turbulence was less after sunset, if airfoil performance was being affected by turbulence.

I've flown hang gliders a lot, and I have experienced many occasions where the air gets very buoyant around and after sunset. This can be caused by various things, but is generally referred to as "evening glass offs", over here at least. I have had occasions where I could glide over flat ground for 30 km. with my sink rate reduced by 50 fpm. I have read of British HG pilots experiencing it as well.

If you really want to do sink rate testing, I think you have to do it around dawn, to avoid generally lifting air masses affecting the results. Even then, Dick Johnson of full size sailplane testing fame has measured 5 fpm general air mass sink in high pressure areas, and lift in low pressure areas.

Kevin

[macboffin]

I can also confirm reduction in sink rates at lower air temperatures ( hence denser air), at dawn and dusk flights with hang gliders, denser air will reduce turbulence effects, the air having thus greater viscosity as well as greater unit mass and inertia. Whilst the effect is noticeable with several square meters of sail which you are closely attached to, the effect on a small model with very limited area by comparison is debateable ; as is the chance of creating accurate airfoils of such small chords and thicknesses, plus or minus say 5%.

[kcaldwel]

I have flown HG and models at sub-zero C temps and I have not seen any remarkable performance difference. I seriously doubt that evening glass offs have anything to do with the temperature or air density changes. The evening glass offs I have experienced were generally quite warm in any case, and seem to correspond with the end of organized thermal activity.

Selig's group's measurements of in Soartech 8 show that good model construction techniques can reproduce airfoils within +/- 0.004", which would be about 0.07% of a DLG chord length, or maybe 0.7% of the max thickness. That is quite a bit better than 5%, and those were not on molded wings. The molded DLG wings could be much closer than that to the desired airfoil.

But none of this gets me any closer to what NCrit value may be representative of DLG flight which mainly takes place within a few hundred feet of the ground, mostly in low wind but thermic conditions.

Kevin

[annihilator]

You would be doing well to see Ncrit values above 8 while flying outdoor. And adding a turbulator at 70% chord shouldnt do anything because the flow has long since been tripped at typical DLG reynolds numbers.

The Ncrit value of 13 for sailplanes is for the full scale ones in high alt situations, not small DLG's in near earth air.

Drela's airfoils are very good at keeping the flow attached in the computer, but in real life the flow will have already naturally transitioned by the 70% chord mark.

And if you are concerned with endurance, minimum power does vary with density. But i doubt a serious difference in flight time could be found between a hot and cold day. But flying at 10K feet vs sea level would show a difference.

[kcaldwel]

Thanks annihilator! That seems to jibe with the real world results.

I wish the UIUC tests of Dr. Drela's airfoils could be published. It would be great information to see how much laminar flow they are achieving at their design Re.

As an aside, evening glass offs aren't a mystery. They were covered quite well in Dennis Pagen's "Micrometeorology" about 20 years ago, for one source. The reduced sink rate is caused by lifting air.

Kevin

[sll914]

My guess as to the evening glass off phenomenom is that around sunset, the air has generally cooled a bit but the ground is still warm from the sun. This makes a layer of warmer rising air close to the ground. The deltaT is only enough for weak lift and is probably fairly uniform over the flying field so that a compact 'thermal' column doesn't ususally form.

On the Ncrit note, I have long wondered what value I should be using for modelling Dynamic Soaring in 50mph winds with violent shear layers and backside turbulence. I'm guessing something on the order of 6 but have never been able to get a positive confirmation... Any ideas?

Spencer

[Brandano]

While colder air is usually denser, it's not an absolute. a sub-zero temperature is fairly common at higher altitudes, or could be caused by a low pressure front. the sudden drop in pressure causes the air to cool, but will also make it less dense. Just a hunch, I might be totally wrong here.

[Flight Engineer]

don't dismiss 1/2 rho.

In big stuff we can often be take off limited by climb performance due to density.

consider that pressure change of one millibar is equal to about 28' density alt but 1 degree temperature change is equal to about 500'.

If you are doing any comparisons you have to take density altitude into account. And thus temperature and pressure.