Revisting thicker skins for laminar flow

HomeBuiltAirplanes.com

Help Support HomeBuiltAirplanes.com:

stanislavz

Well-Known Member
HBA Supporter
Joined
Aug 21, 2016
Messages
397
Location
Lt
It is important to note that the use of a “laminar flow airfoil” does not guarantee that the section’s laminar flow potential is realized. Notably, the laminar flow airfoil on the P-51 did not achieve the laminar flow potential of the section.


BJC
I know this too. Same on speed of productions models of spitfire vs experimental ones...
 

Norman

Well-Known Member
Joined
Nov 28, 2003
Messages
2,933
Location
Grand Junction, Colorado
So, this addresses localized discontinuities (bumps, dips, etc). Is there similar guidance/rules of thumb regarding how accurately the entire profile needs to match the "as drawn" airfoil if the wing is to have the same laminar flow properties as the "as drawn" airfoil? To give an extreme example: the skin of a tightly-stretched Monokote wing might be very smooth in the chordwise direction (along the air's path), but the airfoil shape between the ribs will not be the same as the rib airfoil.
OK, part of this will be from books I just now reread and part from memory because I can’t find Strojnik’s book “Laminar Aircraft Structures” which is really unfortunate because he’s the only author I’ve seen who actually called out a tolerance zone with a wave height. IIRC he said the tolerance height for his wing was 0.002” per 3” gauge length starting at the leading edge (I could be wrong about the number of zeroes after that decimal). However that does not correlate to the 1/50 (0.02) ratio that Bruce Carmichael in “Personal Aircraft Drag Reduction” says is allowable at Re=2.75 million(that’s the lowest Re he goes to). At Re=4.4 million that ratio drops to1/100. Still not nearly as tight as the 0.004 or even 0.006 height tolerance that you often see quoted. This is well within the skill set of you’re average auto body repair shop. Like I said “if you can feel it it’s to big”. Just run your hand over the wing while you’re sanding it. The dust makes a good dry lube so you can feel the contours without much friction. Dips between waves feel like flat spots. When you find a flat circle it with a pencil so you can find it after the dust has been removed. If you have a lot of circles on the wing after the first check it’s probably best to just skim coat the whole wing and sand it down with a long block. Repeat until you can’t feel flat spots. We spent months contouring the U-2 because we mixed our micro in Dixie cups… should have used a salad bowl. The idea is to minimize the velocity surges that waves cause without removing any D-tube structure. The aft facing slope of a wave is far more disruptive to the boundary layer than the front facing slope so just fill in the troughs and don’t worry about about building up above the nominal profile. It’s too late for that by this stage unless you don’t mind rebuilding the wing.





As for Monokote: Apparently it’s fine for models. There’s a paragraph about it in “Airfoils At Low Speeds”. The tolerances must get huge below Re=500,000 and below 60,000 it’s hard to even get the BL to stay turbulent after a trip strip. However on open bay wings the edge of the D-tube will make a corner that will probably trip the boundary layer.
 
Last edited:

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
6,295
Location
Saline Michigan
Do you have a number for the pressure difference
between the inside and outside of a wing?
Just curious ,it must be a fraction of one psi, and vary with construction and airifoil.
.5 psi=72 lbs ft which might be a design consideration in its own right.
Push me for data? I just ran some numbers so we have data.
  • TOWS page 371, for NACA 65sub2 A015 foil. This page plots v/V for all foils with this thickness distribution. Top surface with the wing making Cl = 0.22, shows (v/V)^2 = 1.53 from about 5% to 45% of chord;
  • Making up a suitable example, let's put this wing at 150 knots, which is 254 ft/s. Multiply that by the square root of 1.53, and local velocity is 314 ft/s over what should be the laminar portion of the wing;
  • Let's also say the wing panel between rivet lines is 12" on a side and 0.040" thick;
  • Applying that to Bernoulli's Principal, deltaP = 1/2*rho*v^2. Using rho for standard temperature and pressure at seal level of 0.002378 slugs/ft^2, I get 117 lb/ft^2 or about 0.82 psi. This pressure is pulling the panel away from the wing everywhere from 5%c to 45%c;
  • Then I went to Roark's formulas for stress and strain to get max deflection at the center of a square plate supported at the edges. Formulae are the same for simply supported and fixed edges, but coefficients are smaller for deflection with fixed edges, which will be closer to reality if the skin continues beyond the supported edge and into another panel;
  • Deflection in the center of the 12" square panel 0.040" thick with fixed edges is 0.35";
  • Deflection of the same panel with 0.080" thickness with fixed edges is 0.042";
  • Deflection of the same panel with 0.125" thickness with fixed edges is 0.011" - finally getting there;
  • You can achieve similar reductions in deflections by making the panels smaller - more ribs and false spars and a lot more build complexity and rivet lines;
  • If you have about 100 square feet of wing skin that you will thicken your wing from 0.040 to 0.125" to get laminar flow, you will add 134 pounds to the airplane - is that worth it for the drag reduction you got?
Now if somehow you manage to get 0.82 psi of suction on the inside of the wing at that speed and Cl, you will have an undistorted skin from 5% to 45%. Unfortunately, the wing will have lower Cl as you go out the span, requiring less pressure further out to balance the drop in pressure of the wing - that could be interesting (in a bad way) to manage. And I only selected the Cl where the (v/V)^2 is flat over the laminar region. At lower Cl, the (v/V)^2 starts lower and increases as you go aft. At higher Cl the peak (v/V)^2 is close to the leading edge and drops off the whole way aft, so matching the pressure differences as you go chordwise could get interesting too.
So you would have to adjust the pressure curves applied to the inside of the wing with airspeed, Cl, chordwise position, and spanwise position. I am not sure how you can actually get a vacuum source with that much volume capability from around the airplane (what with the way metal wings leak air), much less adjust it panel by panel.

This is why I said I do not believe you can get there at reasonable weight with thick skins, and why I am skeptical of folks telling us you can balance that pressure with strategically picked vents for the wing...

If you want extensive laminar flow at reasonable weights, you really need skins with much higher bending stiffness - composites with cored skins do the job nicely and at low weights. Composites over massive cores will do it too.

And yes, this becomes a design issue all by itself. As the airspeed goes up, the stresses in the skins and ribs go up too, with the square of airspeed. At my Vdive of 268 knots, I have had to do some design work and then build a stout bird. The beefup for faster airplanes is usually OK, as wing torsional and bending stiffness also have to go up to keep ahead of flutter.

Billski
 
Last edited:

BBerson

Well-Known Member
HBA Supporter
Joined
Dec 16, 2007
Messages
12,807
Location
Port Townsend WA
Everything needed is in chapter 5 and chapter 7 of Theory of Wing Sections, I think.
"Rocking a straightedge chord wise without clicking or jarring is a fairly satisfactory criterion."
"Dust particles are more effective than scratches in producing transition".
"The procedure at NACA is feel entire surface by hand, and wipe thoroughly with a dry cloth."

The problem with wiping is that sometimes the wiping static makes dust stick worse. Ask any painter about dust.
The dust doesn't blow off in flight as far as I know. That's why glider pilots wipe the wings thoroughly.
Airplanes have small wing area compared with gliders, so why bother?
 

Norman

Well-Known Member
Joined
Nov 28, 2003
Messages
2,933
Location
Grand Junction, Colorado
The problem with wiping is that sometimes the wiping static makes dust stick worse. Ask any painter about dust. The dust doesn't blow off in flight as far as I know. That's why glider pilots wipe the wings thoroughly.
That's a good reason to keep a big jug of windshield washer fluid in the hangar.

Airplanes have small wing area compared with gliders, so why bother?
Because the wing accounts for 40 (highly laminar) to 60% (NACA 4 & 5 digit section) of the zero lift drag. A couple dozen squashed bugs can double the zero lift drag of an otherwise perfect 50% laminar wing.
 

lr27

Well-Known Member
Joined
Nov 3, 2007
Messages
3,709
Bill:
Wouldn't the curvature of the airfoil make the plate a whole lot stiffer?

-------
On a metal airplane, the static charge will spread all over the aircraft, or at least the part you're working on and can be easily dissipated by grounding it for a moment.

There are other ways to dissipate static if the surface isn't conductive. For instance, you could use one of those grounding straps they make for electronics assembly and wipe the wings with dryer sheets. Bare feet or leather soled shoes work, too. I'm assuming you're working on concrete, pavement, or grass. If not, you'll need to ground yourself some other way.

There are other electronics industry static dissipation methods, like slightly conductive,was.

I suspect that if there's much dust in the air, the charge will build up again in flight.
 

dog

Well-Known Member
Joined
Dec 29, 2019
Messages
242
Push me for data? I just ran some numbers so we have data.


  • Billski


  • OK WOW and Thanks
    For the first time I have a visualisation of the presure differences on a wing as a continiously variable phenominon vs cord and span and speed.
    I was fumbling an idea around hopeing that there might be a way to exploit the pressure difference and or neutralize it,but had total missed that it was
    continiously variable.
    .82 psi is much more than I imagined, and therefor highlights lift distribution on a wing,and then that the structure is significantly overbuilt on some of the wing.
    For me it realy helps to have a visualisation for the concepts, and theorys and then the math,the last bieng the most difficult without a picture in my head first.Its a totaly wierd sensation having an idea that has been presented to me countless different ways,suddenly take hold.
    I am obliged to you for that.
    David
 

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
6,295
Location
Saline Michigan
Bill:
Wouldn't the curvature of the airfoil make the plate a whole lot stiffer?
The chordwise curve will tend to stiffen the skin in that direction, but with the radius of curvature so large compared to the thickness, the curvature effect is pretty small. And in the spanwise direction, there is no curvature. Yes, the deflections will be somewhat smaller than calculated, but not hugely so.

The effect I describe can be observed in sheet metal airplanes by simply looking out the window - with low angle light - just after sunrise or before sunset, you can easily see the skins bulging with restraint along the lines of rivets. You can also observe that the deflection shape makes little difference in spanwise and in chordwise directions.

The magnitude of deflections and skin stresses is largely determined by the smaller dimension of the panel. This feature allows the designer to use spacing to keep stresses and deflections reasonable. It is also why designers will have one rib spacing at the root, and open up the rib spacing as you go towards the tip.

When I think about panel size, a clever designer who can control vacuum inside the wing could pick a target Cl, then tailor deflections with panel size. This would still be an unlikely path to achieve laminar flow at low weights.

Billski
 
Last edited:

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
6,295
Location
Saline Michigan
Another point in this is that the lower skin also has to be managed, as its skin is being pulled away too. The lower skin has lower velocities and thus requires less vacuum than the top skin to balance it for laminar flow through the front half of the wing. This has to be, right? The pressure reductions must be bigger on the top skin than on the bottom in order for there be net lift between them...So to balance the pressures, you would also need to have separate chambers for top and bottom of the wing. Ugh.

One thing no one has discussed is low speed. If suction on the skins goes up with speed, then a low speed airplane has smaller pressure reductions and better chances of maintaining shape as well as lower Re. Trouble with them is that lower wing loading mean bigger wing areas, which push us to lighter skins and bigger rib spacing.

Billski
 

pictsidhe

Well-Known Member
Joined
Jul 15, 2014
Messages
7,372
Location
North Carolina
Internal chambers can be thin skinned. It's still going be a PITA to put them in. With typical laminar sections having a fairly constant pressure across most of the laminar section, a chamber could be built that is vented at the rear of the laminar section. But, by the time you've built a double skin to do that, you may end up heavier than a stiffer single skin. It would also not be a perfect pressure match, so some skin stiffness is still needed.
 

Norman

Well-Known Member
Joined
Nov 28, 2003
Messages
2,933
Location
Grand Junction, Colorado
It just occurred to me that a bulge would be longer than the gauge length and therefor is a deviation from the nominal shape, not a wave. Yeah yeah waves are localized deviations from the nominal shape but an unintended change to the thickness distribution without introducing small scale waves may be OK. It would basically be a slightly thicker version of the same airfoil until you get close to the spar. The constrained edges of the panel are where the greatest curvature will be so the area midway between the ribs may stay laminar up to the spar in spite of bulging. It may even be able cross the spar if the skin is thick enough to resist that 1/50 wave height/length ratio mentioned in post #43.
 

lr27

Well-Known Member
Joined
Nov 3, 2007
Messages
3,709
Any idea if changing sections depending on how far they are from the nearest rib will trip the flow? The flow won't be completely 2d any more.
 

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
6,295
Location
Saline Michigan
It just occurred to me that a bulge would be longer than the gauge length and therefor is a deviation from the nominal shape, not a wave. Yeah yeah waves are localized deviations from the nominal shape but an unintended change to the thickness distribution without introducing small scale waves may be OK. It would basically be a slightly thicker version of the same airfoil until you get close to the spar. The constrained edges of the panel are where the greatest curvature will be so the area midway between the ribs may stay laminar up to the spar in spite of bulging. It may even be able cross the spar if the skin is thick enough to resist that 1/50 wave height/length ratio mentioned in post #43.
Norman,

OK, I left out two things: The shape of the skin as it goes from one spanwise supporting element to the next, and the fact that the skin also puckers slightly around each rivet while being ballooned out.

Shape of the skin deflection - Assuming the panels to be continuous around the leading edge and across spanwise elements, the skin is then viewed as fixed at the spanwise elements. The skin is first curved concave near the spanwise element it is anchored on, and then convex through the middle of the panel, and then concave again near the next spanwise element it is anchored to. From observation, the radius of curvature appears significantly smaller in the concave portion than it is in the convex region. Consulting Shigley for double canteilever beam under uniform loading, the bending moment is maximum and has maximum slope at the supports, so the curvature must also be maximum here too, so my observation is backed up by beam theory. I would expect this to be disruptive of laminar flow...

The skin always seems to have some pucker or other local deflections around each rivet on the supporting elements. I would expect this to be disruptive of laminar flow too, even if the rivets themselves have been faired in.

Between the concave curvature near each spanwise element and the skin pucker around each rivet, I would expect that the combination would easily trip the flow to turbulent. Not so?

Billski
 

Norman

Well-Known Member
Joined
Nov 28, 2003
Messages
2,933
Location
Grand Junction, Colorado
I would expect this to be disruptive of laminar flow...
Depends on the Reynolds number and the wave proportions I described in post #43. Bruce was a well regarded expert in this field so I trust that he got it right.

The skin always seems to have some pucker or other local deflections around each rivet on the supporting elements. I would expect this to be disruptive of laminar flow too, even if the rivets themselves have been faired in.
Yes, rivets are the root of all evil. Aside from the wrinkling they cause they also leak. If the air leaking out of the rivet hole has any velocity it will screw up the boundary layer.

Between the concave curvature near each spanwise element and the skin pucker around each rivet, I would expect that the combination would easily trip the flow to turbulent. Not so?
The puckering is bad but that trough you're describing probably isn't. A trough parallel to the airflow isn't nearly as disruptive as one at a right angle to it.
 

Norman

Well-Known Member
Joined
Nov 28, 2003
Messages
2,933
Location
Grand Junction, Colorado
Any idea if changing sections depending on how far they are from the nearest rib will trip the flow? The flow won't be completely 2d any more.
If the pressure gradient is different between to adjacent strips of the wing it will cause cross-flow but this should be negligible because the bulge isn't changing the camber much.
 

Hot Wings

Well-Known Member
HBA Supporter
Log Member
Joined
Nov 14, 2009
Messages
6,747
Location
Rocky Mountains
part from memory because I can’t find Strojnik’s book “Laminar Aircraft Structures” which is really unfortunate

I didn't see anything is 'Structures' after a quick scan. I did find the first in 'Design' and the last 2 in 'Technologies'.
 

Attachments

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
6,295
Location
Saline Michigan
Ah, so there you have it. With respect to wing inflation, ribs are not so bad, but spanwise reinforcements are worse. It does sound like there might be a sweet spot...

To get extensive laminar flow over the front 40% of a sheet metal wing, the foil will need to be built using:
  • Continuous skin;
  • All underlying skin support is chordwise;
  • All rivets countersunk and filled/faired;
  • Accurate foil shape reproduction;
  • No chordwise waviness (2" basis) above an arguable number based on your Re;
    • Bruce Carmichael's excellent Personal Aircraft Drag Reduction (I have a personalized and signed copy) shows wave ratio limits of 1/50 at Re 2.75M, 1/100 at 4.4M, and 1/200 at 7M;
  • Chordwise waviness limits are combined build waviness and waviness due to inflation effects.
Since you will not be supporting the skins with spanwise elements until you get as far aft as the main spar, ribs will be doing all of the support forward. This will drive rib spacing to keep skin stresses in bounds. Rib spacing and skin thickness will still be a trade. I suspect that you may need to do some wind tunnel work to establish minimum skin thickness for any given rib spacing while maintaining laminar flow.

Hmm, I seem to recall Prefessor Lesher's airplanes having perfect wings. I saw the wing of the follow on to the magnificent Teal in construction (Michigan Aero Engineering lab had a table along a wall for Prof Lesher's work) with continuous skins and very accurate shaping. Last I heard, the family still has that wing. Did not know the significance of what I was seeing then, or I would have noticed the internal structure. My friend who studied under Lesher mentioned it like I should KNOW how significant the Professor and his airplanes were. Perhaps a close look at Teal when next at OSH is in order.

Billski
 
Last edited:

stanislavz

Well-Known Member
HBA Supporter
Joined
Aug 21, 2016
Messages
397
Location
Lt
No chordwise waviness (2" basis) above an arguable number based on your Re;
  • Bruce Carmichael's excellent Personal Aircraft Drag Reduction (I have a personalized and signed copy) shows wave ratio limits of 1/50 at Re 2.75M, 1/100 at 4.4M, and 1/200 at 7M
How could that number get into fabric covered/unfinished composite male mold flying surfaces ?
 

dog

Well-Known Member
Joined
Dec 29, 2019
Messages
242
[QUOTE="wsimpso1, The other issue in our little airplanes is that with moving air outside the foil and stationary air inside, these airfoils become somewhat inflated in flight.

Billski[/QUOTE]

Still working to understand what goes on in the flow around an airfoil.And relized that I missinterpereted the above quote from Mr Billski
Still haveing trouble thinking it through.
Here goes.
As described air flows around the wing,pressurising
slightly on the "bottom" and creating a bit of vacume on top,meanwhile on the inside of the wing,these pressure differences acting on the wing skins,push the "bottom" skin in and pull the "top" skin out,creating an apearance of inflation,but I think that there isnt a relationship between the airodynamic pressure of the flow around the wing and the actual air pressure in the wing.
Still messes up laminar flow.
 
2
Group Builder
Top