NASA Prandtl-D (Preliminary Research Aerodynamic Design to Lower Drag)

Grey Havoc

The path not taken.
Senior Member
Joined
Oct 9, 2009
Messages
8,388
Reaction score
32
From November of last year, a couple of interesting articles related to the Prandtl-D [Preliminary Research Aerodynamic Design to Lower Drag] program (a few images of the Prandtl-D prototype from the articles are attached):

http://www.nasa.gov/feature/new-wing-shape-tested-in-wind-tunnel

http://www.nasa.gov/centers/armstrong/news/FactSheets/FS-106-AFRC.html (first article in factsheet)

Note that the Prandtl-M proposal for Mars is a direct offshoot of this program.
 

Attachments

Phos

CLEARANCE: Restricted
Joined
Nov 19, 2018
Messages
17
Reaction score
0
AFAIK, this is the most recent presentation Al Bowers has done that's ended up online: https://youtu.be/bCwtcDNB15E

It's interesting that Al Bowers mentions/admits that this lift distribution really only matches up with water birds at one point, as according to an A380 documentary narrated by Richard Hammond (so grain of salt might be in order) that over land birds are effectively span limited by the need to remain within rising thermals and his admission that a span limited plane is still better off with an elliptical span loading.

My understanding of the benefit of this is the span needs to be increased, which would typically be cheating when evaluation a spanload, but the loading near the tips is reduced so the wing actually gets lighter, and drag is reduced by the increased span. Another way to think of it is just adding a non lift generating wingtip device but pointing it straight out.

Ostensibly it also makes the wing itself dynamically stable in yaw? This reduces drag and weight because now you don't strictly need a vertical tail or constant decelerons. His earlier video posited that the B2 flies with its decelerons cracked slightly for yaw stability, is that the reason or is it to conceal the B-2's minimum RCS? The proverse yaw seems a bit foggier, I don't get if he's implying/if it increases drag inboard and reduces it outboard, as he just describes it as generating thrust. I imagine it is, but only if you limit your frame to the portion of the wing featuring aileron and whatever's outboard of that. Or is it generating local drag that reduces total drag, but when the aileron moves down the local drag drops but total drag increases?
 

Viper2000

I really should change my personal text
Joined
Jun 8, 2018
Messages
11
Reaction score
0
Phos,

Bowers answers many of your questions in this paper from 2016. However, contrary to his paper, it is my understanding that Horten had a pretty good understanding of proverse yaw.

This presentation and the thesis upon which it was based may also be of interest. Shifting load away from the tips takes weight out of the wing, & this will reduce fuel burn (or increase payload fraction if you prefer) even if span is constrained, though the lift distributions in these links may not produce proverse yaw.

There is a sort of hierarchy of loads & planforms.

Elliptical loading with an elliptical planform is the simplest arrangement intellectually. Induced drag is always the minimum for a planar lifting system (but note Kroo's work on non-planar lifting systems) of the selected span. The problem with this is that the shape is complicated, so it's expensive to make (or was, in the pre-CNC era), & because the whole wing is at the same lift coefficient, it should theoretically all stall at the same time (in reality, due to Reynolds number effects, it will probably tip-stall).

Many manufacturers select non-elliptical planforms to cut manufacturing cost, & used aerodynamic or geometric twist to get closer to elliptical loading at the cruise design point if they were worried about such things. This requires a bit more thought, because the cruise design point has to be positively selected if it is to coincide with minimum drag.

If the aeroplane is being designed to fly at some sort of economic optimum speed, it may be attractive to deliberately tune the wing for minimum induced drag at the "wrong" (on the low side) lift coefficient to produce a higher value of Carson's speed. Naturally, there's more to life than the cruise design point, & this may affect things somewhat.

The Prandtl / Horten approach is pretty profligate in its use of span, but has the advantage that it may permit the vertical tail to be eliminated. The limiting factor is likely to be OEI asymmetry in multi-engine vehicles, though there are obvious ways to solve this.

It's especially interesting for gliders (at very high L/D circling in thermals is less important), but the current span-limited classes mitigate against it to some degree.

Flying wings do well when assessed in terms of wetted aspect ratio (Liebeck is pretty clear that this is one of the main objectives of his BWB work), & Prandtl-type lift distributions make them look even better, but the poor span efficiency means that it's not really fair to plot them on the same curve as conventional aeroplanes. CoG range is ultimately likely to limit their commercial potential; an asymmetrically loaded Horten or Prandtl wing will naturally yaw & this may cause some very interesting stability & control problems.

Very large spans may also cause ground handling / tip-over problems, especially in cross-winds.

It's a very interesting trade space. I'm pretty sure that tube & wing designs are closer to the manufacturing optimum than any sort of global optimum, but unfortunately the Industry is very conservative.

His earlier video posited that the B2 flies with its decelerons cracked slightly for yaw stability, is that the reason or is it to conceal the B-2's minimum RCS?
AFAIK this is due to boundary layer thickness producing a dead-band, which would increase lag in the control system. I suspect that in "stealth mode" the aircraft will use differential engine thrust for yaw control, but this will also be laggy & may hurt engine life. The lag may also produce some Dutch roll which isn't dangerous but may be nausea inducing (see also the history of the yaw damper on the 707); given the long mission durations, most of which don't require full stealth, it's pretty logical to have more comfortable control laws so that the crew can get better quality sleep in their lawn chairs.
 

Phos

CLEARANCE: Restricted
Joined
Nov 19, 2018
Messages
17
Reaction score
0
It's interesting to know that Horten did understand proverse yaw, I suspect the language barrier is to blame. Thanks for all the info.

Really interesting that the design work about the BWB was for a giant double decker A380 competitor, seems they had yet to learn the lesson that led to the 787.

The issue BWB airliners face is that a passenger compartment (or generic cargo volume) is very low density compared to most things you could build an aircraft to carry, so it's natural to line people up to minimize the parasitic drag they create.
 

Viper2000

I really should change my personal text
Joined
Jun 8, 2018
Messages
11
Reaction score
0
Really interesting that the design work about the BWB was for a giant double decker A380 competitor, seems they had yet to learn the lesson that led to the 787.
Liebeck came out of Douglas (he did some very interesting work on high lift aerofoil sections for AMO Smith in the '70s). I don't know how well Boeing has done at integrating its acquisitions, but I wouldn't immediately assumed that there was or is a single view on the subject.

I suspect that the A380 was selected as the target because this would have been politically preferable to drawing a comparison to a Boeing product. MD-12 experience would have also helped.

The issue BWB airliners face is that a passenger compartment (or generic cargo volume) is very low density compared to most things you could build an aircraft to carry, so it's natural to line people up to minimize the parasitic drag they create.
This isn't obvious.

If you design a minimum T&W aeroplane to carry a lump of lead with a CoG-range constraint & at least two engines (so that OEI asymmetry is a factor), then you will end up with some sort of tail volume coefficient requirement.

There's then a trade study to decide how much moment arm is required, & how much structural depth is required in the fuselage.

Adding a bit of extra volume to accommodate lower density payloads is fairly cheap because the weight reduction from added structural depth will at least partially offset the drag & extra skin weight.

Aerodynamically, there is an optimum fineness ratio for drag. If the shape is reasonably good then pressure drag is small so it's basically all skin friction & perhaps a tiny bit of wave drag.

Practically, people end up designing straight section tubes with nose & tail end caps because this allows the aircraft to be stretched or shrunk to fit different markets.

The problem with this approach is that for minimum design cost the aerodynamic surfaces are frozen, so the vertical tail probably ends up sized by the shortest tube & the horizontal tail by whichever tube has the most critical CoG envelope.

The BWB's problem is that it can't easily be stretched or shrunk because it's tightly integrated, so either it relies upon advanced design / verification techniques to permit the cost of re-sizing to be controlled, or else it has to offer sufficient benefit to offset the penalties associated with over-sizing in most applications.

This potentially poses challenges for the business case as well, because the manufacturer can't use differential pricing to extract maximum value from the market if all the models are the same.

Ultimately the problem therefore comes down to the relationship between design costs vs operating costs. As design tools improve, we may expect that the cost disadvantage of tightly integrated concepts will come down, & their self-evident operating cost advantages will dominate.
 

Phos

CLEARANCE: Restricted
Joined
Nov 19, 2018
Messages
17
Reaction score
0
Oh, I had heard that the frontal area increase was too big a deal for lower skin drag to offset, they must have dealt primarily at more automotive speeds (or been full of it).

This got me looking around for more info on the MD-12, I always got the idea it was a sort of hail mary to try and bump up their valuation just pre buyout, but it seems like they were actually intending to build it.
 

Viper2000

I really should change my personal text
Joined
Jun 8, 2018
Messages
11
Reaction score
0
Cars often have lots of pressure drag, so frontal area is expensive. They also have low fineness ratio, so dropping frontal area whilst maintaining length is attractive. However, if you plot drag per unit cabin volume then the trade is more interesting.

Aeroplanes generally have much higher fineness ratios, & so it's a better first-cut assumption to assume that the zero lift drag is due to skin friction & then add drag due to lift. You can see the general trend for fuselage design on page 63 of Torenbeek's 1982 book. As you can see, the tendency is for aeroplanes to have fuselages of excessive fineness ratio from the perspective of achieving minimum drag per unit frontal area.

The above isn't so say that there is no pressure drag (it's hidden in the assumed value of Cf), but that plots of L/D vs wetted aspect ratio tend to produce nice correlations (e.g. see Raymer) because there is a tendency for everybody to end up with similar levels of disappointment.

In the bad old days, before people really understood how to attack flow separation, designing for minimum frontal area was a very simple solution.

In the same vein, the biplane got a bad name because people didn't understand how to manage interference drag either, & so biplanes were often fairly draggy because the detail design was poor enough to offset their inherent aerodynamic & structural advantages. This was exacerbated by the tendency for many competitions to be held at low level, negating the biplane's superior drag due to lift.

Unfortunately, people are still taught that biplanes are old fashioned & inefficient, rather than the real truth which is that the more complicated a configuration becomes, the more sensitive it becomes to detail design.
 
Top