Sethhunter

I've been promising to start this thread for awhile. The build hasn't started but I'm just about done with the design, so I thought it was time to start the thread. The project is a 1/3 scale Halberstadt D.III. I picked this airplane because it's not often modeled (like my Pfalz DIII and DXV). It has nice proportions for a model and to my eye, looks good. It should not be a difficult build as the wings are untapered and nearly identical. The rib spacing is generous. And the fuselage is box construction (a welcome relief after two Pfalzen!) It looks a bit to me like an Eindecker fuselage, with a small turtledeck that only goes 1/2 way back. Fuselage is nearly constant in width, but tapers in depth to a point at the tail.

This is my first 1/3 scale project. I picked this size because there are increasing numbers of 1/3 scale WWI at my club (which make for impressive group "formation" flights) and this size presents some new (to me) design challenges. Basic stats are: Length about 92", wingspan 114". Weight 40 lbs. Engine FUJI 64. Wing area 4,048 sq in.

Anyone intersted in learning more about the A/C should buy Albatros Publications' Halberstadt Fighters by P. Grosz (pictured below). As usual, the datafile has lots of great info, drawings plates and photos. I haven't decided on color scheme - the datafile describes clear linen, pale blue, and brown/green camouflage. Some unusual details - the trailing edge of the lower wing is straight, but bowed downward, sometimes called the Halberstadt "droop." The A/C was also used as a platform for wireless communication - opening possibilities for some interesting antennae arrangements.

To facilitate field set-up I've designed the wings to come off fully rigged as two "box kite" assemblies. Installing them should be as easy as securing 6 bolts and a pair of drag wires. The elevator is full-flying and will also come off in two panels by removing 2 bolts. Unfortunately, the pieces will still be too big to fit in my Jeep, so I'll need to trailer it - but that's another chapter.

Being new to 1/3 scale, a friend suggested I do a thorough weight & balance analysis before starting. So I just spent a couple days lifting CAD data and building a spreadsheet and calculating the balance point and weight. As a result, I had to do a couple days worth of redesign to lighten the tail because this A/C has a very long tail moment. Had I found this out the hard way, the only option would have been to move the wings, which I hate to do. But all is well: I should be able to plant the CG at 28% mean aerodynamic chord with 2 lbs of ballast (staying below 40 lbs), or 32% at 38lbs/0 ballast. The only change I made to outlines is the elevator size - I increased the span/chord by 15%. Should this prove ample, it's very easy to build a new elevator to scale outlines (advantage of a removable elevator).

That balance spreadsheet was full of useful info. It showed I really should use balsa, not bass for longerons and cross-pieces, and I should use fiber (e.g., kevlar thread) in place of wood diagonals. It also showed the covering is one of the big balance/weight culprits, and carbon, while stronger than aluminum - would not save weight or balance in the critical areas (e.g., tube for the elevator).

As with other projects, should this go well, plans will be available at no charge. If anyone wants a copy of my weight and balance spreadsheet to see how I set it up, just send me a PM.

I have a pile of wood from National Balsa. Hopefully the chips will start to fly in a few days!!

Now for a question for the experts - what fibers have you used to reinforce box fuselages? I has thinking #92 or heavier Kevlar, CA'd at the ends (e.g., in the balsa structure butt joints, and tied/glued where they cross. Any other ideas/suggestions/materials?

John Cole

My Feedback: (4)

Seth, nice!! I like Halbs. Good work with the pre-build calculations.

John

rlukowski

Seth,
I'm on board. Looking forward to following your progress.
Ron

ARUP

I have drawn a set of 1/4 plans for this bird but haven't built it. I think I would build the wing with the trailing edge in a jig. The TE would be a laminate to help ribs hold TE 'droop'. I was thinking of using Spectra line as cross bracing. I did that for some scale CL racers and both RC Bleriots I have built. Indestructible and weighs 'zip'! Good luck! ARUP

Sethhunter

Hi Arup - close to what I was thinking.... I was going to block the tips - front and back, and a few locations along the TE, then glue top/bottom TE pieces to the ribs, forming a lamination. An alternative would be to prelaminate the TE under some cawls. Actually as I think about it, there's another factor to consider - I was going to prebuild the box spar. But pushing the TE down will cause the LE to curve up. Seems like the spars would need some sag too, if the point is to have the LE straight. Either that, or the ribs need to be shaped differently. Think I need to look at the photos more closely[sm=confused.gif]

Anyone want to speculate on the aerodynamics of the droopy TE? This is what I was thinking for incidence:

Top wing: Root: +2 (deg), Mid-panel: +1, tip: 0 (i.e., about .5 inch washout)
Bottom: Root: +2, Mid-panel: +5, Tip: 0

I scaled the +5 degrees (droop) from a photo. I usually set up bipes with equal incidence top & bottom but I know some folks like a little convergence (e.g., lower incidence on top). This design will have convergence with more positive on the bottom. 5 sounds high at first glance but the average for the panel is about half that. So long as the washout is there, I think it should be fine.

I didn't think of using spectra - funny since I've used spectra core lines on my sailboats for years! Spectra (a HMWPE) is known for strength and lubricity, whereas Kevlar (an aromatic polyamide) is known for strength and abrasion resistance). I would have thought Spectra would not glue very well, but be less likely to chaffe against itself than Kevlar. I'll have to fish some of it out of my lines and play around with it. Thanks!

jumpinjan

On a biplane, the top wing has the most incidence, not the bottom. Also, the neutral point is 23% MAC (equivilent monoplane, including area weighting) for biplane wings. I wouldn't do the droop at all, keep it straight, thats my opinion.
I would like to see your spreadsheet, I'm always interested is seeing what other people are doing.
Jan

ARUP

Make the ribs thicker than needed, especially toward the leading edge. After putting leading edge on you can sand the ribs to its level. When satisfied use a shorter 'T' bar sander and pull from front to back toward the trailing edge. You can move over one rib station at a time to arrive at the contour you need without getting some ribs out of whack relative to the others. When satisfied with the top side, flip panel over and block it up so you can do the bottom side. You could check your work by having a 'dummy Master rib' that slips over the spars to compare. Sure would be nice to know how Halberstadt built their wings for when in doubt do it like they did! Maybe another way to do it would be to place LE (staight) and TE (curved on jig) then laminate spars into the ribs. A 3rd way would be to make the ribs a 'sloppy' fit onto the blocked up spars. Glue ribs to blocked up LE and TE. Adjust the spars up/down but straight then glue the ribs to spars using scrap to take up the 'slop'. As far as Spectra line I use it to make kites, too! CA one end to harden it like a needle for 'sewing'. I wrap it around small dowells apply a little CA to lock it and continue to next station. If you look at my little Curtiss in my Gallery that's how I rigged it. Aerodynamically, just 'average out' the droop. Art Chester did the same droop thing on his 'GOON'. I would guess (since I'm no aerodynamicist!) the droop and the upsweep of the ailerons gives the needed washout. It would have been interesting to 'tuft' the wings and examine the airflow, esp. at roots. P.S. There are some really nice drawings in the August '08 WWI Aero journal supplied by George Odenwaller. They were drawn on March-5-67 by the Barcalo Mnfg Co.

Sethhunter

Arup - thanks! These are all great suggestions. There's a photo of the plane that clearly shows the top of the wing curved down as well, suggesting the main spars are bowed as well as the TE. I'm thinking I'll fixture and build the main spar with a curve, and use your idea of building the ribs oversize and sanding I was going to gusset the ribs to the main spar anyway since the ribs will be notched to slip down on the pre-build spar, so a little clearance bewteen the ribs and spar will not be a problem.

Well, I spent the morning researching decalage. The best dissertation I found was NACA TN 269 from 1927. I can forward to anyone interested. Here's my summary of the conclusions:

Affects of decalage on biplane aerodynamics:

Lift of upper wing is generally more than lower wing, even at zero decalage, due to circulation around lower wing increasing airspeed at upper wing, and circulation around upper wing decreasing airspeed of lower wing. Increase of velocity between wings due to venture effect tends to decrease circulation of the upper wing and increase circulation of lower wing, offsetting circulation effect but not as strong as circulation affect. Hence lift ratio (upper to lower) is usually greater than 1.

Tip vortices of each wing reduce the lift of each other. So the presence of each reduces the lift of the other. Total lift for biplane is less than the sum of two equivalent monoplanes. So total lift is decreased and lower wing is decreased more than upper. The lift ratio (upper wing lifting more than lower) increases with angle of attack and can reach 3 or more at high angles of attack (see attached graph). Positive decalage (upper wing having angle of attack greater than lower wing) increases the slope of the lift ratio – that is, shifts the difference in lift even more aggressively to the upper wing. Negative decalage tends to flatten the lift ratio, keeping the lift ratio more nearly constant at different angles of attack.

Positive stagger reduces the affect of lower wing tip vortex on the upper wing, and increases the affect of upper wing vortex on lower wing, increasing the lift ratio.

The notion that positive decalage is necessary for positive stability is not supported by this report. Negative decalage keeps the wings near the neutral lift (ratio) line. See figure 11.


Jan - I recommend some reading on neutral point and margin. Lots of websites... here's one: http://homepages.stmartin.edu/orgs/S...tral_Point.htm
You will see it depends heavily on characteristics of tail. Lots of variables, so it's better to calculate it than assume one value. In my experience, planes are usually designed with NP farther back than .23.

Based on the NACA report, I think a bit of negative decalage is a good idea. It suggests bipes can become "stubbornly stable" during dogfights - my guess because the higher alpha is shifting the load distribution excessively to the upper wing.

Lucky Dog

My Feedback: (1)

Subcribed. Looking forward to your build sethunter.

Cheers.

abufletcher

Oh, boy! I look forward to running out of superlatives...as always!

Here are my two yen worth on "trusting the original." While it's probably fool-hearty to use a completely scale airfoil in a 1/6 scale P-51, I don't think this piece of common wisdom should be carried over to, say, a 1/3 scale Bleriot. On the contrary, in my opinion, the larger the model and the more the model construction techniques parallel the construction of the original (which in the case of most of our WWI models is quite considerable), the more we can trust original as a guide for the construction of the model.

So with a model like the Halberstadt, I would trust the original to a quite considerable degree. For example, I would trust that a scale airfoil with scale incidences will fly it well. In other words, I would personally be loath to redesign the original. To take just one example, if the original aircraft (such as the Sopwith Camel) had its ailerons hinged at the bottom surface of the wing, I wouldn't hinge my model's ailerons in the middle or on top just because common modeling wisdom says that bottom-hinging may induce adverse yaw.

You mention an article on biplane incidence written in 1927 (10 years after the Halberstadt was built). Presumably, this article is talking about the aerodynamics of full-scale biplane design. So following the advice of the article really has nothing to do with "translating" the original into model form. Rather you would be choosing to redesign the original in light of subsequent discoveries in the field of aeronautical engineering. That seems like a slippery slope to me.

I've read and participated in a number of "Halberstadt droop" discussions and I'm satisfied that it was purposeful (vs. those who argue that it was essentially a construction fault). As to what the purpose of the droop was, that's still a head-scratcher. My own wild guess is that the droop results from the designers desire to have a more undercambered wing combined with an understanding of the value of washout. Aerodynamically, it seems like the droop would act like a small flap.

Sethhunter

Hi Don - I agree absolutely. What I like about the 1927 NACA report, however, is it presents a theoretical model (based on circulation theory), and wind tunnel studies on a small model (I don't remember the size off hand) in a 4' wind tunnel, on WWI airfoils. About as good as it gets for applying to our models. Not so surprisingly, the theory and the test data were not in agreement. They ended up critiquing the assumptions in the theory and betting on the data. The chart I attached was experimental. Gotta love it!

I'm not sure I agree with the camber perspective - the photo observation that the top surface of the wing was bowed down too suggests the section was the same throughout, but rotated down (probably pivoted around the LE) around the middle of the panel. I'd guess they were onto the issue that the top wing was doing the heavy lifting (with the structural problems that go along with it) and trying to find a way to balance them more - with negative decalage - while keeping some washout in as well. Just a guess.

abufletcher

I just don't see the rationale for second guessing the Halberstadt engineers. If the wing incidence data (including droop and washout) are available, why not just use that? At 1/3 scale your model should fly pretty much the same as the original. If the original had >1 lift ratio, so would your model. Of course, if your goal is to produce a model that flies BETTER than the original that's a whole different thing. My own ideal would be to end up with a model that has all the flight characteristics and quirks as the original. No worse, no better.

Making design changes also seems questionable considering that we don't really know what the droop was supposed to do.

abufletcher

I seem to remember a photo posted in one of the droop discussions, which showed that the carefully tapered spar was involved in the droop. That was probably over on the Aerodrome.

TFF

http://www.mikerlewis.com/
http://www.aeroconservancy.com/halbfin.htm
http://www.thomasgenth.de/html/relikte_wk1.html
http://picasaweb.google.co.uk/bamf.b...41?feat=email#
http://www.collectors-edition.de/Sal...-3_english.htm

None of the sites are DIIIs but are Halberstadts. Your model is going to be a monster! I think you are going to have to rely on more scale construction with a plane that big; because the plane is getting bigger, the material strength wont be able to keep it together. The engineering is going to start playing a part.

jumpinjan

Seth, I meant to say the Aerodynamic Center (not NP of the aircraft) of the MAC (monoplane MAC equivelent to a biplane ) is 23% for a biplane. For monoplanes it is 25%. Any further back, moments can be sucessfully balanced with trim, to a point.
Jan

cocobear

My Feedback: (2)

Seth, a very good subject. I have been drawn to this plane but frightened away by the wings. As to all of this discussion,it is YOUR plane and you certainly are a master builder, go with your gut. Opinions are opinions but it is your model. I have mine but will retain them. I look forward to this build!!!!

Sethhunter

Thanks for the links, TFF - great stuff. It will take awhile to go through! You are absolutely right about strength - I always start there. Generally I prefer internal strength over "live" rigging but in this case, it was just impractical (I was looking at a fair amount of aluminum and carbon laminate). Still, to keep the field assembly simple, I'm attaching the lift wires to the wing roots rather than the fuselage (so the wings come off as a unit). It's more scale, but it puts signiificant load (hundreds of lbs) through the wing joint to the fuselage - creates high compression (upper) and tension (lower) loads in the wing panels, compression in the interplane struts, and tension in the cables (turnbuckles and brackets). I've used 10G's as a design rule - 6 G's (mild aerobatics) plus safety factor of 1.5+ (margin for material and construction variability). So yes it pushes the design to stronger materials, and significant reinforcement of the strut and cable bracket mounts. The flight envelop will be restricted compared to my 1/4 scale projects.

Thanks Cocobear - I always welcome comments - good, bad or otherwise - they all help!!

jtisch

There is a thread here on RCU about a 1/4 scale DV. I think I posted some photos there.

J

ARUP

I always make the lower wings 'stall' first (unless they have the ailerons). Prolly makes no difference since we aren't inside the airplane to feel the buffet before the wing lets go! RC fliers are more retroactive than proactive (generally).

Nieuport nut

My Feedback: (10)

Seth, have you had a look at Dave Hurrell's DII plan from Traplet? He did a 1/6 scale version with the drooped TE and full scale tail and reported it to be quite an easy flier. It's still available from their online shop.

Martin

Nieuport nut

My Feedback: (10)

Oops, I forgot about Gary Suntherland's 1/4 scale version in Flying Scale Models, Dec 2002. It works pretty well with 4 degrees elevator incidence, and 10 degrees movement either side. Gary specified these two measurements as "critical".

Martin

Sethhunter

Started at last! Starting with the bare board, some wood and a computer! Photos show some of the fuse parts coming off the scroll saw and trial fitting the engine box/firewall together, and the clamp maze while the epoxy sets. Then built a caul to mold curvature into the front fuselage sides (balsa interior, ply outer skin) and put a bunch of weights on the lamination while gluing. This is followed by dry fitting the front fuse together, then building the aft open truss. The red bits are fiberglass caps on the blocks which will anchor the cabane and landing gear wires, all of which get bolted to the bulkheads.

Sethhunter

Next the front fuselage gets glued together. As I pulled the tail together, used an incidence meter on a plank through the main spar opening in the fuse as a reference datum, to make sure the tube in the tail which will hold the stab is aligned. A T-square on top checks the front-back alignment of the stab pivot brackets. So far, so good. Then some details of the "box" at the tail that holds the back end together, and the flying rudder post mounting block. The truss is all balsa with spectra thread to reinforce. Wow this stuff is STRONG and CA works GREAT on it. I did need to go back and increase tension on the spectra by putting little spectra "strops" around the thread and CA'ing in place. I also used thin CA to strengthen the lower Balsa longeron, and put 4 runs of spectra on the bottom to reinforce the longeron. THANKS for the Spectra tip - stuff works great. This fuse sure makes the Pfalz DXV look small! Last photos are the bottom sheeting and then bottom skin (1/64) going on. The dowls are pinning the firewall and cabane bulkheads to the sides. The side skins are 1/32 but probably could have been 1/64. Fuse so far is around 5-6 lbs and balances close to the CG.

Sethhunter

Some recent pictures of progress on the flying stab pivots, control cables, and fuselage sheeting - self explanatory I think. I know some admonish the use of fiberglass, but I like the stuff and find it immensely useful for high stress complex parts.

Sethhunter

and...