Birdmanzak
Well-Known Member
[UPDATE 2010-10-12]
This design has changed a bit since the original post. The competition I was designing it for - a Red Bull Flugtag in Sydney - was postponed to this year. I got in but, after they decided metal was "too dangerous" and "sinks" that it couldn't be used in any entries, withdrew. The plane's still being built - just to see how bad my approximations and shortcuts were.
The details are here: www.wrongbrothers.com
[END OF UPDATE]
[TLDR version]
I'm designing and building a biplane glider. It's got a limited wingspan (8m) and I want to keep the empty mass down to below 30kg [66lb]. I want some of you to review it before I bust out the credit card and hacksaw.
[/TLDR]
- - - - -
Background:
I'm building this for an event. The event is one Dana built a plane for in the "Fun with foam and water" thread. Not that I'm paranoid (actually, I am) but I'd prefer it if "the colour of a stop sign" and "male bovine", or "flu" and "-gtag" aren't mentioned in this thread - becuase I know how much time I spend doing internet research before these events and if Google never indexes those phrases here, then other competitors are unlikely to find this thread (or all it's ground-breaking contents - tongue firmly in cheek).
I've entered before, and also in a similar competition in England. Never very successfully. This time I'm giving myself plenty of time to get the plane right and -most importantly- train the pilot right up.
I'm using the Larry Mauro Easy Riser and Mike Sandlin's BUG and GOAT as examples of similar planes, and stealing a lot of ideas from them (particularly structural).
Training flying will be done in the hill-skimming style of the early Rogallo gliders. Turning will be a bonus, but I want some roll and yaw control to avoid accidents and control the landings. Maximum altitude should never be above 10m [30'] (I know - high enough to kill me
). The final flight will be from a 6m [20'] platform into water. The launch will use a secret and cunning mechanism known only to me, my launch team, and the 60000 people who watched me use it in April. It should take my launch speed to ~8.3m/s [16kt].
I'm Australian so I think and work in metric. I've tried to put Ye Olde measurements in [square brackets] after my metric measurements for those who prefer them. They're approximations and where a discrepancy exists the metric measurement is authoritative. The only exception is for material gauges, because they're only available in imperial units. Bah.
This post is a bit disjointed because I wrote it in fits and spurts - mostly at work and half-asleep.
Software used:
# Excel
# XFLR5 See here - XFoil with lifting line and vortex lattice wing analysis.
# Pro/Engineer
# Pencil and Envelope Mk I
Design goals:
Basic layout:
I want a tailless design. I know a tailed aircraft would likely have better performance but aesthetics and bloody-mindedness dictate this one will be tailless.
Assuming a wing CL of 1.0, a flight velocity of 8.3m/s, mass of 120kg:
Wingspan is restricted by event rules to 8m [26'], which -to me- means I must use a biplane. I expect ~30% efficiency gain from a biplane compared to a monoplane with AR of 2.25 (from Raymer).
I've chosen a 0.25 chord sweep of about 22deg with reference to Nickel (Tailless Aircraft in Theory and Practice). That was while I was looking at the LNV109a so I'll review it with the MH78.
I've arbitrarily chosen a vertical gap of 1.5m [5'] and zero stagger - mostly for contruction simplicity.
Elevons will be mounted on the lower wing for cable routing.
Airfoil selection:
I'm attached to the Douglas/Liebeck LNV109a. I saw the series mentioned in B. McCormick's Aerodynamics, Aeronautics and Flight Mechanics as a good example of a low speed airfoil. Supposedly it's a "turbulent rooftop" airfoil. From memory that means it uses pressure gradients to encourage an early transition so the turbulent BL stays attached longer. Sounds good to me. It's got a small Cm (-0.05) and looks very similar to the Lissaman airfoil used on the Gossamer Albatross. I've also used it twice before. Trying to force it onto this plane meant I needed to use about 7 degrees of washout. I think that's going to reduce efficiency, and make covering difficult. Not to mention how much it will mess with the interpolated airfoils.
I've also looked at the MH78. It lets me cut the washout (so far I've only looked at reducing it to 2 degrees) - with all the benefits that includes. It seems to have better performance than the LNV109a too. I also like the thicker trailing edge from a structural point of view: it should help the foam ribs protruding aft hold their shape.
I've pretty much committed to the MH78, I'm hoping someone will validate that choice.
Pitch stability and control:
I've knocked up a (really ugly) spreadsheet in Excel. It's HERE. It's based on the trim analysis equations in Raymer, takes longitudinal and vertical locations for the CG, basic aircraft and control surface geometry and gives a pretty-looking trim graph that seems to give sensible results. If anyone wants to look it over for stupid mistakes, or just give it a common-sense check I'd appreciate feedback.
The equations in Raymer make it look like plain flap effectiveness drops off after about 20 degrees deflection. Does that sound right? Will I need to limit deflection to 20 degrees or thereabouts for predictable control? I haven't taken into account the loss of pitching moment I'll get if I'm also mixing in roll movements.
Directional stability and control:
I haven't done any work on this. There are eyeballed rudders, some roll control from the elevons, and rule-of-thumb/historical dihedral. The dihedral (5 degrees) is stolen directly from the Easy Riser. As long as the plane's "quite" directionally stable and "sufficiently" steerable I'm happy. It doesn't need to be agile directionally - just good enough to avoid an accident.
The rudders are based on Easy Riser rudders. They're independent and only the rudder on the inside of the turn deflects. I'm also thinking I can use them together as drag brakes if I ever need to.
Structure:
I'm stealing the Easy Riser metal structure almost exactly. The below images are missing any structure between the lower wings - that's clearly not how it'll be built. I'm holding off modelling any of that until I have a better idea of where the CG, and therefore pilot, will be.
I've added a chord-wise member halfway out the wing. I had wanted to make the frame from 1" diameter 0.035" wall 6061-T6 but Pro/E suggests that 0.750" diameter tube will be sufficient. The smaller tube saves about 5kg [11lb] (~14.5kg [32lb] for all the alloy, down from ~20kg [49lb]). My main worry is how well it will survive handling and unexpected bumps or crashes.
The analysis I did with Pro/E was up to about 7.5 G - with the lift vectors pointing 45 degrees forward of vertical and applied as point loads at each the junction of each rib with the spars. It resulted in some ridiculous deformations and a failure but I'm hoping that the high simulated loads and forward component are high enough that inaccuracies in the loading and configuration are overwhelmed. I'll be running more checks, as well as buckling analyses (this one was static only), at least until I get results without a failure.
I want to make the joints in the frame in a similar way to the GOAT. Everywhere there's a hardpoint Sandlin uses a smaller tube inside the main one and bulks it up to a snug fit - I can't remember what with. Tape? I want to cut an axial slit down the side of my sleeves, compress them to fit snugly, maybe epoxy them in place, and then blind rivet them. I'm yet to work out how to get Pro/E to do an FEA of the joint assembly and so far my backup plan is trust to blind faith. Actually, I'll make one up and test it, but I'd rather have an idea I'm going down the right path before I start spending metal. For rivets I'll probably be using CherryMax monel universal head 1/8" blind rivets - unless there's a compelling reason to use something else. The bolted joints will be 3/16" and I'll use an aluminium spacer across the diameter of the tube between the tangs. I'm thinking bicycle brake cables will serve as bracing cables in the frame.
For ribs I'll use Dow Styrofoam HD300F (45kg/m^3 [2.8lb/cu ft]) to make 25mm x 25mm [1" x 1"] square section ribs with glass caps - six in each of the four half-wings. They'll extend from the forward spar to the aft spar and then aft to the trailing edge. The rest of the rib profiles will be made up with white expanded-bead polystyrene foam (11kg/m^3 [0.7lb/ cu ft]. I'll also make the wingtips from expanded-bead foam.
I had originally planned to use the heat-shrink plastic that they store and transport boats in, but it's quite heavy. I'm thinking Mylar is the only thing that will be light enough. Ideally the 56m^2 [602sq ft] of wing skin material will weigh less than 1kg [2.2lb]. I'm worried about tear propagation but it works for a lot of human-powered aircraft, and if it becomes a problem I can always put sticky tape tear panels on it to limit the tearing. I can lose the extra weight for the event day by re-skinning the plane without the tape.
-
So, suggestions? Criticism?
This design has changed a bit since the original post. The competition I was designing it for - a Red Bull Flugtag in Sydney - was postponed to this year. I got in but, after they decided metal was "too dangerous" and "sinks" that it couldn't be used in any entries, withdrew. The plane's still being built - just to see how bad my approximations and shortcuts were.
The details are here: www.wrongbrothers.com
[END OF UPDATE]
[TLDR version]
I'm designing and building a biplane glider. It's got a limited wingspan (8m) and I want to keep the empty mass down to below 30kg [66lb]. I want some of you to review it before I bust out the credit card and hacksaw.
[/TLDR]
- - - - -
Background:
I'm building this for an event. The event is one Dana built a plane for in the "Fun with foam and water" thread. Not that I'm paranoid (actually, I am) but I'd prefer it if "the colour of a stop sign" and "male bovine", or "flu" and "-gtag" aren't mentioned in this thread - becuase I know how much time I spend doing internet research before these events and if Google never indexes those phrases here, then other competitors are unlikely to find this thread (or all it's ground-breaking contents - tongue firmly in cheek).
I've entered before, and also in a similar competition in England. Never very successfully. This time I'm giving myself plenty of time to get the plane right and -most importantly- train the pilot right up.
I'm using the Larry Mauro Easy Riser and Mike Sandlin's BUG and GOAT as examples of similar planes, and stealing a lot of ideas from them (particularly structural).
Training flying will be done in the hill-skimming style of the early Rogallo gliders. Turning will be a bonus, but I want some roll and yaw control to avoid accidents and control the landings. Maximum altitude should never be above 10m [30'] (I know - high enough to kill me
). The final flight will be from a 6m [20'] platform into water. The launch will use a secret and cunning mechanism known only to me, my launch team, and the 60000 people who watched me use it in April. It should take my launch speed to ~8.3m/s [16kt].I'm Australian so I think and work in metric. I've tried to put Ye Olde measurements in [square brackets] after my metric measurements for those who prefer them. They're approximations and where a discrepancy exists the metric measurement is authoritative. The only exception is for material gauges, because they're only available in imperial units. Bah.
This post is a bit disjointed because I wrote it in fits and spurts - mostly at work and half-asleep.
Software used:
# Excel
# XFLR5 See here - XFoil with lifting line and vortex lattice wing analysis.
# Pro/Engineer
# Pencil and Envelope Mk I
Design goals:
Trailerable: it needs to break up into the four main wing sections, cockpit module, and any extras that don't fold flush to the wing for transport. An hour for assembly or disassembly by one or two people is acceptable.
Limited wingspan: the wingspan must not exceed 8m [26.2']. That can't be changed, no matter how much I want it to.
Low stall speed: I can pretty well guarantee a take-off speed of 8.3m/s [16kt] and no more. That should obviously be somewhat above stall.
Controllable: Pitch control is very important. Directional control less so.
Limited service life: The plane will be flown for between 12 and 18 months, no more.
Simple build: I have the tools to do vacuum bagging and infusion (and the skills to do them badly) but I want to stay away from composites for primary structures. I just don't trust my own work, and don't want to bother doing test coupons to validate anything. I'm happier buying alloy with known characteristics and working to that. If I can build most of it with a hacksaw and rivet gun that will suit me perfectly. I'm keen to use foam and a little glass for minor structure.
The low stall speed requirement implies large wing area and low mass. At 88kg [194lb] I'm not a very good choice for pilot but as project manager I'm overruling the weights guy (also me) and putting myself there. I've done my assumed initial calculations with a MTOW of 120kg [264lb], which leaves 32kg [70lb] for the plane. I'd like to get that (and my own) down as much as possible. You can't help with my weight, but if I can get the plane down towards 26kg [57lb] I'll be really happy.Limited wingspan: the wingspan must not exceed 8m [26.2']. That can't be changed, no matter how much I want it to.
Low stall speed: I can pretty well guarantee a take-off speed of 8.3m/s [16kt] and no more. That should obviously be somewhat above stall.
Controllable: Pitch control is very important. Directional control less so.
Limited service life: The plane will be flown for between 12 and 18 months, no more.
Simple build: I have the tools to do vacuum bagging and infusion (and the skills to do them badly) but I want to stay away from composites for primary structures. I just don't trust my own work, and don't want to bother doing test coupons to validate anything. I'm happier buying alloy with known characteristics and working to that. If I can build most of it with a hacksaw and rivet gun that will suit me perfectly. I'm keen to use foam and a little glass for minor structure.
Basic layout:
I want a tailless design. I know a tailed aircraft would likely have better performance but aesthetics and bloody-mindedness dictate this one will be tailless.
Assuming a wing CL of 1.0, a flight velocity of 8.3m/s, mass of 120kg:
Code:
L = CL * q * S
L = W = 120kg * 10m/s^2 = 1200N
1200 = 1.0 * 0.5 * 1.225 * 8.3^2 * S
S = 1200 / (0.5 * 1.225 * 8.3^2)
= 28.4 m^2 [305sq ft]I've chosen a 0.25 chord sweep of about 22deg with reference to Nickel (Tailless Aircraft in Theory and Practice). That was while I was looking at the LNV109a so I'll review it with the MH78.
I've arbitrarily chosen a vertical gap of 1.5m [5'] and zero stagger - mostly for contruction simplicity.
Elevons will be mounted on the lower wing for cable routing.
Airfoil selection:
I'm attached to the Douglas/Liebeck LNV109a. I saw the series mentioned in B. McCormick's Aerodynamics, Aeronautics and Flight Mechanics as a good example of a low speed airfoil. Supposedly it's a "turbulent rooftop" airfoil. From memory that means it uses pressure gradients to encourage an early transition so the turbulent BL stays attached longer. Sounds good to me. It's got a small Cm (-0.05) and looks very similar to the Lissaman airfoil used on the Gossamer Albatross. I've also used it twice before. Trying to force it onto this plane meant I needed to use about 7 degrees of washout. I think that's going to reduce efficiency, and make covering difficult. Not to mention how much it will mess with the interpolated airfoils.
I've also looked at the MH78. It lets me cut the washout (so far I've only looked at reducing it to 2 degrees) - with all the benefits that includes. It seems to have better performance than the LNV109a too. I also like the thicker trailing edge from a structural point of view: it should help the foam ribs protruding aft hold their shape.
I've pretty much committed to the MH78, I'm hoping someone will validate that choice.
Pitch stability and control:
I've knocked up a (really ugly) spreadsheet in Excel. It's HERE. It's based on the trim analysis equations in Raymer, takes longitudinal and vertical locations for the CG, basic aircraft and control surface geometry and gives a pretty-looking trim graph that seems to give sensible results. If anyone wants to look it over for stupid mistakes, or just give it a common-sense check I'd appreciate feedback.
The equations in Raymer make it look like plain flap effectiveness drops off after about 20 degrees deflection. Does that sound right? Will I need to limit deflection to 20 degrees or thereabouts for predictable control? I haven't taken into account the loss of pitching moment I'll get if I'm also mixing in roll movements.
Directional stability and control:
I haven't done any work on this. There are eyeballed rudders, some roll control from the elevons, and rule-of-thumb/historical dihedral. The dihedral (5 degrees) is stolen directly from the Easy Riser. As long as the plane's "quite" directionally stable and "sufficiently" steerable I'm happy. It doesn't need to be agile directionally - just good enough to avoid an accident.
The rudders are based on Easy Riser rudders. They're independent and only the rudder on the inside of the turn deflects. I'm also thinking I can use them together as drag brakes if I ever need to.
Structure:
I'm stealing the Easy Riser metal structure almost exactly. The below images are missing any structure between the lower wings - that's clearly not how it'll be built. I'm holding off modelling any of that until I have a better idea of where the CG, and therefore pilot, will be.
I've added a chord-wise member halfway out the wing. I had wanted to make the frame from 1" diameter 0.035" wall 6061-T6 but Pro/E suggests that 0.750" diameter tube will be sufficient. The smaller tube saves about 5kg [11lb] (~14.5kg [32lb] for all the alloy, down from ~20kg [49lb]). My main worry is how well it will survive handling and unexpected bumps or crashes.
The analysis I did with Pro/E was up to about 7.5 G - with the lift vectors pointing 45 degrees forward of vertical and applied as point loads at each the junction of each rib with the spars. It resulted in some ridiculous deformations and a failure but I'm hoping that the high simulated loads and forward component are high enough that inaccuracies in the loading and configuration are overwhelmed. I'll be running more checks, as well as buckling analyses (this one was static only), at least until I get results without a failure.
I want to make the joints in the frame in a similar way to the GOAT. Everywhere there's a hardpoint Sandlin uses a smaller tube inside the main one and bulks it up to a snug fit - I can't remember what with. Tape? I want to cut an axial slit down the side of my sleeves, compress them to fit snugly, maybe epoxy them in place, and then blind rivet them. I'm yet to work out how to get Pro/E to do an FEA of the joint assembly and so far my backup plan is trust to blind faith. Actually, I'll make one up and test it, but I'd rather have an idea I'm going down the right path before I start spending metal. For rivets I'll probably be using CherryMax monel universal head 1/8" blind rivets - unless there's a compelling reason to use something else. The bolted joints will be 3/16" and I'll use an aluminium spacer across the diameter of the tube between the tangs. I'm thinking bicycle brake cables will serve as bracing cables in the frame.
For ribs I'll use Dow Styrofoam HD300F (45kg/m^3 [2.8lb/cu ft]) to make 25mm x 25mm [1" x 1"] square section ribs with glass caps - six in each of the four half-wings. They'll extend from the forward spar to the aft spar and then aft to the trailing edge. The rest of the rib profiles will be made up with white expanded-bead polystyrene foam (11kg/m^3 [0.7lb/ cu ft]. I'll also make the wingtips from expanded-bead foam.
I had originally planned to use the heat-shrink plastic that they store and transport boats in, but it's quite heavy. I'm thinking Mylar is the only thing that will be light enough. Ideally the 56m^2 [602sq ft] of wing skin material will weigh less than 1kg [2.2lb]. I'm worried about tear propagation but it works for a lot of human-powered aircraft, and if it becomes a problem I can always put sticky tape tear panels on it to limit the tearing. I can lose the extra weight for the event day by re-skinning the plane without the tape.
-
So, suggestions? Criticism?
Last edited:
