Horseman3381
Well-known member
I am a big fan of flying boats, some of my favorite planes are the Grumman sea “Birds” The Albatross, Mallard, Goose and Widgion. The H-4 Hercules is what I consider the epitome of the flying boat and it conforms to my moto of “More motors, more better”.
History:
The Hughes H-4 Hercules (Spruce Goose) is a prototype flying boat build during WWII. It was intended to be used for transporting troops and cargo across the Atlantic where American liberty ships were falling prey to German U-boats. Because of wartime restrictions on aluminum and other metals the plane was designed and built primarily out of wood (mostly birch, not spruce). The plane was not completed until after the war had ended, and only flew once on November 2nd 1947 in Los Angeles harbor where it flew for 26 seconds approximately 70’ off the water for 1 mile. The plane has a length of 218’-8” and a wingspan of 320’-11” which makes it the largest wooden plane to ever fly, and the largest wingspan of any plane that has ever flown. It is powered by 8 Pratt & Whitney radial engines producing 3,000 hp each.
Specifications:
Preliminary Design:
When I started on this project I knew I wanted it to be big. I brought the two images below into AutoCAD 2017 and started drawings up plans. To determine the size of the plane I decided the wings would be removable but the fuselage needed to fit in my car, in one piece, without leaving the back open. This limited the fuselage to 8’-6” long which put the wingspan at 11’-10” and maximum prop size of 8”.
After some rough calculations I estimated the plane would weigh around 30 lbs. (It ended up just over 35 lbs.). I used a 4 blade prop to match what the actual plane uses and to provide as much prop per motor as I could. After looking around online I found the D3548/4 1100KV motor from Hobbyking that works off 3s-5s batteries and with optimum props could produce 4.75lb of thrust. The closest I could get to 8” 4-blade props that would fit on the motor shaft was a 7.5x4 4-blade prop from FMS that they use on several of their 800mm warbirds. These props are not the optimal prop for the motors, but I estimated that the motors would produce between 16-20 pounds of thrust which is more than enough to fly the plane, and barely enough to get it off the ground.
Fuselage:
The fuselage has a single layer box spar running its length with foam formers every 6”-8”. For the most part the plane closely follows the contours of the real plane, however the underside at the front of the fuselage is a flat bottom in lieu of a V bottom. This is because the plane is intended to fly off land 99% of the time and having a flat bottom makes this easier. The bottom back portion of the fuselage, after the step, is a series of geometric bends and the front is done as a continuous bend. I was concerned about the strength of the fuselage and its ability to handle the weight of the plane while landing. To help with this I tried something new to me and filled parts of the plane with the expanding foam.
To determine the best foam to use I did a series of test on different types of expanding foam to determine their density and how they expanded. In addition I did several test on the best way to install the foam. In the end I decided to use Great Stuff door and window, and found that layering it on as you are building in lieu of sticking the straw in a cavity and filling till foam came out made for a significantly better final result.
The full length of the bottom of the fuselage is filled with expanding foam. The nose is completely filled and was done in multiple layers. The side walls of the fuselage are filled to the bottom of the wing from the nose to the step. From the step to just before the tail is filled 1/3 up the side. To make where the tail attaches to the fuselage extra rigid it is filled entirely with foam.
During design I was unsure about where the batteries would need to sit to balance out the plane so I made provisions for a battery hatch in the nose and just behind the wings. The cockpit has a flat top with a hatch like the one on the FT Seaduck. The rear hatch is located where the fuselage is rounded, so that hatch is cut with the hinges to the side. It has backers that are glued to the front and back on the underside of the fuselage skin. There is also a backer on the opening end of the hatch that tucks into the fuselage to keep it closed.
To protect the bottom of the fuselage when sliding across the ground I had originally installed a layer of plastic folder on the flat portion and part way up the nose. However during the first flight the plastic shattered because the foam bottom of the plane was not rigid enough. To address this I lined the bottom of the plane with 3/32” plywood using super 77 and covered it with reinforced packing tape. Leaving the tape exposed ended up causing too much friction so plane wouldn’t move without getting a push. By installing plastic folders to the bottom of the plane alleviated the issue, and with the plywood backing they have yet to crack.
Tail:
The vertical stabilizer is built like a wing on the FT mini planes, but instead of a single piece of foam as a spar it has a wood paint stick, and is curved on both sides. This allows room for a low-profile servo to be installed in the center of the stabilizer with a double control arm to provide a push/pull control wire system on the rudder.
The horizontal stabilizers are structurally the same as the vertical stabilizer. Each side has its own servo to control the elevator, however there is a single control wire on the top of each the elevator that pulsl to move the elevator up and pushes to go down.
Wing:
It took a little bit to figure out how to build the wing and make it so that it can split into 3 sections. It ended up with a 1x3 pine spar that slides into a foamboard shell glued to the bottom sheeting of the wing. There are 2 recessed screw receivers per outer wing section in the spar that are used to connect the 1x3 spar with holes in the bottom of the outer wing using nylon bolts. These keep the wings from pulling away from the fuselage. Where the bolts go through the bottom of the wing the foam was removed and a 3/16” piece of birch wood was installed for strength.
To keep the wings from twisting, the center sections of wing is recessed to allow for extensions on the outer wing portions to slide into them. The extensions are filled with expanding foam to strengthen them while the inner wing portion that receives them is wrapped with 2 layers of reinforced packing tape at the ends.
The main spar of the wing is not perpendicular to the fuselage and where the wood spar runs does not follow the thickest part of the wing. So at the thickest portion of the wing there is a tapered u-box foam spar. The spar is a single layer of foam but is reinforced with spray foam on the inside. In addition, there are no formers in the back portion of the wings. These areas are filled with spray foam insulation to strengthen them. The front of the wing is built with a series formers glued to the front of the foam spar to get its curved shape.
Of all the things on this plane that are overbuilt, the motor mounts are probably the most overbuilt. Each motor mount is a 2” sq. hunk of 1 by oak. This is epoxied and anchored with a screw to a pine 2x2 which is wrapped in foamboard to form the nacelle. These are angled at the end so they are parallel to the fuselage and are hot glued to the face of the foam spar. There are formers glued to each side of the nacelles to help reinforce their attachment to the spar.
I had originally planned to put cowlings over the nacelles, but because the motors were bigger than I had hoped this caused the core of the nacelles to be bigger. The cowlings would have been way out of scale so I opted to leave them off. The ESC’s are hot glued to the bottom of the nacelles where the radiators for the engines were on the real plane. It worked out that their size and shape are what I consider passible as stand-ins. The wiring for the motors and aileron servos were run through the front of the wing prior to installing the sheathing. The engines on each wing are wired together using y-harness’ at each ESC so there is only one plug at each wing joint for the motor control wire. Each motor has its own battery, so there are 4 power plugs at each joint. In order to leave room for all the wires, the center portion of the wing forward of the foam spar is hollow.
The Ailerons are to scale. This proved challenging as it would have looked funny, and I wound venture to guess not have the rigidity needed to function properly, if they were single pieces of foam like on smaller planes. To address this the wings were initially built as if there are no ailerons. Once they were glued the bottom sheeting of the wing was cut at the hinge point and tapered up to the hinge joint to add strength. Both sides of the hinge are reinforced with paper surgical tape to add strength. Because of the thickness of the alerion the control horn needed to be on the top of the aileron, and the aileron servo installed inside the wings with the control arm sticking out the top. The arrangement of the servo and control horn made it impossible to install wire control rods with z-bends. This was because the screw keeping the servo control horn on is not accessible and the distance between the control horns and servo arm was too short to twist the wire enough to install it. In the end I used threaded rod and control linkages.
The pontoons are built with a single layer of foamboard shell filled with expanding foam and the pontoon struts stick into the top of the pontoon all the way to the bottom. Initially the plane had a paint stick running down the center of the pontoon struts for reinforcement. However during a taxi test on the snow I managed to rip them both off. To reinforce them there are some heavy duty aluminum 4” angle brackets with one leg running down the strut, and the other running in the wing towards the fuselage with the portion of the wing where the bracket runs filled with expanding foam.
During design I had intended to fly this off water. However, I remembered Josh’s comment on the original pontoons for the Seaduck creating suction on the water and turning the plane. As a precaution I changed the design of the pontoons to have a step in them just like the fuselage.
When I originally designed and started building the plane I did not include flaps. This is because at the time I only had a 6-channel radio, and was not going to have enough channels to operate them. However, now with a larger channel radio, if I were to build the plane again I would add them to help get the plane off the ground. The plane flys and glides well so they are not needed for landing.
Power Systems:
Because of the size of the plane and it’s servos, instead of using one of the ESC’s to power the plane’s receiver and servos I installed High Voltage servos and use a 2S 1900 mAH LiFe battery plugged directly into the receiver to power the electronics. The plane has an 8 channel receiver with the following channel assignments:
View attachment 109525
To balance the plane the 2S LiFe battery and (2) of the 4S LiPo batteries are in the rear hatch. The remaining (6) 4S LiPo batteries are in the nose. To hold the batteries in place there is a foamboard sleeve. Inside the sleeve are strips of the fuzzy side of Velcro which helps keep them snug. As the plane would likely not withstand any sort of acrobatic flying I am relying on the friction of the batteries in the sleeves to keep them in place.
Painting:
The plane seams and exposed edges are lined with paper surgical tape. I do this on all my planes to help keep the paper from pulling off at the edges and protect exposed foam when painting. It is painted with 2 layers of Rustolium Satin Stone Grey spray paint. I didn’t quite get the shade of grey I was hoping for, but it was the closest I could find.
Testing Incident:
With the plane was substantially complete I took it out to the flying field for a taxi test. This was mid-March and there was still snow on the ground. During the test I went over an unseen snow drift and the plane got briefly airborne. Immediately it rolled right and did a sort of tumble/spin that broke off the pontoons and 2 props. However, the main structure held up well.
Performance:
The plane has differential thrust to allow it to taxi on the ground, however it does not have the power to do so easily. In addition, it takes around 350’ of runway to take off. Once it is in the air it flys great at around 5/8 throttle. It is very docile when flying and requires coordinated turns with the rudder. If you do not disable the differential thrust once off the ground the inside wing tends to dip and the plane starts to drop like it is tip stalling, even at full throttle.
The 1x3 spar in the wings has plenty of strength and doesn’t flex much in flight, however under higher G maneuvers you can see the wing tips flexing where the spar stops (Approximately 12” from the wing tip).
Though the plane is heavy, with the large wing area it glides well and gets some help from ground effect on landing. However it still needs to be brought in under power. In the little stall testing I have done it doesn’t tend to tip stall which surprised me a little as I didn’t leave an under camber on the end of the wing.
I have around 200 hours into design and building the plane. It’s not pretty, but I am super happy with how it turned out. I wish it was a little lighter to make taxing and take off easier, but couldn’t be any happier with how it flies.
History:
The Hughes H-4 Hercules (Spruce Goose) is a prototype flying boat build during WWII. It was intended to be used for transporting troops and cargo across the Atlantic where American liberty ships were falling prey to German U-boats. Because of wartime restrictions on aluminum and other metals the plane was designed and built primarily out of wood (mostly birch, not spruce). The plane was not completed until after the war had ended, and only flew once on November 2nd 1947 in Los Angeles harbor where it flew for 26 seconds approximately 70’ off the water for 1 mile. The plane has a length of 218’-8” and a wingspan of 320’-11” which makes it the largest wooden plane to ever fly, and the largest wingspan of any plane that has ever flown. It is powered by 8 Pratt & Whitney radial engines producing 3,000 hp each.
Specifications:
- Wingspan – 11’-10”
- Fuselage Length – 8’-6”
- Motor – (8) Turnigy D3548/4 1100KV
- ESC – 70A
- Prop – 7.5x4 4-blade
- Batteries:
- (8) 4s, 4000mah Lipo for motors
- (1) 2s 1900 mah LiFe for receiver and controls
- Servos:
- Aileron – Hitec HS-5495BH 6.4kg
- Elevator & Rudder – (3) Corona CS-239HV 4.6kg
- 33 sheets of Foamboard
Preliminary Design:
When I started on this project I knew I wanted it to be big. I brought the two images below into AutoCAD 2017 and started drawings up plans. To determine the size of the plane I decided the wings would be removable but the fuselage needed to fit in my car, in one piece, without leaving the back open. This limited the fuselage to 8’-6” long which put the wingspan at 11’-10” and maximum prop size of 8”.
After some rough calculations I estimated the plane would weigh around 30 lbs. (It ended up just over 35 lbs.). I used a 4 blade prop to match what the actual plane uses and to provide as much prop per motor as I could. After looking around online I found the D3548/4 1100KV motor from Hobbyking that works off 3s-5s batteries and with optimum props could produce 4.75lb of thrust. The closest I could get to 8” 4-blade props that would fit on the motor shaft was a 7.5x4 4-blade prop from FMS that they use on several of their 800mm warbirds. These props are not the optimal prop for the motors, but I estimated that the motors would produce between 16-20 pounds of thrust which is more than enough to fly the plane, and barely enough to get it off the ground.
Fuselage:
The fuselage has a single layer box spar running its length with foam formers every 6”-8”. For the most part the plane closely follows the contours of the real plane, however the underside at the front of the fuselage is a flat bottom in lieu of a V bottom. This is because the plane is intended to fly off land 99% of the time and having a flat bottom makes this easier. The bottom back portion of the fuselage, after the step, is a series of geometric bends and the front is done as a continuous bend. I was concerned about the strength of the fuselage and its ability to handle the weight of the plane while landing. To help with this I tried something new to me and filled parts of the plane with the expanding foam.
To determine the best foam to use I did a series of test on different types of expanding foam to determine their density and how they expanded. In addition I did several test on the best way to install the foam. In the end I decided to use Great Stuff door and window, and found that layering it on as you are building in lieu of sticking the straw in a cavity and filling till foam came out made for a significantly better final result.
The full length of the bottom of the fuselage is filled with expanding foam. The nose is completely filled and was done in multiple layers. The side walls of the fuselage are filled to the bottom of the wing from the nose to the step. From the step to just before the tail is filled 1/3 up the side. To make where the tail attaches to the fuselage extra rigid it is filled entirely with foam.
During design I was unsure about where the batteries would need to sit to balance out the plane so I made provisions for a battery hatch in the nose and just behind the wings. The cockpit has a flat top with a hatch like the one on the FT Seaduck. The rear hatch is located where the fuselage is rounded, so that hatch is cut with the hinges to the side. It has backers that are glued to the front and back on the underside of the fuselage skin. There is also a backer on the opening end of the hatch that tucks into the fuselage to keep it closed.
To protect the bottom of the fuselage when sliding across the ground I had originally installed a layer of plastic folder on the flat portion and part way up the nose. However during the first flight the plastic shattered because the foam bottom of the plane was not rigid enough. To address this I lined the bottom of the plane with 3/32” plywood using super 77 and covered it with reinforced packing tape. Leaving the tape exposed ended up causing too much friction so plane wouldn’t move without getting a push. By installing plastic folders to the bottom of the plane alleviated the issue, and with the plywood backing they have yet to crack.
Tail:
The vertical stabilizer is built like a wing on the FT mini planes, but instead of a single piece of foam as a spar it has a wood paint stick, and is curved on both sides. This allows room for a low-profile servo to be installed in the center of the stabilizer with a double control arm to provide a push/pull control wire system on the rudder.
The horizontal stabilizers are structurally the same as the vertical stabilizer. Each side has its own servo to control the elevator, however there is a single control wire on the top of each the elevator that pulsl to move the elevator up and pushes to go down.
Wing:
It took a little bit to figure out how to build the wing and make it so that it can split into 3 sections. It ended up with a 1x3 pine spar that slides into a foamboard shell glued to the bottom sheeting of the wing. There are 2 recessed screw receivers per outer wing section in the spar that are used to connect the 1x3 spar with holes in the bottom of the outer wing using nylon bolts. These keep the wings from pulling away from the fuselage. Where the bolts go through the bottom of the wing the foam was removed and a 3/16” piece of birch wood was installed for strength.
To keep the wings from twisting, the center sections of wing is recessed to allow for extensions on the outer wing portions to slide into them. The extensions are filled with expanding foam to strengthen them while the inner wing portion that receives them is wrapped with 2 layers of reinforced packing tape at the ends.
The main spar of the wing is not perpendicular to the fuselage and where the wood spar runs does not follow the thickest part of the wing. So at the thickest portion of the wing there is a tapered u-box foam spar. The spar is a single layer of foam but is reinforced with spray foam on the inside. In addition, there are no formers in the back portion of the wings. These areas are filled with spray foam insulation to strengthen them. The front of the wing is built with a series formers glued to the front of the foam spar to get its curved shape.
Of all the things on this plane that are overbuilt, the motor mounts are probably the most overbuilt. Each motor mount is a 2” sq. hunk of 1 by oak. This is epoxied and anchored with a screw to a pine 2x2 which is wrapped in foamboard to form the nacelle. These are angled at the end so they are parallel to the fuselage and are hot glued to the face of the foam spar. There are formers glued to each side of the nacelles to help reinforce their attachment to the spar.
I had originally planned to put cowlings over the nacelles, but because the motors were bigger than I had hoped this caused the core of the nacelles to be bigger. The cowlings would have been way out of scale so I opted to leave them off. The ESC’s are hot glued to the bottom of the nacelles where the radiators for the engines were on the real plane. It worked out that their size and shape are what I consider passible as stand-ins. The wiring for the motors and aileron servos were run through the front of the wing prior to installing the sheathing. The engines on each wing are wired together using y-harness’ at each ESC so there is only one plug at each wing joint for the motor control wire. Each motor has its own battery, so there are 4 power plugs at each joint. In order to leave room for all the wires, the center portion of the wing forward of the foam spar is hollow.
The Ailerons are to scale. This proved challenging as it would have looked funny, and I wound venture to guess not have the rigidity needed to function properly, if they were single pieces of foam like on smaller planes. To address this the wings were initially built as if there are no ailerons. Once they were glued the bottom sheeting of the wing was cut at the hinge point and tapered up to the hinge joint to add strength. Both sides of the hinge are reinforced with paper surgical tape to add strength. Because of the thickness of the alerion the control horn needed to be on the top of the aileron, and the aileron servo installed inside the wings with the control arm sticking out the top. The arrangement of the servo and control horn made it impossible to install wire control rods with z-bends. This was because the screw keeping the servo control horn on is not accessible and the distance between the control horns and servo arm was too short to twist the wire enough to install it. In the end I used threaded rod and control linkages.
The pontoons are built with a single layer of foamboard shell filled with expanding foam and the pontoon struts stick into the top of the pontoon all the way to the bottom. Initially the plane had a paint stick running down the center of the pontoon struts for reinforcement. However during a taxi test on the snow I managed to rip them both off. To reinforce them there are some heavy duty aluminum 4” angle brackets with one leg running down the strut, and the other running in the wing towards the fuselage with the portion of the wing where the bracket runs filled with expanding foam.
During design I had intended to fly this off water. However, I remembered Josh’s comment on the original pontoons for the Seaduck creating suction on the water and turning the plane. As a precaution I changed the design of the pontoons to have a step in them just like the fuselage.
When I originally designed and started building the plane I did not include flaps. This is because at the time I only had a 6-channel radio, and was not going to have enough channels to operate them. However, now with a larger channel radio, if I were to build the plane again I would add them to help get the plane off the ground. The plane flys and glides well so they are not needed for landing.
Power Systems:
Because of the size of the plane and it’s servos, instead of using one of the ESC’s to power the plane’s receiver and servos I installed High Voltage servos and use a 2S 1900 mAH LiFe battery plugged directly into the receiver to power the electronics. The plane has an 8 channel receiver with the following channel assignments:
- Left Throttle
- Right Throttle
- Left Aileron
- Right Aileron
- Left Elevator
- Right Elevator
- Rudder
View attachment 109525
To balance the plane the 2S LiFe battery and (2) of the 4S LiPo batteries are in the rear hatch. The remaining (6) 4S LiPo batteries are in the nose. To hold the batteries in place there is a foamboard sleeve. Inside the sleeve are strips of the fuzzy side of Velcro which helps keep them snug. As the plane would likely not withstand any sort of acrobatic flying I am relying on the friction of the batteries in the sleeves to keep them in place.
Painting:
The plane seams and exposed edges are lined with paper surgical tape. I do this on all my planes to help keep the paper from pulling off at the edges and protect exposed foam when painting. It is painted with 2 layers of Rustolium Satin Stone Grey spray paint. I didn’t quite get the shade of grey I was hoping for, but it was the closest I could find.
Testing Incident:
With the plane was substantially complete I took it out to the flying field for a taxi test. This was mid-March and there was still snow on the ground. During the test I went over an unseen snow drift and the plane got briefly airborne. Immediately it rolled right and did a sort of tumble/spin that broke off the pontoons and 2 props. However, the main structure held up well.
Performance:
The plane has differential thrust to allow it to taxi on the ground, however it does not have the power to do so easily. In addition, it takes around 350’ of runway to take off. Once it is in the air it flys great at around 5/8 throttle. It is very docile when flying and requires coordinated turns with the rudder. If you do not disable the differential thrust once off the ground the inside wing tends to dip and the plane starts to drop like it is tip stalling, even at full throttle.
The 1x3 spar in the wings has plenty of strength and doesn’t flex much in flight, however under higher G maneuvers you can see the wing tips flexing where the spar stops (Approximately 12” from the wing tip).
Though the plane is heavy, with the large wing area it glides well and gets some help from ground effect on landing. However it still needs to be brought in under power. In the little stall testing I have done it doesn’t tend to tip stall which surprised me a little as I didn’t leave an under camber on the end of the wing.
I have around 200 hours into design and building the plane. It’s not pretty, but I am super happy with how it turned out. I wish it was a little lighter to make taxing and take off easier, but couldn’t be any happier with how it flies.
Attachments
Last edited:
