My Carbon Cub Design - A Lesson in Trial & Error

by MikeRobey | September 25, 2018 | (8) Posted in Projects


Introduction

I always liked the nose of the Carbon Cub so this project started out as an excuse to design and 3D print a carbon Cub nose.  I figured the rest would be pretty straight forward design wise and it would have some gentle, stable flying characteristics.  The design of the wings and fuselage was pretty straight forward. I wanted the cute nose, the balloon tyres and the wing struts.  After all without those it isn't a Carbon Cub.  This was going to be easy.  Modelling the nose took me two weeks and three attempts before I was happy with the result.  This was the first indication that maybe this project wasn't going to be that easy after all.


From Trauma to Delight

This story ends with a model that flies nicely but it didn't begin that way.  It took three attempts to get it to fly nicely.  The first version of the design was pretty much to scale. It flew horribly.  On takeoff it would swing one way until the tail lifted up and then the swing reversed itself in the other direction.  I did not see the point in a plane which was such a pain to get off the ground so I was thinking I needed to go back to the drawing board.  That thought was just occurring to me as I looped the Cub.  At the top of the loop it flicked into an inverted spin that I could not get it out of.  Normally with a model plane in a spin, just centralising the controls is enough to allow the model to recover.  Not in this case. My pretty little cub span into the ground.  Surprisingly there was no damage but it was heading for the scrap heap anyway and I was going back to the drawing board for a second attempt.

I noticed that a lot of Cub models have non-scale fuselage lengths.  Most designs seem to lengthen the fuselage.  There must be a reason for that I thought so I lengthened mine.  I thought about the stalling and spinning issue and decided to increase the chord length on the wings as well as increasing the cambered area of the wing at the tips.  At the time I thought what had happened was a tip stall. It turns out I was wrong.  In hindsight I believe I attempted to fly a loop which was too tight causing a normal stall. I also think I had the CG too far to the rear.  However at the time of the maiden of the second version I had not realised this.  I also made the fuselage a little wider and deeper so that it still looked like a Cub even if it was no longer to scale.

Getting the plane to balance at the required CG was very difficult because of the lengthened fuselage.  I really, really hate to add weight to a model.  It was too late in the day to go for a bigger motor and there was not room for a bigger battery.  However, using the fingers under the wing technique, I thought I had the CG over the wing spar which is where it should be.  I was so very, very wrong about where the CG was.  On both versions of the model I had built in 2 degrees of down thrust and two degrees of right thrust.  I have found that to be a good starting point and generally adjust if required.  The first two flights of the second plane were even more traumatic than with the first version!  The swing was now just in one direction but, as soon as the model became airborne, the nose dropped alarmingly causing the plane to bounce back on the ground. I put in more elevator and got it pointing upwards but then it rolled onto its side. I corrected that and we were finally moving away from the ground.  Turns were a nightmare and I could only just keep the plane over the flying field. I managed to get it lined up for a landing, cut the power and landed. Fortunately I did not apply rudder on landing and I was confused by the gentle landing characteristics.  They contrasted so much with the rest of the flight.  It wasn't overly pitch sensitive so I was convinced it was not tail heavy.  I was, of course wrong. I tried a second flight.  I didn't expect things to be different but I wanted to try and observe what was going on.  I was so shocked and stunned during the first flight I had no idea what it had been doing other than scaring the crap out of me.  On the second flight I realised what was happening.  Turning on aileron was not possible.  It rolled nicely but would just keep going straight. Applying rudder caused the plane to dive vertically. It did this instantly so it was not a matter of a little elevator to compensate.  The instant I applied rudder the plane was pointing straight down.  The only way to turn was to bank and yank.

I went home and thought about this. I also chatted with my fellow club mates about it on Facebook.  The first change needed was to remove the right thrust. I had observed when taxiing that the plane would turn right fine but it was almost impossible to get it to turn left.  I normally fly a left hand circuit which meant that, normally I would be turning left. I wondered if that was why it would not do a left turn on aileron alone. I then checked the CG. Prior to the maiden I had used my fingers under the wing to check the CG.  This time I used my CG checking stand and discovered it was 12mm behind where it should have been.  The only way to move it to the desired location was to add 100 grams of lead!  This I did even though I cringed at the thought of doing it.  The result was a CG which was now a little forward of where I thought it should be.  Would these changes do the trick?  It was hard to believe they would given the extreme nature of its behaviour. I was confident they would improve things but it just seemed inconceivable that they would result in my desire of a nice gentle Cub!


The Third Maiden Attempt

This attempt was make or break.  I had wasted enough time, energy and foam board on this thing.  If it didn't fly like a gentle angel then it was for the scrap heap.  Takeoff was uneventful.  There was no swing at all. No diving for the ground and no rolling to the left.   The only surprise was that I still had some up elevator applied when it left the ground and it pitched up at a steeper angle than I liked.  Not vertical but it made the takeoff look clumsy.  Subsequent takeoffs required no up elevator at all. In spite of the lead in the nose the Cub does not nose over on takeoff with neutral elevator.  In the air she flew like a nice, gentle Cub should.  My nightmare had transformed itself into the plane that I had initially expected. It now handled beautifully.  It would turn on aileron alone but needed a little rudder to balance the turn.  Applying rudder did not result in a vicious nose dive.  The only thing that went wrong happened when I tried a loop.  It flicked into an incipient spin at the top of the loop. However, with no correction from me, it recovered from the spin instantly.  Landing was a piece of cake.  Although I had to force myself to apply rudder to keep it straight because, emotionally, I still feared that nose dive.  I had video of the flight and when I watched the loop attempt I realised what I had done wrong.  I had been worried about having enough power now that I had increased the weight so much with the lead.  In attempting to complete the loop I had pulled back too hard on the elevator and made the loop way too tight causing the stall.  The next weekend I tried bigger loops with less elevator and found myself looping to my hearts content.  All I needed to do was a normal sized loop.  Finally I had a nice flying Carbon Cub.  The youtube video below shows the third maiden attempt.


The Design

As stated in the introduction the original goal was a nice flying Carbon Cub with that cool looking Carbon Cub nose.  In addition to that I designed some 3D printed wing struts.  For the undercarriage I designed the balloon tyres. I also hinged the 3D printed landing gear covers to a 3D printed mounting plate which was glued to the fuselage bottom.  2mm piano wire was bent and glued into slots that ran along the front of the landing gear covers and the belly plate.  I designed in attachment points for springs or rubber bands to further add to the suspension. I left them off because they were not needed.  If you are familiar with my previous designs then you know the wing section comes from the Bloody Baron.  Other than a forward CG this wing section is fantastic. Just don't let the CG fall behind the wing spar.


Specifications

  • Wingspan: 1112mm 

  • Length: 900mm (including spinner)

  • Flying Weight: 1257grams

  • Empty Weight: 1057gms (100 grams of lead)

  • Motor: Motor: Emax GT2215/09

  • ESC: Turnigy Plush 40 amp 

  • Prop: 10 x 4.7 slow flyer

  • Battery: 2200 mAh, 3S Turnigy Multistar

  • Max Watts: 212 watts 

  • Max Amps: 21 amps 

  • Max Thrust: 1056 grams 

  • Max C: 9.6

  • Flight Time at Full Throttle: 6 minutes 30 seconds


3D Printed Parts

There is quite a list of 3D printed parts.  The advantage of 3D printers is that aircraft like this can be modelled without leaving out their signature features or resorting to more traditional balsa construction.  Nothing wrong with balsa bashing but I prefer this way. The idea of creating the design electronically and then making it real is such a buzz for me.  I love designing and 3D printing things. The more challenging the better!  At any rate here are the 3D printed parts for my Carbon Cub.

Spinner

I didn't come up with the design for the spinner.  Its my go to spinner and its designed by Flitetest member Pintokitkat.  All I do is scale it to the size required and print it out.  I have used a number of these with sizes ranging from30mm to 45mm and have not had a single failure.  Pijntokitkat clearly did an excellent job.  I print them solid (i.e. 100% infill) for strength.

Control Horns

My standard control horns. I also print these solid.

Battery Tray

The battery tray is angled to keep the battery weight as far forward as possible. The slot at the bottom is to allow the motors wires passage to the ESC.  The horizontal piece just helps brace the tray so that it cannot wobble from side to side. This is to keep the tray firm during battery removal as Velcro can sometimes be quite stubborn.

Canopy/Front Deck

Printed in two parts for ease of printing.  Also printed solid.  CA is used to glue it together.  The tabs at the bottom with the holes are for the barbecue skewer to slide through.

Nose

The nose turned out to be quite a challenge.  This is my third attempt. Its still not totally accurate but its close enough. Can be printed with 10% infill and printing with a raft is recommended to ensure print bed adhesion.

Wing Struts

I needed to print these in two halves in order to fit them in the print area for my printer. To keep the aerofoil shape each piece is printed with a lower and upper which are then glued together.  Small pieces of 1mm piano wire are then used as joiners between the completed strut halves.  They must be printed solid to ensure as mush stiffness as possible.  If I wasn't trying to save weight I would have used thicker piano wire for the full length of each strut (say 1.5mm).

Piano wire stubs are also glued into the end points of each strut. These are then poked into the wing and fuselage to locate the struts and then glued in place.

Motor Mount

This version has zero degrees right thrust and two degrees down thrust which seems to be what is needed.  The Tee pieces are poked through the slots on each side of the motor mount and glued in place with CA.  They add surface area when gluing the motor mount to the fuselage.  Its important to print the motor mount solid to avoid the mount being crushed when tightening the motor mount bolts.

Tailskid

I employed my stock standard tail skid on the Cub.  The only change is I now print it in one piece instead of two. I have no idea why I didn't do it this way to begin with. Printed with 10% infill.

Main Undercarriage

The undercarriage is printed in five parts: the two leg covers, the belly plate and the two tee pieces.  The tee pieces served two purposes. The first to increase the contact area between the undercarriage and the fuselage and hence help distribute the landing shock loads. The second was to provide attachment points for springs or rubber bands which would add more suspension.  As stated previously the extra suspension was deemed unnecessary due to the stuffness of the piano wire.  Given the pounding it took on the first two maidens its stood up pretty well.  1mm piano wire is threaded through the hinge points on each side and bent over and glued at the ends to hold them in place.  The 2mm piano wire under carriage is epoxy glued into the slots on the front end of the side and belly plates. Don't skimp on the epoxy and don't get any glue on the hinge points.  The belly plate was printed solid and the rest with 10% infill.

Balloon Wheels

The hubs are printed in ABS (could use PLA) with 10% infill.  The tyres are printed in TPU with 5% infill.  The tyres take 3 hours each to print but fortunately TPU doesn't suffer from warping and lifting from the print bed.  They work fantastically with just the right amount of compression.  

Main Wing Attachment Brace

I have always been a bit worried about gluing a foam board wing onto the top of a foam board fuselage. If the paper peels away from the foam then the wing might come off. In the past I have simply cross hatched the area where the wing is glued to the fuselage or removed the paper. In a fit of paranoia I came up with this idea. This wing isn't ever coming off!  The U shaped piece is glued into the bottom of the main wing such that the horizontal piece is on the inside of the wing bottom with the side pieces poking out the wing bottom.  The wing is positioned onto the fuselage with the vertical pieces slotting through holes cut into the fuselage top.  Pins are positioned into the centre of each hole in the side of the brace and pushed through the fuselage side.  The stake piece is the positioned on the outside of the fuselage and pushed through the fuselage side. Once the hole has been created on each side the square rod is poked through one side and out the other ensuring the ends are flush with the fuselage side. Once correctly positioned it is glued in place.


The build

The following build guide assumes a basic familiarity with FT style builds. If you are new to FT building the watch one of Flitetest's build videos before proceeding. The build guide for the FT-Spitfire is probably the closest to this one.

The above image shows the fuselage parts all laid out ready for assembly. I like to paint and wing tape everything before assembly.

The first step is to glue one side of the fuselage to the fuselage bottom.  These are B folds.

The next step is to glue the other fuselage side.

The top of the fuselage is then glued to the fuselage sides.  Once in place the fuselage is flipped upside down and pressed against the work table to ensure a flat join. Note the top sits below the top of the fuselage sides (to allow for the thickness of the wing) so when pressing down care must be taken not to damage the top of the fuselage sides by keeping the rear fuselage over the edge of the work table.

The rear fuselage bottom is then glued to the fuselage sides. Again pressing it down on the work table while the glue sets results in a neat flat join.

The rear bulkhead is then glued in place.

The front fuselage bottom is then glued to the sides. The front bottom has a curve so it cannot be pressed flat against the work table while the glue is drying. Instead pressing down and rocking from front to rear while the glue is drying does the job.

The fuselage rear top deck is then glued in place.  Same process of pressing down on the work table to get a good join is employed.

Strips of wing tape are then applied to all the joins to cover them up.

The tee pieces for the belly plate are then glued in place with CA glue.

Same is done with the motor mount.

The top front deck and cockpit window pieces are glued together with CA glue. Note the use of tabs in the design of the front deck.  This is to help with alignment while the glue is drying as well as the overall strength of the join.  The holes in the tabs below the front deck should now be test fitted with the barbecue skewer. File or drill to get a tight but not too tight fit.  Cut and trim the barbecue skewer to the required length (around 100mm) and round each end to make inserting easier.

Now the wings are assembled. Again wing tape is applied prior to assembly.

The first step is to glue each of the spar halves together.

Then each spar half is glued to the bottom of each wing half.  Test fit first and make sure you have the spar halves the right way around.

Each wing half is then folded and glued in the normal manner.  As always allow a good five minutes for the glue to dry.

The wing halves are then glued together. The wing jig provided on the plans allows for around 3 degrees of dihedral on each side.  Slide the wing around a bit to prevent the wing join  bonding to the work table as the glue sets.

I added a bottom piece to the dihedral jig to stop it falling over. I found that happened a lot when using the traditional FT style jig.  Place the jig in line with the wing spar to get the correct amount of dihedral.

Fibreglass reinforced tape is then applied to the wing root join for extra strength.  The tape covering the hole for the servo leads is then cut away.

The slots on either side of the servo lead hole are for the wing attachment brace to poke through. The one on the left side of the image is too wide because of an error in my plans. The plans have since been corrected. A slot needs to be cut into the top of the wing to allow the brace to be put in place. The pins are pushed through the wing to mark the diagonal points for this hole.

A view of the top of the wing shows the pins poking through. The rectangular hole can now be marked out and cut away.

The above image shows the hole cut out.  Apologies for the blurry image.  What can I say? I am getting old and I didn't notice the image wasn't properly focused until too late.  Make a note of the orientation of the cut out piece. Later it needs to be put back in place and the best result will be if it is put back with the same orientation as when it was removed.  The next step is to glue the brace in place. I used UHU-Por but you could use epoxy.  After the glue is dry the cutout piece is positioned back in the hole and hot glue smeared over the join.  Finally fibreglass reinforced tape is used to cover the area.

The next step is a tad tricky to explain. Its quite straight forward to do just not easy to explain. Here goes. The wing is held down in place while a pin is positioned in the centre of the rectangular hole in the protruding tabs of the wing brace. Once correctly positioned the pin is pushed through the fuselage side and then removed. The 3D printed hole maker (I really could not think of a reasonable name for this but you get what I mean) is positioned with its point in the pin hole and its sides aligned to match the rectangular hole in the wing brace.  It also needs to be perpendicular, both vertically and horizontally, to the fuselage side.  When positioned correctly it is pushed through the fuselage side. Once this is done for one side, the cross piece is pushed through that side to hold the brace in place and the process repeated for the other side.

The next step is to remove the paper from the foam board where the wing will glue to the fuselage.

I used epoxy glue to glue the wing to the fuselage making sure the cross piece is correctly placed and glued to the wing attachment brace.  I think you will agree that this wing isn't coming off anytime soon.

The belly plate is positioned onto the fuselage bottom and the area of fuselage bottom covered by the brace is marked out.  The wing tape and paper is removed from this area.

The motor mount is glued in place using epoxy glue. Note the motor is bolted to the mount before gluing.  Also the servos are installed into the wing and fuselage at this time.

2mm piano wire is bent to shape using template provided in the plans. The landing gear leg covers are slotted into alignment with the belly plate and then 1mm piano wire is slid through the holes to hold them in place.  Finally the piano wire is glued to the belly plate and leg covers with epoxy. Don't skimp on the glue but make sure it stays clear of the hinge point between the belly plate and the leg covers.

The main undercarriage assembly is then glued to the fuselage. Lots of epoxy is used to ensure a good strong joint.

The battery holder is next glued into place.  The battery is needed to be there to ensure the holder is as far forward as it can go.  After the glue has dried the battery tray cross piece can be glued in place.

All ready to go or so I thought.  At this point the CG is 12mm too far backwards so after scaring the crap out of myself I had to pull it apart again.

The side thrust was removed by putting a spacer under one side of the motor.  Note that replacing the mount was not feasible.  Any attempt to do so will ruin the fuselage.  Remember how the wing isn't going to come off?  A spacer seemed the best solution even if the spinner isn't perfectly lined up with the nose. I have however corrected the 3D model for the motor mount so the spacer isn't required if you decide to build one. Similarly the nose model has been modified to allow for the removal of the side thrust.  Note the lead on top of the motor mount.  Another chunk of lead was glued to the bottom of the nose piece (seen in the bottom left corner of the image).  If building from Adams foam board getting the CG in the right place will be less of an issue as Adams foam board is lighter.  Whatever happens make sure the CG is in line with the front edge of the wing spar.


The Data

The pdf files for the plans can be downloaded here: pdfFiles.zip

The dxf files for the plans can be found here: dxfFiles.zip

The STL files for the 3D printed parts can be found here: STLFiles.zip

The Sketchup files for the 3D printed parts can be found here: SketchupFiles.zip


Conclusion

This project was intended to be a fairly simple and easy design and build. It turned out to be anything but.  I am glad I persevered and the result is a success.  All that struggling makes a successful result all the more satisfying. My Cub ultimately fulfilled the intended goals and a challenge is always good for the soul.  The final version flies great, lands easily and looks cool.  This one is a keeper.




COMMENTS

Jackson T on September 26, 2018
Well done! Good on you for pushing through the problems, as frightening as they must have been! I normally just give up when a design has issues' like that. Looks like the work well and truly paid off. Nice job!
Log In to reply
MikeRobey on September 26, 2018
Thank you and the work definitely paid off. Love this hobby!
Log In to reply
MiniacRC on September 26, 2018
Wonderful guide and 3D printed tech Mike! I've always loved the thoroughness of your videos. Now we have to get that Hurricane of yours in the FT warbird line!
Log In to reply
MikeRobey on September 26, 2018
No objections on my part! I keep thinking I should do a version 2 without the powerpod and with a more scale nose. I have learned so much about how to design things since the Hurricane. Glad you liked the vid and article.
Log In to reply
jrvalentin62 on May 9, 2020
Hello Mike. Unable to leave you comments on your YT channel (Flight line down under). Hope all is well. Joe.
Log In to reply

You need to log-in to comment on articles.


My Carbon Cub Design - A Lesson in Trial & Error