The AFB (Amazing FleaBike)

HomeBuiltAirplanes.com

Help Support HomeBuiltAirplanes.com:

nickec

Well-Known Member
Joined
Apr 22, 2010
Messages
224
Location
Los Angeles, CA; Houston, TX
An abbreviated, yet useful, test:

Connect three routed struts in a triangle. Connect three conventional solid wood struts of equal volume and lengths. Orient both on a bench and press down from one apex toward the opposite side. Compare the two for strength. Is one more easily broken? How does it break? Where does it break?

The scale of the test triangles should equal the largest triangle of the actual fuselage to be most informative.
 

Riggerrob

Well-Known Member
Joined
Sep 9, 2014
Messages
1,835
Location
Canada
Yes, that lower rear corner if the cockpit cut-out is very heavily-loaded.
An engineering analysis might suggest using a solid sheet of plywood in that corner. Alternately, you could glue doublets inside or outside of the cockpit will. Outside doublets will create little drag at Flying Flea airspeeds.
May I also suggest radiusing the inner corners to reduce stress-risers?
Finally, you may need to glue in a plywood seat back and bottom to transmit twisting loads around the cockpit. The seat-back bulkhead will need holes large enough for access to control cables, etc. but can be cut from remarkably thin plywood.

Finally, a big Thank You for developing a new method for scratch-building airframes. Your new method best utilizes the software and CNC routing skills of the young generation of “hackers” and “makers.” A day or two using CNC tools at their local “maker space” will save them hundreds of hours in their home shop.
You are headed in the right direction.
 

rtfm

Well-Known Member
Joined
Jan 3, 2008
Messages
3,414
Location
Brisbane, Australia
Is the plywood aviation grade ?
Hi,
It's better than aviation grade. So-called "aviation grade" plywood complies with BS1088 which is a British Marine Plywood Specification. It predates the Australian spec considerably. Most of the world manufactures and uses BS1088 marine ply. The Australian standard AS2272 is perhaps the toughest specification in the world and the plywood is just about the best that money can buy. It also is expensive. AS2272 being an Australian Standard has the law behind it, so ply to this spec quite frankly is the best one can buy. The FleaBike is made exclusively from AS2272 plywood.

Regards,
Duncan
 

rtfm

Well-Known Member
Joined
Jan 3, 2008
Messages
3,414
Location
Brisbane, Australia
That's not quite correct. Even with your "built in gusset", the gluing area isn't as much as a gusset glued on the side. But more importantly, gussets on the side load the glue joints in shear, which is what you want for a good glue joint. Your method loads the joints in tension, which is not good.
Incorrect. If I were relying only on the end-bonds, you would be correct. But the structural pieces are sandwiched between two pieces on ply. On the inside is the 1.5mm template, and on the outside, a full 3mm sheet. Those structural pieces aren't going anywhere...

Think of it this way: instead of having to cut a million little corner reinforcing blocks and gluing them in place (and good luck with getting the angles to mate perfectly) I simply include them in each structural piece. QED.

Duncan
 
Last edited:

rtfm

Well-Known Member
Joined
Jan 3, 2008
Messages
3,414
Location
Brisbane, Australia
Other than being cut from 3/4" ply, not solid wood, how is this method any different than a traditional stick, filler and ply joint?
The difference is in (1) precision of fit (2) build time. For example, I cut out the entire fuselage (ie both sides) in just over 4 and a half hours. That was yesterday's job. Today, I will glue them all together. Since this is my first plane using this method, I fully anticipate some hiccups and SNAFUs - but essentially we're talking hours, not days to build a side. And I'm pretty darn certain that my joints will be about as close as is possible to manufacture. Actually, to within a tenth of a millimeter accuracy.

By comparison, when I built my HM293 fuselage from plans, it took weeks of daily work in order to get each member to fit. Every length, every corner, every angle had to be measured, cut, carefully sanded to shape. And some had to be binned because I'd mis-measured, or over cut/sanded. Weeks of work, and then I had to cut every corner block by hand, and hope that each block fit was OK (a lot of them required a fair deal of epoxy filler).

Duncan
 
Last edited:

rtfm

Well-Known Member
Joined
Jan 3, 2008
Messages
3,414
Location
Brisbane, Australia
Yes, that lower rear corner if the cockpit cut-out is very heavily-loaded.
An engineering analysis might suggest using a solid sheet of plywood in that corner. Alternately, you could glue doublets inside or outside of the cockpit will. Outside doublets will create little drag at Flying Flea airspeeds.
Hi. The cabin area is reinforced with a 3mm ply doubler on the inside, as shown below.
AFB fuselage cabin doubler.jpg
May I also suggest radiusing the inner corners to reduce stress-risers?
Do you mean the base of the seat? Because every other corner is clearly radiused. But consider - the seat base and back are separate pieces, so no stress risers possible.

Finally, you may need to glue in a plywood seat back and bottom to transmit twisting loads around the cockpit. The seat-back bulkhead will need holes large enough for access to control cables, etc. but can be cut from remarkably thin plywood.
Actually, I'm going to be using 3mm ply bulkheads throughout bonded to every structural member. That's 12 bulkheads in total (all with suitable cut-outs, of course)

The sides are completely straight, tapering from 360mm at the firewall, to 130mm at the tailstock. So this is a very narrow construction. (And notably, can be built on any flat work surface) I will need all the anti-torsion help I can get, especially since the rear wing attaches to a very narrow part of the fuselage. In addition, I will also be bonding in laterals (ie between left/right), but my CAD skills aren't sufficiently developed for me to draw the fuselage in 3D, and accurately draw in these lateral pieces. This will need to be cut by hand, sanded and fitted. Then I can transfer them to CAD to automate the process for the next build (hopefully).

Finally, a big Thank You for developing a new method for scratch-building airframes. Your new method best utilizes the software and CNC routing skills of the young generation of “hackers” and “makers.” A day or two using CNC tools at their local “maker space” will save them hundreds of hours in their home shop.
You are headed in the right direction.
This is very kind of you to say. Thank you. The proof of the pudding is in the eating, of course, but I'm reasonably confident that this will all work as planned.

Duncan
 
Last edited:

rtfm

Well-Known Member
Joined
Jan 3, 2008
Messages
3,414
Location
Brisbane, Australia
An abbreviated, yet useful, test:

Connect three routed struts in a triangle. Connect three conventional solid wood struts of equal volume and lengths. Orient both on a bench and press down from one apex toward the opposite side. Compare the two for strength. Is one more easily broken? How does it break? Where does it break?

The scale of the test triangles should equal the largest triangle of the actual fuselage to be most informative.
Hi,
Nice idea. Thank you.
 

Dana

Super Moderator
Staff member
Joined
Apr 3, 2007
Messages
9,752
Location
CT, USA
Incorrect. If I were relying only on the end-bonds, you would be correct. But the structural pieces are sandwiched between two pieces on ply. On the inside is the 1.5mm template, and on the outside, a full 3mm sheet. Those structural pieces aren't going anywhere...

Think of it this way: instead of having to cut a million little corner reinforcing blocks and gluing them in place (and good luck with getting the angles to mate perfectly) I simply include them in each structural piece. QED.

Duncan
Ah... OK. I didn't realize you had a skinned structure. You kept saying "template", which usually means a disposable layout aid, not part of the structure. If you have plywood skin throughout on both sides, that's a lot better.

I still think you'd be better with solid wood truss elements, though, it'd be stronger and/or lighter, but what you have may well be "good enough" and certainly easier to buid.
 

Hot Wings

Grumpy Cynic
HBA Supporter
Log Member
Joined
Nov 14, 2009
Messages
7,579
Location
Rocky Mountains
The difference is in (1) precision of fit (2) build time.
I was just trying to point out that your method was similar structurally. Some of ;the others missed that you were skinning both sides.
I just presumed that the advantages of the fit and speed due to CNC cut parts was noticed by all.

I like the idea. It may weigh a bit more ;than conventional fit solid blocks.......but it is also more likely to get built/finished.
 

WonderousMountain

Well-Known Member
Joined
Apr 10, 2010
Messages
2,155
Location
Clatsop, Or
We can no longer keep up with Duncan's build process.
1) Draw frame on Etch-a-Sketch.

2) Pin down & cut Plywood bones.

3) Superglue-Epoxy profile-siding.

4) Post pictures of HBA gloating.
 

rotax618

Well-Known Member
Joined
Oct 31, 2005
Messages
1,074
Location
Evans Head Australia
The HM 14 was a sit in rather than sit on fuse, the one I built was insufficiently stiff in torsion, I had to double the top longerons to stiffen it up. Your sit on flea will have much less torsional strength, with two lifting surfaces at each end and a rudder applying a large torsional force to the fuselage, to my eye the seat area DLAR (doesn’t look about right).
 

pictsidhe

Well-Known Member
Joined
Jul 15, 2014
Messages
8,812
Location
North Carolina
CNC your sticks from hoop pine.
Plane some long stick stock.
Rough cut corner blocks.
Glue corner blocks on to rough cut sticks.
CNC blocked rough sticks to final shape.
Enjoy the kind of fit that takes a long time by traditional methods.
 

rtfm

Well-Known Member
Joined
Jan 3, 2008
Messages
3,414
Location
Brisbane, Australia
We can no longer keep up with Duncan's build process.
1) Draw frame on Etch-a-Sketch.

2) Pin down & cut Plywood bones.

3) Superglue-Epoxy profile-siding.

4) Post pictures of HBA gloating.
Hi,
Ha ha. I've had to obey she-who-must-be-obeyed, and tidy the garden today. But I did manage to cut the missing and/or buggered up parts. So all pieces are present and accounted for. I have a dance lesson tomorrow morning, but in the afternoon, I plan to actually start building. PLENTY of photos to follow, and all confusion will be dispelled, I promise.

Regards,
Duncan
 

rtfm

Well-Known Member
Joined
Jan 3, 2008
Messages
3,414
Location
Brisbane, Australia
The HM 14 was a sit in rather than sit on fuse, the one I built was insufficiently stiff in torsion, I had to double the top longerons to stiffen it up. Your sit on flea will have much less torsional strength, with two lifting surfaces at each end and a rudder applying a large torsional force to the fuselage, to my eye the seat area DLAR (doesn’t look about right).
Hi,
The seating proportions are taken from the HM293. Here are the two designs overlaid. The AFB has a more reclined sitting position, and a more raked windshield. I THINK I've got it right - but I'll soon know...
HM293 with TAFB overlay.png
 

rotax618

Well-Known Member
Joined
Oct 31, 2005
Messages
1,074
Location
Evans Head Australia
Note the complete lack of upper longerons. The lower front and rear seat support bulkheads will have to provide all of the torsional strength.
If you have considered this area of stress concentration, bending, torsion and shear, please ignore my comment.
 
Last edited:

rtfm

Well-Known Member
Joined
Jan 3, 2008
Messages
3,414
Location
Brisbane, Australia
Hi,
I'm with you. However, this is (like all things in aircraft design) a compromise, with "good enough" being good enough. I've taken all the precautions I can with the design, but ultimately, only time will tell. Bear in mind also that this is an extremely small and light airplane.

The thought comes to me, however, that it might just be possible to have a hinged member (red line in the drawing below) which can be locked into place once the pilot is in position. No idea off the top of my head if the fuse is wide enough to allow this... But I suspect this won't be possible.
[Edit]
I've just checked - nope - the fuse is far too narrow at that point. Bugger.

Duncan
AFB fuselage cabin doubler and hinged stress member.jpg
 
Last edited:
Top