Thread Tools
This thread is privately moderated by PeterVRC, who may elect to delete unwanted replies.
Aug 06, 2018, 05:47 PM
Registered User
Thread OP
Build Log

3D printed plane - the "GASB2 Fanjet"


It was TIME to 3D print an RC Model aircraft......

I had been using 3D printers for many years and had seen bits and pieces (snippets) of how the world of 3D printing Aircraft was going. Nothing looked too great or promising in terms of producing a DECENT RC Model that would have everything doe well, and LAST. They were being designed and made 'too marginal', in their aims to save as much weight as possible. Let alone the 'entire modeling world' being so scared of WEIGHT and the ROBUSTNESS required to produce a truly "Fit for purpose" aircraft. The vast majority are 'marginal' and the weight phobia abounding everywhere means they fly like 'feathers' or 'bits of paper int he wind', not truly 'linear with mass and inertias as per full scale aircraft have and do.
So the WEIGHT of a 3D Printed aircraft was not worrying me, it was the actual construction details that would be needed to form a "Fit for purpose" aircraft, and THEN remain within a weight range restriction - even when that is WAY heavier than everyone else is doing in 3D printed aircraft.
In foam models, they are all way too light as they come.... well... in recent years the manufacturers ARE FINALLY putting more 'good stuff' into them, and not worrying about weight as much. Yet they are still not at a point of truly Fit For Purpose and robust etc. But you can buy a foam model (Kit or PNF) and ADD all that you need to make it fully Fit For Purpose (reliable, strong, lasting!), and still not end up too heavy to fly well and properly.

What range of weight headroom does/would a 3D printed model have?

From the numbers (AUW's) I was seeing, they had ROOM (weight room) to be done properly. But they are higher/further along the weight scale than foam planes, thus the total aircraft needs more PLANNING. In foam models you can all but 'do what you want' and reaching a 'too heavy' point is not going to happen. Well, at least once you cross 1200mm wingspan prop aircraft size, or 90mm EDF jet size.
Due to Reynolds Numbers (how air Molecules remain the SAME size for a very small aircraft, or a full scale aircraft) the LIFT a wing can generate rises 'exponentially'per wing area. A 1200mm warbird can carry about 2Kg and fly 'properly', but a 12,000mm full scale warbird of that same aircraft can carry 3000Kg.... even more!!! 10x wing area.... 1500 times the weight carried!! So you don't need that much of a model aircraft size increase, to carry a LOT more weight.
At that base 'material level' - of an airframe, with the strength to cope with its needs, and room for equipment etc - the 3D Printed aircraft version is a heavier item than a foam version will be.

So this is where a 3D printed plane can regain viability.... make a LARGER one.

Someone pointed out the GABS Two Fanjet on Thingverse, so I went and had a look at it.
What??? 600mm wingspan..... 50mm EDF..... hmmm, that is way down in the "Low Lift per AUW ratio" scale!! You really want something at least 1200mm, if not 1400mm or so. To buy you SPACE, for equipment and construction strength.... and weight carrying ability so it CAN fly well with all that stuff done!!.

But another useful thing to look at is TESTING 3D printing aircraft.... having low cost.... 'quick' to do.... and more so as I knew I had all the PARTS to suit that type of 'jet' here already! Thus I looked very closely into it, to see if it was a good choice. Versus just designing my own.

TIME..... Plans ready to go. So a LOT of time is saved.....

Design 'intelligence' - it was very well planned, designed, and 'created'. With a few small shortfalls that could have been done better, but just as it comes it is 'very well done'.

Cheap.... not much plastic (filament).... all the parts I already have from older crashed aircraft....

And his RESULTS of it flying were good too. Plus the "Maths" of all the aspects of it added up and made sense, that it would work viably.
So I got on my way !!!!

....
Sign up now
to remove ads between posts
Aug 06, 2018, 05:55 PM
Registered User
Thread OP

The Design


There are a few main ways to design a 3D printed aircraft. One of the main constants used is "Thin Wall Printing".

When you make something with a 3D Printer it can only lay down LAYERS of plastic. So anything hollow, or with 'arches' etc, "Can't" be printed as the layers above have nothing under them to build upon. eg the top of an arch, 15mm above the 'ground'.
But whilst that seems a big shortfall/issue, there are some great ways around that......

The next aspect, linked to that above issue, is INSIDE a model. Is it hollow?? Is it fully solid? What things do you need 'inside' and how are they going to be formed and supported?
The easy way is using "Infill", which is where the inside if filled with a certain amount of plastic. Make it a honeycomb shape so it is strong, yet not 100% plastic which would be heavy and cost more. You can set a % of area to be Filled, say like 20%, and then the internal filling shape (whether straight lines, circles, honeycomb) will be 20% plastic and 80% "air". You can make a strong object even with as little as 5% filling. It comes down to specifics in the design - what it will really need.

But in 3D printed model aircraft, where weight limits CAN be reached too easily, you can;t afford to "fill" the model. So they tend to be made as an empty SHELL with just SOME internal stuff, as required to it to be workable. Bays for a battery.... servo pockets etc.

Which brings about the main aspect of printing a model aircraft - the "Walls" of the printing.
Each layer printed will do a 'perimeter' and then whatever is needed inside of that. The Perimeter can be one run of extrusion (plastic fed out as a 'noodle'), or many. Typically you would use 2 or 3 or 4, to give the outer surface more strength.
But when weight matters - like for this task - you have to compromise. So most designers are using ONE perimeter wall. Single Perimeter. This single 'line' of extrusion typically being 0.4mm. So that is the thickness of the outer surface of the aircraft.
That is not much!
Not much when 3D printed plastic is somewhat weak versus other materials.
It is actually quite strong, yet not so great when in large flat areas - like this will need.
Full scale aircraft use THIN aluminium alloy sheet mainly, so that is THIN...... but it is also a much stronger material really.

The biggest issue for 3D printed PLASTIC aircraft is "melting" !!!
The most used, suitable, plastic is PLA. It 'plasticizes' at about 60deg C, which might seem a lot hotter than the typical environment we will operate in. But it is NOT.
When the temperature outside is "25deg", that is in the SHADE. That is the AIR temperature. The general air temperature around. But in SUNLIGHT it is higher, and that amount higher is relative the to the ABSORPTION of the surface, which is largely controlled by its colour - or to be more specific, it REFLECTIVENESS of the sunlight. White reflects more than Black..... thus why darker things (like cars etc) in the sunlight get HOTTER than lighter things.
Then also, INSIDE something, where the heat is trapped and somewhat 'compounds' can get much hotter.
So 25degC AIR temperature can make a plastic PLA model not that far away from 'plasticizing', which means it gets soft and 'floppy' then.
This is a BIG issue that cannot be overcome....

The next issue is STRENGTH.
PLA is quite 'hard', which is sort of good for strength. But something 'hard' is typically 'brittle'. So when it breaks it 'snaps'... not just bends.
Aluminium and alloys will mainly bend. But this 3D printing is going to make something that will snap, crack. Once its strength limit is reached it will go 'bang' with that crack, split, snap.
But as long as something is "Stronger than forces it will encounter" that point of breaking won;t be reached.... so......

Finish.
The finish on a 3D printed surface can be quite good..... even very good. But it won't be matching an Injection Molded surface like on your lunch box, Tupperware container, etc. So it will at least b a little bit 'rough'
Leave it raw and you will only get the colours possible that filament comes in.
Paint it, and the 'darkness' of areas will affect how hot they could get.
So the finish on a 3D printed model aircraft leaves a bit (lot?) to be desired. Though from a bit of distance (2 metres?) it is probably 'good enough'.

Some models are designed with 'frames' internally. Some are not.
This GABS2 Fanjet uses a simplistic, yet quite intelligent, method of no frames - by using some of the inherent results that a SLICER Program will, or can, include in a model. A "Slicer" being the program that takes a 3D CAD Model and forming the plans/paths of the LAYERS that the 3D printer will need to do to form that model in the real world.
Those ideas he used were quite impressive, so I liked that ingenuity he used.
It all just added MORE ideas into my 'pool' of 3D Printing techniques - especially for model aircraft printing!


.....
Last edited by PeterVRC; Aug 06, 2018 at 06:29 PM.
Aug 06, 2018, 07:04 PM
Registered User
Thread OP

Design changes


When I looked over the Fanjet design I was very impressed with all the specific details of what and how he had designed it! Left just as it comes it COULD have been great..... but it was not quite.....

It was designed to use a specific 50mm EDF unit, so the dimension of the mounting for that was SPECIFIC - and there is no way to alter that!
But I had a different 50mm EDF unit.....

As a 3D model, overall this was a great model! But often someone will design a model in stages, and have 'lower level' modeling which is then 'processed' to be better in its final form. But the processed model is 'messy' - yet a great model result - which makes it all but un-editable later. So whilst the designer might have the lower level model versions that CAN be edited easily, the end result model can be unviable to edit.
This is what happened in this model.....

This meant that I could not EDIT his model to suit my EDF unit. Not without some major 3D Editing efforts. Sigh.
But I decided to do that anyway - plus I had an extra plan/project.....
The plan was to build a BIGGER version next. One that is more structurally and AUW viable. A 70mm version almost twice the dimensions.

You can see in the pics below how nice the model has been made/formed. Then also what MESH was created to be able to output that nice model. A very complex mesh! It is almost certainly a model that had a 'normal' much cleaner mesh and then was 'turbo smoothed' by software to form a more detailed and smooth model. This 'messy honeycomb' look of a mesh is typically what a turbosmooth type process will create. A designed would never create that 'manually'.
This meant I had to 'cut off' sections I did not want, then totally recreate them myself to suit what I wanted/needed, then re-join them to his quite complex model.
The original design had he EDF a a 'flowing' part of its fuselage/wing structure. I could have made mine the same, but to be 'universally flexible' for future modifications to suit ANY EDF unit type, I decided to make it as a "Pod" - where the EDF nacelle unit is raised a bit above the fuselage/wing. And this was important for future SCALING UP of the design, again so that ANY EDF could easily be edited/tailored for.

....
Aug 06, 2018, 07:15 PM
Registered User
Thread OP

New EDF Nacelle/Pod


I had to cut out a large area of the 'top' of the original fuselage/wing and nacelle piece, so that I could replace it all with my own design parts. A large area was cut out to leave a big 'hole' there. Then I built that up to be a 'stand' the the 'Pod' (nacelle) that would sit on top of that.
The nacelle was built from scratch (Using a Spline that is then "Lathed" to form the round nacelle) to suit my 50mm EDF, but it used all the exact same design idea of the original because they were all done in a great manner!

Note how my Mesh is CLEAN..... no Turbosmooth.... it does not even need that at all anyway.
At this Clean level, cutting the top cover off is easy, as it just follows 'straight' lines. Doing dimensions specific work is easy too.

So the Fanjet now had a Nacelle "Pod"......
I had to leave the ahead area "depressed" as that would have been a lot more work to remove/edit, but also it is quite fine to be there as it is anyway. It sort of looks cool, but will also allow better inlet airflow too.

....
Aug 06, 2018, 07:29 PM
Registered User
Thread OP

All the pieces


So I pressed on to Print all the pieces.....

Getting back to the Single Perimeter and INTERNAL requirements.
The designer opted to use a VERY SMALL Infill percentage - like 1%, 2% - to cause the Slicer to add in 'ribs' to the design. If you use a very small infill percentage, and of Straight Lines (Rectilinear), then you will only get a 'few ribs' internally.
A slicer places these 'ribs' like a grid pattern. If you have more % then there are more ribs closer together. Get the percentage 'right' and you can have just ONE rib, set to be at the middle of the piece.
If you move the Model on the Slicers 'bed' (on screen) you can alter where the rib occurs. So you can set it to be under a required support area etc.
It is a quite ingenius use of the 'results of the maths' of the Slicer - to form structures required, without having to have made those internally in the 3D Model itself, by yourself. it is quite limited, but with good THOUGHT you can make great use of that too!

PLA printing forms the strongest bond when it is printed at around 220-230degC.... so that is what you do to get the strongest possible Thin Wall of the single perimeter, 0.4mm, surfaces.
YOU also use NO cooling fan (turned off in the software), to cool the plastic as soon as it is printed, so that it can form that best and strongest bond to the layer under it.

All the printed pieces were exact sizes and fits of 'holes' for tubing used, which was somewhat surprising - such accurate tolerances, which is partly linked tot he AWESOME Tevo Tornado printer too!! Plus the surfaces it forms are extremely smooth and great quality! Some of that is due to the high heat level used, which allows good flow of plastic to 'meld' together smoothly, but you still also need an ACCURATE printer as a part of that.
Any 3D Printer will/can do it, but it is a matter of HOW GOOD that surface will be!
Note that I had a picture earlier of the designers RED Fanjet..... and that is pretty poor 3D Print quality. Partly due to lower heat used, and partly due to it being a less 'detailed, accurate, printer. Probably more like some I3 Prussa clone (as per my Geeetech I3 was).

....
Aug 06, 2018, 07:41 PM
Registered User
Thread OP

Assembling a 3D printed Aircraft


Hmmmm..... how come almost "EVERYONE" says that CA adhesive is great, best, for assembling 3D plastic parts!!!?????
Or for any uses....

It is just RUBBISH. One of the worst glues there is!!
It is not fast to cure, like they say..... even with Accelerator it is still so-so.
But it has one GIANT issue that makes it the WORST glue to use!!
It is BRITTLE!!!

CA forms a very brittle crystalline structure. Like with glass, as long as you use it where it is SUITED it will last and not break. But when it breaks(!!) it BREAKS!! With a bang'... crack.. snap....
It has no 'softness' to absorb shocks. Exactly as per glass.
This is what makes it USELESS for joining a 3D printed Aircraft together!!

The best glue to use is EPOXY!!! Hands down!

But, I got sucked into the hype of "CA is best for this" and I bought some more... and some more (different) Accelerator.... and used it for a bunch of the Fanjet parts joining....

DOH!!! Of course it was the same rubbish as per always!!!
You just pick up the plane and it 'cracks' as the forces put pressures on the CA joints. It doesn't break apart, it just makes cracking noises! If you hear a crack then SOMETHING has happened!! Molecular bonds have been BROKEN! Not all... just some.... but that is not good!
Every time you handle the plane it makes some cracking noises! Some more molecular bonds are broken!

So the remaining half of the plane I did using Epoxy!!
Epoxy also FILLS gaps quite good - way better than even Thick CA does.
And then I ran over all the CA joints with Epoxy 'rammed' into the seams and also smoothed over them. To help 'fix' those poor joints.

No more cracking noises!!

There is ONE benefit of CA... well, a potential one....
It WILL BREAK in a crash and thus allow the parts to separate and survive better. But that is of little use if the glue COULD BREAK under even less stress than a crash - like IN FLIGHT!! And that is certainly possible!

Thus it is just not truly Fit For Purpose!!!
Epoxy is!!!
Aug 06, 2018, 07:57 PM
Registered User
Thread OP

Parts.... Wiring....


Part of the 'great design' was the designer included 'tubes' and channels for all wiring runs through the aircraft!
They all 'formed' correctly when printed - though they were marginal in size, as there is not much room to have larger more spacious tubes etc.
This is where a LARGER plane will be able to do all this sort of stuff with ease, in such a large internal volume available.

My 11 blade 50mm 'something brand' EDF fit perfectly into my edited design.... though the room in the cockpit area for the ESC, RX and battery is a tight squeeze. But it is tight in this sized foam type aircraft too!

Again, the SMALL size of this plane made all of that quite marginal.
BUT, it did all fit..... and the end AUW was 600g. A fraction over the designers number of 577g.

I have a 800mm Wingspan BF-109 that is 640g with FPV fitted..... and that actually has LESS Wing Area than this Fanjet. (longer narrow chord tapered wings). And the BF-109 flies fine,
This Fanjet actually has a LOT of Wing Area for its span! So 600g will be no issue for it at all.

Though when I pick it up it FEELS HEAVY!! It makes me feel/think "Will this even fly!!!????".
Many of my seemingly 'overly heavy' models make me think that... and then they fly totally fine anyway! "Air" has a lot more substance, molecules, than it 'seems', and thus can carry more weight than you might imagine.

Air is basically the exact same thing as Water - a substance all around and flowing... made of molecules that are 'solid' items in effect. We are just 'wading', moving, through it..... and aircraft do too..... with 'vanes' that guide/align travel through it.....

So anyway, 600g seems fine really.

I also do not like the 'feel' of the "glued together plastic pieces" which don't form one solid structure like a foam molded aircraft has. So it makes me feel like it could break apart on a harsh landing, and more so in a crash! And it possibly - probably? - will.
In a way it is far more like FULL SCALE Aircraft than a foam one is! So it will ACT like a full scale if it is treated harsh, or crashes!! That is not optimal for a model to be longer lasting, but I guess the aim is NOT to treat it harsh or crash it !!

....
Aug 06, 2018, 08:03 PM
Registered User
Thread OP

One downside of Single Perimeter


When the pieces printed, the Single Perimeter, 'surface', being just 0.4mm, meant that it is somewhat like tissue paper rather than alloy sheet. It can 'sag' a fraction. So the surface is not perfectly formed. BUT.... either are the surfaces of Full Scale Aircraft with their thin alloy sheeting!
So whilst foam will typically have a 'perfect' surface, and looks great(!), the Single Perimeter 3D printed surface is not perfect. I guess you could loosely say 'it is more true to scale'...... hmmmm.....

In a larger model, which has that exponentially greater lift level, you will be able t use a thicker 'skin' Two, maybe even three, primeters. 0.8mm.... 1.2mm thick. That will also buy it more resilience to HEAT.... "melting"!!

That is the purpose of the next test..... Fanjet 70mm!!! With TWO perimeters, and more internal structures. To lead the way for a scratch built design of a 90mm jet - using far more structural design than any current 3D Printed plans/models around are using.
I have not decided on what jet that will be.... but it will be a accurate scale military jet of some kind. Most likely something I do not currently have one of.... and at this moment a T-33 is looking high on that list! The same as the RC lander size.... 1600mm wing span, 10S 90mm..... seeing I have all that stuff from my crashed RC lander T-33 anyway!!!


Quick Reply
Message:
Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion Sizing up planes (Gasb2 in this case) and expected max weight. charentejohn 3D Printed Planes 4 Apr 28, 2018 05:28 AM
Build Log 3D Models of Planes for 3D Printing (Polikarpov I-185) LuftwaffeSS Electric Warbirds 33 Oct 20, 2016 11:50 PM
Discussion 3d printed plane vincent123 Balsa Builders 7 Feb 23, 2016 12:27 AM
Build Log 9D- the 3D printed 3D plane sqrt18 3D Flying 17 Oct 20, 2013 01:04 AM
Discussion 3D printing: The world's first printed plane Louied Slope 2 Aug 02, 2011 02:19 PM