Thread Tools
This thread is privately moderated by PeterVRC, who may elect to delete unwanted replies.
Aug 06, 2018, 11:29 PM
Registered User
Thread OP
Build Log

3D printed plane - No.2 - Larger


The GASB2 50mm Fanjet is just a first test of 3D printing techniques, and what results they give. To show SHORTFALLS in 3D printing.
I already know that a Single Perimeter surface has its 'costs'.
This larger "70mm" Fanjet is aimed to test out TWO perimeters..... 2x 0.4mm = 0.8mm.
Theoretically this means 'double' the weight of single perimeter.... but at least this design is almost totally 'empty' internally. eg no complex design/layout of ribs etc.

I quickly did a scaled up version of the 50mm.....
70mm / 50mm = 1.4 : 1
But no.....
70mm 6S = 1.6Kg thrust (in-plane), versus 50mm at 450g thrust in-plane..... That is 3.5 : 1.......
But no.... you won't make a plane 3.5x larger!

Next... the Funjet this is all based on has a 830mm wingspan for its 70mm Fanjet version.....
Hmmmm.... I want more lift.... so I settled on 1.8x in scale. A 1080mm wingspan.

I don't know what that will mean in the final flying result, but I am predicting the greater size and mass will make a more penetrating and 'linear' flying aircraft. It is not aimed at speed, just to fly well and be resilient!
Sign up now
to remove ads between posts
Aug 07, 2018, 01:10 AM
Registered User
Thread OP

What happens when you scale up a model....


When you scale up a 3D Model EVERYTHING gets larger (scaled up). That means servo pockets.... wiring tubes and channels.... joiner rods and spars....

It is still best to scale up the model, but then you need to EDIT and alter things back 'down' in scale to suit what is required in total.
Because this is a very messy mesh model I basically have to 'cut out' any things not right, and then replace them with new modeling of the correct dimensions.
I looked into rebuilding the entire wing(s) but they are quite detailed in shape and thus quite hard to do! I decided it it best to keep those, but edit the parts that I need to change.
That will require a fair amount of TIME.... but so would designing one from scratch anyway!

Before editing the parts, I exported them to object to 'Slice' and see what the weights would be. Once edited that will hardly change anyway.
The TOTAL plastic parts weight is 1.6Kg..... hmmmm....
Costing approx $15 in plastic.

The EDF, ESC, RX, Servos (2) and Battery would be something like 500g region..... so 2.1Kg AUW. Hmmmmm......

I have some 70mm jets heading from 1.6Kg and upwards a bit more.
They have less wing area and are "smaller" really. Well smaller wings (area).....
I think that at 1080mm Wingspan.... with a very long chord at the root..... and ending up a bit more typical chord by the tips..... it will be fine to carry 2.1Kg.
Though really, you would more typically aim for the 1.8Kg region.

Once again, this 3D Printed Plane will be operating in the "truly scale" regions! LOL.
"Heavy" and fly it like a full scale aircraft is flown - "properly".
Its main issue will be its LOW speed capabilities.... or lack of....
ANYTHING flies fine when FAST !!!

I have begun the editing, and whilst there are not that many things (items) to do, they are a bit messy and time consuming to do.
I have already added a full length Wing SPAR, which the 50mm does not have.
And that is one thing (aspect) I have decided MY scratch built 3D Printed Aircraft designs will have - a proper/adequate STRUCTURE system, which in most cases will mean carbon fiber spars where required. For EG, on a BIG plane that will have fuselage length spars, wing multiple spars, tailplane spar.... anything it needs to STAY intact 100% for sure!
Once you get to larger sizes they can carry a LOT more weight easily. So by the time you hit a "90mm" sized EDF jet you do have more leeway where LIFT will be able to surpass MASS (weight) issues. At this 70mm size it is still a marginal thing - to have strong structure (two perimeters) AND keep the weight to acceptable levels.
Acceptable TRUE TO SCALE levels... not "feather-weight", overly light, typical RC models type weights!

I might begin some modeling of the 90mm jet also, to roll out some parts weights QUICKLY - to get a clue of where that would head.
In a 1600mm span 90mm EDF jet (which is big for 90mm really), you could carry 5.0Kg or a bit more. My 1500mm EDF F-86 Sabre is 5.0Kg and flies totally fine - operating still "too light" really. It could probably fly fine at 6.0Kg anyway....

.....
Aug 07, 2018, 10:12 AM
Registered User
Thread OP

Test No.1.


I decided to do a 'test' of Slicing the 70mm Fanjet parts in both a Single Perimeter and a Two Perimeter version, to see what the weight differences are.
Two Perimeters does NOT mean the weight will be DOUBLE that of a Single perimeter version....

The Two Perimeter set came out at 1.7Kg and the Single perimeter set came out at a fraction over 1.0Kg
700g lighter..... which is almost all the weight of all the electronics/power additions. EDF, ESC, RX, Servos, Wiring..... and the Battery.....
Hmmmm...... that is a LOT of weight that could be saved.

So the next thing to test was PRINT a Single perimeter Wing piece - to see how that looks.......
7 hours later.....

For a start, the surface has 'collapses' in it. Imagine a typical FABRIC covered WW1 biplane's Wings..... where the fabric, doped to seal it, shrinks and pulls into the space between Ribs. That is what the Single Perimeter does.
This is not some 3D Printer setup created issue - it is what it will do, because the plastic is so thin and it has no support to stop it trying to 'shortcut' between two internal ribs (no infill).
The 50mm Fanjet did the same thing!! And its gaps between ribs is even less than this larger version!
On that model I had printed some parts at 200degC and others at a more typically used 230degC. The hotter temperature method causes it a bit more - as you might expect from hotter plastic.
I had also printed some with the Cooling Fan ON, and others with it OFF. When it is ON it will cool the plastic very fast, and then the plastic has no time to 'move' as it cools. It makes for very precise placements...... except.... even with it ON there were still depressions across those Wing pieces anyway.
Thus, there is no way left to prevent it....

I am yet to print a Two Perimeter version to see how that goes...... but that will be next.

You could just IGNORE this 'Wing Depression' resultant and it is still going to produce almost as much lift as it would if were a 'perfect' surface. So it is not a fatal issue.
But in VISUAL terms.... it looks pretty blah.
On aircraft that would have 'level' surfaces - or at least a much, much, lesser 'wave' in any surfaces - it just looks bad.
No amount of ribs will stop it occurring.... more ribs would just make it occur less in depth, and more often. More wavy overall - more waves, of less amplitude.

I have to wait to see if TWO perimeters makes its extra WEIGHT worth it...... maybe it will still not be a 'perfect' surface anyway.

This above issue has no link to how a Single Perimeter surface will be more affected by HEAT in sunlight.... so that is the bigger issue it will have versus Two Perimeters.

....
Aug 07, 2018, 09:12 PM
Registered User
Thread OP

Test Wing piece


It is a little bit hard to see in pictures, but the depressions in the wing are NOTABLE.... too much. Too much to be acceptable.
The internal 'ribs' are also weak and deformed.
In the 50mm version they were already a little bit like that, but acceptable..... at this larger size it is useless.

I don't think that even two perimeters could save that from occurring. It stems from a deeper problem than just that Single Perimeter thin surface.
Though it is also possibly linked to NOT COOLING the plastic right away that it is laid down, with the Cooling Fan - seeing that is OFF. But it is still not ONLY from that....

I am going to test print a 3Dlabsprint Mig-15 Wing section to check how their internal ribbing system works. I suspect I will have to design and add that same form of ribbing/support into the Fanjet wing. You then use ZERO percent Infill and all internal structures are created in the 3D Model - not by the Slicer adding "stuff" that just happens to be useful. Which in this case what the Slicer adds is NOT truly useful.
I sort of like/prefer the idea that YOU add what YOU want and NEED, specifically. Then you know it was done right/best, with no need for 'luck' of what the Slicer 'maths' happens to maybe make of use.

....
Last edited by PeterVRC; Aug 07, 2018 at 09:19 PM.
Aug 09, 2018, 08:30 AM
Registered User
Thread OP

Next test....


To learn a bit more, I downloaded some of the 3DlabPrint aircraft test pieces - with specific interest in the Mig-15 90mm EDF jet.
I printed that wing piece and saw what they were doing, in construction terms, to make the wing surface maintain its 'curve' almost perfectly. It was a simple enough idea of a set of internal ribs, that are set at a 45 degree diagonal (not square to the wing), with a 40mm spacing between each rib.
As it turns out, that wing piece is 'excellent'.... EXCEPT, it has slight 'flattening' (short-cutting) of the surface of ONE grid area. That is not an issue at all... almost a zero.

I decided that I would need to re-design the GASB2 70mm Wing internals to have user built ribs - not the "Slicer created" simplistic versions the designer had used in the 50mm version. Even the 50mm version would benefit a LOT from doing these ribs for it.
But HOW to make these ribs in 3D modeling!!!????
I pondered.... and pondered.... but could not work out HOW to do that 'easily' with some 3DSmax inbuilt tools.....

Googling did not help..... until I bumped in to a thread where "Localfiend" was outlining how he made his 3D Printed aircraft, and a quick message to him got a link to the information to do it all ! So that was a life saving bit of help from "Localfiend"!! (Thanks!!)
Well..... except it used a different editing program, with a tool not exactly available in 3DSmax.... but the WAY it worked (what the tool had done) told me what other tools to use to do it all anyway! It was so simple......

When you build stuff internally into a 3D Printed object, which you want 'hollow', you can't design the internals to actually TOUCH.... join... the outer walls. They have to 'stop' 0.4mm before the walls/surfaces. This is because the Single Perimeter you will be using will cause that outer wall/surface to be the extrusion WIDTH more inwards - and that is 0.4mm in this case.
If the internal object is too close, or touches, the outer wall/surface, then the Slicer will create a totally different way they 'join' - a totally useless end result!
This means when you make the Ribs, they need to be 0.4mm smaller than the surfaces they are intended to touch once printed.

To make something be 0.4mm away from a COMPLEX shape... curves, tapers etc.... the best way is to make a SHELL of that same main object that is 0.4mm inside the real object. That will TRACK the outer walls to create that 0.4mm smaller object. Unfortunately the Shell Tool in 3DSmax has limitations in how 'smart' it is, and some trickier areas cause it to make errors there. But in this case those were easy enough to manually fix......

So once you have the smaller object (an inner wing), you can create a grid of planes that are 0.1mm thick. This is the smallest thickness that can be printed (with a 0.4mm nozzle) - but you could make these 0.15mm.... 0.2mm.... if you wanted more strength. But in testing the Mig wing piece, that was strong enough at 0.1mm internal ribs etc.
You set up a GRID of the planes, with that grid set to 45deg. I actually aligned it to have ribs head into the corner 'points' on the wing, so that was some odd value not far off 45deg.
Also, have that grid fully 'enclose' the inner wing object.

Now 'Cookie Cut' it with a "Boolean Object" and that will leave ONLY the plane portions that were INSIDE the inner wing object. And that means they are now 0.1mm RIBS!! Task done !!
Go back to the real wing piece (not the smaller inner one)... show both that wing piece and the ribs set.... and you will see those nice ribs in place within the wing. Then join those two objects to be one object with those two parts in it (outer wing and ribs).

Oh... I made mine with 30mm spacing, to improve the support that will reduce any attempts of the surface to sag in. This sagging is what will occur when HEAT softens the plastic, so it will help fend off the issues of heat..... at least better than 40mm spacing would.

I just need to add in a SPAR.... some wing joiner 'lugs'.... the servo bays - which will be modular so they can be changed to other servo shapes easily... and wiring tubes....
All those things are quite straight forwards to design and add - no real tricks needed to add those.

I did a quick Simplify3D test 'Slice' of the wing outer model and it looked good. Everything worked as planned.
Note that the weight and plastic/cost shown in the pics (top left) is not right as the final wing will have a LOT of lightening holes put into all of the ribs! Once all the internal bits are added (servo bays etc again) then the remaining 'complete' rib areas can be edited for their lightening holes. But you don't know where those need to go until all the rest is done. Though I could do a test of adding holes to them ALL, and THEN add the extra parts and see what the Sliced end result looks like. It could very well be perfectly fine to do it that way and that means adding holes FIRST is a far simpler automated process to do those ALL, than having to choose exactly where to put them (and not others) after. So I will do that 'holes first' test too.....

....
Last edited by PeterVRC; Aug 09, 2018 at 08:46 AM.
Aug 22, 2018, 06:43 PM
Registered User
Thread OP

3D printing issues....


I have been unable to "assemble" all the required 3D Model pieces into a 3D Printable outcome, so far.....

The strategy seems simple.

1) Build a Wing SHAPE - you only need the Outer Skin of it all.

2) Make an Inner Shell of that Wing, to use as the template for a Rib set that will be 0.4mm smaller than the Wing Skin. Remembering that for it to 3D print correctly the internals 3D MODEL need to stay 0.4mm away from the Wing Skin, as the plastic Extrusion thickness will bring those to match and bond when Printed.

3) Make a Grid of flat Planes, to "Cookie Cut" with the Wing Inner Shell, so then you have a set of diagonal Ribs that are that 0.4mm away from contact in the 3D MODEL.

4) Make a set of "Holes" - which are really solid objects that will again "Cookie Cut" their shapes into the Ribs, thus forming holes.

5) Make up various required internal items, such as tubes for Spars and Joiners, and also "Support Shapes" that will allow objects that would otherwise be 'hanging in space' (and thus not turn out) as the 3D printing layers build upwards.

Now, the required steps SOUND simple enough.....
The issue is that ACHIEVING them, in 3DSMax, has been full of shortfalls and problematic results!

Almost every STEP works out fine on its own. But you can't viably combine them all.
Cookie Cutting Ribs... fine,
Cookie Cut their holes.... fine.
But add the 'tubes' (Spars, Joiners) and those do not "join" to the Ribs properly - but they do IF you do that before the holes - but then the Holes can't 'Cut' properly.
By trying various 'tricks' of manipulating the 3D objects, in many ways, I can ALMOST form the End Result, that WORKS.... but lots of ways LOOK fine, but won't work when put into the 3D printing "Slicer" program - it builds 'flawed' structures, in various ways per the various ways you 'trick' 3DSMax to achieve the listed goals. So NO WAY works....
It is a bit grey and tricky to 'write' the issues and how they unfold (it would take a LOT of paragraphs!). I pretty well know WHAT they are not doing right, but can't find any way to FIX or prevent that.

So I have all these nice pieces..... that look right, nice, and can even be made to form a good LOOKING wing model. But you can't 3D print it without it having flaws that make it unworkable!
So I am still puzzling over WHY.... trying to ferret out the parts of the 3D MODEL that the SLICER does not like or cope with.

The real underlying issue is the COOKIE CUTTING. They do that task and it LOOKS fine in 3DSMax, and the result would even be totally workable in most uses of that Model. But the real underlying geometry (?) they produce is 'flawed' and then the Slicer program has a bit of a fit with that and thus it produces a 'flawed' printing outline/map.
If you MANUALLY made all the Ribs, and their holes, then it would be fine - but that is an all but impossible task of complexity. You NEED to be able to 'Cookie Cut' and make that task 1000 times easier and quicker. But if you use their object 'splicing/cutting' tools, you get these flaws - which will most likely be FIXABLE by manual digging and editing, but that too is a huge task then!

So I am still investigating WHAT and HOW to get through this issue.....

....
Last edited by PeterVRC; Aug 22, 2018 at 06:59 PM.


Quick Reply
Message:
Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Question Plans for a larger full fuse 3d plane tanner331 3D Flying 1 Jan 03, 2017 03:43 PM
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