Firefox MIG-31 movie aircraft

Hi Robert,

I did have a look at the small nozzles ages ago and couldn't find a match, closest visually was the F-16 but I was looking for size matches at 1/24th or 1/48th, Just did a quick search to jog my memory and saw one I hadn't considered - the F-5, it does look like it could be right size at 1/48th scale, but can't see a variant with petal markings.

Don't put a lot of time into finding the parts as I will have to recreate them anyway since it's unlikely there will be a scale match for them that will work at 1/24 or 1/48.

Jon
 
I think I have worked out what the small nozzles are, the obvious choice for a fast plane - SR-71 Blackbird at 1/72nd scale.

Realised I had a Italeri/Testors 1/48th scale from years ago and providing the scale of that is right it looks correct.
 
What material are the large scale planes built with now? As I think I said before I could do cross-section slice drawings, couldn't guarantee airworthiness as they did seem to make some changes to the RC versions they built for filming!

Meant to ask this a while ago, there are some what I think must be kit parts around the lower engine section, does anyone recognise them?View attachment 697372
The width denoted by the red line is 90mm or 3.5 inches. There are 2 parts on the inner top surface, look like half a spark plug to me, but might be 2 parts put together. Then one on the inside edge and bottom middle which looks vaguely familiar, maybe on the Millennium Falcon? Then another part top right and bottom left which makes me think of a fuel tank with straps?

It's not a big deal as this is 1/12th and it's unlikely the parts are available in the smaller scales, will just have to model something similar.

I would be using 1/8"-3/16" ply for the formers, foam in between them, to spot on, i would use a few thin coats of resin to seal it all, sand it smooth, and then make a fiberglass mold from it to pull from, then most likely foam wings and verticals with wood ribs for added strength, thin ply skins, super light coat of resin to seal it all and 3/16' ply for the internal formers and resin coat them for strength and fuel resistance lol :)
 
Just a quick update, getting my Formlabs clone up and running has proven to be a bit more complicated than I thought. The resin trays have a PDMS (clear silicone) layer that seems to be not the greatest solution to the need to have a "forgiving" starting surface for the prints.
I think with mine this was exacerbated by possible glue contamination from the perspex being repaired. Had to do a thorough clean out of the tray and discovered the silicone layer was breaking down. So after an evening of web researching found a thread on the B9 creators forum (another SLA printer) talking about alternative materials to the one used as it seems quite UV reactive which causes the silicone to fog, bit of a problem for a printer using a UV laser?!?
They suggested a couple of silicones, one I had found as it's sold locally by a supplier of other resins, and one which seemed a better formula but only sold from Australia. Pricing on this was sensible, shipping not. Bit more hunting and found what looked like the same stuff sold by a company in Ireland. Price was sensible, so ordered some. First attempt was a failure as without thinking poured the two parts into a tall mixing cup which caused bubbles, and due to the fast setting time they didn't rise to the top and pop.
Did it again this evening, this time - careful slow pour down the side of the cup and gentle stirring. Result - no bubbles at all, and amazingly clear, but it is specifically designed as a special effects replacement for glass/ice.

So hopefully will have high res master parts in a couple of days, stay tuned!
 
Hi jon just saw yourpage here am new here ; am amodel maker /sculptor I make everything from life sizefigures ; to hobby models am very interested and very scared about getting into this 3d modelmaking / make your own world ; and I was wondering if I could ask you what 3d printer you use ? and is the materials as the result looks great too me
 
Last edited by a moderator:
hi jon just joined here and saw your project ; tried to send you before but I think am having prob;s with working out this web page ; anyway I wanted to ask you what kind of 3d printer and materials are you using ; it looks great and I was thinking of giving the 3 d modeling a go
cheers
Paul
 
Hi Paul,

Currently up to 3 printers, lol.

Original was a Reprap Ormerod 2 kit, good way of learning printers when you have to build it yourself. Have rebuilt the Z axis drive as the original design relied on printed gears, and no matter how I tried I couldn't get them as accurate as I wanted, so changed to a direct drive which gives perfect 0.04mm layer resolution. Printed PLA to begin with, but it's not a great material to work with. Have used Colorfabb XT and NGEN also ASA which prints really nice. But I was trying to print prototype parts near the max size of the bed which still resulted in warping no matter how well it was stuck. Worked out you can't get away with the fact that the bottom of the print on the heated bed is at least 50 degrees hotter than the upper layers, it shrinks more at the top when it cools it literally pulls the part off the bed. I'm being picky here but when you are trying to mate pieces together and the bottoms are not flat it's annoying.
Then with a suggestion on the Reprap forum I picked up a 3D Systems Cubex printer second hand, it has a large print volume and is virtually enclosed as standard, but prints pretty average due to the slicing software and some odd printing behaviour. Converted it to run from open source hardware, added an AC powered 750W heated bed and reinforced and sealed the enclosure. Printing with ASA or ABS works perfectly as the enclosure gets up to about 60 degrees so the shrinkage doesn't happen. In fact ABS parts self release from the bed when the enclosure cools down, quite a scary cracking sound the first time it happened!
I am now just finishing off setting up a Formlabs Form 1 clone, picked it up cheap as it wasn't printing right. Have worked out the main reason it was having trouble was simply because the tray was way off level with respect to the build platform, and the original silicone coating in the genuine Formlabs tray was also way off, such that only the rear right corner was making contact.
Only variable I will need to fine tune now is the settings for the non-standard Makerjuice resin, it's supposed to be close to the Formlabs grey so hopefully won't be that difficult to dial-in.
I will probably sell the Ormerod as anything small and detailed will print on the Form 1 and everything else on the Cubex.

Jon
 
Steeper learning curve than I thought with the Form printer, made more difficult by the fact I bought resin which needs a setting not provided by the software for a Form 1 only Form 1+.
Trying the stock black resin setting was nothing but failure, incomplete outlines or infill and a load of junk bits left in the resin which messes up the next go if you don't strain the resin. Not that difficult to do actually after I got some 400 micron filters from a paint shop.
The open setting mode was only available with the special recent version of the software which recognised the printer but rightly decided it was not an official printer so refused to print :(

As luck would have it some guys on a google group just in the past couple of weeks produced a piece of software which lets you run a full calibration and re-flash the firmware where the values are stored, then the software was happy with the printer. :thumbsup

Must have run about 10-20 small tests which gave me a feel for the laser power and speed settings, but was getting weird results in the first cm or so of the print. Turns out there is some squish and stretch with the low layers, maybe more so than the norm with the resin I am using as it has some flex. Also the stock setting to make sure the resin tray returns to level was too high for mine causing a small "klonk" for the first 50 layers or so, this actually seems to be an issue with the official printers.

Have got a working setting for 0.1mm layers, tried some 0.05mm layers for the 1/48th engine parts which kind of worked, but think the Makerjuice SF resin with a bit of flex is not the right material, so have some FunToDo F1 blend coming, which is said to work with the stock software setting, fingers crossed!

This was last but one test print of a 1/24th canard with panel lines, laser power was slightly low which caused the iffy support part, the part starts where the panel lines do.

Firefox 1-24th canard.jpg

Just tried it again with about 10% more laser power and came out perfect, well 25% of it did, the software lost connection so the print job didn't finish loading for some reason, will hold of starting print jobs until they complete uploading from now on.
Even at the "rough" 0.1mm layer setting the layers don't show up at all with this resin, admittedly it's best case scenario as it doesn't change thickness much.
The panel lines might be slightly oversize but maybe not a bad thing as the paint finish will reduce them a little.
 
Last edited:
Realised I hadn't posted this, found a very good colour match, Peugeot dolphin blue. I have seen this colour in the flesh so to speak and taken photos and compared to the studio model and it's almost an exact match, maybe just a bit less metallic in the mix so it appears a bit darker for the scale effect.

Peugeot_RCZ_Blue metallic.jpg

I think the way to go might be to paint the whole thing with this then varying coats of a 50-70% grey on the sections to get the right shades.
 
Last edited:
Took a lot of time faithfully copying the detail pieces for the wing leading edges, didn't help having a couple of software crashes that lost me about 3 hours work at one point, getting carried away and forgetting to save. Updated the one plug-in that was causing the crash seems to have stopped it happening again.

Firefox leading edge pieces.jpg

Tried doing the inset lighter panels as dots as the original photo etch, but at 1/48th the smallest diameter I can make them that will print (0.3mm) looked totally out of scale. The inset parts on the top edge might be a bit deep, will run a test print to see how it looks.
Had good results with the FunToDo resin at 0.1 and 0.05mm layers, had to work on a custom setting as it was over-curing way too much on the stock setting. I can only guess the laser is different to the normal Form 1, or possibly the different type of silicon I am using in the tray lets considerably more UV through.
The FunToDo resin is pretty brittle. One canard I printed when left on a windowsill facing the sun more on one side warped as if it was expanding when curing. Trying to straighten it snapped it pretty easily.
The Makerjuice SF resin by comparison doesn't print as well for small detail but even with thin pieces is almost impossible to snap.
The 0.05mm layers prints have proved the panel line sizing is right, they just register enough, and it can be controlled by adjusting the exposure setting for the resin. Will post some pics tomorrow.
 
Pics of the 1/24th and 1/48th the test prints.
Firefox 1-24 canard test.jpgFirefox 1-48 canard test.jpg

Couldn't quite get it to focus on the 1/48th one but close enough. Discovered the resin gets less brittle when it is UV cured for an hour, thought it would go the other way.

After trying to get a reasonable likeness found an artist on cgtrader selling a Clint Eastwood model for a sensible price, highly detailed, actually too much as the original file was 550mb!

clint-eastwood-3d-model-3d-model-obj 2.jpg

I did tweak his nose a bit, was just a bit too bulbous wide, stood out more when I had to reduce the polygon count to be able to use it.

So after many more hours built a helmet and body

Clint 2.jpg

Did a little bit of photoshop on his face eyebrows and a bit of facial toning.

Clint pilot.jpg

The body is a posed game character model Nate from Fallout 4 exported from Xnalara, needed some tidying/tweaking, as in where did the last three hours go, lol. Hand models need to be combined with the control sticks or they are going to be too fragile to print. Now I know I can print it will add some detail into the seat.

Getting close to the finish now.
 
Last edited:
Doing final checking that my panel line sizing is going to be ok for the 1/24 version, I went over the panel lines before a bit of sanding and think I went a bit too hard as they look a bit overdone, think I need to lay some paint and lacquer on so I am not fooling myself.

Firefox 1-24th wingtip.jpg

This is quick 120 grit sand and blast of primer. There are a couple of blips and the brim needs trimming off.

Discovered something interesting when test printing (this is number 4 ;) ) First one was ok, but with 0.12mm layers the ABS was, well tearing is the best description I could call it and the rear edge was just too thin to even print in places where the panel lines top and bottom nearly meet.
Changed the layer height 0.15mm, which is about the max so that the horizontal lines resolve, and this actually looked a lot better, also dropped the print speed and temp. This is printing in a sealed enclosure to keep the temp high to stop the warping/de laminating you can get with ABS, so the less heat you put in the part the better.
Was still having problems with the rear wing edge actually being able to print, so after a bit of searching tried adjusting the extrusion width setting in the slicing software. Essentially when you have a 0.5mm nozzle the actual extrusion width is around 0.58mm, so the default auto setting is 0.6mm. I realise now it sets the resolution of the print. Reducing it to 0.55mm made a surprising improvement, the rear edge of the wing resolved better and the print looked sharper.

Going to run the same part on the Form for 1/48th now.
 
Thanks, it is a strange addiction, you especially doubt your sanity when you are moving something 0.02mm and say to yourself "that looks better", lol.
 
Well that turned out pretty well for a first try and the panel lines look spot on.

Firefox 1-48th wingtip.jpg

I stopped it before it had finished as I thought it was over curing, turns out it was just a tiny layer glitch that had stuck to the side, too much power on the early layers I think.

Also made me realise that a couple of the panel line cuts were a bit misaligned and deeper than the rest which is what was I was seeing on the bigger print.
 
Having done various other test prints on the Form printer I have come to the conclusion I will have to do 1/36th scale rather than 1/48th. The front canards are literally like wafers at 1/48th it they are scaled properly, also the panel lines are much more difficult to reliably resolve. I am being a bit of a perfectionist over this, but I want you guys to have a kit that doesn't disappoint.

Also the screen models were 1/36th or 21 inches long so it is kind of fitting.

Firefox 1-48th vs 1-36 canard.jpg

The 1/36th part was pre-UV curing and scrubbing which will get rid of the particles clinging to the surface.
I was kidding myself that the 1/48th stuff was working, the panel lines would literally disappear with a coat of paint. The parts like this will probably be better in the Makerjuice resin as it feels more like standard plastic and is not brittle, and seems to have less post UV cure warping.
These parts print fast enough that I will probably include printed parts rather than castings, as I can print 8-10 in not much more time than 2. Also with them being so thin even tiny bubbles in the resin would mess them up.

Think I might have to go with a photo etch master for the engines as even at 1/36th I'm struggling to get the results I want. The thin relief parts are a single laser pass, this resin actually cures too easily, I should try with the Makerjuice resin - might get better results.
Firefox turkey feathers.jpg
Middle one of these is close but still rough looking. Excuse the fluff, pre-uv curing they are tacky and pick up everything, especially when you drop them on the carpet!

Tried a paint test on a 1/24th tailplane piece
Peugeot Tuanake Blue KDL 01.jpg

This is Peugeot Tuanake Blue code KDL (Dolphin blue is the name the gave it on the Peugeot RCZ over here) with a gloss clear coat in daylight, hard to reproduce it correctly but IRL it looks right

Firefox paint test.jpg

Excuse the obvious mess up on this, used a bit of standard masking tape and the high solvent clear coat melted though.
This is silver base on front black on rear with the blue and gloss clear coat on top, all out of rattle cans. I think a mixture of grey, silver and black pre-shading of the panels will emulate the look pretty well.
Also in low light it goes near black as the original does.

There is some pre-ringing (and a bit of underextrusion) in the prints on my Cubex printer, it causes a few micron step before a panel line. I happens under deceleration, tried printing slower and it still occurs, don't think I will even get rid of it unless I printed so slow it would take about 20 hours to just print a tailplane!
The practical solution is to over-extrude about 10% then just sand until you get a flat surface. ABS wet sands real nice, you can use quite high grit and the surface doesn't get messed up. It means the 2-3 people so far who want a 1/24 or 1/18th kit will have a bit of manual labour to finish the parts.

Here's a pic of the Form clone in action
Form 1 clone.jpg

I moved all the hosted pics in this thread to google drive after Dropbox got locked out from too much usage, have gone back through the thread and fixed the links.

Main folder with all pics


If you go in there you will find Firefox.mp4 which is a screen recorded fly around of the model for you guys to enjoy.
 
Last edited:
I was going to say that you should go to 1:32 but if the screen models were 1:36, well, make them studio scale. Great work!
 
Glad you're still working on this.It looks great.
1/36...Mmmm. This I can live with!:). I've very nearly completed my Cygnus model and will be looking for something to occupy me over the winter. A firefox would be good. Can you put me down for one-please?
Thanks,

Steven.
 

Your message may be considered spam for the following reasons:

If you wish to reply despite these issues, check the box below before replying.
Be aware that malicious compliance may result in more severe penalties.
Back
Top