1/18 Scale P-51B 3D Print Build

[DeanKB]

Bloody hell. An actual 1/18th scale Mustang, printed from a printer at home, using custom made plans. 

 

Seriously impressive.

[Hans]

Looks great, Jay! 

[LSP_Kevin]

Incredible, Jay. Just incredible.

 

Kev

[brahman104]

So much to digest here!!!! Truely incredible Jay!

 

Craig

[MikeMaben]

Very nice. When you finish , could you scale it down to 1/35 and sell some to us ???

 

 

[Rocat]

Amazing work!

[Maxim61]

On 12/6/2023 at 6:02 AM, BiggTim said:

Interesting little tidbit - I wanted to try this "Accura 60" resin for myself, so I contacted a sales rep to find out where I could buy some. It retails for 2635 USD per 10 kg jug, and that's the smallest quantity available! So, if you're unhappy about Shapeways ridiculous prices for items using this resin, at least now you know part of the reason why. Needless to say, I did not buy any. 

 

Tim

 

I can tell you from experience that 10 kilograms of resin will print a lot of stuff. Shapeways is still very overpriced and even more so with their cheap resin parts.

[Uncarina]

Jay, I am just catching up here, and I have clearly been missing out! With respect to the Accura 60 material, could the cost be divided among several modellers? Alternatively, could you avoid Shapeways and purchase directly from the source?

 

Cheers,  Tom

[airscale]

amazing Jay - I made a jig at this stage too, but nothing like as sophisticated as this - this is just next level!

 

such a pleasure watching every challenge get bested and this is going to be a true masterpiece when complete

 

Peter

[Paulpk]

Jay, great progress, just a pleasure to watch this new technique evolve! Your gig is an excellent idea. I think without it you would regret it!

[JayW]

13 hours ago, Uncarina said:

With respect to the Accura 60 material, could the cost be divided among several modellers? Alternatively, could you avoid Shapeways and purchase directly from the source?

 

Hi Tom - I am going to try some different stuff before I delve into Accura 60.  Or Shapeways.  Report out soon (as soon as it warms up enough - it is so cold right now I cannot keep my house or office warm enough.. .

 

12 hours ago, airscale said:

amazing Jay - I made a jig at this stage too, but nothing like as sophisticated as this

 

Hi Peter - yessir I know.  At the time though, you were not in the 3D print business.  Had you been, my guess is that you would have produced a similar jig.  Anyway, yours did the trick, didn't it.  Either way, it was clear to both of us that some sort of jig was going to be necessary.  Thanks for looking in!

[SwissFighters]

On 1/16/2024 at 8:16 AM, JayW said:

I have much to show.  Progress seems slow, but actually it is not.  

 

In order to continue with fuselage construction including a good cockpit, and in order to assure this model P-51 is going to look straight, with wings accurately attached, I know I must create a major assembly fixture for the forward fuselage (forward means between the firewall and the tail section).  My forward fuselage will be in two pieces - one from the firewall back to station 200 (this is where the fuselage tank ends), and one from station 200 to the beginning of the tail section at station 248.  The main reason for this is the size limitations of my 3D printer.  The short aft portion (sta 200 - 248) will not need an assembly fixture; I will 3D print practically the whole thing, like I did the engine cowl.  

 

Actual P-51 fuselages were manufactured in much the same way, or any fuselage for that matter.  Here is a photo of some major assembly tooling for the Lope's Hope forward fuselage at AirCorps Aviation:

 

 

Peter Castle gave me a whole lot of pictures including this - thanks Airscale!  Note there is a heavy frame and plate for the firewall interface, and way in the back a similar frame and plate for the station 248 interface.  Also note the longerons and some of the fuselage frames are in place.  And there are posts coming up from the floor to support the lower longerons.  My fixture must have the same stuff.

 

The upper and lower longerons are the major structural elements of the fuselage. plain and simple.  Everything starts there.  That will also be the case on my 1/18 model.  So first thing for me to do was to design, in Rhino, the longerons.  In real life they are H-section heavy extrusions, machined to be lighter the further aft you go.  And they have tension fittings on either end.  I would attempt to duplicate all that where the longerons are visible.  Many many things attach to the longerons, so it was a long drawn out project with much research into the configuration - fuselage frame locators, simulated cockpit floor brackets, all manner of other location tabs or flanges for other stuff (like the engine control quadrant, the flap actuator torque tube mounts, landing gear uplock fittings, windshield frame tabs.....) - endless it seemed.  And I probably missed some stuff.  But here are completed longeron details in Rhino:

 

 

3D printed:

 

 

 

See that little nested white block on the top longeron?  I fractured that one and repaired it with CA'd plastic.      The real forward fuselage longerons are longer - they go from the firewall to the manufacturing splice at station 248.  Mine will end at station 200 as I already mentioned.  As they are, they barely fit into my 3D printer.  Glad mine is as large as it is; many are smaller. 

 

In order to produce those longerons, I had to model many of the parts that interface with them, so the locator features could be included.  Here are some of those parts in Rhino:

 

 

 

Took quite a while.  You have already seen the windshield parts.  I will remind everybody that before any of this modeling could be done, I first had to generate exterior surfaces using ordinate data from the original drawings - something I have been toiling at for months now.  That's where it all starts.

 

Now for that major assembly fixture.  It must have the following features:

 

1.  A firewall plate with four longeron locators (two upper, two lower).

2.  A station 200 frame plate also with four longeron locators.

3.  A center platform that controls the four points where the wing attaches to the longerons. 

4.  Two intermediate supports/locators for the longerons (they are long and flexy and need those additional supports).

5.  A flat floor plate to assure no warping or bending, with attach holes.  The attach holes for the firewall plate and the station 200 plate must be slotted to allow fore/aft movement such that fuselage side assemblies can be extracted and re-inserted into the fixture during build-up.

6.  A centerline slot in the floor and matching flanges on the tools to assure good alignment.

 

And it must be very dimensionally accurate if I expect the model to fit together - for the already completed engine cowl to properly attach, the aft fuselage to attach, the wing to attach, and the radiator intake/exhaust area to attach (you have seen some of these parts as well).  I cannot rely on pre-molded parts like in an actual kit to fit everything together, and felt I could not make equivalent parts.  So I will build up this part of the fuselage like they did the actual airframe.  And hoping I get great accuracy with 3D printed fixture parts.

 

Here is the forward fuselage major assembly fixture in Rhino:

 

     

 

It consists of two end plates and frames, two center support bases, two intermediate detachable support plates, and a stiff mounting floor. 

 

The canted firewall plate and frame (note the four longeron locator posts and skin panel flanges): 

 

 

The station 200 plate and frame (also note the longeron points and flanges for skin panels):

 

 

And 3D printed.....

 

Half the fixture parts fresh out of the resin vat still dripping resin:

 

 

Cleaned, cured, painted.  Note the flanges on the undersides that will fit into a full-length slot on the floor: 

 

 

The center pieces mounted to the floor (no end plates yet):

 

 

The floor is made of .08 thick plastic sheet, and will itself mount onto long .25 x .75 inch evergreen plastic bars (due in the mail any day now).  That should give me a robust flat floor that doesn't flex.  Right now it's a bit flexy.  As noted in the picture, this fixture has locating posts for the wing, located with great accuracy.  Critically important.  There are holes in the lower longerons that match up to those pins.  

 

All the fixture parts in their glory, bolted onto the yet-to-be completed floor:

 

 

And viola - with the longerons!!

 

 

All appears to be in order, and I believe I have the accuracy I need to build up a proper forward fuselage.  Now I can begin fabricating individual frames and other elements and attaching to the longerons using the integral designed-in locating features.  Next post you should see some of that.

 

And a sneak preview of how this jig will work for me:

 

 

A very pumped modeler here.  Hoping I can make my dreams come true with this approach.  Some of you may suspect this is a bit overkill.  I think not.  This assembly method is going to allow me to do alot of very cool things.  You will see.  Stay tuned, and keep warm all you folks in the Northern hemisphere!

 

 

 

 

 

 

Thanks for sharing Jay. I'm really curious to see how this works out! I wonder, with what I presume will be quite considerable bending moments around this section of the fuselage (how heavy is the nose section?), will the longerons be strong enough in themselves to carry that load? I assume on the prototype, the longerons were necessarily some of the strongest, stiffest components which could therefore carry stressed skins. Will the model skins carry any load? I'm pumped to see how you resolve these issues! Keep up the great work!

[JayW]

12 hours ago, SwissFighters said:

I wonder, with what I presume will be quite considerable bending moments around this section of the fuselage (how heavy is the nose section?), will the longerons be strong enough in themselves to carry that load?

 

I assume here that you might be concerned about the strength of those H-section resin longerons, being as spindly and flexy as they are.

 

Well - on a real airplane with semi-monocoque structure like a Mustang, tension and compression loads are carried by longerons and frames.  Longerons are kind of like the chords on beams, with frames spaced at intervals to keep the longerons in place.  While shear loads are taken by the skin, stabilized with stringers where needed.  Like the vertical wall of the beam.  Stringers can also share tension loads.  Matters not; the Mustang fuselage has no stringers.  What loads?  The engine torque for one, along with thrust from the prop.   Other than that, critical loads come from flight - high G maneuvers including inverted flight.  The tail imparts loads on the fuselage, and also weight of the engine multiplied by the number of G's.  It is that kind of loading that sizes fuselage structural components like longerons. 

 

For the model - it's all about stiffness and "abuse loads".  The weight, although considerable for a model, just isn't going to account for much concern with the fuselage.  My H-section .015 inch gage longerons and .02 skins will take the weight of the nose, and the load from the tail gear with very healthy margins I expect.  No, for the fuselage I would be far more concerned about handling loads - grabbing it just below the cockpit cutout for instance - the weakest part.  Already though, I know from experience if I am going to pick up one of these very large models, I do it with two hands placed on robust regions like the nose and tail, or under the wing roots.

 

Lastly - I was blown away on my Corsair project just how robust 3D printed parts can be once bonded in place and doing what they are supposed to do.  The main landing gear diagonal braces, for instance, were just as spindly and fragile as could be alone as single parts.  I worried about them.  But once incorporated into the model, attached to the gear posts and wing spar - wow what strong parts.  I worried no more.  I expect the longerons, once supported by frames and with the skins bonded to them, will be plenty stout.  But you will never see me grabbing the finished model by the flat sides under that Malcolm hood.  No sir.   

Edited January 17 by JayW

[SwissFighters]

7 hours ago, JayW said:

 

I assume here that you might be concerned about the strength of those H-section resin longerons, being as spindly and flexy as they are.

 

Well - on a real airplane with semi-monocoque structure like a Mustang, tension and compression loads are carried by longerons and frames.  Longerons are kind of like the chords on beams, with frames spaced at intervals to keep the longerons in place.  While shear loads are taken by the skin, stabilized with stringers where needed.  Like the vertical wall of the beam.  Stringers can also share tension loads.  Matters not; the Mustang fuselage has no stringers.  What loads?  The engine torque for one, along with thrust from the prop.   Other than that, critical loads come from flight - high G maneuvers including inverted flight.  The tail imparts loads on the fuselage, and also weight of the engine multiplied by the number of G's.  It is that kind of loading that sizes fuselage structural components like longerons. 

 

For the model - it's all about stiffness and "abuse loads".  The weight, although considerable for a model, just isn't going to account for much concern with the fuselage.  My H-section .015 inch gage longerons and .02 skins will take the weight of the nose, and the load from the tail gear with very healthy margins I expect.  No, for the fuselage I would be far more concerned about handling loads - grabbing it just below the cockpit cutout for instance - the weakest part.  Already though, I know from experience if I am going to pick up one of these very large models, I do it with two hands placed on robust regions like the nose and tail, or under the wing roots.

 

Lastly - I was blown away on my Corsair project just how robust 3D printed parts can be once bonded in place and doing what they are supposed to do.  The main landing gear diagonal braces, for instance, were just as spindly and fragile as could be alone as single parts.  I worried about them.  But once incorporated into the model, attached to the gear posts and wing spar - wow what strong parts.  I worried no more.  I expect the longerons, once supported by frames and with the skins bonded to them, will be plenty stout.  But you will never see me grabbing the finished model by the flat sides under that Malcolm hood.  No sir.   

Thanks Jay, yes, you assumed correctly! I guess I didn't want to appear rude by using terms like spindly and flexy but that's what they are at that scale. I really appreciate you sharing the rationale behind the work. 

[JayW]

Last I posted, I stated that my next sub-project would be the windshield and its surround.  And in order to do that I needed to complete my forward fuselage jig, load the upper longerons, and use that combo to build up that windshield subassembly. 

 

What the jig lacked at that time was a good stout floor support structure - heavy members to eliminate flex.  That was done successfully with .25 x .75 inch section beams, and now the jig is working to perfection.   Here it is with the longerons loaded, and the largely completed windshield and surround assembled on it:

 

  

 

You are going to see alot of that jig in future posts.  That is going to assure me of a properly integrated vitally important forward fuselage.  

 

So that windshield.  Another task was to try to improve on the transparencies.  Was having lots of troubles with the ones I have printed up already, with Nova3D clear. So I pulled out the wallet and got me some "Resione G217 clear", which a guy on U-tube couldn't say enough good things about.  So did another.  And indeed this stuff is more well-behaved than the Nova3D.  It isn't as sticky, it cures with shorter exposure times, and seems to have less discoloration (darkening).  And, it holds detail better.

 

 

What you do not see there is days and days of failed prints, and flawed parts for one reason or another.  I have a pile of rejected parts!  What you do see there is the result of exhaustive sanding and polishing and clear lacquer paint layers.  There - all the transparencies for the cockpit enclosure, including BTW the aft windows.  Well, minus the Malcolm hood.  I made one with the new resin, and it is much better than the test parts I made a couple months ago.  But it isn't quite ready for prime time yet.  Another post soon.  The front bullet-proof flat glass in that picture has already been installed into the windshield surround assembly.  Can you see it?  It's pretty clear!

 

So as you can surmise, I have declared a limited victory on 3D printed transparencies over a vacu-formed part or parts.  These are not perfect, but they are pretty darned good.  And the accuracy and added detail is much more than a vac-formed part could give me, pretty sure.  Next post, you will see among other things the top and side glass bonded into the windshield frame, to accompany the front glass already there.  My aim is to win over the doubters - hope I can.

 

So here is the windshield surround assembly freed from its jig:

 

 

 

The assembly consists of details you have seen already but unassembled - the cowl, now painted window frame, the shroud (with handholds and defroster nozzles added, and painted), and the painted IP support frame.  Two oxygen system related gages have been added, the decals provided by Peter Castle's (Airscale's) Lope Hope decals, which are second to none.  There is more to come for this frame.  That spindly uber-fragile windshield frame is considerably more robust now.  I like. 

 

So what else - oh, the instrument panel!  It was not very difficult to Rhino model one, and print it up.  Here it is overlaid on top of another of Peter's Lopes Hope decals:

 

   

 

I initially had some doubts the holes and dials would match up - but they do.  Perfectly.  Nice going Airscale.  Painted up:

 

 

If y'all really like it, give Peter the credit - it is the decals that really make it special.  So this is a 3-piece laminate.  The front is my 3D printed part, a middle .005 inch thick clear plastic piece, and a back .02 inch thick white plastic sheet with the decal applied to it.  The challenge is to bond them together without contaminating the dials.  Mission accomplished although with much hand-wringing. 

 

Also, I began the N3B gunsight.  This is a challenge with all the mirrors - here is a shot from the parts catalogue:

 

   

 

So off to the races.  Rhino design took some time for sure.  And the parts are soooo small.  Here:

 

 

In front is the mounting bracket.  On the left is the "head assembly".  On the right is the gunsight body.  And center rear is the lower mirror support fitting. And here is what I have so far:

 

 

 

Parts fit together pretty well.  What you see here is a testament to the latest technology of affordable 3D printers.  It continues to amaze me.  Still to come is the very dinky lower mirror, an electrical wire, and the familiar 45 deg slanted glass.  Next post!

 

In anticipation of the soon coming windshield window installation, I used some of my rejected parts to make another windshield assembly for the purposes of fabricating the .005 inch thick exterior strips:

 

  

 

It consists, of course, of the four glass parts and the spindly window frame, painted with a coat of primer.  Spindly no longer.  I even added globs of CA to the joints on the interior side to make it still more robust.  There will be fairly forceful burnishing and other manhandling things going on during the process of making the exterior strips, so it needs to be strong.  Better done on this mock windshield than the one I will use for the model itself.   Hopefully you can see, BTW, how nicely the individual window panes fit into the frame, providing nice little grooves or valleys for the exterior strips to match up to.  The advantage of digital design and manufacture.  Looking forward to that work, I think.  

 

Next post you should see a completed windshield/surround sub-assembly, perhaps even skinned.  Then it will be on to fuselage side panel assembly - frames and skins and longerons - all done with the new jig.

 

Take care until next time.  Thanks for looking in!

 

 

Edited January 31 by JayW