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

[JayW]

Hello all!  Recently I purchased a 3D printer -  Elegoo Saturn 2 - which has the potential to transform my modelling.  Those of you who follow me know that my recently completed 1/18 scale build (F4U-1A Corsair) benefitted from a good bit of 3D printed parts almost all of which I designed myself but had printed by a capable fellow in the UK.  Since then, I now have the means to do my own printing, if only I can get good enough at it.

 

So my next project, if all goes well, is to create a very accurately shaped P-51 in 1/18 scale.  At this point I am undecided whether to do a B/C variant (razorback) or a D variant (bubble canopy).  My first choice would be a -B with a Malcolm hood modification, followed by a -D as a second choice.  I would prefer not to do a straight up -B razorback (or maybe better known as a "greenhouse" canopy version), as Peter Castle (Airscale) has done his incredible Lopes Hope in 1/18 already, and I would prefer not to tread on that hallowed ground.  The only available kit I know of a P-51 in 1/18 is from 21st Century toys or whatever it turned into over the years - a D model.  They are long out of production but can be found on E-bay for inflated prices.  I actually have one, but I will not be using it because it has too many inaccuracies.  Years ago, though, I heavily modified just such an animal resulting in "Miss Velma":

 

 

 

That satisfied my goals at the time, which was to take an existing large toy model and do my best to make it look real.  Miss Velma, like many P-51 models in many scales, has shape inaccuracies, and can be greatly improved upon, and that is what I intend to do.

 

So I have three important ingredients in which to do this:

 

1.  Aircorps Library - they have a full set of engineering drawings for the P-51 variants, including "ordinate" drawings, which are tables of point coordinates that define the shapes of the outer surfaces of the aircraft (and even some inside shapes like throats of intakes).  These ordinate drawings will allow me to create surface definition in any scale very accurately.

 

2.  CAD modelling software - I have been using Rhino 7 for a couple years now, and it is indispensable for layouts, part definition, and 3D printing.  I am no expert user, but I get by.

 

3.  3D Printing.  This will be the first time I will have tried to make 3D printed parts the primary components of the build.  Previously I have used it simply to augment an existing model.  This is my weak link.  I am a bit of a newbie when it comes to 3D printing.

 

To start, I am in the process of converting point data, obtained from vintage engineering drawings, into Rhino surfaces - hopefully for the entire airframe. 

 

Here are examples of the point data vintage 1940's:

 

 

 

 

There is also wing data that I have not shown.

 

I can produce points in space from those tables, and connect them with curves in Rhino - like this:

 

 

Very laborious and monotonous, I can say.  Not too surprisingly, the Rhino model is showing what looks alot like a P-51.  It better - the good folks at North American back in the '40's used the same data to build the tools used to manufacture the airframe.

 

From there, and after some time-consuming schooling on primary surfacing, with classes available on YouTube, I used those curves to create a bunch of patches:

 

 

Clearly I am not finished.  Those patches can be combined to make a single compound-curved surface, or larger surfaces.

 

I also have wing lofts, done a different way:

 

 

P-51 wings have single curvature which is to say they are defined by straight line elements.  Which is to say that you can have a contour at the inboard end, and a similar smaller contour at the outboard end, and you have fully defined the wing surfaces simply by lofting between them along straight line elements.  That is what I did above.  You see two lofts - the basic wing, and the inboard extended leading edge - a feature of the -D model, but not the -B model.  That extended leading edge is also single curvature. 

 

Far and away the most difficult detail to surface so far has been the carb air intake below the prop spinner.  After days of applying everything I learned in most of 14 classes I took on primary surfacing techniques, and episodes of tearing my hair out and occasionally walking off in a huff, I got this:

 

  

 

 

It is not perfect, but any inaccuracies are measured in hundredths of an inch at full scale - which is infinitesimal at 1/18 scale.  So, unlike a whole lot of P-51 models out there, this "smiley face" detail is deadly accurate.  BTW - the Tamiya 1/32 P-51 has a very accurate smiley face, and I am convinced their designers had access to the point data that I have used.  Next post you will see a test part of this detail, in 1/18 scale. 

 

I also have a complete surface for the prop spinner, done months ago:

 

 

 

That was a breeze compared with the other surfacing efforts to date.  Just a contour curve rotated around the thrust axis.

 

In theory, if I can have a fully defined set of surfaces for the P-51, defined in Rhino in full scale, I can scale to 1/18 and create just about everything from them.  A pretty exciting prospect, and easier said than done!

 

Next post I will show the results of some test parts.  I hope this build generates some interest; it's going to be a big challenge. 

 

P.S. - if any of you Mustang experts out there have any information on the Malcolm hood - like shapes and cross sections, well that would probably clear the way for me to do the B-model!

 

   

 

 

Edited September 22, 2023 by JayW

[JayW]

I needed to do a reality check on this ambitious dream I have here.  To get a better idea what I am up against.  How well did I model the engine cowl surfaces, including the "smiley face", that I have created so far?  How well is the printer set up (exposure times, etc), and how well have I designed my supports and is my part orientation for printing acceptable.

 

So I designed and created some test parts.  Already I had a part defined for the cuffed propeller, done a couple years ago just for kicks:

 

       

 

I could go on and on with this effort.  It was a ton of fun.  To make a long story short, Airscale Library has some Hamilton Standard data on prop blades including the one used for P-51's. Data is available for cross-section shapes at various prop stations from root to tip, and their twist angles.  Create curves in Rhino of all these shapes, at the right angles, loft between them, do some tweaking, and you get what you see. 

 

Then I had to define other test parts using the cowl surfaces I created over the last few days, which I described last post.  And also the spinner surface.  Here:

 

 

And after many hours of 3D printing I got these parts:

 

 

 

Except for possibly the prop blades (they are practically perfect), these are merely test parts.  Parts I make for the model will be much more detailed. 

 

So this was a test.  What did I find?

 

The printer is probably set up about as well as I can do it.  No print failures, and part quality looks good.  I certainly expected no less; I have already produced some parts for the Corsair.

 

The "smiley face" part has excellent surface smoothness.  no lumps, no ripples.  What a victory!  But it didn't print that great.  More work to do on that score....

 

The cowl surface is actually pretty good.  There is one seam between patches where I didn't do a good job of getting it smooth (I can fix that), and some edges are a bit ragged as can be seen on the smiley face part below.  I hope I can fix that with a locally more dense array of supports on the 3D print file.  I am a bit concerned about that.  It could possibly be addressed with some putty though.  Here:

 

 

 

Fit between the two parts is pretty good, but I was expecting perfection.  The 3D printing process is not yielding theoretically perfect results.  Any imperfections will show up when trying to mate two parts together.  And I do not like those ragged edges.  Hope I can get that taken care of.  I should say now that I intend to skin this model in aluminum litho sheet.  So the final fuselage parts (wing too) will be under the defined surface by .005 inch to allow for the thickness of the litho sheet.  And that will cover up some imperfections in the 3D printed parts.

 

And speaking of imperfections:

 

 

The spinner halves didn't fit together worth a sh_t.  And edges are ragged similar to the cowl parts.  The spinner will not be skinned so it has to be right on.  Again - more supports locally should help with most of the shape problems.  Also, the prop blades didn't fit in their holes at all.  The holes need to be a few thousands of an inch larger, just like holes for a bolt need to be slightly larger than the bolt diameter.  Instead of being exactly the same diameter.  I am a dummy.  Also, I am finding that 3D print parts might be just a tad expanded - so the prop blade roots might be a tad oversized, and the holes a tad undersized.  On the bright side, these parts look very cool, and have the potential to be transformed into excellent ready-for-primetime parts.  I will be doing some substantial redesign, and will probably make the prop blades and spinner my first effort at production parts.  A stand-alone unit that will gather dust for a long time waiting to be installed on a complete fuselage.

 

I combined test parts to see how it looks:

 

 

 

I am both encouraged, and discouraged.  It sure looks like a Mustang - that's great.  But fit-up is going to be a challenge.  This is not going to be as easy as I envisioned.   But we all experience learning curves, and I hope to experience that here.  Stay tuned - I will post progress on a greatly improved prop before long.  Meanwhile I have alot of surfacing work to do on the computer. 

Edited August 31, 2023 by JayW

[Jim Barry]

Fantastic! Can't wait to see more! 

[Warbird Kid]

Exciting new project! Can't wait to watch the results of this masterpiece in action! 

[airscale]

here we go!

 

Fantastic start Jay, the learning curve is going to be steep - as you have found out you think you have won when you have designed it, then a whole new world of pain opens up printing it 

 

Really looking forward to this and the many adventures (& lessons) to come

 

Peter

[easixpedro]

Oh, this is going to be slick. The former engineer in you must be in hog heaven! 

 

Looking forward to more

-Peter

[Thunnus]

Very interesting, Jay!  A brave new world you are entering into!

[geedubelyer]

This is likely to be an enjoyable ride.

Don't forget to factor in the ability to create kits or sell files of this. I can't believe you'll be the only one to desire a 1/18th scale Mustang and you might be able to recoup some material costs along the way.  

Thinking outloud and along those lines, do the two marques (B/C) differ greatly? Would it be possible to design in the ability to offer either version by producing modular and interchangeable components?

 

Cheers,

Guy

[Alain Gadbois]

Great project!

It might seem counterintuitive, but the imperfections in the spinner front part around the lower edge will be minimized if you print it the other way around. Of course, the contact points will be on the outer surface, but these will be very easy to sand off.

 

Alain

[LSP_Kevin]

Fantastic project, Jay! My experience so far has been that resin 3D printing is not particularly reliable when it comes to dimensional accuracy, and post-curing can make the the resin swell ever so slightly, compounding the problem. And the ragged edges around support attachments is something that still plagues my prints, and was a big problem with the BOMARC parts. Hopefully we can both solve that one with practise.

 

Kev

[Gazzas]

Wow, Jay!  I am impressed and will be following. 

[TAG]

[Citadelgrad]

This is fantastic, Jay.

 

i am eyeing the Elegoo Saturn 3 as we speak, and one of my first projects would be a prop spinner for my Tempest.

 

 

Needless to say, i will follow this with intense interest. 

[MikeMaben]

8 hours ago, geedubelyer said:

                     ... do the two marques (B/C) differ greatly?

 

Nope, Bs were built in California , Cs were built in Texas.

 

Alistair may be able to help with the canopy .

 

 

[Cap'n Wannabe]

That is incredible.