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

[Jim Barry]

Those chains for the tail wheel door are legend. When I think I have it tough, I think of that effort you made. 

 

[JayW]

On 2/6/2024 at 6:41 PM, Oldbaldguy said:

have you considered printing the chains?

 

Yes I have.   It will not be very long before I try it out.   

[JayW]

I have been developing the fuselage side panels in Rhino, including the Malcolm hood additions with the roller bar.  In so doing, it became more and more apparent to me that my rollers, and the width of the bar, appear too small.  I have been struggling to figure out what diameter those rollers are.  First, I am not aware of a P-51B/C Malcolm hooded model in any scale that actually has the rollers.  Second, I am not aware of any technical data out there that would tell me.  So all I can do is study all the pictures I have, like this (notice the rollers along the sides of the fuselage):

 

  

 

Comparing roller size to other items in there, where I have an idea what their dimensions are, and I first came up with a roller diameter of .75 inch.  Pretty small.  Well, too small, it was beginning to be apparent.  So I did some more measuring, and revised that diameter to .875 inch.  I played around with 1.0 inch rollers in Rhino, and things were just not integrating well.  So I am convinced a 1.0 inch roller is too large.  Hence 7/8 (.875).  Who knows?  Being a British field mod, they may be a metric size.....

 

Anyway, some of you might think I am splitting hairs, and you might be right.  Here I spent a couple days changing everything affected by the larger rollers and roller bar:

 

 

 

It looks better to me.....

 

So it occurred to me that it might be wishful thinking to print up a couple of side panels either with the roller bar incorporated into them, or locating provisions for the bar, print up the Malcolm hood along with its side guides, and have it all fit perfectly to the already completed windshield.  As exacting as parts can be when digitally defined and 3D printed, alot parts are in play and tolerances are going to add up.  I need a good match-up.

 

So I decided to make some test side panels, and do rev "A" of the side guides for the hood (I did some months ago but new larger ones required for the larger rollers):

 

 

 

Here are the side guides bonded to one of my Malcolm hoods (that I will not be using, but now makes a good test part):

 

 

Pretty nice looking for a test part....    Now to get those test side panels onto the jig and longerons:

 

 

 

These are simplified test panels, so I cannot permanently bond them.  Just taped.  My finished windshield is on there too - taped in place.

 

A couple things I wanted the test to tell me:

 

1.  Will the hood fit onto and slide on the rollers? 

2.  Will the rollers retain the side guides such that the hood will not just fall off?  

3.  Will the hood fit to the windshield properly when closed?

 

OK numbers 1 and 2 - When Rhino modeling the details, I took a guess how much clearance I needed between the track surfaces on the side guides, and the rollers.  Too tight and the side guides will not fit onto the rollers, or if it does the hood might not be able to be slide fore and aft.  Too loose and the hood will fall off; it is too flexible to just stay in place on its own.  Here it is snapped onto the rollers:  

 

 

 

It wouldn't move!  So I carefully pried it off and gave edges of the rollers and the track some light sanding, and tried again.  It grudgingly was able to slide some, but the more I moved it, the easier it got.  So victory.

 

Now for the uber important number 3 - fit-up to the windshield.  Boy did I get a surprise:

 

 

 

 

A full .03 inch too low!  WTF??  Also too low at its aft end:

 

 

It is dragging on the top of the station 146 frame, where there should be about .03 inch clearance.

 

Normal tolerances are not responsible for that.  Something is wrong.  What is it??   Why, why, why??  It's supposed to be perfect.  So I began measuring everything.  Where is the error?  Is the windshield too tall?  Nope, its right on.  Are the side panels mislocated?  Nope - they are fine.  Is the hood too short?  YES!  Too short.  By about .03 inch.  There is the problem.  But why?  For some reason the part doesn't match the Rhino model.  Did it shrink when being printed?  Don't know yet.

 

Fortunately, I have another couple of hood transparencies that I am not going to use, and I had printed a second set of side guides incase I ruined one.  So I shimmed between the side guides and the transparency with .02 plastic, and retested with a now slightly taller Malcolm hood.

 

Got great results:

 

 

 

 

Wow!!  What potential.  Am I glad I did this test.  Now I have to figure out what is what with the Malcolm hood transparency.  I will do some critical measurements on all remaining parts (I have 5 total, none of which are clear enough for prime time) and see if that dimensional error is consistent. 

 

Now I am fixated on the Malcolm hood.  I now have two 3D printed burnishing surfaces - one male and one female:

 

   

 

 

They were not hard to design and print.  About half a day's work.  The transparency fits nicely into or over either, and I can sand and burnish and polish away. 

 

Next step is to understand how I got that .03 mismatch and make corrections, whatever they may be.  Then, I will have Shapeways do my next transparency, in Accura 60 resin.  The Malcolm hood transparency is just too difficult for me.  I have come close, but no cigar.  The quote is about $45 for one, $72 for two.  I'm going to do it, and hope that is what gets me over the finish line.  When the parts come they will require sanding/burnishing/polishing with the new tools.  

 

In parallel with that activity, I will finish designing the forward side panels.  Lots of interior items that will interface with them - some I will incorporate into the panels, some I will just incorporate locating features.  To worry about - control pedestal, flare gun port, electrical boxes, throttle quadrant, and pulley wheel brackets. 

 

Fun times.  Stay tuned please!    

 

Edited February 9 by JayW

[Neo]

Wiw 75 for 2 thats salty pricing considering you did all the design work and those parts are small

[Timmy!]

Glad to see this canopy coming together.  Functionally it a great indicator you nailed it!  Shapeways is probably a good call from the time is money point of view.  Looking forward to see the results. 
 

Timmy!

[SwissFighters]

13 hours ago, JayW said:

I have been developing the fuselage side panels in Rhino, including the Malcolm hood additions with the roller bar.  In so doing, it became more and more apparent to me that my rollers, and the width of the bar, appear too small.  I have been struggling to figure out what diameter those rollers are.  First, I am not aware of a P-51B/C Malcolm hooded model in any scale that actually has the rollers.  Second, I am not aware of any technical data out there that would tell me.  So all I can do is study all the pictures I have, like this (notice the rollers along the sides of the fuselage):

 

  

 

Comparing roller size to other items in there, where I have an idea what their dimensions are, and I first came up with a roller diameter of .75 inch.  Pretty small.  Well, too small, it was beginning to be apparent.  So I did some more measuring, and revised that diameter to .875 inch.  I played around with 1.0 inch rollers in Rhino, and things were just not integrating well.  So I am convinced a 1.0 inch roller is too large.  Hence 7/8 (.875).  Who knows?  Being a British field mod, they may be a metric size.....

 

Anyway, some of you might think I am splitting hairs, and you might be right.  Here I spent a couple days changing everything affected by the larger rollers and roller bar:

 

 

 

It looks better to me.....

 

So it occurred to me that it might be wishful thinking to print up a couple of side panels either with the roller bar incorporated into them, or locating provisions for the bar, print up the Malcolm hood along with its side guides, and have it all fit perfectly to the already completed windshield.  As exacting as parts can be when digitally defined and 3D printed, alot parts are in play and tolerances are going to add up.  I need a good match-up.

 

So I decided to make some test side panels, and do rev "A" of the side guides for the hood (I did some months ago but new larger ones required for the larger rollers):

 

 

 

Here are the side guides bonded to one of my Malcolm hoods (that I will not be using, but now makes a good test part):

 

 

Pretty nice looking for a test part....    Now to get those test side panels onto the jig and longerons:

 

 

 

These are simplified test panels, so I cannot permanently bond them.  Just taped.  My finished windshield is on there too - taped in place.

 

A couple things I wanted the test to tell me:

 

1.  Will the hood fit onto and slide on the rollers? 

2.  Will the rollers retain the side guides such that the hood will not just fall off?  

3.  Will the hood fit to the windshield properly when closed?

 

OK numbers 1 and 2 - When Rhino modeling the details, I took a guess how much clearance I needed between the track surfaces on the side guides, and the rollers.  Too tight and the side guides will not fit onto the rollers, or if it does the hood might not be able to be slide fore and aft.  Too loose and the hood will fall off; it is too flexible to just stay in place on its own.  Here it is snapped onto the rollers:  

 

 

 

It wouldn't move!  So I carefully pried it off and gave edges of the rollers and the track some light sanding, and tried again.  It grudgingly was able to slide some, but the more I moved it, the easier it got.  So victory.

 

Now for the uber important number 3 - fit-up to the windshield.  Boy did I get a surprise:

 

 

 

 

A full .03 inch too low!  WTF??  Also too low at its aft end:

 

 

It is dragging on the top of the station 146 frame, where there should be about .03 inch clearance.

 

Normal tolerances are not responsible for that.  Something is wrong.  What is it??   Why, why, why??  It's supposed to be perfect.  So I began measuring everything.  Where is the error?  Is the windshield too tall?  Nope, its right on.  Are the side panels mislocated?  Nope - they are fine.  Is the hood too short?  YES!  Too short.  By about .03 inch.  There is the problem.  But why?  For some reason the part doesn't match the Rhino model.  Did it shrink when being printed?  Don't know yet.

 

Fortunately, I have another couple of hood transparencies that I am not going to use, and I had printed a second set of side guides incase I ruined one.  So I shimmed between the side guides and the transparency with .02 plastic, and retested with a now slightly taller Malcolm hood.

 

Got great results:

 

 

 

 

Wow!!  What potential.  Am I glad I did this test.  Now I have to figure out what is what with the Malcolm hood transparency.  I will do some critical measurements on all remaining parts (I have 5 total, none of which are clear enough for prime time) and see if that dimensional error is consistent. 

 

Now I am fixated on the Malcolm hood.  I now have two 3D printed burnishing surfaces - one male and one female:

 

   

 

 

They were not hard to design and print.  About half a day's work.  The transparency fits nicely into or over either, and I can sand and burnish and polish away. 

 

Next step is to understand how I got that .03 mismatch and make corrections, whatever they may be.  Then, I will have Shapeways do my next transparency, in Accura 60 resin.  The Malcolm hood transparency is just too difficult for me.  I have come close, but no cigar.  The quote is about $45 for one, $72 for two.  I'm going to do it, and hope that is what gets me over the finish line.  When the parts come they will require sanding/burnishing/polishing with the new tools.  

 

In parallel with that activity, I will finish designing the forward side panels.  Lots of interior items that will interface with them - some I will incorporate into the panels, some I will just incorporate locating features.  To worry about - control pedestal, flare gun port, electrical boxes, throttle quadrant, and pulley wheel brackets. 

 

Fun times.  Stay tuned please!    

 

Looking at that last picture could it be that the hood is back to front? It appears the highest point is not in the middle of its length - made me think you might have popped it in the wrong way round...

[airscale]

Fabulous Jay - such precision, skill and ingenuity

 

I am getting to relive my P51C build through you and loving it

 

Peter

[JayW]

20 hours ago, Neo said:

Wiw 75 for 2 thats salty pricing considering you did all the design work and those parts are small

 

Salty indeed.  But I compare that to the hours upon hours I have spent on my previous efforts.  And it is beginning to appear that they (Shapeways) have the magic clear resin - Accura 60 - that will give me the clarity and exactness I  need, and am not quite getting from the two resins I have used.  And that stuff is by all accounts prohibitively expensive to procure oneself except perhaps for mass production purposes.  It's worth it to me if I can get a real good canopy.  This project is one of those where cost is a secondary consideration, if that.  

 

7 hours ago, SwissFighters said:

Looking at that last picture could it be that the hood is back to front? It appears the highest point is not in the middle of its length - made me think you might have popped it in the wrong way round...

 

I wish.  No, its placed right.  If you think about it, the forward end has to be considerably smaller than the aft end.  The forward end, its outside edge (not inside),  must take on the contour of the windshield, even a bit inside it.  The aft end, its inside edge (not outside), has to be proud of the fuselage contour so that it can slide aft.  The Malcolm hood has no hooked cam track to lift it out of the hole when opening, or plopping it back in the hole when closing, so the back end must clear the fuselage at all times.  It just simply slides fore and aft.  The Corsair BTW, the later ones after the birdcage versions, has just such a track shape, where the malcolm-style canopy lifts out of the hole to clear the fuselage when opening.  Much more complicated than the Malcolm hood arrangement.

 

Another dead giveaway is that the stiffening bead on the aft end is much more pronounced than at the front end.  And the picture may be giving the wrong indication - the high point for my hood is indeed about in the middle.  Just sayin'.  

 

Lots more to come with the Malcolm hood stuff.

Edited February 10 by JayW

[JayW]

I would like to report out on my analysis of the .03 height mismatch I discovered on my Malcolm Hood test parts, FYI.  This:

 

 

First, I looked for an error in the Rhino modeling.  It has been a few months since I created the Rhino hood model, and I had two versions of it, where I hoped to find the height error.   Nothing.  All is as it should be.  So next I needed to do some careful measuring of the several parts I have printed up over the last few months.  To do that exactly, I made a simple little jig to place the part in securely such that its height can be measured with calipers, and compare to the digital model.  Extracting dimensions from a Rhino model is easy-peasy.  Measuring the part, not as easy.  Like this:

 

 

 

Note that it controls the width of the part.  Important, as the width will affect the height.  And I found that with remarkable consistency, each part was short (the height dimension) by about .03 inch.  Forward and aft edges.  Some of my parts were made from the Nova3D clear resin, and some with the Resione G217 clear resin.  Didn't matter - about the same shrinkage.  Length (fore/aft) was about right, so no shrinkage or stretch in that direction.  But shrinkage in height.     Something is going on in the printing process which is causing that shrinkage - but I am baffled what it might be.  Wish I had my old hood part made from gray resin, to check it too, but alas I threw it away months ago.

 

Were I to use one of my own parts, assuming anything more I print will have the same anomaly, I would just make adjustments in the side guides, or rails to account for it.  I shimmed the test parts to solve it - my real fix would be a bit more sophisticated than that.  But I want a Shapeways part, and I do not know if the Shapeways part will have the same shrinkage.  Probably not.

 

So what I am going to do is go ahead and order the Shapeways part made from my unaltered Rhino model of the hood, and when I get it back, I will do the same measuring exercise with the same little jig.  And depending on what height measurements I get, I will make whatever alterations necessary in Rhino for the side guides, and print version 3 of them.  Hopefully the Shapeways part will be perfect and I will not have to do any re-design at all.  Not optimum, but I don't know what else to do. 

 

Thanks for putting up with this tedious exercise.  Next post I hope to report on a Shapeways part.   

Edited February 10 by JayW

[geedubelyer]

Jay, how is the hood orientated during printing?

Is it effectively horizontal?

Is it worth attempting a print at an angle? 

I know nothing about 3D printing but intuition suggests there is some minute collapsing or shrinking occuring between layers 

[JayW]

24 minutes ago, geedubelyer said:

Jay, how is the hood orientated during printing?

Is it effectively horizontal?

Is it worth attempting a print at an angle? 

 

Like this Guy:

 

 

Last I printed some parts, I oriented one part vertically with forward edge up, and one with aft edge up.  The fwd edge up version worked best, because the supports had a more robust edge to attach to and left less local deformities (my aft edge has a stiffening bead, whereas the fwd edge doesn't).  

 

So nothing going on between layers - the length of the parts are good.  Honestly I am baffled.  One edge is supported, and one is just in mid air supported only by previous layers.  But both appear to have shrunk in height by about the same amount.  3D printing does involve a bit of growing or shrinking, but not this much.

 

Tilting the part is an option, but (a big but) with the bulbous shape this canopy has, any tilt will increase the angle of one side of the bulge while decreasing angle of the other side.  In 3D printing, the more vertical a layer stack is, the better.  Anything off vertical and unsupported leaves the next layer with some unsupported overhang, and the potential for deformity.   So to minimize overhang angles, such that I would not need to place supports inside the edges (and marr the surface), I oriented the parts as you see. 

 

Thanks for your interest!  I now wonder how Shapeways plans on orienting the part and the supports.....  I told them I don't want supports inside the edges. 

Edited February 10 by JayW

[Trak-Tor]

Fabulous!
I do appreciate the detailed progress report.

[JayW]

6 hours ago, airscale said:

I am getting to relive my P51C build through you and loving it

 

You know of course Peter that I refer to your 70+ page build thread all the time to see how you did it.  You used your "mad skills" at scratch building for most all of your build.  For this one, I am concentrating on 3D printing (cheating!!), unapologetically.  Your Lope's Hope is IMO the world-wide benchmark for a model P-51.   Thanks for looking in.

[JayW]

Calling all P-51 experts - so many of you.  Who can tell me more about this item in the P-51B/C cockpit (LH side by the seat):

 

 

AN3089 signal light with several filters.  And a mounting or stowage clip "A-2542".  I do not believe I have ever seen these items before.  The government spec for the light:

 

 

Looks like some hand held item that is stored along the sidewall of the fuselage in the cockpit.  Anybody have any pictures they could share?  I ran out of patience searching the web.

[Antonio Argudo]

could be this? sorry no full view