"Very nice sandy beach ... but the sea is really far !". French Aeronavale (Navy) Potez 25 TOE. 1934 "Pink Cruise"

[KevinM]

Looks really good from here Hubert esp. the PE.I think your upper cowling looks pretty good just short nothing that shim stock could not fix the height  look good.

[belugawhaleman]

Truly impressive work Hubert!

[CANicoll]

Wow, I am still in awe of your 3D printing capabilities.  Nice work!

[HubertB]

18 hours ago, KevinM said:

Looks really good from here Hubert esp. the PE.I think your upper cowling looks pretty good just short nothing that shim stock could not fix the height  look good.

 Thank you Kevin, but the difference is really significant, about 2.5 mm, so that's 1/10 of an inch. Of course I could fix this with shim stock and Milliput, but then, what would I have said about the poor design of a kit that would come with this difference ? Because it was poor design from the guy who drew the new part  

And, actuallly, sorting it out at the source, i.e. revising the design was a one-hour affair on the computer this morning ... The new cockpit cover is in the printer right now, with other parts I modified, like the lower cowling ...

Hubert

[HubertB]

18 hours ago, belugawhaleman said:

Truly impressive work Hubert!

 

15 hours ago, CANicoll said:

Wow, I am still in awe of your 3D printing capabilities.  Nice work!

Thanks guys. As for my 3D-printing capabilities, I would argue that a truly capable guy would have avoided the design pitfall that sent the slicer software crashing again and again for four days .... ... Well, you live to learn, don't you ? That's a mistake I will not repeat, now that I'm a smarter a&$$

We'll see tomorrow, when the current print is over, if I have really gained in knowledge. Yesterday's print was not the success I expected, honestly.

Hubert

 

[HubertB]

Where would a man be without Duck Tape ?

I continue working on the assembly of my Potez. Whilst the prining of the new parts is under way, I have started splashing some primer on the cockpit assembly, but frankly, ther is not much to show here.

And I have decided to tackle other bits, like assembling the main wing. Lukgraph has molded the main wing in three parts, a center section and two outer sections. They are molded plain, and therfore quite heavy. A string joint will be needed to esnure the whole wing stays straight over time. I have drilled some 2 mm dia holes in the mating profiles of the three sections, inserted some brass tube of the same dia, and expoxied everything together.

I wanted to make sure my assembly would be straight and true, and remain so whilst the epoxy setlled. This is where I resorted to that ubiquitous saviour of household men, namely Duck Tape 🤣 😂 🤣!

I think my work should be OK. What do you think ?

More when there is more.

HUbert

[HubertB]

 A small update ...

Me and my big mouth 🤣 😂 ... Before boasting about using duct tape, I should have triple checked the alignment of the wing's 3 oarts. When I removed the tape, it was to see that there was a slight vertical mismatch between the center section and the right outer wing .

After careful use of my JLC saw, adn the application of some brute force, I managed to pry apart the two offenders. Now to redo the gluing, with a multiple check of vertical alignment.

Yesterday's printout came OK, but for the hollowed-out landing gear struts. Start again ...

THis is what the Potez looks like with modified parts. The new resin I used, Elegoo's 8K "red clay", came out sharper than the grey resin I used before. Just look at the radios. The difference is small, but noticeable.

and the Potez with the new parts taped in place ...

 

The radiator and lateral cylinder banks are only loosely tape-fitted. The adjustment is in fact perfect.

Happy me . Now about regluing this wing and continuing painting the cockpit...

Hubert

[BlrwestSiR]

Amazing work Hubert. You make designing and printing your own parts look so easy. 

[DocRob]

You put yourself under some pressure with such an ambitious project, Hubert. I really like the results of your printing efforts, albeit the described methods and software usage are just plain Chinese to me. Bummer with the landing gear struts. Finding long drill bits with the needed diameters will be difficult, but maybe you could push a heated steel rod through, especially when there are at least some remains of the projected channels left. As a better guide, you could drill as deep as possible with the bits available. You could also slice the struts and epoxy some rods in.

Cheers Rob

[HubertB]

5 hours ago, DocRob said:

You put yourself under some pressure with such an ambitious project, Hubert. I really like the results of your printing efforts, albeit the described methods and software usage are just plain Chinese to me. Bummer with the landing gear struts. Finding long drill bits with the needed diameters will be difficult, but maybe you could push a heated steel rod through, especially when there are at least some remains of the projected channels left. As a better guide, you could drill as deep as possible with the bits available. You could also slice the struts and epoxy some rods in.

Cheers Rob

Than’s for the comment, Rob. 
To explain the process for everyone :

1) you create a 3D part with a specific software. I am using Alibre Design, but the most commonly used is Fusion 360, which has a fee licence for hobbyists.

2) When the part is designed, you export a printable file. The oldest format is .stl, which results in very  heavy files, and I am now using the .obj format

3) the export file is uploaded in a slicer software. The function of this software is literally to slice your part in a number of layers, which the 3D-printer can then use to print the part. The printer will print each layer in succession, with a UV projection that will polymerise the resin (for an SLA printer such as mine). 
 

4) in the slicing software, you prepare the parts for printing. Basically, you decide of the orientation of the part, and of the supports that need to be added to help « hold » the part during the printing process. The printing takes place from the bottom of the part, up. Only the platform on which the part will be printed is inverted and will progressively lift as the printing goes. The platform is immersed in a vat with the UV-sensitive resin, and the bottom of the vat is clear to allow the UV projection.
To get better results, and avoid visible printing lines, when the human eye is trained to see in two horizontal and one vertical axis, you tilt the part at an angle in the slicing software. The supports can generally be automatically created, but, most of the time, with a bit of experience, you feel where more supports are needed. The worst you can have is either a part insufficiently supported, that will deform, or sometimes detach from the platform. Especially as the lifting of the platform has to fight a suction effect from the viscous resin it is immersed into.

5) when you are happy that you have a printable part, or group of parts, you order the slicing software to slice it. I have set the layer height on my printer at 20 microns, so that means 50 layers per millimeter of vertical height. You end up typically with a print that is going to be made of 5000, 8000, 15000 layers. And each layer takes a few seconds (2.5 for UV exposure, plus the time for the platform to lift a few fractions of a millimeter, and lower again), so that is why typically a print takes a few hours. For me the print runs were between 8 and 13 hours.

6) When the last layer is printed, you have to wash the parts from the uncured resin that still adheres to them - remember the resin is more or less viscous. For most resins, the washing is done in isopropylic alcohol. It takes a few minutes to wash the parts. Before being usable, they then need a « post-cure », which is done by exposing them to UV-light for a few minutes, in a curing machine, or leaving them in the sunshine.

Et voilà !

In the case of the struts where I want a hollow center for the brass rod, inclining them for printing may result in resin « sticking » inside the tube you progressively print. That is why I ended printing them vertically, so that the center hole is aligned vertically layer after layer. Even then, you have to remember that you may have a suction effect that will keep the liquid resin trapped in the middle of the hardened resin part ( like the old magicians’ trick of having a glass of water upside down with a a sheet of paper keeping the water in the glass). So you need to design ventilation holes. All of this is a bit empirical, and the small size of the parts makes the process more complicated, as the size of the holes is limited, when the surface tension of the resin may fight against it flowing easily through the holes …

The resulting resin is very hard. It dan be drilled, sanded, milled, sawn, cut, but I am not sure it will react positively to a heated rod. And of course, I tried slicing the struts in halves in the design, but this is where I ended up with a part that was sending the slicing software into fits …

I am making progress with each new printing trial though. So it is just a matter of time before I get a satisfactory result. But printing a 50 mm long strut vertically takes 10 hours, plus the time redesigning the part, and setting it in a printable way in the slicing software …

Hubert

[PanzerWomble]

.........& I thought getting to grips with my Silhouette machine was 'cutting edge '     😳*

 

 

 

*repeats to self not going down this rabbit hole 

 

 

Bravo to you though Hubert , this will be a build for the books ! 

[HubertB]

14 hours ago, PanzerWomble said:

.........& I thought getting to grips with my Silhouette machine was 'cutting edge '     😳*

 

 

 

*repeats to self not going down this rabbit hole 

 

 

Bravo to you though Hubert , this will be a build for the books ! 

Well, the Silhouette is, literally, all about using "cutting edge technology » .

Just saying …

Hubert

[HubertB]

Who said "Repetition is the mother of skill" ? Once again, my tribulations have proven him (or her) right !

I can now say that I have a number of successful prints of the damned landing gear struts. I launched a new print this morning, with four copies of the struts, two vertical and two at different angles. One vertical print fell off from the plate, but the other three are a success. Before curing them, I did try to get a rod in each strut, along their whole length, and it worked !

"A small step for modelling, but a giant leap for the Potez project". I am in a quotes' mood tonight 😂. More seriously, it means I can envision building a strong and scale-accurate landing gear for the Potez 25 TOE.

Time to celebrate 🥃 🥃 🥃 (after those, forget about using any pointy or sharp tool, and trying to glue teeny weeny PE parts - btw, I was mentioning in another thread my close-to-maniac habit of cleaning the bench at the end of every session. This allowed me to find two really really really tiny PE levers that I thought had flown on an orbit between Mars and Jupiter, among the very few shavings and dust particles on the bench )

Hubert

[BlrwestSiR]

Hubert, thanks for the description of the various steps involved. We recently got a Bambu Labs A1 which is a filament printer. Our youngest has been having fun downloading and printing files that are ready made. Hopefully he gets into the design side of things and he can learn how to make his own stuff.

[KevinM]

I am happy Hubert you finally had success with the prints and yes the red clay has a better presentation.The instruction on the printing I would have to be standing right there as you describe it and perform the magic.I am keeping my eyes on this so await the next installment Sir.

[HubertB]

8 hours ago, KevinM said:

I am happy Hubert you finally had success with the prints and yes the red clay has a better presentation.The instruction on the printing I would have to be standing right there as you describe it and perform the magic.I am keeping my eyes on this so await the next installment Sir.

Thanks a lot Kevin. I am happy to report that I have been able to assemble a reinforced landing gear. With 1.5 mm brass rod for the lower axle, 1.2 dia in the four front V braces and 1 mm in the rear diagonal struts, it is plenty strong, and nothing to compare with the resin-only first model. And it still looks the same, and scale-accurate 

Update pics soon.

Hubert

[Peterpools]

Hubert

Absolutely incredible work and your attention to accuracy is inspiriting.  Of course, I have a bit of a hard time following some of your issues and thanks to the photographs, have a good idea f the work that has gone into the conversion.

 

[Peterpools]

Hubert

Just finished catching up and am completely blown away by the effort, time and knowledge that is required to print at your level. I wish I were a lot younger and would have loved to learn printing but being almost 77, knowing my limits keeps the frustrations and costs down. 

I also was amazed at the difference of the detail and sharpness of the printing between the two resins. 

Of course, your initial question of whether or not we have major reservations on cutting up an expensive model and in today's ages, that seems to be most kits and all the time, as there is always the worry, something will go so very wrong, and the kit then qualifies for the bin.

I've been an OOB builder for decades and it so greatly helps keep my nerves from going into overdrive.

 

[Landlubber Mike]

Those custom printed parts look amazing Hubert!  And thanks for outlining how the process works!

[DocRob]

Good, that the printing of the landing gear worked this time with the reinforcement inserted.
Thank you for the detailed description of your printing process. I may have to look into that myself. It´s not that I´m completely out of the dark, as I produced 3D animations in Lightwave, when I was young. I had a "real" 486 processor by this time and rendering the results took weeks. I learned to design 3D objects then, but there was no need for accuracy, it only had to look good.

Cheers Rob

[HubertB]

Measure twice to print once, and then 3D printing woes ...

Well, I expected far more progress than what I have been able to achieve over the last 10 days.

I have started preparing the fuselage for assembly. When checking the fit of the various components, I realised that :

1) the cutout in the front corners of the main fuel tank extensionwere not enugh to ensure the fitting of the nading gear. SO back to the (virtual) drawing board.

2) but also, I had not designed properly the landing gear, with the rear diagonal struts ending some 4 mm too far forward than they should. To make matters worse, I had shortened the "wheel base" even more when I assembled the various parts of the landing gear with the reinforcing brass struts  ! So, guess what, back to the drawing board again, this time with a different approach that will allow a correct geometry.

Fine enough, but then I have to tell you more about 3D-printing finesses

There are many variables you can adjust when 3d-printing with resin. To explain a bit more the way this works, once the slicer has produced the big file with all the thousand of slices, and you import it in the printer for printing, you can launch the printing operation. As i mentioned, the supporting plate for the printing is upside-down, lifting progressively, layer after layer. The bottom of the resin vat in which the plate is immersed has a transparent film bottom, and the "image" of the slice to be printed is projected with a UV-led through the film, triggering a photo-ploymerisation of the resin, the layer to be printed becoming solid. In the process, the printer then lifts off the plate, to break the suction from the resin on the film, and lowers the plate again, with just the new layer thickness difference.

When setting a print session, you can adjust:

1) the layer thickness

2) the exposure time for the UV projection

3) the exposure time for the "bottom" layers, that will stick to the printing plate. You want to ensure a strong adhesion, that will fight the suction and tensions of the printing process

4) the rest time after exposure of one layer, i.e; .he time before the printer lifts off the plate

5) the delay after the rest time, before the actual lifting of the plate,

6) the lifting speed of the plate, for the bottom layers and the ordinary layers

7) the speed at which the plate is going down again

8 the area, shape and thickness  of the "raft" which is the part that will stick to the plate, and hold all the supports for the printed part

9) the orientation of the part relative to the printing plate

10) its distance from the plate, which will determine the base length of the supports

11) the density of the supports, which is then automatically calculated. Plus you can then add more supports if you tink it is necessary

12) the shape of the contact point between the "raft" and the support, its diameter

13) the diameter of the support beam, at its "bottom", middle and tip

14) the shape and diameter of the contact point between the support and the printed part

15, the depth of the contact point "in" the printed part

17) the shape and density of cross braces between the supports, to ensure they do not bend or break as the printing process takes place, with its suction and tension forces to counter-act....

I am sure I am forgettig a few more parameters, but, if you count the multiple data in the above list, it means you have about 30 or more parameters you can adjust for the printing process ! Which means about as many reasons for a print to fail.

Then there is the type of resin, its colour, its viscosity ...

Then there is the cleanliness of the printing plate, its "grain' to ensure a proper adherence of the first layers ...

Then there is the transparency of the film, and whether it develops its own stickiness - the isopropylic alcohol you use for cleaning can induce a stickiness to the film

Then as I found out, you have outside factors like the air temperature, the relative humidity and the resin temperature that will also have an influence ...

I mention this last point because the temperatures have lowered recently, and the humidity climbed like a rocket, as we got the successive storms coming from the Atlantic.

So, to make a long story short, I have had 10 days of successive print failures, sometimes with 2 print runs in the day. And all of this using the same - new -resin I had used for my previous prints . The problem was that the supports failed to « hold » and seemed to delaminate and break away from the raft. All the tweakings I tried, and there are a few possibilities, as you can guess, resulted in the same miserable fails ...

Yesterday, in despair, I came back to an older version of the slicer software, just in case, with the original parameters, and  using  a resin that is supposedly 2 years beyond its "use-by" date. And it's a 95 % success !

So, where do I stand to date ?

I have started painting the cockpit structure. It was wooden, so I used oils on a sand base. The result is a bit too uniform for to my liking, but there is no way I am redoing this now, especially as the cockpit openings is larger than on some other biplanes, but not that large in the end.

I have started adding reinforcements for the inside, for all the areas that will be carrying either the weight of the wing, or of the whole finished aircraft.

This is when I decided that I did not like the PE parts from Lukgraph. The problem for me lies in the fact  that the brass sheet is quite thin to begin with. After the etching process, you end up with too thin and too frail-looking bits, which do not resist the painting process to top it off. The solution to that ? Print some new parts like the throttle of the trim wheels (the latter wrongly placed on Lukgraph's instructions). Assming of course you can print them satisfactorily ...

This what the painted cockpit looks like as of today. The front bulkhead is an add-on not yet fully painted, that, like on the real aircraft, bears the structural loads of the wings and landing gear.

And a close-up on the PE parts

 

 

Among the multiple fails, I had some parts that printed well, to replace the PE components. They are in dark grey or red clay colour in the pic :

 

A lot sharper and with a more "solid" feel than the flimsy PE parts your can see on the right of the pic for comparison.

But wait ! This is a French aircraft. And on French aircrafts, the throttle in the "off" position is pushed forward, not pulled backwards like on British or US planes. Who wrote "Measure (think) twice before commiting to printing ?" Some smart a#$$ I am sure .

So back to the drawing bord. That's a two-minutes mod, so no big deal, but now I have to print those again.

TBC

Hubert

 

[KevinM]

Lots of parameters to the printing I think you scared me from that venture Hubert.The wood of the copit looks just fine by me.

[HubertB]

To be fair, the slicing softwares have a database of printers with pre-set parameters most suitable for the said printer. But some factors can influence the printing, and this is where you start playing with the parameters. 
In the end, I think the resin reacted badly to humidity and /  or air temperatures, to loose the tensile strength that allowed the supports to resist the suction forces when lifting the plate and print.

Hubert

[DocRob]

Excuse me for not digesting all mentioned parameters, Hubert. I had no idea, what it takes to get decent printed results and you only described the slicing process. I know the feeling, when you thought you did everything right, but no success, but I guess, the thorough way you do it is more than a bit of pioneering.
I like your wooden framing, but if you want to let it look less uniform, you can always add some matted or semi matte clear tones in orange or yellow shades or rub in some brown pigments.

Cheers Rob

[HubertB]

2 hours ago, DocRob said:

Excuse me for not digesting all mentioned parameters, Hubert. I had no idea, what it takes to get decent printed results and you only described the slicing process. I know the feeling, when you thought you did everything right, but no success, but I guess, the thorough way you do it is more than a bit of pioneering.
I like your wooden framing, but if you want to let it look less uniform, you can always add some matted or semi matte clear tones in orange or yellow shades or rub in some brown pigments.

Cheers Rob

Thank you Rob. I do not want to scare anyone away from trying 3D- printing. Truth is it can be very easy - both machine manufacturers and slicing software developers work hard for that - and you can get a success  (almost) everytime,

But it can also be extremely frustrating when using the same parameters results in a failure when it was a success the day before.

I think you should read from all my errands that I am not as good as some said in this thread  . Ruffling my feathers to send me back to the ground is not a bad thing , even if the landing may look more like a crash sometimes 😂🤣😉.

Anyhow, I am now back to good printing results. FWIW, Anycubic’s « Translucent green » resin has proved to be both a high resolution and a fail-safe product, again, as I used it a few years ago for printing my Travel Air Mystery Ship.

I intend to try a bit more modulation in the cockpit and follow your ideas. Right now, I shot some primer on the newly-printed cockpit-détail parts like the throttles or trim wheels. Modulation trials comes this afternoon.

And I designed and printed yet more new parts for the Potez. I was wondering how Lukgraph had managed to print the wing and cabane struts hollow, in order to inset a reinforcement rod. Well, they have not. The holes are only about 8 mm deep. I managed to drill a complete hole in a 60 mm long strut, but it was tedious, frankly, and I was lucky to be able to - almost - meeting holes when I was drilling the strut from both ends.

And then I realised that Lukgraph had the struts wrong anyway. They are angled outwards. Lukgraph says it should be 27 degrees. A drawing in the « Potez 25 » book shows 39 degrees, and is wrong, but checking on pics and the Potez factory drawings. I found the right angles, both from the front and from the side. And the print this night worked great, including see-through holes for the reinforcing rods.

So far, I have designed and printed some 30 different parts for this project. i have still to design the pilot and passenger windscreens and print them in clear resin.

More later.

Hubert