Playmobil Boeing B-17 "Flying Fortress": The making of. #2

A fiberglass Odyssey.

We ended the previous post with an ambitous goal: 

Turning the playmobil B17 wooden model into a master and producing plastic copies out of  it.

Going from top to bottom.

Having taken this fatal step, several problems surfaced.

If the goal was to produce the playmo B-17 in plastic, a wooden master was not appropriate.

First of all, the balsa wood model was 1 mm thick. This was not thick enough to give the toy an adequate strength once it was turned into plastic.

Secondly, to make the interior "habitable" for the Playmobil guys, the balsa wood  frames had to be removed. This would destroy the whole airplane.

A self-supporting fuselage was required.

Also the balsa wood model did not provide the needed strength to make silicone molds.

And it was required "something" that could transfer different parts of the model to make different molds.

The solution: fiberglass sheets left over from a motorbike repair. 

1.- The B-17 Master: A Fiberglass Fuselage.

We started with the most complicated and important part of the plane: the fuselage.

The theory was simple:

Laying fiberglass sheets over the wooden model and then treating the surface until you get the right thickness and finishing for mold making.

The practice was less so.

The sequence of operations to replicate the wooden model with fiberglass sheets involved:

• 1. Sealing the wood with bi-component Polyurethane. 

• 2. Finishing the surface with putty, sanding and sealing it again.

• 3. Applying release agent on the model surface.

• 4. Applying 2 coats of gelcoat to get a smooth and uniform surface.

• 5. Over the gelcoat, applying successive layers of fiberglass sheets impregnated with polyester resin until the adequate thickness is reached.

• 6. Once the polyester resin has set, separate it from the model.

• 7. Use the face of the fiberglass sheets in contact with the wood (negative of the master) as a mold to create the final master. This involved repeating steps 3 to 6 again.

Stages of fiberglass sheeting

And that was the best case scenario...but plans only come together for the A-team.

Vaseline is useful depending on what it is used for…

As a release agent it leaves much to be desired. Following the advice of a supplier, we applied vaseline to the first half of the wooden fuselage we wanted to replicate with fiberglass sheets.

And we proceeded with the whole fiberglass sheeting process.

When we tried to demold it, everything had glued together. And the whole wooden fuselage was shattered. 

The destructive effect of vaseline

This made it impossible to copy the second half of the fuselage with fiberglass sheets.

The leftovers had to be patched as if they were the remains of a galleon, with wood grafts and putty.

So we had to repeat the wood sealing and surface finishing processes.

And then redo the fiberglass sheeting process for the symmetrical half of the fuselage…this time using wax as release agent.

The snowball was picking up speed.

One last thing:

Working with fiberglass is far from healthy and comfortable: it requires the use of PPEs, it stinks up the whole room and it stings like hell.

Despite all of the above, we finally got a fiberglass master of the fuselage with provisional wing and stabilizer inserts.

Fiberglass master (front) & wooden model (behind)

Fiberglass B-17 belly with engine inserts (left).
Fiberglass fuselage provisional assembly  (right)

Considering all that had happened, we decided to look for another way of manufacturing wings, stabilizers and inside components.

But first, there was still something to do.

2.- Fiberglass surface finishing

The surface finishing of the master was critical.

It had to show the peculiar feeling of good plastic toys: smooth and even, yet slightly satiny. 

This is what it had to be transferred to the silicone molds.

The process required for that implied:

• a) Pre-sanding the entire fiber glass master with 240-grit water-based sandpaper.

• b) Spraying the entire surface with grey primer (to check for flaws).

• c) If large flaws were found: Apply epoxy putty and sand it with 240 grit sandpaper.

• d) If small flaws were found: Spray them  with filler primer and sand it with 600-500 grit paper, water-based.

• e) Repeat from b) until all imperfections are removed.

• f) Finally, apply lacquer to harden the surface.

• g) Apply a thin grey primer

• h) And inspect. If faults were found (and there were always plenty of them) , go back to ...e)

• i) Finally, apply very light coats of lacquer to produce a slightly satin effect to the touch and eye.

Fiberglass finishing process: intermediate stage

Fiberglas finishing process: final stage.

3.- The B-17 master: Wings and inside components.

At this point the project had stopped being fun and had become ... something else.

The prospect of making wings, stabilizers and aircraft inners (bulkheads, equipment, weapons, etc) with the fuselage method was discouraging.

As said before, we decided to rethink our work methods.

We had access to CAD software (Catia V5) and we decided to "invest" in a 3D printer to speed up design and manufacturing. 

If you are familiar with CAD design, you will then know that CATIA is not an "amateur" software. It is very powerful but it has its ways.

Fortunately, one of us had some experience with CATIA. This allowed us to resume the work quite straightforwardly.

In a few weeks time, first CATIA models were ready to print (Browning machine guns and plexiglass bow dome).

Ball turret CATIA model: full assembly

Radio operator room CATIA model: full assembly.

Bomb bay CATIA model: full assemby (top)
Navigation & bombardier room CATIA model: full assembly ( bottom)

Cockpit CATIA model: partial assembly.

The wings deserve special attention.

For wing 3D printing we built a CATIA model using a point cloud (coordinates) from the real  B-17 wing airfoil (NACA0018/NACA0010).

This point cloud was downloaded from the NACA website and imported into CATIA, where the length and tapering angle of its surface were adjusted.

Due to their size, the wings had to be printed in sections. And then more problems would arise.

Wing & engine CATIA models: partial assemblies.

4.- 3D printing.

3D printing kicked off well.

We bought a BCN3D Sigma (a filament 3D printer), hoping that it would be a plug & play device.

And, at first, it didn't disappoint us.

The aircraft interior components started coming out of the printer with relatively easiness, but at their own pace.

Printed parts pre-assembly into fiberglass fuselage.

A taste of things to come..

We used 2 types of printing filaments: ABS and PETG.

ABS (black or grey in the photos) sands well and produces a very fluid filament. However, it has a lot of warping and it gives off toxic gases.

PETG (white in the pictures) has much less shrinkage, but it sands worse and it was wearing out the printer brass nozzles. This is the filament we used to print out the wings.

Partial assembly of 3 D printed wings .

However, a filament 3D printer is not a resin 3D printer. Back then we hadn't even heard of it.

That involved another "funny" process of printout surface treatment.

Only this time the process was not long and tedious... but tedious and long.

These were the process steps:

• 1) Pre-sanding with 240 grit sandpaper dry.

• 2) Applying XTC epoxy resin.

• 3) Spraying with grey primer ( for defect detection)

• 4) Applying epoxy putty on defects

• 5) Sanding with 500-600 grit sanding paper

• 6) Repeating until desired results.

And so on with more than 190 component types of different sizes and shapes.

Final finishing of some 3D printed parts.

Complete assemby of navigation, cockpit & bomb bay compartments.

5.- The 3D printer crashes.

Nothing was going to be easy along this project.

The relative ease of printing referred above included typical problems such as warping, nozzle clogging, nozzle wear and tear, etc.

After roughly 2 years, an increasing loss of steps in the printing process showed up. It became evident during wing printing. 

Mails and phone calls were sent to BCN's customer service department, who showed interest but did not solve the problem. 

We started replacing printer components: guides, limit switches, etc.

But nothing seemed to solve the problem.

About to pack the printer and send it over to BCN for inspection, we contacted a guy in our town (Madrid, Spain).

He works with everything related to 3D printing, including repairs.

And he solved the problem in less than 7 days.

It was purely mechanical.Changing one drive shaft and adjusting the tension of the drive belts did the miracle.

We could continue...

Next post: The playmo B17 manufacturing process. Techniques employed.