Playmobil Boeing B-17 "Flying Fortress": The making of.#3

The Playmo B17 production process. Techniques employed (I)

In the previous post  we described the playmo B-17 master making.

In this article we begin to describe the process that consumed more than 95% of money and time: the small series production method.

The production research and development  were carried out in parallel to the master creation.

One of us focused on mold development  and resin injection techniques. The other of us focused on the master.

Many parts of the master were unmade when the test production started.

This caused a feedback process that led to redesigning many B-17 parts as manufacturing methods were refined.

From the beginning we sensed the complexity and scale of the enterprise. But we fell short again. No step in the process would give us any slack.

The initial snowball had company. And at least one of them was sure to run us over.

Views of the mould manufacturing process.

1.- Process overview.

One thing was clear from the beginning:

The only non industrial way to make a quality plastic toy involved using plastic polymer resins.

Which didn't clarify much either.

Production using 3D printing was immediately ruled out due to:

• Eternal printing times.

• Unreliability of the process.

• Cost.

• Limits on the print sizes.

• Poor quality and resistance of the prints.

• Need for printout post-processing .

• Colour application processes.

Despite of all the hype that had gone along with 3D printing it continued to be a prototyping process . Althought thanks to new technologies, it had become formidably cheaper. 

So after countless hours of blog reading and youtube watching  we started the first experiments with resin and silicone moulds.

And we came to the following conclusions:

a) The dimensions, complexity, thickness and detail of the parts to be manufactured advised against pouring the resin by gravity into the moulds.

The resulting parts were disastrous: incomplete, full of bubbles and fragile.

b) An injection process was required. And there were two possibilities:

• Pressure injection.

• Vacuum injection.

In pressure injection, the resin is pushed into the mould. Mould and resin are then subjected to an overpressure of 3-4 atms in a pressurised chamber.

This removes the bubbles from view but not from the part.

The technique shows several drawbacks:

• The air does not disappear. It remains pressurised inside the resin part at 3-4 atms. All that pressure is only retained  by the cured resin around it. Thus in thin parts these pressurized micro air bubbles are potential fracture points.

• When injecting resin into a large mould, with long and thin passage channels, the air  present before the resin enters the mould prevents it from filling properly.

• A pressure chamber of the size needed to fit the moulds we were building required to be certified by the competent authority .It was prohibitively expensive and dangerous: it was literally a pressure bomb.

In vacuum injection, the resin is sucked into the mould that has previously been introduced into a vacuum chamber, in which it has been done... well, just that, a vacuum.

This vacuum allows the complete filling of the mould and eliminates most of the air in the set resin part, improving its aesthetics and strength.

It is also safer to operate than an pressure chamber: If something goes wrong it implodes. And it has to withstand a pressure of 1 atm., at most. This makes the equipment considerably cheaper.

However, there was a serious drawback: The resin has to set very slowly so that the air it contains has time to leave the mould.

This brings us to another drawback: the need for venting. 

The air in the resin has to find its way out of the mold. So it is required to place ventilation pipes in the parts of the mould where air is more likely to accumulate (higher parts and corners).

Venting ducts in multi part mould.

This last requirement involved  higher resin costs as well as time to install and clean the ventilation ducts.

Besides, a slow-setting resin is more expensive and slows down the production process.

These factors do not matter when the model is composed of few parts. But the  Playmo B-17 is made of more than 250 resin parts, so it is a big problem.


Despite these drawbacks, we decided on the latter method. The pros outweighed the cons.

In time, we thought, we would be perfecting the system and solving the problems.

The goal was to start as soon as possible and see if the process was feasible.

2- Resins.

With resins there were also two alternatives:

• Polyurethane resins.

• Polyester resins.

Polyurethane resins are more expensive but they have the following advantages:

• Lower brittleness.

• Lower shrinkage on solidification, which translates into better fits.

• No odour once set.

• Faster setting speed, which has an impact on production speed.

• Higher number of formulations.

However, they also have disadvantages compared to polyester resins:

• Shorter pot life.

• Shorter life of silicone moulds (20-30 injections).

At first we opted for a slow setting Polyurethane resin (with  long pot life) for the larger B17 parts. It had a higher viscosity but its pot life of 10 minutes allowed us to inject large parts before setting and it would give the air time to escape from the mould.

As we gained experience in the process, and, above all, with the emergence of colouring problems (which will be discussed in the next post), we changed to a fast setting resin, with a pot life of only 90 seconds.

3.- Silicones.

Except as a sealant, silicone is quite expensive. Wherever you put it in...

With silicones, history repeated itself: there was 2 candidates:

• Tin cured silicone rubber.

• Platinum cured silicone rubber 

The first one was cheaper but with a higher shrinkage and lower durability. It was also easier to work with as it is less sensitive to contamination with dyes and clays. Moreover it has longer pot life.

The platinum cured silicone lives up to its name: it's very expensive. It is also very sensitive to contamination and has shorter pot life, but with negligible shrinkage and, in theory, more durable.

Given our initial lack of experience in mold manufacturing and the expenses we were already running into, we opted for tin-based silicone.

Tin-based silicone also had a working time of 90 minutes, which was very useful given the size of some molds and the need to degas all the silicone.

4.-Silicone moulds.

With the mould making process we crossed our Rubicon. This is when we realised the true dimension of the project.

This second snowball was even bigger than the first one.

When it stopped rolling it had a mass of 108 kgs. of silicone, distributed in 30 moulds....

A view : shelf crowded with Playmo B17 moulds.

Along the playmoB17 project we used two part silicone rubber moulds where the halves of the mould form the cavity for shaping the components.

Given the number of pieces to be reproduced, whenever possible, we made molds in which several parts of the same color were casted at the same time.

Two part mold: clay stage. See grey master components.

Open two part silicone mold: after resin injection. 

This entailed the difficulty of balancing the resin flows inside the mould, so the design and  choice of the parts to be injected involved a careful study.

Nevertheless sometimes it was necessary to restrict vents or feeds in order to increase the resistance of resin flows in those parts of the mould that were filled first.

5.- Mould boxes.

In order to reduce the use of silicone and eliminate deformations during mold handling, we furnished the moulds with mould boxes.

These pieces also contributed to the process of mold creation and allowed a better fit between the two halves of the molds.

Playmo B17 belly: mould boxing.

After disastrous tests with fiberglass and polyester resin, we decided to use chipboard strips for mould box manufacturing .

The chipboard strips were stacked allowing a leveled contact plane between the two mold halves. That plane was in fact  the parting line of the two silicone halves.

Playmo B17 belly: mould clay stage.

In addition, the strip width (25 mm) allowed us the insertion of adjustment guides for the 2 mould halves.

Playmo B17 belly: silicone mould half.

6.- Mould duration.

This remains the great unknown.

The sources consulted  give an average of 20-30 uses for the silicone moulds.

This depends on multiple factors: quality and degassing of the silicone, storage temperature, method of demolding and type of resin used.

It seems that the heat given off during resin setting is what wears the silicone the most.

A concentrated mass of resin that gives off a large amount of heat over a small silicone area is lethal.

In our case, the only advantage of working with large and thin parts is that the amount of heat produced during curing is distributed more evenly over a larger silicone surface, avoiding the appearance of hot spots.

On the other hand, the degassing of the silicone avoids the formation of interior cavities that are breakage points when demolding the parts. 

And the demolding process, due to the design of the molds, is not very aggressive with the silicone.

So fingers crossed ....

However, the essential component to make the project a reality was still missing: the injection vacuum chamber. 

This will be the subject of the next post.