Technically Speaking 1: Mould Insert Testing

​About Getting Technical Updates

Part of our mission is to educate people. To help people see the problem and potential of plastic. To help people access the potential of this material we need to give them the tools and the knowledge to be a part of the solution. This update marks the first in a series of on-going technical deep-dives summarising our experiences so that others can learn from them.

Technically Speaking #1 : Mould Testing

This first update discusses our recent experiences whilst exploring our new injection moulder's capabilities, and the research we've done around the modifications of the basic moulds we have in order to expand those capabilities. I'm breaking this update into two sections based on the two main materials we're working with at present: LDPE and PP.

The Injection Moulder

We acquired our injection moulder with a financial donation from the Friends of Lancaster University, an alumni fund providing money for projects which might not otherwise be funded but which provide access to new and exciting experiences and opportunities for Lancaster University students. We acquired the machine itself from a Precious Plastic inspired Austrian team called Plasticpreneur. The frame is made entirely from 18 and 21 mm plywood which makes it strong but lightweight, and therefore portable. It also features two big improvements over the standard Precious Plastic design - a rotary injection mechanism and a chamfered injection nozzle:

​The rotary injection mechanism provides us with additional mechanical advantage, reducing effort we have to expend to perform a successful injection. The chamfered nozzle means we don't need to screw our moulds to the output / hotend. Instead, we slide our moulds under the output and press the mould against the hotend using a spring-loaded platform for a compression fitting, reducing the time taken when injecting.

The Moulds

Moulds become expensive very quickly because they need to be cut from metal (usually steel or aluminium) and the machinery and skills required to do this is not trivial. We decided to start simple and purchased three basic moulds. Each mould is constructed from hardened 6 mm sheet steel. Moulds consist of a top plate (with a 45 degree chamfered injection port and 4x M6 bolt holes), a base plate with tapped M6 holes, and a mid-plane with a shape cut from the sheet steel (a circle or hexagon).

---

​Polypropylene Testing

Polypropylene is one of the main source materials we are looking to use for the products we aim to produce. It is relatively easy to mould, with a low-ish melting point (~180 degrees C) and has a great surface finish. It's one of the plastics which is not currently accepted by the kerbside recycling system in Lancaster.

Our initial tests showed that we needed to be up at about 230 degrees on the injection moulders controllers in order to get a successful injection. Finished products are very tough, with a slightly matt finish. Melting PP does come with a slightly unpleasant chemical smell when it is melting so air extraction / venting is preferable.

​With the initial testing successful, we looked to see if we could customise the moulds we have by placing inserts into the mould to create custom shapes within the build envelope of the cut out. In the below designs you can see a red insert placed inside the steel mid-plane which yields the yellow plastic part when injected. This technique would allow us to make new, custom parts without the hassle and expense of having new metal moulds made - providing we can find a material to make these inserts out of.

We tested a number of different materials for inserts and looked at making positive and negative insert styles. Compatibility with PP is more limited than it is with LDPE, due to the higher temperatures involved.

PP with Acrylic Inserts

​Acrylic can withstand the temperatures involved here allowing us to make parts with generic profiles. The cross below was made using two pieces of 3 mm acrylic glued together to create a 6 mm thick insert. You can see the thin line halfway up the part which denotes the halfway point between the two layers of acrylic.

​Thin pieces of acrylic required for making engraved impressions were less successful; the material did not penetrate the tight corners of the insert yielding poor resolution of the final part. The insert itself warped under the heat and was more brittle afterwards, failing completely during the removal of the final part.

​​PP and PET Inserts

​One of the exciting ideas we had was injecting recycled material into custom 3D printed mould inserts. This would allow us to create some really intricate shapes due to the flexibility or 3D printing. Given the current availability of 3D printers, this is something we were keen to try out.

We had previously tested injecting LDPE into a 3D printed part made from PET (the same plastic used for water and fizzy pop). PET has some heat resistance to the temperatures involved but it cannot do this for long. We did produce one successful test and one failed test using a PET 3D print. We feel this warrants further investigation.

​PP with PLA Inserts

See the disappointing results injecting LDPE into PLA moulds for more details on why this wasn't possible.

​PP and Wooden Inserts

​We tested a few wooden moulds with PP to see how badly the plastic adheres to the wood, since heat resistance isn't an issue. PP adheres very strongly to laser cut plywood. We tested using sealant and mould release on the inserts with no significant improvement. It might be worth soaking inserts in oil or water to see if this improves their suitability, but care must be taken with water due to rapid heat expansion and evaporation in the enclosed space of the mould. 

​PP and Nylon 6 Inserts

​Sadly, Nylon 6 cannot be laser cut like acrylic and wood. Instead, we needed to rout parts from a large piece using a small CNC router. Without the necessary experience however, we had issues creating these moulds and needed to revert to testing small sections of the material in moulds to check adhesion. Nylon 6 seems to work well for PP, and was not bonded to the part at all. It is worth noting that due to PP's lack of pliability, removing parts from the mould might be difficult, but this material looks very promising.

The small section of Nylon 6 pictured below was dropped into the circular coaster mould and you can see that we were able to extract the piece after injection. The piece was surrounded and hard to extract but it was undamaged by the temperature of the PP.

---

​Low Density Polyethylene Testing

Our source of LDPE is industrial waste from a manufacturer of wheel nut checkpoints for trucks and industrial machinery. As a result, the material is almost all brightly coloured. It is also soft, flexible, easy to mould and work with and has excellent matt and gloss surface texture. It is a great material to work with and we've found it extremely versatile.

Initial testing revealed interesting surface textures  and low melting points which we will need to research more systematically to get right.

​LDPE with Acrylic Inserts
​
LDPE works really well with Acrylic. For initial tests we used engraved 2 mm acrylic inserts and placed these inside the mould cavity so that plastic was injected down onto the insert. The engraving shows up as a negative of this pattern. The material creates very good detailing and excellent gloss surface finish on the sections which meet the gloss acrylic.

LDPE with Multi-layer Acrylic Inserts

Since the acrylic we had in stock was only 3 mm thick and our mould cavities are 6 mm thick, we also tested inserts with multiple layers of acrylic bonded together. Although these worked OK, we do see some poor edge definition on these parts which might be due to the angle of the laser at the bottom of the cut.

LDPE with PLA Inserts
​
We tested inserts 3D printed in PLA with the LDPE we had available. The LDPE is too hot for the PLA despite its lower melting point and the inserts warp and fail. 

​LDPE with PET Inserts

3D printing parts with recycled PET is difficult as the consistency of the material is not perfect. However, it does have just enough heat resistance to be able to deal with LDPE's injection temperatures. We managed to get a single injection from this gear part, although it had significant over-spill due to shrinkage of the PET (final part was more like 5.1 mm high). Even so, this part came away from the mould cleanly and without adhesion.

​LDPE with Wooden Inserts

We tested several inserts made from laser cut 6 mm birch plywood. We also tried coating these inserts in mould releasing agent but this seemed to have very little effect. Parts did come away from the moulds cleanly but surface finish was not ideal do to the rough texture of the inserts.

​LDPE with Nylon 6 Inserts

Finally, we tested several Nylon 6 inserts with LDPE. Nylon was suitable for use with LDPE too and showed no signs of adhesion between the two materials and, due to the flexibility of the LDPE parts were easy to separate. Again, Nylon 6 inserts will need to be routed out with a CNC router because the material cannot be laser cut with a CO2 laser cutter.

---

​Conclusions

We've successfully performed preliminary testing of inserts to modify our basic moulds. We found that LDPE is compatible with acrylic, Nylon 6, wood, PET and PLA. PP was only compatible with acrylic, wood and Nylon 6 but also had some partial compatibility with PET printed parts.

We are keen to keep the waste generated by our activities to a minimum. Nylon 6 and PLA would represent the use of virgin plastics, whilst we can use recycled PET and acrylic. We are therefore keen to test Greencast or recycled acrylic to reduce the impact of this kind of testing and production. We are also keen to see whether wooden inserts could be used if we soak them in oil / water prior to injection. Surface finish for wooden inserts might be improved by finishing the mould itself by sanding and coating with polyurethane varnish.

Finally, we are keen to test larger, multi-layer 3D printed inserts and see how well these injection mould. We will also investigate SLA resin parts.