Our injection moulding tooling was finished a few days ago, and the moulds were moved from the workshop to the production floor where they were setup on an appropriate-sized injection moulding machine.
I’m happy to announce our first “T0” test shots were produced yesterday at the factory in China and are on their way via FedEx to me!
As you may recall from my previous blog post on injection moulding, T0 is the name given to the very first test shots after the injection moulding machine has been setup with a new mould. They’re used mainly to confirm with the customer that the overall product meets their expectations and it’s expected that there may well be cosmetic and technical issues.
Firstly, let’s look back at one of our prototypes to see see how that looked:
This was printed by a 3D printing bureau in Toronto on a Stratasys Objet 500. The material used is called “Digital ABS” because it mimics the look and feel of real ABS plastic, but is ironically not actually ABS (unlike the material used in low-end 3D printers). Whilst many of us are familiar with 3D printers like the Makerbot, Reprap or Printrbot, an Objet is an industrial grade printer used by designers for prototypes, props, and other parts requiring tight tolerances and cosmetic finish. They cost about $250,0000 each and work in a slightly different way to most consumer level 3D printers because they have a moveable platform that goes down with each layer, and they use a powdered support material. The good news is that the prototype and the first output from the injection moulding machine look similar — if there were any major differences, that would be somewhat unsettling!
Our 3D printing bureau was concerned about the ‘wall thickness’ on our parts because it really stretched the capabilities of their printer. This is a challenge for a product designer, because what works well for injection moulding is sometimes on the fringe of what you can successfully 3D print — even with high-end equipment. What you can’t see from the photo above is some wall warping in the prototype models. You can see that our logo is almost transparent due to the plastic only being ~0.50 mm thick at that section. You will also notice that even a very capable 3D printer leaves ‘artifacts’ such as the ridges on the top of the prototype case. At the time, we couldn’t have the “lens” that fits into the circular holes in the top of the case printed as the bureau wasn’t doing a run of it and it’s extremely expensive to set up just for one customer.
So, what do the first test shots from the injection moulding factory in China look like? Here you go:
The first thing that struck me was that they were black, which I wasn’t expecting. This is the raw state of our “resin” (ours is a mix of ABS and polycarbonate) without any colouring additives. Our next shots will be in our final colour, which is a slightly off-white/very light grey and we will have to work with them to get the shade to our exact liking.
The second thing that struck me was they look much more ‘solid’ than the 3D printed version, which is a relief and a design expectation. A wall thickness of 1-2mm works well for injection moulded parts, and we’re at the lower end of that limit.
Now I’ll walk you through the things that will get fixed as we move to T1 tooling in the next few days.
If you look closely, you’ll notice round ‘swirls’ on the top of the test shots. This is a leftover artifact from creating the tooling (moulds) and in this case is from the CNC machining. When we give the factory the green light that the T0 parts are sufficiently good, they will start with polishing and texturing of the moulds and these marks will disappear in that process. Why not polish and texture the mould before the first run? Well, you technically could — but once you remove tooling steel you can’t put it back, and it’s also easier to see marginal areas of the part without the high cosmetic finish.
The next issue is a little harder to see. We have three parts to the Wimoto enclosure:
#1: A “type A” or “type B” main body. Type A is the one with the slot at the bottom for the Grow, Thermo and Water sensor elements.
#2: A bottom plate that screws into the main enclosure with itty bitty screws.
#3: A clear, acrylic “lens” that is incorporated into the top of the enclosure.
In the next photo I’ll try and highlight the current problem with the acrylic lens component:
If you look closely, you’ll notice that the circles of the main enclosure (the part in black) and the clear circles don’t form a tight fit — they should. This is referred to as shrinkage in the injection moulding industry and is a common issue, so don’t worry. It’s often hard to predict exactly how the resin will behave in a new mould until tests are done and the system is tuned. Tuning might include adjusting the temperature that the resin is heated up to before injection, the pressure used to force the resin into the mould, or even removal of slightly more material from the mould to make the part slightly bigger (a mould is a negative image). Our factory is confident this is an easy fix and they will also be polishing the mould to provide a highly transparent finish so that the lens is optically c;lear (it’s cloudy in the test shot).
It’s important for the lens to fit extremely well to make the device weather resistant, whilst also providing access to the sensors (light and humidity). The humidity port is the tiny hole in the lens on the very left:
We place a tiny (1mm*1mm) piece of membrane behind the humidity port to allow water vapour to travel in and out whilst protecting the interior of the Wimoto. The lens itself is held in place with “sprues”, which are little plastic posts and holes that get melted together to permanently bond the lens and the main enclosure together and make it water tight.
Finally, something we can’t see in these pictures, is the screw holes. Because the screws are tiny (1.2mm threads), it’s not possible to use an ejector mechanism to remove residual debris from the holes created during injection moulding. So, this will be performed by hand at the factory once we go to full production mode. The factory is sending me samples of the screws for approval, and I’ll take some pictures and post them. You’ll also be glad to hear that we’re sourcing tiny screwdrivers to include in every order so that you can remove the base plate and change the batteries easily! Whilst there seems to be a trend in consumer electronics to not provide user replaceable batteries, we felt strongly that you shouldn’t have to throw away your device after a few years of use!
It will take a few days for the samples to arrive via FedEx and I’ll post another update when they get here as well as show you how the final products will look.
Please provide feedback or questions in the Comments section! I’ll go back and answer any remaining previous Comments today when I get a second. If you missed the blog post on the moulds themselves, it’s here — it’s interesting to go back and look at the moulds and try to figure out how they put them together to produce this!