Fishy Blog

17 days continuous immersion, 10 boil washes and still going strong

Everyone knows that nylon is hygroscopic right ?

It absorbs water from the atmosphere until it reaches an equilibrium point and that property can cause issues when 3D printing using nylon filaments. But what happens when you print a 3D form then immerse it in, lets say, a washing machine simulation?

The following photo shows a bunch of control forms, then the same design of form made from the same batch of the same recycled nylon monofilament. The second bunch has been immersed for 17 days continuously, exposed to sunlight (through a jar) and boil washed 10 times with a generous helping of a non-biological washing detergent called Surcare. OK, its not the harshest detergent in the world, but I hope you agree that this test scenario represents a good first go for any clothing related applications.

17 days and 10 boil washes
Left; Control form made of recycled nylon fishing nets and 3D printed. Right; Same material after 17 days immersed in detergent/water mix and including 10 ‘boil washes’

Some things to notice;

There is a slight bleaching of the original so we know that the detergent has been doing its job, but otherwise the nylon is showing is legendary resilience. There is no delamination, no cracking, no warping or visible deformation due to water ingress or absorbtion.

There is one key change though; flexibility.

Check out the clip.

As you can see the second form is far more easily deformed under pressure. It springs back very well when pressure is released but there is a definite change in its deformability.

In fact, the ease of deformation rose rapidly to a peak after a couple of days immersion and stayed around the same after that suggesting that the material had reached some kind of equilibrium state.

Obviously 17 days and 10 boil washes isn’t enough to prove definitely that 3D printed nylon forms can be used in clothing applications, or indeed fisheries, marine or water management applications, but its a damn good start. And I’ll repeat my statement from a previous post that this is all done using a sub-£400 printer and pre-commercial filament.


Fiddling and fettling

While the gross parameters of recycling used nylon monofilament fishing nets are now pretty well known and understood, efficient production of high quality 3D printer filament is a work in progress.

Tweaking the extrusion speed, the temperature, the material preparation and presentation and the way that the resulting filament is handled as it cools are all in play. They vary with ambient temperature and humidity, and can change radically if a speck of errant dust makes its way past our quality checks. Knowing where to focus effort is an engineering calculation, driven by methodical collection of data on each test run and analysis of the trends that data might show.

Aesthetics and engineering are only two parts of developing a new filament, we also need to understand how customers might use the material, what sort of lifespan it might have, how it might be disposed of, recycled or reused. Fishy Filaments is a Circular Economy business model, so we have a responsibility to understand the material’s full life cycle.

In progress right now are;

Extrusion consistency optimisation (how do we make the most, best filaments from the material available)

UV/fresh water soak test (what happens when a printed form sits in sunlight when wet?)

Long term water penetration test (what happens when a 3D printed form sits in fresh or salt water but in the dark)

Detergent immersion and repeated wash cycle test (what happens when you soak a 3D printed form in washing detergent and subject it to repeated heat cycles that replicate a boil wash)

That said as we get better at producing filament and printing different types of form with it we are also producing nice items like these, just because we can.

This slideshow requires JavaScript.


Moving from aesthetics to engineering

The previous few posts have been primarily concerned with exploring the aesthetic potential of the recycled nylon from Cornish fishing net. I hope you agree that this was useful, and as I’ve repeatedly stated these are only explorations. Its up to the customer how they use the material. Its my job to open doors not close them.

So as production testing advances and the process is refined behind closed doors (I’m not going to tell you what I’m doing to improve the process so don’t ask), I’m shifting from look at feel and towards empirical quality in engineering terms.

For those who already 3D print in nylon there will be little news here as regards the technical challenges and a quick look at the result will inform the informed about the kind of testing that I am doing. For the unititiated there are some underlying challenges related to the material science behind nylon production, use and re-use that have to be addressed to the satisfaction of a technically motivated customer. Prime amongst these is the hygroscopic nature of nylon.

A hygroscopic material absorbs water, not like a sponge into open voids, but into its molecular structure. Not all plastics are hygroscopic and some physical forms promote the attribute where others reduce it, in general the higher the surface area to volume ratio the more sensitive the material will be. Nylon 6, the main constituent of monofilament fishing nets, is hygroscopic and has been shown to absorb up to 9.5% by weight of water. So simply allowing nets to drip dry prior to processing isn’t enough. We do need to dessicate the materials arising from them at some points in the recycling process.

Of course water isn’t just present in the sea (shocker !), its also present in the air as humidity, can be released by the materials when they are heated and is present on everyone’s hands, sweaty palms or not. So it is important to know where water is, how it interacts with the materials, how to mitigate its actions and when is best to mitigate its actions.

You could go nuclear on it and dry everything to within an inch of its life and live in a bubble. But apart from the detrimental health effects on workers, if you stuck the whole process in a zero humidity clean room, that isn’t going to be the best or most economic use of energy or capital.

Long and short of all this dull, repetative, behind the scenes work is that the Fishy Filaments recycled nylon filament made from used fishing nets with no additives and no harsh chemicals used in its processing can definitely be used to build extremely tough and resilient 3D printed forms.

Test bed evolution as customer service

Any R&D exercise is only as good as its test environment and I have previously written on the modifications carried out to the baseline Wanhao Duplicator i3 that I use for materials development testing.

First thing to say is that I chose this model for several reasons that go beyond simple cost and include the ability to modify the machine component by component. Only by changing hardware parameters can we explore how a wide variety of customers will experience the end-product. That could be done through recruiting a test panel with access to a wide variety of printer hardware but in the first instance we need to be able to provide a performance envelope so that we are not asking that panel to take too many risks with their machines.

The first mod that I carried out was to build the printer a cabinet. This was a cheap Chinese server cabinet, so 100% metal-built and flameproof, but more salient is its ability to block drafts and maintain a constant temperature.

Second mod was the replacement of the hot-end with and all-metal product from Micro-Swiss. This was done to simulate higher-end printers but also on the assumption that it would allow more reliable printing at the elevated temperatures that the Fishy Filaments recycled nylon requires. Actually it didn’t perform well at all. I found that the aluminium barrel (the piece that replaces the PTFE tube found in most hot-ends) promoted deposition of nylon rather than preventing it. I can’t say whether that is due to friction, surface scoring due to repeated cleaning, heat distribution or any other reason, but what is apparent is that the PTFE system works better overall, even at 275°C. I have now swapped back to the original hot-end, albeit with a new heating element and thermister.

However it isn’t perfect and at that temperature, probably due to differential expansion, the fit between individual components of the hot-end isn’t really close enough, so it allows leaks of molten nylon to seap through threads.

The third mod is relatively minor and is difficult to assess definitely. I replaced the plastic extruder plate and spring loaded lever arm with a machined metal version, again from Micro-Swiss. I was concerned about filament slip as the extruder assembly heated up and could see the possibility of flex within that component contributing to that issue. I can’t say whether the replacement is much better or any worse on its own, but I don’t intend to swap it out at present.

The fourth mod was removal of all cooling fans from the extruder assembly. Which on the face of things sounds crazy given the other issues and modifications carried out. Why would I deliberately make the extruder assembly cooling less efficient. Well, its difficult to get a machine rated to 260°C to operate at 275°C reliably and the fans just get in the way. Airflow fluctuates as the extruder travels around the bed, heat is sucked away from the surface of the hot-end and the leaky material flow needs frequent attention to keep print quality high. All in all its just easier not to have fans getting in the way. Most makers of printers rated for nylon recommend no cooling on the print itself as a means to promote inter-layer adhesion, so I’m just taking that advice a step further.

The fifth mod is swapping out the covering on the aluminium bed for a borosilicate glass bed. I noticed that bed levelling was getting increasingly difficult through testing as I pushed the existing system to 110°C. Bed warping is not unknown, especially at the lower end of printer specification, as the cheaper bed heating elements are close to being a point under the centre of the bed rather than a trace element covering the whole bed. You end up with a central hot spot with bed levelling adjustment screws located as far away from that spot as possible. Getting a true level surface becomes a bit like juggling jelly as you try to balance the bed temperature at the time of adjustment with the inevitable finger burn of levelling a bed running at over 100°C.

Covering the bed with a glass surface provides a means to avoid that warping, reducing levelling time and reducing the frequency of levelling. It also provides the opportunity to remove the surface from the printer, making printed form removal and tape renewal far easier, especially when you have a cabinet around the printer.

None of this will greatly shock the 3D printing officionados out there. All the techniques have been tried and tested by others but what I’ve ended up with is a test rig that simulates aspects of printer design that different printer makers have addressed in different ways. Only by fully understanding the design decisions made printer makers can we provide advice on how to cope with variations away from the manufacturers nominal use profile. If there were any field of manufacturing where deviation from the equipment maker’s instructions were likely it is 3D printing, so to me this attention to detail lies at the heart of being a good materials supplier; anticipating the issues that your customers might encounter and finding potential solutions before they get there. Or if you prefer, this is the old adage ‘Walk a mile in another person’s shoes’ at work in defining a customer experience.


Second Pass Dye Tests – Deep Blue Delight

Encouraged by the first pass results on our recycled nylon filament made from used Cornish fishing nets, I’ve dyed some 3D printed test forms in exactly the same way. Same everything within an acceptable margin of error.

This slideshow requires JavaScript.

Middle ring is the control and is not dyed.

In strong white light the orange looks a little bit odd with the underlying green-blue body colour, but the blue is a very attractive deep royal blue with just a hint of green where the dye penetration is less complete. Not shown here but in strong Cornish sunlight the blue is especially pleasing.

If anything the colour intensity on both forms is better than the simple filament tests and there are obvious geometrical reasons why that might be.

A non-obvious effect of the dying process is that the forms have become more flexible when compared to the control. Nylon filaments have a reputation for absorbing water so we’ll have to see if that flexibility remains after they have dried more completely.

Just to re-iterate; These tests are all about colour not print quality, so photographic evidence is never going to be as representative as the naked eye and my opinions are subjective. Also the print quality isn’t the best, but that’s mainly because I’m using a super-cheap printer on a material that it isn’t rated for and operating the machine well outside its design specs.

But all in all I’m really happy to have confirmed the findings of the first pass and then gone on to demonstrate that there is real potential here to expand design horizons. I have a couple more dye tests to do but next week is Cornwall Chamber of Commerce Business Week, so maybe I’ll see you there 😉

First Pass Dye Tests – Another Door Opens

I’m not going to say too much about the technical aspects of this photo except that the dyes used are off-the-shelf, suitable for domestic use. The dye penetration was about 30% of the cross section, so not complete, but not just a surface effect either. All three lengths of filament came from the same batch and run so started as exactly the same colour (the middle filament).

First Pass Dye Test

Obviously nylon takes a colour-fast dye. Some of your clothes are testament to that fact and nylon-centred 3D printer makers Markforged have already shown how they apply that attribute. My research question here was whether it was worth chasing that fact as a saleable attribute in our finished, Fishy Filaments, recycled product.

Dying isn’t something that 3D printing enthusiasts do much, largely because most filaments are made as opaque, containing relatively expensive pigment additives alongside the original body colour. In most plastics that ‘natural’ colour is a neutral off-white. Adding pigments can alter the engineering qualities of 3D printing filaments, as well as the colour, so many of the more technical materials are only sold as ‘natural’ colour.

Pigments don’t mix in the same way as dyes and because they operate by reflecting light off quite large particles they tend to be more UV resistant. Dyes tend to operate at a much smaller scale, molecular rather than particulate, and are a bit more sensitive to environmental exposure. Sometimes the way that a dye is uptaken into a material is a little unpredictable and can be influenced by factors such as surface texture and material density.

Put it another way the two main 3D printing materials, PLA and ABS, often compete on the basis of UV resistance to colour fade and that’s great if you actually want permanence and bulk standardisation. But colours aging is not necessarily a bad thing, especially in highly functional materials used in difficult environments or where a designer would like to take advantage of wear to highlight some aspect of their design.

Controlled reactions to environmental exposure are already used in engineering applications to provide information about part degradation. And who from the 1980s can forget the fashion for distressed denim. Moreover dyes like alazarin, and others, are used in medical tests to provide rapid diagnostics. So the ability to soak a dyestuff into the surface of a 3D printed form could have multiple high value applications.

I’m not a dye chemist, but I do know that dye chemistry is a multi-billion pound industry that employs many cookies much smarter than I.

I’ve only tried one brand of simple domestic clothes dye and it worked well first time. The simple fact that we can open the door to providing a high quality recycled 3D printing material that is open to 4D applications (controlled colour change with age/environmental exposure) can only be a positive and it encourages me to take these tests a little further.

Whether the applications meet the material and a match is made is impossible for me to say. Materials development is all about opening doors to imaginative designers.

Fishy Filaments has just opened another door.


Wall Thickness vs. Colour

These are things that won’t mean to much to those who aren’t actively 3D printing themselves but trust me, they are essentials in the test program. We’re really getting into the R&D framework that I call Creative Metallurgy here so feel free to check it out.

Wall Thickness vs. Colour;

The recycled nylon is translucent which means there are two elements to how ‘colour’ presents itself; through the reflected light and the transmitted light.

Reflected light is a function of the photon absorbance of the material itself (or pigments within the material), so is independent of wall thickness. Designers of both forms and materials who use opaque materials don’t have to consider transmitted light.

Transmitted light can be a direct transmission i.e. light passing through from its source with specific wavelengths filtered on the basis of the material properties, or it can be internally reflected and refracted, and so impacted by the design of the form and the way that the form is printed. Not only do the pigmentation and absorbance impact the final perceived colour, but the refractive index also impacts the overall printed outcome.

Not all translucent materials are attractive as their thickness increases, but if the world of glass and ceramics tells us anything it is that most people find translucence appealing when the form is light, delicate or decorative, but less desirable when the form is primarily functional.

René-Jules Lalique, creating his glassware in the early 20thC, was the master of the balance between transmitted and reflected light.

This slideshow requires JavaScript.

So why am I saying this ?

Well, the recycled fishing net is translucent. We already knew that. The question was how would that translucence manifest itself in a 3D printed form? There was no way to predict that because the material has been ‘aged’ by repeated exposure to ocean and weather. It has worked hard in its first life, so how its second life would look was always going to be a bit random.

The good news is that the recycled nylon retains good colour, even at extreme blends, and the really good news is that it exhibits both transmitted and reflected colour separately.

What does that mean?

It’s really difficult to photograph without studio equipment but the colour, when printed, looks like the kind of dual-colour that antique Lalique glass shows. It has that layering of colour, with a green-ish body but with a light blue ‘bloom’ when viewed at an angle. Depending on the printer and the form printed it also has a silvered quality due to internal reflective surfaces.

I’ve tried to photograph it but it really doesn’t show well here.

This slideshow requires JavaScript.

The circular raft is one layer thick, flexible and very resilient overall, but weak in tension parallel to deposition direction.

What does this all mean for Fishy Filaments?

We knew that the material had non-technical aesthetic qualities in terms of the story of its first life, but the fact that it prints with a highly decorative finish shifts the balance of what markets might be approached. I’m not going to say too much about that but the message here is that some materials can be more than simply means to solve physical problems. Just ask any jeweler, sculptor or architect.