Blog

The insides of a Platypus Pens Model 20 fountain pen

The Model 20 has more parts than any other Platypus pen, and those parts are made of a range of different types of plastic selected to optimise the function and appearance of the pen. There are so many parts that I decided to make a set of custom trays to held the parts and allow me to easily confirm that I have all of the parts that I need to assemble the pen.

Part 1 is what I call the ‘driver’. It has a crown-like ring of teeth that engage with the top of the bladder, which has indents for each tooth. It allows for the bladder to be twisted and compressed while still letting you unscrew the pen body while it is full of ink. That lets you most reliably see the ink level and thereby let me have a minimal ink window which, in turn, maximised the amount of lead ballast that I could fit into the grip.

Part 2 is the ‘plunger rod’ and it has two distinct threaded portions. The bit nearest the driver has a four-start coarse thread that guides the twisting and compression of the bladder. The upper portion has a two-start left-handed thread that is what the blind cap runs along as it is wound out and in. Part 2 is made of a plastic type that is fairly hard and wear-resistant, PCTG (not PETG). The plunger rod is reinforced with a stainless steel machine screw that runs down its length to a nut captive in the driver.

The plunger rod installation is designed to allow easy replacement. However, even though it did wear noticeable during the 100,000 cycles of testing (see here), but even so it was still fully functional. I now doubt that it will even need replacement.

Part 3 is the inside of the blind cap and it has three threads. The inner thread marries with the two-start thread on the plunger rod and it is retained by a washer under the head of the stainless steel machine screw in the plunger rod.

The second thread of the blind cap inner is a short tapered two-start thread with the same pitch as the plunger rod. That thread engages with minimal turn into a female thread at the end of the body liner. That locks the blind cap in place and prevents my friend, Peter, from breaking another  pen by twisting the blind cap in the wrong direction! Part 3 is made of plain PLA, which is harder than most other 3D printable plastics (but sometimes brittle) and so it is wear-resistant when paired up with the PCTG of the plunger rod. (I use melted beeswax as a lubricant there and on the cap threads because it stays where it is put and wont give you greasy fingers.)

Part 4 is the decorative finial that screws into the third thread of the blind cap inner (that I didn’t explain). That thread is left-handed and fine pitched so that it can hold firmly and it can be removed to give access to the bolt head in case that the drive system needs to be disassembled. I’m not sure that it will ever need to be disassembled, but maybe the plunger rod will need replacement after 20 years of use!

Part 5 is the blind cap outside. It is purely decorative and has no threads.

Part 6 is the ink window and ink reservoir. It’s made of either PETG or CPE. The CPE is a modified PETG that is supposed to have improved layer adhesion and be less brittle. I’m not sure that I can tell the difference between the two types of plastic and so I will probably use up both rolls. The ink window and reservoir are printed in a manner that makes them almost fully waterproof, something that is not always easy with FDM 3D printing. Almost watertight: not exactly leaky, but I found that without further treatment ink can creep into various tiny channels and voids in the walls and then effectively stain the parts. To prevent that I seal the inside of the parts with low viscosity epoxy that runs into the channels and voids to prevent the ink from doing so.

There are two threads in part 6: the male threads to engage the body; and a female tapered thread to engage the white tube that holds the nib and feed. The body thread is pretty ordinary as it is nearly just a 10mm metric thread. The female thread is tapered to maximise strength of the junction between  the nib tube and ink window. (Yes, I broke a few while placing the lead ballast and so I beefed the joint up even though it is well-secured by epoxy and the grip when the pen is finished.)

The nib tube, part 7, is made of ABS plastic so that it can safely resist the heat while I heat-set the nib and feed to each other. The ABS is also useful in being well-behaved while I ream out the tube for a close fit with the feed. The fat region at the bottom of the nib tube in the photo is a temporary extension to help align the reamer.

Part 8 is the cap topper and it has a fine right hand thread to engage with the top of the cap inner while I put the clip in place. The joint is glued with epoxy, so don’t try unscrewing it!

Part 9 is the cap liner. It has the capping thread, triple-start, a waisted portion that engages with the end of the grip to give a reliable seal, and a thread to take the cap topper. The bottom half of the cap inner is made of a relatively soft PLA, Polymaker PolyTerra, and that gives a good feel for the cap threads and helps seal with the grip. The top is made of a stronger plastic (same manufacturer, so the same grey colour) so that there is no problem holding the tension of the clip in the long term.

Part 10 is the outside of the cap. No threads or tricks.

Part 11 is the body liner. It has a female thread to marry up with the ink window thread, and a special four-start female thread through which the fat portion of the plunger rod runs. Part 4 is made of plain PLA for its hardness and excellent wear characteristics.

Part 12 is the body outsides. It has the male triple start thread for the cap.

Part 13 is the grip which has a subtle triple start thread that acts to make the grip ‘grippier’.

Part 14 is the bladder which has a pentagonal star-like cross section and is made out of the very resilient 3D printing filament called TPE. The bladder gets glued to the nipple at the top of the ink reservoir with polyurethane glue (think of the original ‘Gorilla Glue’). The bladder can be removed for replacement if that ever proves necessary but its perfect survival of the 100,000 cycles of stress testing suggests that it is likely to last a lifetime.

That’s it for this post, one photo, one link, and lots of words. It will not do well with Google, so I am glad that you found it!

Seams are one of the shortcomings of many 3D-printed objects—at least for objects printed using the FDM printers like those I use to make the Platypus pens. I have previously written about how I can often avoid them by printing pen parts that consist of a single helical extrusion, a process that is often called “vase mode” printing. Those parts look great but they are not very strong by themselves, and so my pen parts are made in multiple layers glued together with marine epoxy. That makes them ‘plenty strong’, but having to effectively laminate all of the parts is inconvenient and can sometimes limit the designs that I can work with. A better way would be better! At least sometimes…

Enter the scarf seam (not a neck-warmer)

I first devised a way to minimise the visibility of seams more than a year ago and have been using it in a limited way with good results. I called the results a ‘scarf seam’ because of its resemblance to a woodworking method for joining timber end to end, the scarf joint (or sometimes ‘scarph’ joint). Scarf joints can be a simple matched pair of tapers or can have complications that help to hold them together without glue, or while the glue sets.

The picture shows one of many scarf joints used in the reconstruction of a historic sailing ship by the Sampson Boat Co. The construction of the Tally Ho is nearly complete and a long series of YouTube videos document the process. I enjoyed watching them very much and look forward to watching it sail!

The scarf seam that I devised for hiding seams in 3D printed parts is an adaptation of the tapering layers that I use to start and end helical vase-mode extrusions. I realised that if I followed the upward taper with a downward taper then I would have the 3D printing equivalent of a scarf joint. I tried it out and found that it worked remarkably well.

If you have a recently made Platypus Model 10 pen then you will have scarf seams on the grip below the first bulge. You may not be able to see it as, after all, it is designed to be hard to see! There are also scarf seams on parts of the section liners of both models 1 and 10, but they are even less visible as they are glued inside the grip.

How it works

The magic of the scarf seam in the extrusion of the outermost perimeter is that it allows the printer nozzle to be moved away without any need to change the extrusion rate or the height of the nozzle. That results in almost no blob or gap at the seam. Internal perimeters and fill are printed before the nozzle moves back to the outer perimeter still without any vertical movement, and the next scarf seam begins. In that way the seams start and end with zero height and zero extrusion (or very close to it) and, ideally, it is all but invisible.

So you want to print a scarf seam…

It is worth noting that I was able to make scarf seams only because of the unusual way that I generate the g-code that drives my 3D printers. The custom programs that I write to generate the g-code algorithmically give me full control. Most people who print things with their 3D printers work differently.

Conventionally g-code is generated by first modelling a 3D shape using a CAD program or similar. That model does not have any seams or layers. That 3D model is then loaded into a program called a ‘slicer’ that divides the shape into 2D patterns that we call slices or layers, and then makes the g-code that tells the printer how to reproduce the slices. Seams are inevitable because each layer has a start and an end, and the nozzle has to be raised between one layer and the next (unless, of course, the slicer is set to ‘vase mode’). If your slicer program cannot make a scarf seam then you cannot get a scarf seam when you print your model. Until very recently no slicer was able to make scarf seams.

Enter open-source software

Last November I decided to share scarf seams with other 3D printing enthusiasts, with the intention that someone (not me!) would be able to incorporate them into a standard free slicer software package. I posted demonstrations and notes on it on Reddit and a couple of forums dealing with 3D printing and slicer software. Eventually the idea was taken up by a developer of the slicer OrcaSlicer and it is now part of the latest beta release of that software! 

You can read about the scarf joint seam in the OrcaSlicer release notes here and see the first scarf seam YouTube video, by fellow Australian Michael Laws (not me, Michael Lew), by clicking the thumbnail image below.

I am grateful to the person known as @Noisyfox doing so much work to go from the basic idea to working slicer code.

As I mentioned in the previous post , the bladder is both the core and the Achilles heel of the early twist-fill fountain pens, at least from the perspective of today. The rubber tube or sack was not sufficiently robust to withstand the repeated stresses of being twisted and untwisted. The problem was compounded by the fact that the rubber would become stiff and embrittled over time.

Twist-fillers were not the only fountain pens that suffered from the failure of ageing rubber. Replacement of the rubber bladders, tubes, or diaphragms is an everyday occurrence for collectors of vintage fountain pens. Sometimes the replacement is easy (as I found with the bladders of old Conway Stewart pens), but sometimes it is very difficult (as with the Vacumatic Parker 51 that I broke…).

The Platypus Model 20 bladder

Polyurethane (TPU)

The Platypus Model 20 bladder is made of polyurethane, a type of plastic that is often called TPU in the context of 3D-printing. Polyurethane is remarkably robust, chemically stable, and resistant to many solvents. Polyurethane’s flexibility and stability make it remarkably useful. It is widely used in many different roles, from stretch fabrics (Spandex, Lycra) to components of car suspension and skateboard wheels.

Polyurethane comes in many variants and different degrees of hardness. The TPU filament used for printing the Platypus Model 20 bladders is softer than many TPU filaments, with a technical Shore hardness rating of 82A, as compared to the most common 95A.  The particular filament used is Filaflex 82A, made by Recreus in Spain.

In contrast to rubber, polyurethane does not suffer from embrittlement much at all. It has even been reported that polyurethane becomes less brittle over time [1], albeit at elevated temperatures. It has also been reported that the presence of water hinders embrittlement from developing with ageing [2], a finding that might be relevant to polyurethane ink bladders.

I do have some limited long-term experience of 3D-printed polyurethane. I still have the tree frog that is the first ‘flexible’ print that I ever made. You can see in the photo that it is not a great print (I like to think that I could do much better now), but it sat on a windowsill for about 7 years and has proved itself both stable and an excellent dust collector.

Polyurethane is not perfect

Not every property of polyurethane is perfect for its role as the bladder for the Platypus Model 20 fountain pen. Unfortunately polyurethane is slightly permeable to water. Not disastrously so, but even a little can lead to drying out of ink if it is left in the pen for long enough. The next blog post in this series will detail my studies of water loss from various bladder constructions and the ways that I have made the problem minor enough to be unproblematical for normal fountain pen use.

References

1. Abbas Tcharkhtchi, Sedigheh Farzaneh, Sofiane Abdallah-Elhirtsi, B Esmaeillou, F Nony, et al.. Thermal Aging Effect on Mechanical Properties of Polyurethane. International Journal of Polymer Analysis and Characterization, 2014, 19 (7), pp.571-584. ff10.1080/1023666X.2014.932644ff. ffhal-01191410f
2. Possart, W., Zimmer, B. Water in polyurethane networks: physical and chemical ageing effects and mechanical parameters. Continuum Mech. Thermodyn. (2022). https://doi.org/10.1007/s00161-022-01082-y

Platypus pens is now offering its pens constructed from a PLA filament that contains 50% terracotta (by weight). The pens look like they are made from terracotta, and feel like it too.

Terracotta infused PLA filament makes a truly lovely pen, but it also offers a few disadvantages for the user and for the penmaker.

For the user

The surface of pens made from the terracotta filament is slightly rough and that means that ink cannot be simply wiped away with a tissue. A wet cloth is usually sufficient to clean the pen after filling, and an occasional purposeful cleaning may be necessary. What’s more, the rough surface is relatively prone to marking when scraped against hard edges, and when scraped against soft surfaces it is slightly abrasive. Cap posting is not recommended for terracotta pens.

Terracotta pens are likely to develop a patina after long use, something that some people might not like, but others will certainly enjoy.

For the designer

Terracotta filled PLA has a matt finish with no sheen or shine. Just like a terracotta pot. The patterns originally developed for Platypus Pens exploit the shininess of some PLA filaments to exhibit sparkles and play with light reflection. They do not work so well with terracotta. Therefore a new pattern has been designed, cleverly called ‘Pattern 4’, a pattern that uses shadows instead of reflections. I call the pattern ‘dimples’ for reasons that may already be apparent. (Dimples works pretty well with the shiny filaments as well and so it is being offered for all pens.)

For the maker

The terracotta filament comes out of the extruder differently from other PLA filaments. It seems more fluid than normal PLA at the optimal printing temperatures and it’s heavy and so it sags during printing when given half a chance. What’s more, it doesn’t weld as well to the non-terracotta filaments used for the contrast bands and so each of the bicoloured components has a weak point. That is not at all a problem for the completed pen because the weak joint is reinforced in the epoxy-glued sandwich of inner and outer parts, but I suffer a much higher failure rate among the components made from terracotta. The component rejection rate is probably about double that of components made of normal PLA.

Why use terracotta at all?

So if terracotta has all of those disadvantages, why bother making (or buying) a terracotta pen? Simple: the terracotta pens feel GREAT!

The micro texture and slightly porous nature of the terracotta feels distinctly different from the  ordinary PLA Platypus pens of otherwise identical design—different in a way that many people find to be better. It’s hard to describe the feel, so perhaps I can just say that it is less ‘plasticy’. I’ve never held a Visconti Homo Sapiens pen, but I suspect that the feel of their lava powder-infused resin might be similar to the terracotta powder infused PLA.

Should you choose a terracotta pen? Not if you would find the extra care needed to keep it free of ink stains, or if you do not like a pen to shows signs of its long use. However, if you want a pen that feels wonderful in the hand and that is unlike any other that you will see then, yes, give it a go.

A pre-review on YouTube

Mick L (Michael Lampard) has just released a YouTube video where he talks about his terracotta Platypus pen (prototype).

Platypus Model 1 and Model 10 pens have a shared distinctively elegant and simple shape. This blog post describes the shapes of those pen models and explains why they were chosen.

The body and cap of both models have outlines that are formed almost entirely on the intersection between two circles. They consist of a symmetrical pair of arcs truncated at the top and bottom (see the diagram below). Once that form was chosen the technical design is based on the diameters of the circles, how far apart the circles lie, and where the arcs start and end relative to the centre-line of the circles.

The cap liner obviously has to be tapered to fit inside the outer shell of the cap, but its taper is straight and slightly narrower than necessary simply to fit within the cap. Its shape is deliberately arranged such that while you are capping the pen the cap threads become aligned with the body threads before they engage. That eliminates the possibility of cross-threading, which could damage the threads in the softer cap liner material.

The cap liner also has a small step inwards about half way up. That step acts as a stopper for the widest part of the bulge at the end of the grip, and effectively seals the nib into a small volume to minimise nib drying.

The grip section of the pen has a straight taper for comfort and to give the writer plenty of options as to how to hold the pen. For the Model 1 pens the grip ends with a bulge and then a narrow portion right at the base of the nib. The bulge is what seals inside the cap, and it is set back from the nib so that the writer’s fingers are comfortably held back from the relatively short nib. The nib of the Model 10 pens is longer and so it is not desirable to move the user’s grip so far back from the nib, and thus the Model 10 grip ends very conventionally with a slight flare. Of course, that flare is what engages with the sealing step within the cap.

There is only a very slight step between the threads and the patterned body and the step is smoothly tapered. That means that the pen can be held far from the nib, if desired, without the threads and step being very noticeable. The body’s coloured band is placed between the threads and the step in order that it is neatly covered by the cap band when the cap is in place.

All in all, the shape of Platypus fountain pens is determined by a mixture of functional considerations and aesthetics. I like how they look and I hope that you do too.

In some previous posts I have discussed the pros of 3D printing of fountain pens, but alongside those advantages there are some important downsides. This post discusses them and the strategies to mitigate them that have been used in the design and manufacture of Platypus pens.

In order to understand the difficulties for fountain pen making that come with the pen being predominantly 3D-printed, we need to have a clear idea of the 3D printing process.

If you are interested in learning more about 3D printing then you can read some of my other blog posts, and I recommend the video series about the basics of 3D printing created by YouTuber Tom Sanladerer.

The layer by layer process intrinsic to 3D printers allows for a remarkable range of object shapes and designs to be formed, but that flexibility comes with a downside: 3D printers are quite slow. The components for one Patypus fountain pen take about four hours to print altogether. (And then more time is taken to assemble and finish the pen, but that is not the topic of this blog post.)

Rapid prototyping and development

Despite their slowness 3D printers are widely used in the process called ‘rapid prototyping and design’. A 3D print might takes hours to complete which stands in contrast to a few seconds to injection mould an object from plastic, but the 3D printer is ‘rapid’ because the tooling for the injection moulding process might take weeks or months to make or procure. It’s the total turnaround time from design to inspection and testing that is rapid with a 3D printer, and that short turnaround facilitates design iteration and optimisation prior to the product being manufactured by more conventional methods.

The 3D printers that sparked the development of consumer-level 3D printers were called RepRap in recognition of that role: the “Rap” part is short for rapid prototyping. The “Rep” part of RepRap is short for self-replicating. The first RepRap machine was able to print many of the components of itself, which were then used to make a second RepRap machine which printed the parts for a third, and so on. My second 3D printer—the one on which I print the Platypus pens—is constructed with parts printed by my first, and the first was repaired and improved with new parts printed by the second: a nice RepRap circle.

It may not need to be said, but Platypus pens are the product of rapid prototyping and iterative design.

Limitations of FDM printing for pen production

The layer by layer process of FDM 3D printing brings some important limitations to the objects that can be printed, and those limitations are relevant to fountain pen production. Here is a list of some. Each will be dealt with in turn below.

• Surfaces are not smooth

• The dreaded seam and pimples

• Weakness at the layer lines

• Threads are difficult

Surfaces are not smooth

The roughness of layers is often considered to be a limitation of 3D printing and a deficiency in the finished objects. However, Platypus pens are designed so that the layer lines contribute positively to the quality of the resulting pens: the pens would not be produced with completely smooth surfaces even if it was possible. To a large extent, the unusual and eye-catching character of Platypus fountain pens comes from the layer lines. See the blog post Layers of Sheen.

Even the un-patterned grip section of these pens benefits from the layer lines as they make the grip more ‘grippy’ and mean that even sweaty hands can hold a Platypus fountain pen without slipperiness.

The dreaded seams and pimples

That problem is typically combatted by having the filament retract suddenly just before the layer change and then un-retract when the nozzle reaches the start of the next layer. The retractions and un-retractions almost always leave visible defects on the surface of the object, defects that are called seams where they are vertically aligned and pimples where they are randomly scattered over the surface.

Platypus pens are designed to almost entirely eliminate this problem. Most of the components are printed in a single layer and so there is no seam or pimples. Well, not exactly one layer, but the extrusion path is a single helix. Single wall helical printing is often called ‘vase mode’ printing, and it eliminates seams and pimples by removing the need for filament retractions and un-retractions.

There is one component in the Platypus pens that is printed in the conventional layer after layer mode: the section inner. It does have a seam at the thread that connects to the pen body. You may need a magnifying glass to see it because the part was printed with a high quality printer fitted with a direct drive extruder and the printing settings are optimised to minimise the seam.

Weakness at layer lines

The layer by layer nature of 3D printing usually leads to uneven strength. The plastic is strong in the direction of its layer plane because it consists of a continuous strand of extruded plastic, but the object can be weak at the layer junctions because the plastic of adjacent layers is incompletely welded together.

That inter-layer weakness is not much of a problem for thick-walled objects and for solid objects, but for a long thin-walled thing like a fountain pen it can be critical. Furthermore any weakness is especially problematical when the object is printed with a single wall, as are most parts of a Platypus pen! What to do? Well, the solution to this problem is conceptually easy: make an epoxy-bonded sandwich of nested components. The strength of the sandwich is much, much higher than the strength of the two printed layers.

The cap of a Platypus pen has an outer layer made with shiny, colourful filament and has an interesting surface pattern and the plastic. That layer, by itself, is pretty weak, but that doesn’t matter because it is sandwiched together with a liner made of a different plastic selected to suit its structural and functional roles.

The body is made of three parts glued together with epoxy: the outer thread and band which slip into the end of the body proper, and the inner threaded liner that extends almost to the end of the body. After being sandwiched together those components make structures that are much stronger than the sum of their individual properties would suggest.

The section has a single wall helically printed coloured grip on the outside and a conventionally printed inner. The only part of the pen that is entirely one piece is the section thread that goes into the body. It is made of a modified PLA that has strong inter-layer adhesion and is not brittle. It is relatively thick wall where it meets the grip and is more than strong enough when the pen is fully assembled.

Threads are Difficult

A process of rapid prototyping and iterative design was used to come up with the several strategies employed to help make robust and effective threads for the Platypus pens.

Strategy 1: the male and female parts of the thread have different layer heights. That reduces the possibility of the threads interlocking on the layer lines and improves the feel of the threads in use as well as reducing wear.

Strategy 2: use a thread cross section that ‘knows’ that the thread is being 3D printed. My usual 3D model design software is the open-source CAD program OpenSCAD, and there is a very useful thread library available for it that includes a huge range of official thread designs. They work well as long as the thread size is large relative to the layer height and precision of the 3D printer. Unfortunately the threads of standard nib units are very fine and so I had to develop my own software for threads. (Yes, it involved a process of rapid prototyping and iterative design!) On a platypus pen only the male section to body thread is a standard design (M10x1 for Model 1 and M10.5×1 for Model 10).

Strategy 3: reduce thread wear by ensuring that the male and female thread parts are made of different plastics. The layer by layer construction of 3D printed threads means that they are never perfectly smooth and so a thread with close tolerance is going to the prone to both galling (where two surfaces freeze together from friction induced ‘cold welding’) and wear. In the Platypus pens the threads are made of dissimilar plastics, one harder and one softer. In use the harder plastic distorts the softer during the initial running in of the thread and the result is a good fit with minimal ongoing wear. Using a softer PLA in the cap liner not only helps prevent cap thread wear, but also minimises the possibility of the cap scratching the body surface when the cap is posted and un-posted.

Strategy 4: minimise the risk of thread damage from cross-threading. The cap liner is shaped with an inner taper that guides the section during capping so that the cap and body threads are aligned before they come together. It is only possible to cross thread the cap and body if you cap the pen without the section installed, and why would you do that? The body to section thread can be cross threaded if you are inattentive—as could that thread on any pen—but it’s a fairly hard to do because the section has an unthreaded lead-in that goes into the body before the thread engages.

Perfection!

So, given how effectively the difficulties and disadvantages of 3D printing have been overcome or minimised, Platypus pens mut be perfect, right? No, not right. If you look closely enough you will be able to see some minor flaws in any Platypus fountain pen. Those flaws come from the nature of 3D printing which cannot compete with the perfection of computer-controlled lathes and modern injection moulding for precision and repeatability. Squeezing molten plastic out of a nozzle like toothpaste out of a tube necessarily involves some imprecision and the occasional minor glitch.

Platypus pens are skilfully designed and carefully hand made to a high standard and you can expect them to be very comfortable, eye-catching, unusual, and, above all, to write well.