Experiments With Flex

[Topgunbai]

6 hours ago, duckbillclinton said:

Has anyone ever noticed,  as soon as the tines are open, ink flow on a wide stroke is no longer just controlled by capillary action, but also relies on the ink film (bubble) created by the paper, 2 tines, breath hole (optional), and feed?  As soon as this thin film ruptures, railroad starts.  So to successfully maintain a stable ink film (bubble), you will guarantee a continuous wide stroke, and you can continuously writing this wide line till you run out of ink or paper.

 

Very good point. I noticed this quite a while ago, and as a result I began playing with the ink, that is, adding surfactant into it, so that a stable ink film can be maintained. Adding surfactant can have side effects such as severe feathering, but nevertheless it is a rather cheap way of solving the railroading problem.

 

I've been working on partial differential equations, finite element analysis, etc., and I understand that a thorough description of the mathematics and physics behind flex nibs will require a ton of advanced tools, which is exactly what makes it impractical for most hobbyists - we would need supercomputers (or at least machines like a 64 core, 128 thread workstation) for high-accuracy simulations of ink flow (including ink-air exchange) in the feed, minimal surface generated by flowing ink and moving tines, and the force inside the tines. Such simulations were definitely not possible in the 1930s, yet people still make great flex nibs at that time.

 

Good mathematical theory and detailed numerical simulations will of course be very useful in factory designs of flex nibs, but when playing with flex nibs as a hobby, we need much cheaper techniques to get things done, such as the good old trial-and-error.

[duckbillclinton]

23 minutes ago, Topgunbai said:

 

Very good point. I noticed this quite a while ago, and as a result I began playing with the ink, that is, adding surfactant into it, so that a stable ink film can be maintained. Adding surfactant can have side effects such as severe feathering, but nevertheless it is a rather cheap way of solving the railroading problem.

 

I've been working on partial differential equations, finite element analysis, etc., and I understand that a thorough description of the mathematics and physics behind flex nibs will require a ton of advanced tools, which is exactly what makes it impractical for most hobbyists - we would need supercomputers (or at least machines like a 64 core, 128 thread workstation) for high-accuracy simulations of ink flow (including ink-air exchange) in the feed, minimal surface generated by flowing ink and moving tines, and the force inside the tines. Such simulations were definitely not possible in the 1930s, yet people still make great flex nibs at that time.

 

Good mathematical theory and detailed numerical simulations will of course be very useful in factory designs of flex nibs, but when playing with flex nibs as a hobby, we need much cheaper techniques to get things done, such as the good old trial-and-error.

 

It's not just partial differential equation, minimal surface from the field study of discrete differential geometry also plays an important role on ink film.  From my previous post's photos, my sample design 1 uses crossflex mod, when the tines opened, the shape (geometry) of the ink film is a very complicated but also very simple SINGLE surface.  Minimal surface problem contains a large variety of allowable function sets (the simplest example would be parabolic functions, hyperbolic functions, and etc.).  My assumption would be:  If the conditions for forming an ink film did not fall under an allowable sets of minimal surface functions, the film will not form, or an existing film will rupture.  This turned out to be false!

 

Like another pen pal pointed out from the previous posts, flex nib design can also be done with trial and error.  In fact, when I came up with a plausible flex nib design model (or say flex nib modding methodology), I did not need to predict or solve any of the possible minimal surface related functions.  Instead, I built and observed my designs during use, and found that my overly concern of minimal surface problem is NOT necessary.  In fact, in an ideal world, with ideal ink, ideal paper, ideal nib, and so on, an ink film will ALWAYS form by following the minimal surface principal, it's a question of whether the film would form at the orientation as intended, meaning that, at least one edge of the surface is contacting the paper. In fact, in real world, the ink film indeed could form at the wrong place and resulting railroading.  We have seen it all the time.  When a nib's middle slit is too hydrophobic, opening up the tines will create an ink film for just a split of second, it will then retract upwards back to the pen feed by capillary action, the film had never touched the paper.  So, on the other hand, if ink film can be formed at the right place, then here we go, it's just a thin film equation problem.

 

[duckbillclinton]

6 hours ago, txomsy said:

I think there are many ways to skin a cat.

 

Understanding in depth the physics and the mechanics is one.

 

Trial and error is the other. And my guess is that that one is the way the original nibs came about. That and luck (good or bad) in the sense that the quality of the materials they used rendered their steel less "rigid" than modern ones (where that "rigidity" has been largely sought after for a long time). Then you just try intuitively different design approaches until you get it.

 

 

Knowing the science behind a proper flex nib design, can help us quickly resolve any problems we encountered in flex nib modding.  If a problem has several KNOWN contributing factors for failing, the results of failing is a combination of them, and it can be a very large number.  This could be way too many trial and errors, not to mention it also require the testing need to be done in a systematic way with good discipline.  For me, sometimes I could lost track on my test cases, and waste a lot of time to recall.  In a case of a problem having UNKNOWN number of contributing failing factors, random distribution dictates, you have a good chance will NEVER find the right answer(s) by trialing. 

[duckbillclinton]

2 hours ago, Topgunbai said:

 

Very good point. I noticed this quite a while ago, and as a result I began playing with the ink, that is, adding surfactant into it, so that a stable ink film can be maintained. Adding surfactant can have side effects such as severe feathering, but nevertheless it is a rather cheap way of solving the railroading problem.

 

I've been working on partial differential equations, finite element analysis, etc., and I understand that a thorough description of the mathematics and physics behind flex nibs will require a ton of advanced tools, which is exactly what makes it impractical for most hobbyists - we would need supercomputers (or at least machines like a 64 core, 128 thread workstation) for high-accuracy simulations of ink flow (including ink-air exchange) in the feed, minimal surface generated by flowing ink and moving tines, and the force inside the tines. Such simulations were definitely not possible in the 1930s, yet people still make great flex nibs at that time.

 

Good mathematical theory and detailed numerical simulations will of course be very useful in factory designs of flex nibs, but when playing with flex nibs as a hobby, we need much cheaper techniques to get things done, such as the good old trial-and-error.

 

Further add to my above reply, solution(s) found by trial and errors is(are) usually special case solution(s).  It does solve our problem, but at the same time it limits our view on the problem.  A general solution is more ideal, as it enables us to have a large variety of different nib designs.  Indeed, finding the general solution is hard, and applying it is beyond an average home DIY hobbyist's capabilities.  For me, I am absolutely sure, the solution I found is not the true ultimate general solution, it's far from it, but at least, it is a wide set of special solutions with clearly defined rule set, and it has enabled us to mod a flex nib with broaden design selections.

[Topgunbai]

51 minutes ago, duckbillclinton said:

an ink film will ALWAYS form by following the minimal surface principal, it's a question of whether the film would form at the orientation as intended, meaning that, at least one edge of the surface is contacting the paper. In fact, in real world, the ink film indeed could form at the wrong place and resulting railroading

 

I've experienced this phenomenon too, i.e. the ink film fails to touch the paper. At that time I was thinking that it was solely the shape of the tipping ("baby's bottom") that caused the problem. Now I realize that the geometry behind may be more complicated.

[MarioR81]

Have anyone tried this on a cheaper pilot nib? (Kakuno, Metropolitan, etc?). I have a kakuno fine I really like that could really use some flex easing but I'm concerned that I may completely ruin the nib. I can get some line variation out of it but have to press hard enough to make a deeper mark on the paper.

[duckbillclinton]

55 minutes ago, MarioR81 said:

Have anyone tried this on a cheaper pilot nib? (Kakuno, Metropolitan, etc?). I have a kakuno fine I really like that could really use some flex easing but I'm concerned that I may completely ruin the nib. I can get some line variation out of it but have to press hard enough to make a deeper mark on the paper.

 

Indeed, tried with great success.

[MarioR81]

1 hour ago, duckbillclinton said:

 

Indeed, tried with great success.

 

pics are much appreciated

[duckbillclinton]

13 hours ago, MarioR81 said:

 

pics are much appreciated

 

Well, I have made 2 pilot (compatible) steel nib mods, but gave both to some high school students as gifts.  Here are 2 photos taken by one of the kids (not very good quality).

 

It's a WingSung 698, uses the same nib design as Pilot's steel nib collections, pen feed is also a direct copy but cheaper made of pilot design (unlike WingSung's other cheaper Pilot nib compatible pens, those use a different feed design, much cheaper to make),  tip size originally is European pen maker's EF, and I grind it down to true Pilot EF with right oblique.

 

After all the nib modding and tip grinding, line width variation achieved is 0.35mm to 2.5mm, not too aggressive, but it writes nicely.  Sadly, the kid didn't take any writing sample photos.

 

 

[InesF]

On 7/16/2021 at 7:04 AM, duckbillclinton said:

wow, i m shocked.  not much response from the community, so is it worthy to write an article on flex nib mod...

Hi.

Don't be shocked. It is summer time, many of us nerds enjoy sunlight outdoors - you can't get tan from computer screen backlight...

 

I'm looking forward to your theory lession and will read it with pleasure!

 

May I start with a first question: how to influence nib geometry so that the tines will open with the front slit remaining in V shape (the more narrow opening at the paper side)?

Or is such a behaviour not necessary?

[duckbillclinton]

4 hours ago, InesF said:

Hi.

Don't be shocked. It is summer time, many of us nerds enjoy sunlight outdoors - you can't get tan from computer screen backlight...

 

I'm looking forward to your theory lession and will read it with pleasure!

 

May I start with a first question: how to influence nib geometry so that the tines will open with the front slit remaining in V shape (the more narrow opening at the paper side)?

Or is such a behaviour not necessary?

 

Hmm, hope I read your question correctly...  Assuming the material used, cold work process, and nib sizing (say #6 nib) stay the same, then the curvature of the nib plays an important role on tine opening angle, flatter construct (starting from near the breath hole down to the tip) will reduce tine opening. 

 

Just imagine we have a small strip of flat paper card, let's cut a slit in the middle (similar to a pen nib), then we press it down to a hard, flat, and smooth surface, the slit is not going to open much. 

 

As contrary, a real life example, I have 2 #6 Bock titanium nibs, very disappointing experience (comparing with the good old almighty OMAS Titanium nibs), they flex very little and too easy to spring, so I took the nib out, carefully using hand tools to increase the curvature, the results are quite good, the tines open with wider angle (wider width also), feels stiffer and no longer spring even I press down hard (but not too hard, obviously). 

 

Greater curvature will increase the tine opening angle, and also increase the resistance for bending, so the nib will become stiffer.  This also means opening up the tines required more strength, so I use hand tool again, to make the tip tilt up a bit (similar to the classic and most-pleasant-to-write Sheaffer integrated nib), that solved all the problems. 

 

*I will post a picture of this modified Bock #6 titanium nib later on.*

 

The newer gen of Japanese Platinum 3776 nib is exactly the type that comes with a flat construct, the tines can barely open even with a very hard pressure.

 

The other parts of a nib's geometry play less important role than curvature.  Longer slit cut will increase the tine opening width, but will not affect the angle (think of the small strip paper card example).  Thicker nib construct increase nib stiffness and the chance to spring.  Wider nib shoulders affects nib's front section's stiffness (from breath hole to the tip).

 

I will add more if I can think of anything that is missing.  Still though, hope I got your question right, lots writing here, LOL

 

Also, I recommend you Google search and read Amadeus W's Fountain Pen Designs website, very informative.

Edited July 18, 2021 by duckbillclinton

[MarioR81]

8 hours ago, duckbillclinton said:

 

Well, I have made 2 pilot (compatible) steel nib mods, but gave both to some high school students as gifts.  Here are 2 photos taken by one of the kids (not very good quality).

 

It's a WingSung 698, uses the same nib design as Pilot's steel nib collections, pen feed is also a direct copy but cheaper made of pilot design (unlike WingSung's other cheaper Pilot nib compatible pens, those use a different feed design, much cheaper to make),  tip size originally is European pen maker's EF, and I grind it down to true Pilot EF with right oblique.

 

After all the nib modding and tip grinding, line width variation achieved is 0.35mm to 2.5mm, not too aggressive, but it writes nicely.  Sadly, the kid didn't take any writing sample photos.

 

 

 

Cool, how's the ink flow? Also how do you make the vertical slit? The dremel I have makes a slit too large. 

 

 

[duckbillclinton]

37 minutes ago, MarioR81 said:

 

Cool, how's the ink flow? Also how do you make the vertical slit? The dremel I have makes a slit too large. 

 

 

 

Ink flow is very good, the feed has been heavily modified.  Pilot's feed design is one of the best in the industry for modding, it's top notch and bar none.  The feed contains 2 grooves, 1 groove for ink flow (capillary action), another groove for air breathing, so widen both with a Razer (I also use a 0.15mm micro diamond cutter disk by hand to polish and rough the groove, the same cutter disk is also used on a dremel to cut the narrow slit), it will result the best possible ink flow (please be gentle though, don't go too wide on the grooves, you can't bend the rules of capillary and air pressure).  The finished pen is a super flex, (nearly) Never skips nor railroads.

 

Once I start writing my article with a new thread, I will give full details on every tool I use and the tricks to get things done properly.  There will be a lot of surprises, I promise, but it will be a while before I redo all my experiments with verification.

Edited July 18, 2021 by duckbillclinton

[MarioR81]

cool thanks. Looking forward to it.

[txomsy]

Regarding feedback...

 

I once was in a meeting with a handful of Nobel Prizes in Chemistry. Not that I am anybody, 'cos I ain't. At the time I was deep into Computational Chemistry, and over the dining (or was it lunch?) table I took to ask them what they though about Theoretical Chemistry.

 

Their answer was pragmatical: most problems are easier to solve in the lab with due ingenuity, plus you would have the experimental confirmation. I've posed similar questions to great names in Physics and got similar answers, which at the time I found odd, since we were building a computer Grid to treat (among other things) the data from the Large Hadron Collider at CERN and expected they'd value more Theoretical Physics. And, when I talk about Bioinformatics or Biocomputing I always insist that any prediction shouldn't be trusted until experimentally confirmed.

 

Where I am getting at is that we are now in sort of a historical anomaly. Prior to the metal age knowledge of metals was likely limited. For a few thousand years you would have a myriad of artisans devoting their whole life to working with metals with their hands and trying to better each other, trying all sorts of errors, giving them form with their own hands, testing them, building on their received knowledge and expanding it.

 

Then came the age of the machine and automation, now all that practical, experiential (and experimental) knowledge is vanishing and we are left with knowledge about the machines to make what was deemed at one point the best option for whichever reasons, and only a handful of artisan smiths who play every single hour with alloys, experiment, test them and know all their ins and outs.

 

Surely if one were to ask an artisan iron smith that had specialized in small, flexy sheets of steel, working every day at making flexy iron thingies, that person would likely know most there is to know or at least have the intuition for it and get at it in a snap.

 

Missing an expert in the field, we, amateurs, are left with either theoretical physics/engineering, or with trial and error until we build up enough knowledge to match an expert of yonder.

 

In less words, while advanced physics may be a good approach, good old trial and error coupled with good intuition may result in favorable results as well (and maybe faster).

 

OTOH, I suppose you have already gone over Amadeus' ingeneering blog https://fountainpendesign.wordpress.com

 

Seems to me he has already done much of this work as well.

[duckbillclinton]

1 hour ago, txomsy said:

Regarding feedback...

 

I once was in a meeting with a handful of Nobel Prizes in Chemistry. Not that I am anybody, 'cos I ain't. At the time I was deep into Computational Chemistry, and over the dining (or was it lunch?) table I took to ask them what they though about Theoretical Chemistry.

 

Their answer was pragmatical: most problems are easier to solve in the lab with due ingenuity, plus you would have the experimental confirmation. I've posed similar questions to great names in Physics and got similar answers, which at the time I found odd, since we were building a computer Grid to treat (among other things) the data from the Large Hadron Collider at CERN and expected they'd value more Theoretical Physics. And, when I talk about Bioinformatics or Biocomputing I always insist that any prediction shouldn't be trusted until experimentally confirmed.

 

Where I am getting at is that we are now in sort of a historical anomaly. Prior to the metal age knowledge of metals was likely limited. For a few thousand years you would have a myriad of artisans devoting their whole life to working with metals with their hands and trying to better each other, trying all sorts of errors, giving them form with their own hands, testing them, building on their received knowledge and expanding it.

 

Then came the age of the machine and automation, now all that practical, experiential (and experimental) knowledge is vanishing and we are left with knowledge about the machines to make what was deemed at one point the best option for whichever reasons, and only a handful of artisan smiths who play every single hour with alloys, experiment, test them and know all their ins and outs.

 

Surely if one were to ask an artisan iron smith that had specialized in small, flexy sheets of steel, working every day at making flexy iron thingies, that person would likely know most there is to know or at least have the intuition for it and get at it in a snap.

 

Missing an expert in the field, we, amateurs, are left with either theoretical physics/engineering, or with trial and error until we build up enough knowledge to match an expert of yonder.

 

In less words, while advanced physics may be a good approach, good old trial and error coupled with good intuition may result in favorable results as well (and maybe faster).

 

OTOH, I suppose you have already gone over Amadeus' ingeneering blog https://fountainpendesign.wordpress.com

 

Seems to me he has already done much of this work as well.

 

 I am speechless.

[duckbillclinton]

On 7/17/2021 at 10:35 AM, Topgunbai said:

 

I've experienced this phenomenon too, i.e. the ink film fails to touch the paper. At that time I was thinking that it was solely the shape of the tipping ("baby's bottom") that caused the problem. Now I realize that the geometry behind may be more complicated.

 

My previous mentioning of thin film equation is completely wrong and irrelevant.

[txomsy]

I think the equations are likely a bit more complex. One needs to consider not only the flexibility gained, or the ink flow characteristics, but also the durability of the solution: if flexing a modified nib may result in undue stress, the metal may suffer fatigue and sooner or later break or suffer plastic distortion.

 

For that one also needs to consider the temper of the metal and the distribution of stress lines through the nib. Amadeus' blog has many interesting tips.

 

From the little I know about simulated annealing, steel can be tempered to be fragile yet strong or flexible but deformable, with a large range of variations, and a good nib probably needs to hit the sweet spot in the line for optimal performance. The further one departs from it, the easier it will be that the nib will break or deform, maybe at the first use or maybe after some time or only after some abuse, but that needs to be considered too. It might be that not any steel nib is susceptible for flex modification.

 

It will also depend likely on the thickness of the steel sheet. A thinner one will be more flexible but also likely less resilient. A thicker one will require more pressure and possible be more resistant. It is important not to forget that today's nibs were selected to be strong and most resistant to deformation in order to compete with ballpoint pens.

 

Stress lines will depend on nib shape. This may have little impact at the beginning, when the metal is still strong, but if the cuts are too deep or located in the wrong place, stress will be redirected and concentrate on possibly suboptimal points, leading to increased metal fatigue and, after some time, damage.

 

If I were to approach it -and had the time, which I don't- I would start thinking of alternative designs, testing them in practice and when one seemed to work, I'd be modeling it in the computer to see the distribution of stress lines in an attempt to predict its resilience. Then, I'd have to use (close to abuse threshold) every day to confirm if it can stand use and how long. Probably what nib-makers of old did, in a time when pens were used 10-12 hours a day 6 days a week by all users that would result in prompt feedback. Now it would take one of us a lot longer.

 

The reason for starting with design is that the info is readily available, one needs only look at flex dip nibs to find out suggestions for designs, whereas testing different alloys might have been an option for nib-makers of old (who could just order a sample sheet of metal and use it in their own manual nib-making presses to get test nibs, but is mostly limited to available nibs today. Plus, in the old times, it would be easier in that one would have the tactile experience to know how a good nib should "feel" in the hand as opposed to suboptimal materials and nowadays this experience is rare.

 

In hindsight, I think that the current status was driven by stock market forces: in order to compete with BPs, the best approach for FPs would have be to lower costs as well, so that schools would continue favoring them. At the time most probably thought it would undercut profit margins and lead to company decapitalization by investors, leading them to increase margins to remain attractive and instead then forced them to adagpt to new users grown with BPs by strengthening the nibs. Which is what we have now. Nails.

 

One may wonder if a cheap FP would have allowed them enough added sales to be justifiable, possibly aided by strong marketing campaigns touting the beauty and elegance of FP line-width-variating scripts over the dullness of BPs as a distinction trait; but since the market was already saturated with FPs before the advent of BPs (because there was no other sensible option), it is doubtful they could have increased sales by reducing prices.

 

One can also wonder why did makers not keep a line of flex pens for the knowledgeable, but I suppose they did until most users were so used to BPs that demand for flexibility was negligible or could not justify keeping the lines.

 

This leaves us with modern nibs engineered to be hard, strong and rigid.

 

There have been suggestions of reducing modern steel nib thickness to increase flexibility. This might be an option, but it is too difficult to master and do accurately from an amateur point of view. Which leaves us, again with modifying the overall shape in a way that increases flexibility but maintains bouncyness and springyness long enough to be practical.

 

The problem with ink flow may be dependent on the nib metal alloy and feed material (and surface tension) used, or may be attacked by modifying the ink. Ideally one would prefer a solution that requires no ink modification. And possibly that can adapt to most inks. I'm thinking of the Eversharp manifold/regulator nibs as a possible solution, but that is even more difficult to achieve for an amateur.

 

Booofff... too long a post already. Who said s/he wanted discussion?

[AAAndrew]

As for problems with the ink film reaching the paper, I cannot speak to the science of the ink, but I can speak from a lot of experience with dip pens having trouble with ink that won't reach the paper. What I've found is that the vast majority of the time, the issue is the surface of the underside of the nib has some kind of interfering coating, usually oil, that prevents full flow of the ink. This usually happens when the underside is touched by a person, and their skin oils on the nib will prevent the ink from flowing well down the slit. Even though a fountain pen is not reliant on the hydrogen bonding in the ink to keep the surface tension as it gathers on the back of the nib, like a dip pen, instead the feed literally feeds the ink to the back of the nib, down the slit and to the paper. If the back of the nib has a hydrophobic coating (like skin oils) on it, even in a fountain pen, this will prevent the ink from flowing down the slit all the way to the point like it should. The way to solve this problem is, obviously, to remove the oil and not touch it with bare skin while working on it. 

 

Just my 2-cents from lots of experience with highly flexible dip pens and ink flow problems. 

[MarioR81]

1 hour ago, AAAndrew said:

As for problems with the ink film reaching the paper, I cannot speak to the science of the ink, but I can speak from a lot of experience with dip pens having trouble with ink that won't reach the paper. What I've found is that the vast majority of the time, the issue is the surface of the underside of the nib has some kind of interfering coating, usually oil, that prevents full flow of the ink. This usually happens when the underside is touched by a person, and their skin oils on the nib will prevent the ink from flowing well down the slit. Even though a fountain pen is not reliant on the hydrogen bonding in the ink to keep the surface tension as it gathers on the back of the nib, like a dip pen, instead the feed literally feeds the ink to the back of the nib, down the slit and to the paper. If the back of the nib has a hydrophobic coating (like skin oils) on it, even in a fountain pen, this will prevent the ink from flowing down the slit all the way to the point like it should. The way to solve this problem is, obviously, to remove the oil and not touch it with bare skin while working on it. 

 

Just my 2-cents from lots of experience with highly flexible dip pens and ink flow problems. 

 

I write with dip pen nearly every day and I can attest to this. And I'd add that flow issues happen when the ink is too thick. I have a small ink well that doesn't seem to hold water as well as a screw cap bottle so I may need to add a bit of water to thin out (currently using W&N calligraphy ink). Fountain pen inks seem to be too liquid for dip pens, at least the ones I've tried.