# Fluiduino

A fluidic transistor that I designed and printed.

This blog post is a temperature check for how interested the community would be in an open source, 3d printable, fluidics hardware platform. I want to make sure that this is something that the maker community wants because this isn’t for me, it’s for you. So if you won’t use it, I won’t make it.  First, let me wet your appetite by listing cool projects and applications of fluidics, most of which can be printed on a 3d printer and then I’ll tell you how it works and what I want to do with it.

Awesome projects:

Educational. Teach electronics and logic via the fluid/electric analogy which is extremely strong.

3d printable pneumatic stepper motor made from plastic parts only.

Rocket propulsion control systems.

Steam punk steam robot that can function while on fire.

Robust space travel.

Wood wind magic key piano that opens up a trap door with wind.

Programmable digital computer with no electronics.

Programmable fountain with no electronics that you can interact with.

Microfluidics, biological control. This could be an open source project in its own right.

Smart meta-materials(speculative).

How it works:

The electric/fluid analogy is very strong. Electrons flowing in a wire is much like water flowing in a pipe and in fact, the two systems follow many of the same laws.  A pressure difference is like a voltage, a long narrow tube provides resistance, a membrane is like a capacitor and a coil of pipe is like an inductor. You can send power with electricity and also with fluids(i.e. hydraulic power systems). These are all analog components but the analogy between fluids and electric circuits goes even deeper.  Just like with electric circuits, you can make fluidic circuits digital.  The principle here is that streams of fluid interact with each other. That is, two streams will divert each other, not pass through, so that you can use a control stream of fluid to divert a power stream.  You can use this to build a fluidic amplifier/transistors, logic gates, flip-flops and even a digital computer.

Fluidic amplifier

I’ve taught electronics labs before and let me tell you, electronics are not intuitive to most people. Fluid flow is much more intuitive. One of my main motivations for an open source fluidics project is to make educational kits for high school and college students so that they can gain strong understanding for how electronics work by understanding how the more intuitive fluidic systems work.

Electronics are awesome. I love my computer and I love my Arduino.  But there are some applications that semiconductors are bad for. In particular, you can’t light your laptop on fire and expect it to function. Fluidic computers work just fine when on fire, provided the fluidic device is made of something like ceramics or metals and the fluid is something like gas or oil.  Indeed, you can make your power stream be rocket thrust, very hot, and use a perpendicular small control stream to divert the power stream, thus being able to steer a rocket! Nasa has a nice blurb about his here.

Rocket control

Fluidic rocket control.

Generally fluidics is practical in harsh conditions. Yet another example is space travel.  There are tons of charged particles and other radiation flying around in space that damage semiconductors.  Indeed, semiconductors have to be radiation hardened to withstand these harsh conditions. Because fluidics use fluids instead of electrons for their operation, they can be more reliable in such situations.

Fluidics is just plain cool. The number of steam punk applications are endless. You can make a robot that moves and thinks with steam, just steam! There was someone who made a desk that would play notes via a pipe organ when the drawers were opened. If you open the drawers in the right sequence, a trap door would open. It was made entirely from wood with no electronics. Cool huh?

Fluidic pipe organ desk with a secret compartment

Fluidic pipe organ desk with a secret compartment

One speculative application that’s on the forefront of science that I’m particularly interested in is digital meta-materials.  Digital meta-materials are materials that are made of bits of matter, put together like legos to make something that is functionally more powerful than each individual block.  In this case, I’m imagining fabric with a vascular system that responds to temperature and pressure  via fluidic logic.  I don’t think that 3d printers can beat traditional manufacturing in what traditional manufacturing does well already but I do think that 3d printers can win, hands down, when it comes to complex intricate objects like printing an intricate fluidic vascular system. The philosophy here is that we should be looking for applications of 3d printers that were not possible before.

What I want to do:

I’d like to start an open source fluidics project. Something like Arduino, easy to use, but with fluids. Maybe we could call it the Fluiduino? My primary motivations are educational and inspirations.  I want students to have a deep understanding of electronics and I want makers and hobbyists to easily be able to make rocket control systems, steam robots and programmable fountains. What I want from you is feedback as to whether an open source fluidics platform is something you’d be interested in playing with in the future? I’m not doing this for me, I’m doing this for the community as whole so if you’re not interested, I won’t do it. If however, you feel as inspired as I do, I’d love to find people to work with on this project.

-Sebastian Spiegel

1. Tim Sullivan

So cool! You should definitely take this idea further. I know the possibilities of this thing are endless, but I would definitely like to try making some digital logic device, some kind of fluidic integrated-circuit – maybe an adder? It would be awesome to have some open hardware principles applied to this, and just see what the community does with it.
Just some ideas for mr.Spiegel or anyone else trying his print:
– make it square/hexagonal. If you can attach them and have standard geometry, im sure people would try to make other functional blocks.
– make a “pipe” block for connecting other blocks together
– provide a consistent way to give pressure. Like a mini, printable, open-topped water tower?

Best of luck with your project

• independentindustries

Thanks! I really like the idea of making them into puzzle pieces. Some design problems have to be worked through though since each component needs power and ground and perhaps even a clock, all of which must have low impedance which means wide cavities. That is, I think I need three layers, top layer being power, middle being the logic and the bottom layer being ground.

2. Nate Burley

I agree with Tim: It’s a great idea. I’ve actually set up a company (currently just yours truly) and one of my first projects is to build kits for educational purposes. I’m onto my second set of prototypes and have had some early successes! The focus of my kits would be geared more to allowing students and researchers automate chemistry techniques by assembling breadboards before investing in customised kit. My main lines of attack would therefore not be in competition with this project.

I struggle with the open source ideal to be honest as I desperately want to do this kind of thing for a living; I’m fed up of busting a gut for people I don’t respect, who build companies which give too little back to the world on the community. However, I ache to share what little I know already and to be part of a community which shares in return, in order that the individuals prosper as a result.

What do you guys think? Is there room for such a guarded approach to open source? Room enough for profit and philanthropy?

• independentindustries

@Nate
I share your sentiments and I struggle with the same questions. With open source, I get the feeling that the value in your product doesn’t come from the product itself but with the relationship you have with your customers. Do they trust you to produce something of quality, give guarantees or provide some other service? Anyone can make their own Cola and sell it but somehow Coca Cola sells the majority. Why? Perhaps its that the value isn’t in the product but in the reputation and that can’t be stolen.

Do you support the product? Fedora Redhat is a company that makes money with an open source project and it does it by providing support.

Isn’t it the case though that you can patent your product and choose the license? For instance, you could make it free for individual use but not for a company or institution. Perhaps you could just patent a key component that every such kit would require? But to be honest, I don’t like this strategy. Property rights on ideas feels like a weird idea to me.

I like the open source model and think that even though it limits what business models will work, it benefits all of us more in the end. Let me know how your experiments with open source go and post a link to your product so I can check it out.

-Sebastian

3. Nate Burley

Ah gosh I just wrote a big post and my browser crashed!

I’m with you on creating value through great service and community management. I also think there’s scope for having small, collaborative, agile-focussed, communities outpace larger corporations when it comes to innovation. We’re seeing some of this in the 3D printer field at present. However, I’m struggling to see how to get something small and sustainable off the ground without external funding. Right now, I’d be able to actually put more serious time and effort into this little side show of mine, if only I didn’t have a day job too…!

I wouldn’t call what I have anything close to a product at present heh! I have produced my phase 1 prototypes which allowed me to test some ideas using SLS printed parts. They weren’t great due to the powder which is contained within the internal channels (had to be manually scraped out) and the fact that the parts do seem semi porous too (thus not fully waterproofed). Those parts did have functional internal micro channels with a 1 mm ID though, and they were capable of housing externally sourced valves which functioned in situ.

I’ve avoided FDM parts as I was worried about the requirement for solid supports – though water soluable supports do interest me and I will look into that (I have access to a 3D touch); I wanted instead to work towards a point where I could literally remove a working piece from the print bed without worrying about laborious post-process.

My second round of parts are SLA parts as a result, and they’re looking much better. They’re a lot more expensive, so I’ve spent more time optimising their geometry to reduce the cost of each part (I saw a reduction of ~80% having done this!). I do have some residual wax / grease support inside the internal lines but I think this might be flushed out with a little careful thought. Again. these parts use externally sourced valves (though the ones in this round are a fraction of the cost of those used in the previous phase).

Shapeways is slowing me down as I’m using them to produce the prototypes and need to collect my parts in to groups to reduce shipping cost. It just takes so damn long to receive them!

I’ll let you know when I’ve tested this round…

• independentindustries

You should be able to be able to print channels without using support material, fully hollow, on an additive 3d printer, that is, a home 3d printer. Here’s how: instead of using a circular or cube cross-sections of the pipe, use an eye shaped cross-section, that is the intersection of two circles of radius 1 located at x=-.5 and x=+.5, respectively, or something like that. This way, the overhang is never close to horizontal.

By the way, yes I think you’re right in that our goals are a little different. It sounds like you’re doing microfluidics, used to automate (bio)chemistry, while I’m doing old school fluidics, used for mechanical control systems and computation. Microfluidics is awesome. I wonder if the future will have microfluidic sensors for testing water quality, air quality, personal medical tests, etc… that is, I wonder if microfluidics will one day be as ubiquitous as electronics are today?

As for getting an organization off of the ground, I really have no practical experience, as I’m still finishing my PhD, but I’ve been thinking very hard about alternatives to the investor model of business. The basic problem in starting a company is to get a group of people to cooperate on a single goal. Financial capital can certainly make this easier but I wonder if it’s necessary? I just had a nice conversation with a friend of mine about equity by sweat. That is, you start the organization and start logging hours of work, labeled $t_1$. If other people join you and work as well, then they also accumulate hours of work. The amount of ownership of the company that each person has is proportional to the ratio of that person’s work and the total hours worked by living individuals. So your portion of the profit from the company would be $Profit \frac{t_1}{t_1+t_2+...+t_n}$. The point would be to incentivise individuals to cooperate on projects they believe in while also preventing any one person who hasn’t put the hours in from getting a disproportionate amount of ownership of the company. Even the worst cheater could at most get equity limited by the lifetime of hours an individual can work.

4. aprophetforhope

hey!

Well I definitely have two working prototypes. One has a externally sourced one way valve which works like a charm. The second was to test the size of channels I could get down to… 0.5 mm ! the 1 mm and 0.75 mm channels were also clear. I need a couple of bolts to hold the pump head together before I can test that properley but just holdung it tighly I’m confident I already had it working just now!

5. Gene Hacker

I certainly have interest in it, I’ve been trying to make a collection of 3d printable/laser-cutable fluidic elements for a fluidics experimentation kit.

What I have been trying to make and I suspect you might want to do similar is to make a bunch of breadboardable fluidic elements. IE just a bunch of fluidic elements with tubing connectors. Tubing is nice, as you can easily add delays and inertances(the fluidic equivalent of inductance).

Though to be a true experimentation kit we need stuff like flow/pressure sources, pressure sensors, and flow rate measurers which are going to be a bit harder to print.

Printing also presents some problems, as many printing processes produce parts with low surface finishes. Low surface finish could make fluidic bistables and turbulence amplifiers(essential for acoustic sensors you might find in a steam punk robot) highly unreliable.

Fluidics is not necessarily a great way to teach students about electronics, elements like fluidic amplifiers really aren’t analogous to transistors and even the most basic of fluidic elements are highly non-linear

Other than that, making an arduino equivalent with fluidics is certainly going to be difficult.

6. novelchip

I also want to make a kit to build fluidic circuits (or microfluidic) mainly using air as fluid , since with liquids you allways have the problem of bubbles. I have built resistors, diodes, transistors, capacitors each one the size of a lego block. I can assemble circuits on a LEGO plate exactly as we do with a bredboard in electronics. You can find the instructions of building such blocks on instructables…look for “microfluidic LEGO bricks” . I am looking for people like you with similar interests to share ideas and come up with applications that you can not do with electronics or electromechanics….build them and showcase them rise interest in fluidics.

7. Ciber

If anyone has seen the Strandbeest kinetic sculptures by the Swedish artist Theo Jansen, they use a set of pistons powered by compressed air to do binary logic and keep track of their number of steps. Think of how complex the behaviors of these things could get if said logic was miniaturized and 3d printed.