Beam Counter: a Rubber Block That Counts!

Rubber doesn't usually have a lot of memory. Contrary to a sheet of paper, when you deform a block of rubber, it just springs back. But by changing the structure of the rubber, we can change its mechanical properties. We can make a metamaterial out of the rubber that does have memory. Specifically, we can make a material that counts how often you compress and decompress it. It works with thin beams that buckle and bend left or right, acting as our material bits. We design these beams so that they bump in to each other to perform a simple operation similar to counting. As can be seen in the video, every time I compress and decompress, one more beam is flipped to the right.

Read the paper here, or for free on the ArXiv.

Overview

• 3D printer
• PLA or PETG filament
• 2-component silicone rubber (we suggest you use Moldstar 30)
• Mixing container and stirrer
• Cutting pliers
• Needle nose or other gripping pliers
• A small sharp knife, scalpel or razor
• Vacuum pump (optional)
• Talcum powder (optional depending on the rubber)

Safety kit

• Band-aids
• Safety glasses
• Gloves (recommended by the manufacturers of Moldstar 30)

What rubber did you use?

The samples in the paper are made of Zhermack Elite Double 32. This is a platinum cured VPS rubber. This rubber is quite popular (in the scientific field) for various reasons but it is difficult to source as its primary market is the dental industry. Dental suppliers don't like selling this stuff to non-dentists which makes it not really DIY friendly.

What rubber do I suggest you get?

Moldstar 30 from Smooth-on is much easier to source and works quite well. Possibly better, but that's hard to gauge. You don't have to degas this rubber either as it is a little less viscous than other rubbers. Vacuum equipment can be a little expensive so that's also a nice feature, specifically for the DIY friendliness of this material.

What should I look for in a rubber if I want to use something else?

Some rubbers are softer and some rubbers are harder but that is not really something that will make or break the functionality of the beam counter. It is most important that the rubber is not visco-plastic and or visco-elastic. What that means is that the rubber is more like a rubber band than for exaple liquorice or gummy bears. Both are soft but the candy slowly creeps back after deformation and the rubber quickly springing back. You likely won't find these properties on the back of the box of any silicone rubbers, and if you do find them, its probably a measurement of those properties for the individual liquid compounds that you mix, instead of the properties of the final solid material.

Can't I just 3D-print the counter out of a flexible filament, resin or other soft material?

In principle yes, and we included a file for that! (the 4-counter_PART and 10-counter_PART files)

However, when trying this ourselves we found most materials not to be very suitable for the task. The best results that we got were with Filaflex 70a from Recreus and that kind of worked, but is really difficult to print. The beams plastically deformed a little too much during compression so that the counter hangs on one cycle. Possibly you can get it to work, but it hasn't been worth our time yet.

If you have a rubber printing process that works really well though, get in touch ;)

This step should be easy. We suggest printing the mold (either the 3- or 10-counter file) out of PLA because it's easy to print and it is also quite brittle, which will help with "deconstruction" later. We printed the samples using Cura with Ultimaker S3 and S5 printers and standard 0.4mm nozzles.

We suggest printing the samples with a brim to improve adhesion and reduce the chance of the print failing.

The internal pins can be challenging to print and you will need to make sure your printer is not over-extruding. If you want really pretty parts, using a smaller diameter nozzle (0.25 mm for example) can help for these features. Just printing a lot slower is also probably a good idea if you run in to problems.

Why are the walls so thin?

The walls of the molds are made thin so that it is easy to remove the sample later and because the molds are single-use only. The wall thickness chosen is pretty much at the sweet spot for: ease of deconstruction, non-arachne slicers of yesteryear, and the walls not buckling under their own weight.

Cleanup

Did the mold come out all right? There are a lot of retracts on the mold so you might find there to be some stringing on your part. You should try to clean that up as much as possible or try change your slicer and printer settings to reduce this stringing.

Did the dimension come out all right?

You can measure the dimensions of the mold to check. The walls should be 0.4mm thick, the thickness of the thin beams should be 1.32mm and the thickness of the think beams should be 3.72mm. If the pillars in the middle of the thick beams are very bumpy and is there no hole visible in the middle, then you might be over-extruding or oozing too much filament.

It should be easier to measure these dimensions on on the rigid mold than on the soft part that comes out of the mold at a later stage, which is why we recommend measuring the mold.

The flexure used to compress the samples takes quite a bit longer to print. It requires supports and the easiest way we found to print it is with supports everywhere except on the point where the "plunger" connects to the "main moving bit". Add a support blocker here (see figure) as otherwise it will be very hard to get the supports out. Most printer should be able to bridge that small gap.

For the material of the flexure, we recommend using PETG. It doesn't plastically deform as easily as other materials but PLA and ABS should work just as well given the range of the flexure.

We tried printing the supports out of dissolvable PVA supports but that's really not worth the hassle. Getting the normal supports out with some needle nose pliers should be very doable.

Depending on whether you want to compress the 3- or the 10-counter use the regular or WIDE crushinator file.

While your printer is working on the flexure, you can go ahead and cast the beam counter.

Mixing the rubber

Follow the instructions on the bottles of your chosen rubber. You will likely have to mix the two compounds 1:1. Put some elbow grease into stirring. If you are using a transparent beaker we recommend putting the white compound in first and then the colored one. It doesn't matter for the final product but it makes it easier to see whether you properly mixed the rubber at the bottom of the container.

You don't have to use disposable cups when working with rubber, as compared to resins as you can just pull out the cured bits out when they solidify and make a colorful coaster.

The approximate masses and volumes for the different beamcounter molds are:

10-counter: about 85ml (or about 110g for moldstar),

3-counters: about 30ml (or about 40g for moldstar).

Vacuum degassing the rubber

After the mixture is complete, it is recommended to vacuum the rubber to get rid of small bubbles. If you don't have a vacuum pump and vacuum chamber you can skip this step. You will be more likely to get bubbles which can be a problem if they end up in the wrong places. This effect can be mitigated a little in pouring as discussed later.

When you pull the vacuum, the surface of the mixture will bubble and come up before dropping back down again. You can keep the mixture under vacuum for a minute but don't wait too long as it has the rubber is starting to cure.

Pouring

I recommend slowly pouring a constant jet of rubber into either end of the counter and letting the rubber slowly flow in to the beam-sections. When you pour over the beam-sections you are more likely to trap air exactly where you don't want it!

We recommend pouring from up high with a thin stream of rubber, especially if the mixture has not been put under vacuum. This supposedly pops bigger bubbles as they move from the bulk liquid to the thin stream.

Once all the rubber is in, and if you overfilled a little, wipe over the top of the mold with something flat like a credit card. In my experience it's better to have a slightly concave instead of convex top surface.

Don't worry about cleaning up just yet

Did any silicone drip on your counter top? Don't worry. When it solidifies it will be much easier to clean up as you can tear away the solidified pieces.

Are you still worried? Just work inside of a container like a bucket or on top of a big plate.

This is the deconstruction bit. We will break open the mold to remove the sample. Make sure the rubber has completely cured. The bottle the rubber came in should mention the curing time but this is greatly influenced by the temperature at which the rubber cures. Check the state of curing by poking the rubber left in the beaker, or the rubber that you mistakenly poured on the floor or table.

Can't I re-use the mold?

Maybe you can but it won't be easy. The mold doesn't have "draft angles" that professional mold makers would use to make it easy to get the part out. While researching I hardly ever made two samples with the same dimensions twice, so it wasn't worth my time making a more re-usable mold.

You could also definitely improve the mold by making it out of multiple parts that fit together. Feel free to remix the files to make that yourself! I haven't yet and it would probably require a resin printer to get the tolerances tight enough for the rubber not to leak out. You could use some Vaseline/petroleum jelly to fill in the gaps but that only gets you so far.

Cut the flash off

There might be some thin bits of rubber on the edges of you mold that you want to remove. These thin bits are often called flash. You can choose to remove the flash after demolding by grabbing them and tearing it off, but if you have a sharp knife it might be easier for you to cut them off while they are still in the mold. Put the knife at a shallow angle and cut along the wall to remove the flash.

If the flash is particularly thin you can also rub it off with your fingers while it is still in the mold.

How do I deconstruct the mold?

Again, using some elbow grease. But before you start, put on some safety glasses. Flying bits of sharp plastic are best kept outside of the eyes of you and your loved ones.

To deconstruct the mold ,I like to make a cut in the perimeter wall, then rip the layers apart with my hands or needle nose pliers. If you have your 3d printer setup so that the layer adhesion is less than stellar, this step should be a lot easier and. After most of the perimiter is gone, you should be able to rip the backside off at which point the sample should come out. Try not to rip pieces of rubber out of the sample, but if you have to, don't rip them off of the beam-elements but instead from one of the "bases".

Did you break the pins in the thicker beams? You can likely push them through by hand, another pin or a paperclip.

Cleaning up

All the rubber on your countertop and equipment should also have solidified by now. You can pull them off and discard them. The cured silicone is inert and not dangerous. It can be recycled, but when thrown away will likely end up in a landfill. Take your time to take out your free colorful coaster from the bottom of your measuring cup instead of throwing it out.

Be careful

This is the bit where you are most likely to cut your fingers. Please don't.

If you do, that's where the band-aids are for.

To finish the beam counter you will have to cut slits in all the thicker beams. Before you start cutting, make sure you correctly orient the beam counter as shown in the picture. You want the left-most beam to be a thin one and the rightmost beam to be thick one. Then make a cut on the right side of the thick beams terminated by the cylindrical holes.

Support the bits you want to cut

You want to support the thick beam with your fingers around the part that you are cutting. Don't worry too much about cutting the thinner beams as, unless your knife is crazy sharp, you wont easily cut through unsupported beams.

Depending on the rubber, this post-processing step may or may not be important. To make the rubber less sticky you can coat it in talcum powder. You need very little, but make sure you get it between all the beams and in the slits you cut in the previous step.

I like to put some powder in a small container, dip my finger in and rub it over sample. You don't need a lot. Once you are done, you can tap the counter on the table to get the excess powder off.

Try not to breath in the talcum powder and maybe wear a mask. It's probably safe, but there has been some controversy around the safety of talcum powder in the past.

Now for the last bit, you can put the sample in the holder.

Make sure the sample is aligned straight up. To align the sample to the front of the counter (so that the beams don't hang on the back surface) you can flip the holder with the beam counter over, put it flush to the table and pull the flexure open a little. Release the counter and flip it back over.

Now comes the moment of truth. Simply press and decompress.

Depending on the tolerances and other factors, you might have to apply a little force when decompressed, for the beam counter not to reset.

Is it not resetting by itself?

You should be able to reset the counter by forcing the flexure open a little but that will depend a lot on the rubber and the tolerances. To "speed up" I like to add a little gravity assist by tilting the whole thing so that the thin beams (the memory beams) flip to the starting state.

If that doesn't work for you, you can always reset the counter manually by coming in with your finger.

Is it not working at all?

There are a host of different issues that can cause the beam counter not to work. The dimensions of the beams might be the first thing to check, second is if there are any visible bubbles in wrong places. If you tried a different material it might be too gummy. Feel free to share your experiences below.