Time warp
March 16th
On my way to the 8th SDX meeting, I had a breakthrough.
I’ve been trying to search for a better version of a side-glow fibre optic cable because the one I had was not working as intended when I remembered that one has tiny score marks where the light shines through. So I thought, why not test this, to make sure I make the right purchase? I felt a little worried about taking out my model and scalpel on the train, but nobody was around and I was quick and careful. The result is obvious – even lower down the length of the fibre optic, the light is very bright where I scored it.
This means, and I will need to test this, but I think if I got the 0.75 mm wire-reinforced fibre optic cables I found on eBay, and used sandpaper down the length of it, it will be a pleasant glow throughout. Unfortunately, I can’t try this out right now because I don’t have sandpaper, nor any more length of fibre optic or LED accessible. But according to this website, my idea might just work. The cable I found on ebay (seemingly the only wire-reinforced one on offer, I wonder why there aren’t more available) is more expensive than the one I had, and because it’s so thin I will need double the amount. Or perhaps the narrow waterfall will only need one, while the thicker one can have 2-3. That means about 30-40 cm used per product.
This video suggests sanding down the top of the LED to connect the fibre optics more snugly, and use a black shrink tube, with gentle heat. I probably won’t be purchasing black shrink tube, but it shouldn’t be necessary if the whole unit is on the inside of the product within specially designed holes.
And here, someone is suggesting using black Sharpie at the ends of the fibres to diffuse the end light and make the sides glow brighter. Then another person suggested white paint first, then black, to reflect, but not allow light through from the outside. I think fibre optics might be something to work with in my 4th year project. But that is an idea for another day.
Also this gentleman shows how boiling water can create a permanent bend in the fibre. I’m not sure how I would achieve this – perhaps create the shape of the bend by 3D printing or laser cutting just where the bend would be, and keeping it as a forming tool. He also wrote an Instructables page on it. He says: “as a rule, don’t bend these fibers to a radius less the 5X the fiber diameter”. For the 0.75 version that would be a 3.75mm bend.
At the meeting the prototype I made was well-received, and I got some constructive feedback and ideas on how to proceed.
We tested our products on an empty shelf, and I have to say I feel I was right – my (comparatively) tiny single product would look very lonely with the large white space surrounding it, and I will need to source lots of natural materials (mostly rocks to go with the theme we already liked previously, possibly laser-cut tropical leaves) to combat this. I will also design a bio which should look attractive and should be easy to read, not overpower but compliment the product.
I was a little taken aback by the news that the product and display should be finished by April 1st – I still have a lot to do, but I enjoy the creation, and I now have most of the processes figured out.
What I will need to do this week:
Order more fibre optic cables (research what size and type I need)I have now done this – purchased 10 meters of 0.75 normal fibre optics, without the wires – I figured it wasn’t too expensive, and if I will permanently bend them, I won’t need the extra material of the wires.
Sand down fibre optic cables and test- Order black TPU filament?
- Test for best blue wood stain
- Test fibre optics with heat shrink tube – flatten multiple 0.75mm side by side?
- Design and test backing with Polybak laser cut layers
- Test blue vinegar/steel wool stain on kapok?
- Find best way to avoid having to solder – pre-made LED with USB?
- Solder more circuits?
- Finish CAD design
put USB cable on the back corner(it is too thick and in the way, takes away from the display quality)redesign fibre optic cable holes(different size and more oval)better way to keep button from being pushed in- a way to take off top layers and be able to open up the electronics
Convert the CAD layers to SVGsCreate final Lightburn page layouts- Laser cut, stain and glue pieces together (make sure there are no glue marks)
- Design, decide on material and engrave/cut bio and business cards
- Calculate all material and making costs exactly! (business card, bio, etc.) (write list of materials)
All materials used for a single product:
- 8 (?) sheets of 20×30 2mm plywood
- 0.4 sheet of polybak?
- About 50 grams of Gorilla wood glue
- X grams of moss
- About 100 ml of glycerin (for preserving moss)
- About 10 grams of coarse sand
- About 0.01 grams of kapok fibres?
- About 0.2 grams black TPU filament?
- 3 x 10cm 1.5mm or 8 x 10cm 0.75mm fibre optic cables?
- 50 mm transparent heat shrink tube?
- About 100ml vinegar and steel wool for stain?
- If electronics soldered:
- 2 pcs LEDs
- 0.5 pcs USB cable
- 1 pc button switch
- 250mm electrical wires
- 1 pc 47 Ohm resistor
- 300mm heat shrink tubes?
- X gram soldering wire
- If no electronics soldered:
- pre-made LED lights?
March 17th
After some rough calculations, material costs would amount to roughly £10, which sounds great considering we discussed a £50 product price.
However, my hours roughly: Moss collection and preserving: 0.5 hours (if in bulk); Laser cutting: 1 hour; Gluing: 1 hour; Soldering: 1 hour
That is about 3.5 hours per product, roughly £35, and that is extremely optimistic (knowing my perfectionist tendencies, I would probably need 3 times the amount of time). So material costs + my time = £45. If I gave a donation of about £3 per product to WWF, that would make my profit all of 2 whole pounds. Which doesn’t sound fantastic if you ask me. Of course, I rounded it all up and I’m sure I could automate or bulk or outsource some of these activities, but it’s looking like the price I was told people would pay for it is a lot less than it is worth, objectively. Hm.
If I rounded it down, and calculated about £5 for materials, and 2 hours to make it efficiently, that would be about right. But this means I will need to find efficient ways.
Some ideas to do this:
- Purchase large sheets of 2mm plywood and arrange the pieces so that there is the least amount of empty space, and the least amount of time taken up by checking if I have all the pieces, and putting them in order
- Purchase all materials from wholesale or receive for free (hand-collect, off-cuts, upcycle etc.)
- Outsource 3D printing, laser cutting, soldering, gluing duties
While this all sounds great from a business perspective, to me it feels depersonalizing to the product, making it a means to an end as opposed to a unique, hand-made object designed to provoke positive mental health and environmental action. I think if the product truly takes off and people are interested in it, this will be a really difficult balance to strike. But I am excited for the what future may hold for Chrysalis.
My quick experiment, whereby I cut two equal pieces of the 1.5mm fibre optic cable, sanded down the entirety of one of them, and put them both next to an LED light seems to prove my theory (see the cable on the right whcih was sanded being a lot brighter)
March 18th
I have completed the redesign on CAD:
- Smoothed the edges, added two extra tabs behind the button for strength, the furthest one attached to the layer below
- Added a sunk-in plaque (I seem to have lost the previous file but there weren’t many changes)
- Put the USB cable in the back
- Smoothed inner compartment for the electronics (this might end up to be too large
- Sunk the LEDs further in (closer to the fibre optics for extra brightness and stability of the cable connections)
- Redesigned the strain relief (this was necessary because the previous one didn’t fit in the new angle, but I will need to test printing it – I sunk it into that inner lining instead of having it stick out this time)
- This of course meant redoing all of the layers
Due to the layered nature of the design and the fact that the electronics are inserted into the top layers, I have not been able to find a way to make the these layers detachable, to be able to access the electronics.
I tested a simple “cage” idea, using thin cardstock strips temporarily attached to the model with double-sided tape. I liked the effect, so I created it in CAD as well, to try to test different shapes, but it is difficult to form curved shapes.
I also tried leaf-shaped strips, which made the backing look more like a dome. I can’t quite decide which one I prefer.
I used some small pieces of Polybak, pierced holes in them and used a small bit of the 3mm aluminium wire to see what it would look like if I used that to hold the strips together. I think pushing the strips into rectangular holes of the right size through most of the base (16mm) and using some wood glue will be enough to hold them.
The Polybak sheets are 30×45 cm, and the first design I created for the leaf-shaped strips are 2x 180mm, 2x 170mm and 1x 160mm.
Because I edited the container for the electronics in the base, and deleted the last layer as that was not needed (of course some of the holes had to be changed) I was able to eliminate one of the eight Lightburn files for the plywood layer sheets. There will also be some empty space left (I am basing the arrangement on the 20×30 cm sheets I currently have – in the case I use up all of these, I will be looking into larger plywood sheets which will be able to save me even more space).
I tried printing the new version of the strain relief, with the previous settings I used. It did not go well – I think 3D printing will not become an everyday hobby. The constant unexplainable issues are driving me crazy.
Anyway, I sliced the train relief with different settings the second time – I used the in-built Super Quality, with the layer height being 0.12, and the print speed changed to 25mm/sec – this made the printing time nearly 1 hour. For a single, tiny strain relief. Thankfully, as it was printing, the time was reduced from 56 to 32 minutes. And the result is a hundred times better, though not perfect. The tightness was perfect, it was quite difficult to insert the USB cable.
I worked on a few more Polybak dome designs, this time in a geometric style. First I cut out a lot of 3x3cm triangles and put them together with masking tape. When this didn’t work very well, I tried 5x5cm triangles, and after a little research, I also used 5×5 pentagons. I think this shape works quite well.
I cut out the insides of the 5×5 triangles and pentagons, to see what it would look like because we wanted to make sure that all sides of the diorama were still visible, but I’m not sure I like the effect.
Also played around with some rough sketches to find a unique shape.
And here are a few more images I found online of different geometric dome shapes.
March 20th
How time flies!
Today I received the 0.75 mm fibre optics.
Agenda for the next three days:
- Finalizing dome design (CAD, card models, best way to fix to model – holes and glue? SVGs, Lightburn)
- Finalizing blue wood tint
- Laser cutting all layers (testing laser on Polybak?)
- Testing hot air gun on heat shrink tubes
- Testing fibre optics permanent bending (hot water?)
- Gluing layers, staining
- Soldering parts
- Gluing moss, sand, kapok fibres
- Gluing dome
The business side:
- Designing final logo
- Designing, cutting/engraving bio
- Designing, cutting/engraving business cards
- Collecting more rocks (branches?) for display (and moss for models)
So I made another frame design – I compared all the ones I have made out of paper and cut out holes, and I don’t know if either of them fits. What makes them look worse is the framework style – but if I was only able to use more eco-friendly materials like plywood (as opposed to acrylic, which is plastic, and also toxic to laser cut), that would not be transparent. This would mean I either have to cover up the back (we agreed I would not do this), or I would need to create a framework, but that doesn’t seem to work too well visually.
The problem is that until I decide this, I should probably not be engraving the layers. Unless I attach it to the outside.
I worked on a more basket-style geometric dome on CAD which my friends seemed to prefer, so I designed it for laser cutting and transferred the SVGs to the Lightburn documents. I’m hoping that the 6mm snapfit will be enough to hold them into the base, with some wood glue.
I will need to test cutting out a shape with a 600/60 setting etch for folding the plywood in half without breaking it, so that I can keep at least some parts together without needing glue. I could, of course, make the whole thing out of a single piece, if the folding works, but that would require a single large piece of plywood.
It’s nice that even with the “cage” all of the forms fit onto the 7 sheets of plywood.
Alright, I caved before I even cut the thing, and put the whole cage on one page. It just makes more sense – it would use less glue, and it would be much easier to hold the shape together with folds.
Now that I look at it, I noticed that the piece with the latest dome design looks almost like a treasure chest… Which is a lovely idea, but would require a lot of time to redesign, and it would need a backing with no holes to be more realistic.
Deleted a few tiny hill pieces from the back, because they won’t be necessary and would take away from the front large hill in height.
March 21st
I made a new test batch of wood stain (named No1), with a light green tea (80 celsius, the bag taken out within a few minutes), about the same amount of distilled white vinegar, and about 2cm of the ultra-fine steel wool, in a small glass jar. I used a small foam brush to apply it on a piece of leftover plywood, and placed it on my oil heater to help it dry sooner (8am). I will keep monitoring this to see any changes in the colour.
A few days later I removed the steel wool, which will hopefully ensure that the mixture won’t get rusty.
I figured because experience has taught me that this stain does not last long in the jar, and should be used up as soon as possible (after about half a day/full day of seeping) I created a plywood sheet laser cutting design full of only the pieces which need the stain (14 copies of the three pieces) to make it a lot faster and easier. I will need to try and look into whether laser cutting vinegar, tea and steel wool-stained wood are safe – I don’t imagine a lot of findings, but I would like to be as safe as possible.
In earlier laser cutting tests I have found that drawing a small piece in an empty corner of the work area on Lightburn (fitting onto a 20×30 plywood sheet) using similar colours to my regular ones but darker versions so they will be on different layers (grey for framing, blue for engraving, green now for folding, and red for cutting), and disabling the actual cut pieces allows me to quickly test the settings before getting to the real job.
I made this test grid to test a few folding methods and settings (600mm/min, 60% power, as well as 750/50, both of these with a single and a double line, because having a more curved edge due to the two bends might help) to see which would work better for the cage frame.
I would like to experiment with living hinges, but I do not know what is the best way to design those – I cannot easily create that shape in Fusion360, and simply copy-pasting a living hinge design into Lightburn would not show me what it would look like when used. Most of those designs are intended for a single fold, and not a spherical 3 dimensional one.
My wood staining journey continues, and after about half a day of waiting I have stained a full sheet of plywood. We shall see how that works out when laser cutting it.
While it was still quite wet I paced the sheet on my oil heater, and later realized that this created a noticeable bend so I placed some heavy objects on top of it to smooth it out, and this only resulted in some deeper colouring where the radiator touched the wood. Hopefully they won’t be permanent.
Before laser cutting I checked the laser head to see whether it needed any cleaning, and indeed, it was absolutely filthy with wood soot. I used isopropyl alcohol and a few cotton buds to clean this up, then continued onto testing folding patterns.
My first test of folding grids did not work out very well, but it seems the setting 750/75 for the bends works rather nicely. It will not hold if you force it to break, but it helps to not have to fiddle with the awkward angles of the “cage”.
I tested some living hinges, but in such a small area it did not work at all and rather just broke in half. I would still like to try working with them – I think cutting slightly larger shapes than a single line will help it to bend.
I don’t know if I quite like the design of the cage framework, but what’s nice about it is that the overall shape. What I don’t like is more the inner squares. I can play around with those, but leave the tabs, and holes on the unit, untouched, and it will still fit without problem.
My first test used settings that were not quite as successful, and probably due to the slight bend in the material some areas didn’t go through enough and broke on the top edges.
However, the second try (with settings 750/75) was an absolute winner.
Without gluing, however perfect, the pieces fella part at the seams.
Unfortunately wood glue dries very slowly and was unable to attach the top small triangle and square shapes, so I will need to find a quick and eco-friendly way to do this.
The first layer I cut was actually one which both the already cut “cage”, and the newly purchased 0.75mm fibre optics fit in, so I tried both. The fibre optics were perfect (I made sure not to leave a gap around this time) but I made a small mistake with the cage – the right side front bottom tab is mirrored, and does not fit into the hole underneath it. So I cut it off and will glue on a new tab, but I will probably re-cut it just to make it attractive. Definitely need to remember to fix this on Lightburn.
After I broke a thin piece because it didn’t cut at all at the curve, I tried a different tactic for the next sheet. Unfortunately I have now wasted two sheets of plywood trying to find the perfect balance of cutting time and power. It seems no matter what I try, it just will not cut through fully. My first try was 850/55 with two passes, then I did 850/70. I really thought it could compete with 525/100 at least, but I was wrong. I will just have to try and reuse those sheets by cutting small pieces out or using a scalpel, because I cannot tell if it cut through if I don’t take it out, but also if I put it back in it will shift and will not be a perfect cut. Such a headache!
So now I will do 2 passes of 750/75.
This one actually worked nearly perfectly, but one of the curved corners still didn’t cut fully through so I will change it to 750/80 for the next one.
This barely worked, and the second time the machine, or maybe the honeycomb underneath the plywood kept making strange noises. There was also a large soot mark at one corner but there was no indication of any danger or burning.
My next try will be a single pass 95% power but much slower. Maybe 450. Or maybe that’s still not enough.. Damn it!
425/95 is the one to go with!
Nope. This is getting ridiculous.
415/98. YES! Like butter.
Also the engraving needs to be stronger – 750/20.
As I was cutting the final sheet (5 pieces which did not cut well or broke), the limit alarm went off midway. I was able to cut three smaller pieces, but the large framing layers 0 and -1 did not finish… this means another wasted plywood sheet, and me scratching my head as to what could have happened, because I have not changed anything midway through the cut, and the framing was perfectly fine. There was a lot of smoke, but Lightburn doesn’t mention that…
To do for tomorrow:
CLEAN THE LASER HEAD!Fix cage sheet(left side where the button is, the tab is towards the wrong end – does not fit into the hole!)Did this today – seems that the flying wood dust particles are scratching the inside of the protective orange acrylic cover on the laser head.
- Finish cutting Fix sheet (0 and -1)
- Cut the stained sheet (14×3 pieces – MIRRORED! Put on honeycomb stained side down just to be safe – I may have to find a way to elevate the honeycomb as gases might get trapped underneath)
- Glue, solder
- Collect rocks
- Continue working on logo, bio design
March 22nd
Today I worked on glueing the pieces together. I did not glue the top ones above the left fibre optic to the main piece, nor the bottom three layers because I have not done the soldering yet but it will be helpful to show it tomorrow at SDX.
Things to do (next prototype)
- Bring the LEDs down a little so that the fibre optics can have some more vertical space and will stay there – possibly use glue at a point where it doesn’t touch the LED
- Bring the large waterfall LED and fibre optic top hole closer to the front – it is too much of a bend for the stiff fibre optic to stay in the valley
- Add holes for vertical layers -7b and -7d behind the button for extra strength
- Use small paintbrush to “paint” the little area on the finished plywood piece behind the fibre optics with wood stain (Do NOT use the foam brush – too large and unattractive marks outside the lines – will have to cover fully with moss)
- Fix frame structure and stability (bent edges will break apart, angles will be difficult to glue)
- Learn to solder quickly
- Collect, preserve more moss – test using green food glue for vibrancy
- Test sanding, permanently bending fibre optics (hot water?)
- Collect rocks etc. for display
- Design bio, business cards (3D pop-up?), cut/engrave (wood stain?)
- Update my Instagram with fresh photos and product creation stages (or make a new one)
So I just created my final brand name (Krisalys Creations) and a new Google account/e-mail address (krisalys.creations@gmail.com). I edited the bio according to the store’s guidelines.
I am planning on making a new Instagram account where I upload everything I’m working on, although the one I currently use only has old photographs that I heavily edited, nothing really personal.
Worked on some quick favicon designs. It’s playing with the letters K and C from Krisalys Creations, and the shape of a chrysalis.