A Bucky Lamp WIP

[Yojevol]

Background

​

It must be 20 years ago that we bought this bucky ball lampshade for our bedroom although I didn't know it was bucky at the time. I would lie awake looking up at this new feature fascinated by its design and construction. How could such a complicated structure be made for such a modest price? (BHS Lighting Dept). Studying it from the comfort of my pillow I realised it was made up of 5 and 6 petalled florets and what’s more it had the regular structure . Each 5-petalled floret is surrounded by 6-petalled florets and each of the 6-petalled is surrounded by 3 x 5's and 3 x 6's. This arrangement, being reminiscent of a modern football, gave me a lead in to study the ‘ball’. It turns out that it has a formal name - wait for it - ‘Truncated Icosahedron'.



The common-or-garden plain icosahedron; one of the platonic solids (the Greeks were into these things). It has 20 faces (icosa = 20 in ancient Greek and hedron = faces) which all fit together perfectly:-

To get to the truncated version all the pointed peaks are truncated (chopped off), like this:-

​

Thus the familiar football appears. It has 20 hexagonal faces, 12 pentagons and 60 vertices. This structure is hugely important in the 21st century because carbon atoms can spontaneously coalesce into this molecular form (Carbon60) with an atom at each vertex . It is commonly found in soot. The beauty of it is that it can be elongated into tubes by adding in extra rings of hexagons to form immensely strong carbon tubular fibres.

Truncated Icosahedron as a moniker doesn’t roll off the tongue too easily so it has acquired the nickname of ‘Bucky Ball’. Therein lies another story:-
Richard Buckminster Fuller was (he died in 1983) a renowned American architect who developed the geodesic form of construction (think Eden Project) which is based on the basic icosahedron. His name is also applied to the many molecular forms which are known as ‘Fullerines’

About the same time as acquiring the lampshade I was beginning to get adventurous in the use of TurboCad. So why not, I thought, just for the satisfaction of it, produce a bucky ball in CAD?
I spent a lot of time fruitlessly hunting around for some geometrical clues to get me underway. I just couldn't see how to get started, but then it struck me; just do what I normally do when detailing a bit of woodwork, ie. simulate in CAD what I would do in reality in the workshop.

In this case, if I were making a Bucky Ball out of card, I would start off like this:-
Keeping HexA flat on the worksurface, simply fold B and C up to close the gap and glue. →

It’s not so easy in CAD because you can only move one item at a time and you don’t know where to rotate it to.
So a few construction lines are required to determine where the adjacent corners meet. Now B and C can be rotated up, one at a time, to the desired position. This provides the basis for the whole Bucky Ball construction.↓

This was my ultimate interpretation of a bucky ball in CAD →​

An Idea Develops

Perhaps inspired by @pjm699's design, I started to think about the possibilities of making a bucky ball in wood. The obvious hurdle to over come would be the making 32 facets with sufficient precision to fit them all together accurately. It was only after I kitted myself out with a CNC router that I contemplated the means to make them accurately and in quantity.
My thoughts went back to the design. I wanted more than just a ball with holes in it.
Back to the lampshade in the bedroom; make a bucky light; hang a bulb in the middle.
No, too easy, too boring. Have separate lamps shining out of each hole. No, still too easy.
Have something that draws the eye INTO the ball.
Yes, Yes; a light feature that draws attention. Now that would be a conversation piece.

I'll leave the details of the individual lamp unit design hanging in the air for the time being and concentrate on the woodwork.

The basic bucky ball needs, to my mind, a little something to make it more distinctive and attractive, that little bit of finesse. So I'll introduce a cock bead between each of the facets in a contrasting wood.


Other details are:-

1. 2 sizes of lamps; one for the hexagons and a smaller one for the pentagons
   
2. lamp orifices edges rounded over.
   
3. top and bottom pentagons will be lamp free.
   
4. ball will be made in 2 halves, top and bottom. so that it can be split to allow access to the electrics inside
   
5. 2 halves will be joined together with an internal threaded rod connecting the top and bottom pentagons (hence the plain panels).
   
6. woods will be American Black Walnut and ramin
   
7. support, will be a stand or hang from the ceiling, yet to be decided.

Prototypes and Jigs
I've spent many weeks getting to the stage where I'm confident to actually start manufacture. During this time I have experimented with procedures and jig design. These will become evident as I go through the job.

Work-in-Progress
This will be a true WIP. I'll report as I do it. It may turn out to be a long drawn out process.
There's always the possibility it won't be successful. Even if it is successful I may not actually like it or I may be refused permission to deploy it but, whatever, the journey will be worth the effort

Brian

 

[Orraloon]

A very educational post. All the best with the wood version.
Regards
John

 

[Yojevol]

Well it's time I got going for real on this project starting with preparing the ABW for the 20 hexagons and 12 pentagons

About 3 years ago, definitely pre-pandemic, I made a dining table for a client of my daughter's interior design business. The slabs came in 3m lengths of which only 2.5m was required so I 'won' 7 lovely bits of surplus timber. Here are the two I selected for this job:-
The smaller piece had already been used to supply some material for the development phase (see below). Mostly I used plywood, as in the pic, but It was beneficial to try out the ABW in the latter stages.
The larger piece had to be ripped to a width of 90mm and the sliced to 7mm thickness, I did this by the plane/rip, plane/rip sequence to give me 7 pieces with one planed face and one sawn face. They were then brought down to 6mm on the drum sander:-

Finally they were chopped to length with the mitre sawwith some material left over which will inevitably be needed later on.:-

Time taken 1 hour

Brian

PS. Here is a selection of trial efforts at making the polygons including a trial gluing:-

 

[Yojevol]

Today I prepared the billets for placing in the polygon cutting jig.
First of all boring a 20mm location hole in the middle and then 3 holes for holding down screws:-

All done in 30mins including setting up the pillar drill

I tried a first one in the cutting jig but it was obvious that it was not going to be reliable for the full

production run. The 3 holding down screws were not man enough so I introduced a central thumbscrew which holds the whole assembly fast to the baseplate:-

As you can see the jig is based on a dividing-head indexing plate The 2nd to outside ring of holes has 60 holes so, being divisible by 5 and 6, can be used for both the pentagons and hexagons.
A small dowel pin through one of the 3 screw holes stops the billet rotating with respect to the index plate which is held in place by 2mm rod, aka a drill, through to the baseplate. The baseplate is located on the saw to give a polygon side length of 47mm.
I've managed to make one satisfactory hexagon, so it's 1 down, 31 to go.
Brian

 

[Richard_C]

Good stuff, thanks.

I've always found bucky balls fascinating although hard to get your head around - hexagons and pentagons but seen as an open frame it's a load of triangles . There was a whole sub culture of buckyball desert dwelling in the USA.

Just for interest and not helping your build at all, here is a picture of the big ball designed by Buckminster Fuller for the Montreal world fair in 1967. The building inside it is a much later addition.

Also link to a site with all kinds of formulae for diy 'outdoor' versions which others might find interesting. If my garden was a little bigger I would be tempted to build the bamboo version to use as a runner bean support instead of a boring straight row. Not quite got room.

www.desertdomes.com/index.html

You've probabaly already found the buckminster fuller institute site, here it is:

www.bfi.org/about-fuller/big-ideas/geodesic-domes

 

[Yojevol]

Today I rattled off the cutting of all the hexagons and pentagons. I was amazed at how quickly they could be done, averaging 2 minutes each. This photo was taken just as I was about to make the last cut:-
-

Having got them it's difficult to resist laying them out to show them off:-

Job done, but this is not how I originally envisaged. As indicated in the introduction, I thought I would be using my Denford CNC machine. However after a few trials I was not happy with its reliability. Occasionally I would get a side that was not quite straight. I haven't investigated the cause yet but I suspect that one of the travel guides is sticking slightly (I have recently done some refurbishment work on these guides). There's no room for errors on this job so I turned to my Metabo mitre saw to see what that could offer. This is the method I tried first:-

A fence angled at 60° and the saw set at at 60°. However this was difficult to get the required accuracy and was prone to building up a cumulative error as I progressed around the 6 sides. What was needed was a method of rotating the workpiece that was independent of anything else. A turntable would do the job, but how could I get it to rotate through exactly the desired angle. At this point I recalled that a dividing head index plate was lurking somewhere in the nether regions of my workshop. I acquired it years ago thinking that it would come in handy one day. Well its day did arrive and I'm very pleased I didn't throw it out in a recent cull of underused tools.
I now realise that a plywood turntable would do the job and be comparatively easy to make. It would just need a bit of care in drilling the 5/6 indexing holes. It would also avoid the possibility of sawing into the steel index plate (I had to modify the saw's vertical limit stop to avoid that reliably for this exercise).

Next step - boring the lamp holes which will be done on the Denford

 

[Yojevol]

Change of plan. I decided that boring the lamp holes was not sensible at this stage; better to keep the reference 20mm hole and use it to carryout the final procedures on the outer shapes of the panels, ie, add the cock beads and mitre the edges.
The beading is being made from Venetian blind (ramin) slats. The process is:-
1. strip off the paint from one side of 6 slats on the drum sander.
2. Round over both edges using a 2mm router cutter on the spindle moulder.
3. Cut off both edges on the bandsaw to give 2 x 8mm strips from each slat.
4. Cut the strips in half lengthwise to give 12 lengths of beading
5. Gang up the 12 strips and produce a mitred end using the disc sander
6. Chop off the prepared ends in a jig on the mitre saw set at the mitre angle (30° for the hexagons)
7. Repeat to produce 120 beads for the hexagons and 60 for the pentagons.

Here are some pics of the process:-

--------The material thru the stages-----------------12 lengths on the sander-----------------------and on the mitre saw

The mitre saw jig is interesting in that the length stop (that's the bit held in place by the thumbscrew) has its end cut at 30° to be parallel with the saw blade so that the cut-off pieces are all the same length. However the workpieces themselves have to be reversed so that the cuts are angled in the opposite sense to the first end.

The next stage is to glue them on to the polygons and for this I developed 2 jigs to hold them all in place:-

Each consists of a central boss to hold the panel and 5 (or 6) cauls which are pressed on to the the cock beads by a rubber band. The cauls can be held back by a dowel through a notched block. These blocks also keep the cauls flat on the base.
These jigs enable me to trial fit all the beads and then glue each bead individually without disturbing the adjacent ones.
The panels are mounted front face down in the jig such that they rest on the strips surrounding the central boss. Thus when the cock beads are pressed down onto the base they end up a consistent 2mm above the panel face.
I'm using CA glue which means they have to remain in the jig for only a few minutes.
I've now finished all the hexagons. Here are a few:-

You begin to get an idea of how the cock beads will look.

Now for the pentagons

Brian

 

[Yojevol]

Having got all the cock beads attached, the next stage was to shave most of the ramin material away to establish mitres on all the

sides. My CAD model informs me that there are 2 relevant mitre angles between plates,18.69° for a hexagon/pentagon joint and 20.91° for a hex/hex. So starting with the hex's, set the spindle moulder angle using my Trendy new toy:-

20.9 will have to do​

When I was prototyping I carried out this stage by setting up a false fence and table with just enough tool protruding to make the cut and ran the workpiece through by hand. That was OK for a few but I felt I needed something a bit more secure for the 180 production cuts. Also it meant that only the aris at the top of the cock bead was in contact with the fence after the cut. So I made this sled,

the front edge of which controls the depth of cut. The position of the workpiece is fixed by a spigot up through its centre hole. This worked fairly well until the cutter was called upon to cut uphill. ie. against the grain. I had 2 failures (eating in to my scrap allowance) before I realised what was going on. The cutter was catching the grain causing the plate to rotate which then caused it to split in half. To avoid this happening again I ground away most of the material on the disc sander, where the grain would be troublesome, just leaving a slither to be removed by the router cutter. I also went to a smaller diameter cutter and made sure the thumbscrew was really tight.



For the pentagons I made another sled as the dimensions are different and I introduced a bit if course abrasive belt bonded to the sled to help grip the workpiece. It all went to plan from then on.

----------

----------

----------------------------------------

Job Done
Brian​

 

[Yojevol]

Goodness me, it's 3 weeks since my last post here. Things on this project haven't gone too smoothly recently. That we're now in the open season for mince pies hasn't helped.
It was my (revised) plan to complete the panels by cutting the central lamp apertures on the Denford CNC router. The first one was semi-successful - the cut was very ragged. I decided that a downcut spiral tool was needed. Getting hold of one took a few days, but after that I did several good cuts. After a few days mince pies break I came back to the job but to my dismay the cuts were off centre as can bee seen here:-

I spent ages trying to work out what had gone wrong, but never did come up with a definitive reason. By this time my scrap allowance of hexagons had evaporated so I set to to make 4 replacements - more time lost.
Having got an acceptable batch of 20 hexagons, it was time to turn to the pentagons. I decided to not risk the CNC again and went for a good old fashioned boring method on the pillar drill using an adjustable bit:-

This method progressed without further incident.

Today I started off thinking I would try putting a few panels together using sticky tape. I taped 5 hexagons together on the flat then drew the free ends together to for a bowl. It was very easy to do. Encouraged, I attached the next layer of 5 pentagons, then 5 more hexagons, another 5 hexagons, 5 pentagons, 5 hexagons and the last pentagon. Job done - easy, peasy:-

There are a few minor gaps but they will pull up with a little persuasion (or filler).
Brian

 

[Yojevol]

LAMP UNITS
Now that I've finished the basic woodwork I need to turn to the lighting aspects.
Each panel, apart from the top and bottom

pentagons, will have a lamp unit; that's 30 in all. The basic construction will be like this prototype:-

The plastic ring is cut from a length of drain tubing and a length of LED strip stuck on the inside.

Having cut the rings from the tube using the bandsaw, they are then trued up on the lathe:-

The plywood ring not only supports the plastic but also acts as a length guide.

Next I need to make the rings slightly larger in diameter so that the LED's

can be equally spaced around the inner wall.

To do this I slit the wall on the bandsaw and glue (CA) in a 5mm bit of wood, ABW of course, which is then trimmed back to match the plastic.

And here they all are ready for the LED strips

​

The next stage is the boring bit. I may be a little while.




Brian

 

[J-G]

Excellent Brian, particularly taking account of the LED spacing -- I might have taken a different (more difficult) approach of boring out the plastic tube!

 

[Yojevol]

 

[Yojevol]

Progress has been made since my last post but it's been a bit of a roller-coaster not helped by the winterval and open season for bagging mince pies with all the trimmings.

The next phase was to produce the lamp backplates and front covers out of 3mm acrylic sheet. This was to be done on my Denford CNC. Having prepared all the g-code instructions to produce them in a batch production process, the first operation was to test it out with an 'air' cut, ie having the router perform a few mm above a drawing just to confirm it's doing what I intended and then a real cut on some thin MDF (no spare acrylic to waste):-

Right. We're now ready to go.
I got the acrylic screwed on to the sacrificial table top and set the offsets (CNC shop talk), Just to be sure I decided to do another air cut. Unfortunately at this point Denny decided to go on strike; the X-drive (right/left tool motion) seized and refused to budge.

This was serious. I had to resort to dismantling the X and Z (up/down tool motion) drive and guide systems to find what the problem was. It turned out to be one of the nylon guides was a tight fit on its guide rail. These were renewed by me some months ago with self-made items. It just needed a small amount of nylon removing, with a home made abrasive file, to get it running freely again.
It's a bit of a mystery as to why this should have happened. The only thing I can put it down to is the use of the silicon spray lubricant I had been using earlier to improve machine's motion. Maybe this had caused the nylon to expand slightly.

Having got the travel system reassembled I gave it a powered test. All seemed to be fine. Next I reinstalled the router itself (a Porter-Cable unit). Further testing threw up the next problem, The router wouldn't run when energised. Oh no!

Time to go and finish the last of the mince pies washed down with a glass of port.

Next day I went through the power supply system methodically:-

1. Can I hear the power feed relay operating?
   
2. Is router feed fuse okay?
   
3. Is there a voltage across the router input terminals?
   
4. Is power reaching the router on-board manual toggle switch?
   
5. Is power coming out of the switch? - , no matter how the switch is set.

So there was obviously something amiss with the switch. I prized it apart and the problem immediately presented itself; it was full of wood dust. 18 years worth with no obvious way to get in.
The switch innards were the usual collection of small bits of metal, plastic and tiny springs. I gave up trying to reassemble it and went and ordered a replacement. To get on with the job I merely connected the switch wires together thus getting power to the router motor.

Off we go again. But, not so fast.
It turns out that acrylic sheet is tricky stuff to machine. After a few minutes the tool had bunged up with molten plastic. I had previously experimented with this tool (the recommended single flute end mill) on some scrap plastic sheet (in hindsight, probably not acrylic) with no problem. I had researched speeds and feed rates and thought I should be in the right ball park. I decided to buy a new larger diameter tool thinking that a larger flute would be beneficial. I experimented with speeds and feeds and ended up going with 5000rpm, a feed of 800mm/min and a depth of cut of 1.5mm.
I also abandoned the idea batch cutting and did each backplate individually. This would reduce the risk of the whole batch going down as a result of another problem. Here are some pics of the machine in action:-

As you can see I was struggling to get the last one out of the available material.

Having now got some experience I decided to go ahead with a batch

production of the front covers which are a simple circular design:-














So after much heartache I ended up with 30 kits of parts to make up all the lamp units. Here is a sample of the 2 sizes:-

Note that the 3 tabs on the backplates have been doubled in thickness. This was necessary because the nearest suitable woodscrews I could find, at 25 x 3mm, are too long. They would otherwise protrude right through the hexagonal/pentagonal panels. Here is one fully assembled:-

Time to apply some juice:-

Night Night
Brian

 

[Yojevol]

The next stage is assembly, so here we are on our kitchen table where it's a lot warmer and tidier than the workshop:-

And fully assembled, at least the wood elements, with a couple of coats of spraycan acrylic lacquer:-

Taking advantage of a nice bit of sunshine this afternoon.

Brian

 

[Sporky McGuffin]

Ye gods - an infinitybuckylamp? You, sir, just won at Instructables!

Quite marvellous. I am greatly enjoying this.

 

[Yojevol]

Ye gods - an infinitybuckylamp? You, sir, just won at Instructables!

That's it Sporky. Hopefully I will achieve my idea as proclaimed in my introduction:-
"Have something that draws the eye INTO the ball.
Yes, Yes; a light feature that draws attention. Now that would be a conversation piece."

I achieved this today after sawing up a ton of dead elm logs - from one extreme to the other:-

All the lamps fitted to the bottom half. A bit of soldering to do now!
I had a bit of trouble with the screw pilot holes which I had pre-drilled to a common pattern. They worked for the bottom layer of 5 hexagons but for the next 2 layers I had to re-drill them to get the back plates to nestle in neatly. A couple of mishaps (drilling right through) caused me to 'invent' this drll depth stop:-

Hopefully the top half will go more smoothly

Brian

 

[danst96]

Wow! Nothing else to say but wow. I doff my hat to you good sir.

 

[Fred48]

@Yojevol
Really enjoying this thank you. Amazing.

 

[penyrolewen]

I amazed at your skill and patience. Love it.

 

[Yojevol]

Both shells now fitted with lamps and wired up:-

And from the outside (protective film still in place):-

And energised:-

Brian

Edit - I've just realised that all the LED light rings have been installed upside down. That's why the images appear to be in pairs.
Oh dear, a hard decision needed