Highback chair (Part 4: Finishing touches... Sort of)

Time for a new ... chair.

 

This seems to be a repetitious thing for me, but as it is they make for quite a nice handy prop when photo-ing my ladies within my tight quarters make shift studio.

 

I actually had a few other designs in the pipeline before I was supposed to get to this, but certain inspiration got the better of me, and I finally decided to take some of my new toys to town.

 

namely the planer, and CNC router.

 

Part 1: The pain and joy of CNC

 

now these two were discussed already in the update of one of the previous things, but this time it would be quite a bit more intricate.

 

First off I decided that I would try to make my own glueboard stock.. so I chopped up some dismantled fig Spruce paneling, glued them together, and fed them through the plane, untill I was left with a 4mm thick plank of future grief...

 

wood162

 

Although I now need to back up a second... since the actual design work on this thing started months ago, when I realized that I could in theory create a oriental backrest motif for a chair.

 

to see what I was after, just think anime/visual novel figures, silver hair, china dress, and you'll get there soon enough.

 

So what is CNC?

 

The acronym is simply Computer Numeric Control, so the theory is, numbers go into machine, machine chews the numbers for lunch with a side order of anticipating despair, and machine maybe spits out a usable piece -- perhaps resembling what you thought up in your initial routing design.

 

electro030

 

My router of choice is a cheap DIY kit which you can find on either Ebay or as mine Amazon Warehoused marketplace after I got fed up with sellers that didnt have the kits even though their listings claimed otherwise... going straight to the source from china for me isnt the best option, because of delivery times and possible import complications.

 

These kits come in all different sizes and flavors, so first of all you need to pay attention to the listings for what you are actually getting.

 

By numbers mine is a CNC 3018 (By its designer it was called the 3018 Woodpecker I think) which translates to a work area of 300x180mm with a Z stroke of 45mm with the cheap chuck.

Adding the ER-11 collet holder reduces the Z range to 40mm because its physically longer, but you can offset the motor a bit to compensate.

 

And to be honest... its not really meant to do what I am about to do with it. not in a very precise manner that is. The main issue is the X/Z axis sled that at this width, the sliding rods are perhaps not up to the task of keeping things in check under heavier loads.

 

My tool of choice for CNC routing design is something called GCAM SE, which is a freeware pain in the butt, but even though I might have a minor penchant for masochism, that was not the main decision parameter.

 

Now if you say home brew CNC (or 3D printing), there is really one name that will pop up, and that is Fusion 360 from Autodesk of AutoCAD fame... and the program is good, really good apparently. With people like this relocated swede doing a overgrown glockenspiel with it.

 

Another one is Easel, which doesent quite have the hype ... but the one thing I dont like about either, is they are web-based.

 

Another is the end user license agreements, which in fusion's case mean at least in my interpretation that you were open to Autodesks auditing any time and if they found out you were making say ... bootleg playboy bunny key-chains with it... they might just pay you a friendly afternoon visit.

 

But probably not... but it is an option.

 

anyways, back to the simplicity at hand...

 

GCAM is as simple as they get, and it sort of assumes you know what you are doing, which as you know is a challenge for me occasionally... the doing, not the knowing.

 

wood171

 

It has quirks, and using it to your own best preference will require fiddling.. in my case I've found it best to work out in my head what the tooling will take away and build in the offset manually. It takes time, but the more you do it the more it starts to be a basic instinct of sorts ... and hopefully not in the murderous way.

 

Again, this is something that Fusion360 would do automatically if you remember to tell it what you are using as a end bit... but we are not using Fusion.

 

So now we have our program, and we have our raw work piece, and nothing for it to screw it in place onto a piece of sacrificial base material which in turn is rigidly fastened to the moving base plate of the machine.

 

wood163

 

Load up the program to the controller software, press cycle start, and the the agony can begin:

 

wood164

 

Three hours later we get this:

 

wood165

 

On first glance it looks good ... too good actually for the purpose of the following info, but thats because I've put the broken off pieces from where they had snapped off.

 

This is where things get a bit technical... the reason being I was an idiot and forgot a few basic rules, namely once you cut off a piece from all corners it will come off. If given a suitable opportunity the now loose piece can lodge itself in between your routing head and remaining work piece, which happened here... causing the machine to skip a few times, and losing its zero-point (the point on the work piece where all given numbers relate to when you first begin).

 

I did manage to re-acquire the zero-point relative well but it still was off enough to warrant scrapping the piece.

 

Another was the two small tangs in the middle circle snapping off which I could salvage with superglue... but as I mentioned, there were other issues.

 

Again Fusion360 has built in features for this ... but were not using fusion...

 

The solution is to add support tabs between the routing lines so if you know something might come off the tabs will assist in keeping said something in place, but will require manual removal later...

 

The program also left a bit to be desired as to how the machine went about its business, so some manual tweaking was needed... so it was out with the notepads.

 

wood168

 

Using just the first routing layer as a base I used the control visualizer to check my progress as I went through the motions.

 

The first major alteration was making the first few cut passes very shallow, so the machine could first make a trench for itself that it could follow more easy when the cuts became more deep... too heavy at first and the cutting head would just wobble about due to the structure of the machine... and slightly cheap linear bearings that don't know are they fine or not.

 

The second tweak was to tweak the routes so the machine would track back what it had done instead of going back to the start of the cut to go deeper which should save some time.

 

and thirdly the addition of the support tabs..

 

wood169

 

So nothing to it but to make more raw material, so I threw the planer several scrap pieces of 18mm glue board I had, and 100 liters of saw dust later I had a nice selection of 4mm stock to pick from...

 

So in theory we were ready to try again... and the rewards were near instant failure.

 

wood166

 

Its worth mentioning that CNC machines do not have a sense of humor.

They require precise instructions.

 

So if you give them crap, they will give you crap back. (Look up CNC fails on youtube to see some ultraviolence in this field...)

 

The upper left one was a result of spelling G01 as GO1 so the machine did not lift itself enough for the next cut and plowed into the work piece quite thoroughly and right into the sacrificial wood underneath.... I actually thought I had binned the machine by this point, but the flex in the rigging actually came to use here.

 

Another favorite is decimal dividers ... here we by default use a comma ... and while the control visualization will understand a comma as a divider the actual machine will not, and skip the faulty command, with possibly amusing results.

 

The one in the middle is a weird one, since its a behaviour quirk of the machine, and maybe not a programming error by itself.

 

Long story short to make back handed radius turns the machine wants to see some kind of X/Y movement before it ... otherwise it just skips the command and.. you guessed it, goes through your work where you really didn't want it to.

 

on the right you can see the first run where I had regained the zero point, but as a result the machine was cutting too deep the first few passes and as a result wobbled about quite badly.

 

This is called debugging. This is why in actual factories they do simulated runs, and foam-core runs so if something does go wrong, you have the ability to adjust before doing your actual work piece.

 

Time for the last few debugs and tweaks.. namely adding some more height to the support tabs...

 

wood170

 

So on the fourth time ... exactly 2 hours 44 minutes and 18 seconds from pressing cycle start...

 

wood167

 

we only had one bug but overall a very presentable piece that needs a bit of cleaning up around the edges.

 

I still have a slightly more coarse routing bit I could try to see if the cut would be neater, but I am pretty satisfied with the result so far.

 

But I think we can continue from here, since making the frame around this should be about the same as what I did the with the bar stool's awhile back, but I still need to design and route the ornamental pieces for beneath the seat as well.

 

One could argue that laser cutting could have been an option, but the scorched brown sides of the cuts were not a visual effect I was really after... also a big enough laser cutter that uses actual eye safety instead of impromptu sunglasses is much more expensive.

 

Part 2 - Let the machine do the work

 

Continuing onward with the project there was still a few design elements to be made.

 

This time the work was a bit more simple, but because it had been months since I had done the previous big design it took a bit of remembering about how to coax the free cam software to deliver what I wanted.

 

But in the end I got there with the patterns for the seat base:

 

wood172

 

These were then transferred onto the 4mm board stock, with the usual unintentional practice efforts...

 

wood173

 

So now we have our ornamental things done, so now its more or less off towards the meat of the project, namely the frame pieces themselves.

 

On this front I've had some practice in the past, with my little bar chair project for my defunct ahnenerve diodrama, so I decided to use the same basic structure from there. Namely 12x12mm main beams, and 10x10mm connecting pieces.

 

This time however instead of trying to chop my fingers off with the table saw to make these, I took the spruce paneling and only rough cut it to square sticks, and then I thinned them to the desired dimensions on the planer, and the results were quite ... profound compared to the last time I was at this...

 

wood174

 

So after this I just had to chop these up to the desired lengths:

 

wood175

 

The parts inventory consists of:

2 back legs

2 front legs

4 long side connecting pieces

and 6 short side connecting pieces -- six because the rear motif requires its own end-caps

 

And with these cut, I have the option of showing off the rough idea:

 

wood176

 

So now we have a bag of bits, but the challenge is getting these bits together. Now I wont go into detail about the joining procedure since I've gone through that in the previous high chair thing, but since one thing I do remember about that time, was the annoying nature of having to measure and drill the holes more or less precisely, and despite my best efforts I ended up having to fill a few resulted gaps.

 

This time since I do have a machine that can repeat things as many times as I wanted, I decided to let it do this tedious bit...

 

So I had it drill itself a series of holes for some locking pins that would then cradle the sections of the frame for drilling and slotting operations.

 

wood177

 

Actually I had to do it with a machine, since I couldnt just drill a hole for the motifs, because they are about 3x4 mm square sections, so this worked out quite nicely in terms of fitting:

 

wood178

 

wood179

 

The only remaining annoyance is that I still need to manually drill the holes at the ends of the connecting pieces, but some bother had to remain I guess... but still roughing up the pieces together without any real attention to keep the more pretty sides to the camera it looks like we might actually get this done some day...

 

wood180

 

Part 2.5 - Let the machine do the work ... eventually.

(AKA: The last pieces of the puzzle)

So after about a year of different obstacles I finally managed to get enough effort together to machine the last missing pieces of this puzzle.

So why did this take so long?

Motivation, thats why.

Nothing kills it better than seeing someone else have a ready made thing in the background of their photographs to re-evaluate your own efforts with absolute scrutiny.

and when after some internal debate and getting fed up looking at the pile of unfinished bits, I finally get enough willpower to try and finally put it all together... except my little cnc engraver thing decided it had had enough... and required a logic-board swap.

So after finding and sourcing the required arduino bits to ditch the slightly "iffish" chinese modified electronics, and tuning it up ... tuning it up some more ... and then finally having to go through the drive gears, it was ready to tackle the final bits I needed.

wood215

wood216

 

The final real hurdle in setting up the routing was to make it clear to myself that I didnt need to be within a thousandths of an inch with the two groups of holes, Since the longer legs had to be done in two different operations due to their length.

So with this, I could stress less about the positioning of them in the back legs, where the big motif piece goes into. But I did make the lower frame piece's holes oval so there is some play when it comes to time to try and put it together.

Amusingly when I left off, I basicly had enough raw pieces to make up to three of these, but due to some hick-ups and whatnot stuff getting misplaced and re-purposed, I barely had enough for the one...

 

Part 3 - Fitting, finishing, and hiding crimes...

 

So now we had more or less finished pieces to start gluing them together... which was easier said than done again, since before we could get to the magical stick together substance there was a little bit of cleaning up to do. mainly with the main motif, with its rather .. hairy disposition.

wood217

So nothing much for it but to stick a piece of 120 grit sandpaper on a piece of 1mm shim strip and go to town. I also used some manicure sanding sticks for this, but the flexible but eventually stout metal actually came in handy in some more tight places.
 

The rear motif was a case of deciding, enough was enough, since there was some minor issues where the CNC engraver had stepped out of line a bit, but nothing really jarringly obvious.

wood219

About a mornings worth of fiddling, everything was ready to start gluing together.

wood218

I didnt take any in between photo's since putting the main pieces together was a all hands on affair, and the glue was drying the whole time, so you'll just have to make do with the first glued together pieces:

wood220

So that was about 32 different spots that had to be glued together for this thing...

The remainder wasn't as daunting, but what needed a bit of effort was the cross brace I wanted underneath the seat cushion to keep this thing more or less in check.

This time its been slotted on both pieces, and then pinned together with two matches so it doesn't flex.

wood221

The final glue phase was putting the back and front together with the stuff that was meant in between, and besides a minor need to get clever with my clamps to get the thing straight it was all together:

wood222

Some more fiddling was in store, namely evening out all the slightly jarring edges, and evening out the intended 1mm steps of the inner pieces versus the main legs.

for some added No Guts No Glory spice, I took out a cheap wood engraving kit I had gotten, and decided to add some fake line details to the decorative pieces, since their meant to be separate pieces and not a big slab of wood, as they are now...

wood223

The flashbombing doesn't help to see them, but they are there...

I didn't go full Banzai with the main motif, and only added the horizontal lines and outer ring lines. also on the edges I did the horizontal lines, though in hind-sight I probably should have done a few of them in the other direction, but wood really doesn't like to be carved across the grain so I wasn't going to push my luck.

wood224

Now its starting to look like something:

wood225

A quick test fit with me elf who was still hanging around the photo-room, and it looks promising, and as expected the main motif is going to be well hidden by any full sized DD that sits on it... oh well. The inspiration for this was also like this so there is that.

wood226

The cushion is going to be a bit more thicc than the one I borrowed from my bar-chairs for the quick mock up, but that will come after surfrace treatments, which are going to require some experimentation to see that the motif doesn't start to misbehave with the lacquer.

 

Part 4: Finishing touches... Sort of.

So finally its time to get this thing colored, and upholstered.

For the stains I went with my usual supply of cheap brand Mahogany and Beech stains, since I wanted a bit of a two-tone effect.

Unfortunately I watered the Mahogany one a bit too much, and the difference isnt as pronounced as I would have liked, but still its there, once you look for it.

wood227

Letting it dry over night, it was time for the varnish, and I gave it two coats so far. I might add a third one later on, if I think I need the extra glossies, but I was pretty happy with just the two coats:

wood228

 

For the seat cushion I decided to utilize some stuff that I got as a freebie from the fabric shop I visited during the summer, namely a strip of faux leather, and some 20mm thick foam padding.

wood229

I measured the seat base, and as the original inspiration called for a slightly bigger cushion than the frame, I added a few millimeters all around for the box-pattern.

Unfortunately even though the material was a bit too thick for this sort of thing, it still puffed outwards with the stuffing, so might have to redo it someday.

wood230

 

And after plopping it on the frame with some hot-snot, it also came apparent that the white color might be a bit too contrasting for the chair. So that also stacked up for a re-do some day.

wood231

 

But in anycase, I am now calling this over year long project done. So one thing off my to-do list.

 

Thanks for looking.

Edited October 31, 2019 by finnleo

This is super cool! It's interesting to see how these types of wood pieces are made. And the design looks good as well. I look forward to seeing the progress.

Thank you very many.

 

There will be progress, but its going to be slow progress unfortunately.

 

I'm back to working two jobs that take up most of my week day's, and even though I was planning on designing the routes for the decorations underneath the seat base last weekend, cleaning up my living room took top choice with the remaining energy I had.

 

Its also a practice in writing for a change, so the lengthy nitty witty stuff is because of that... but I thought it would be informative if someone is interested in one of these machines, about how easy/hard they cant be to handle.

Wow that is amazing! it's very interesting to see the complexity of work like this. Thank you for sharing I can't wait to it finished.

thank you very many.

 

Its still going to take a bit of time, although I did manage to make some progress this weekend, but not as much as I wanted ... but that's what I get for piling up too many things and some tiredness to the weekend mix... but I'll compile what little into a minor update soon...

You need to start selling stuff there's a Serious lack of 1/3 furniture, and the stuff you make is amazing!

heh... well, it would be nice... but time and economics are an issue.

 

No spare time at the moment (though that could change in the following months, but thats its own story), and trying to set a cost where you get a reasonable return for your effort and raw materials, when yja is occasionally stocked with 5-10$ simple sofas or such...

 

its evil I say ... eeeevil.

A small update to the first post, after a rock-and-roll jump-start of this pile of bits after a years hiatus.

Wow, the amount of work put into this piece is spectacular! I love the design on the chair back, it's so intricate! 😊

Thanks.

And again a update to the first post in terms of actual progress.

Impressive!