A Goban 19*19 - Part One

Now the prototype is out of the way it is time to start on a full board with (hopefully useful) process improvements. As before I will be describing in horrifically exhaustive detail the entire process so that hopefully other wood workers can copy the design and style to do their own (should they wish)

Plans

The plans were not an issue last time so I just generated new ones for the larger board. Two additional plans produced were a duel strip plan to help with the star point drilling and a half size strip to allow the horizontal cuts to be made.

the plan designed to assist with star point placement

Cut List

For this project I worked out a well resourced cut list.  There is a lot of spare wood because there are always things that go wrong and I don't want to have to fiddle about with trying to make something work when quantity will help oleviate many nasty issues.

I did spend some time considering wood choice this time as the prototype choice was too dark (well it wasn't at the start but it certainly ended that way). In the end I went for unsteamed Beach with either the caulk approach or American Black Walnut for the lines.

White Beech

9 boards of

18mm deep 140 wide and 500 long

3 boards of

18mm deep 100 wide and 600 long

Black Walnut

5mm deep 100 wide 420 long

The initial 'large' wood order

I have a lovely UK source of nicely )accurately) planed wood which at woodshopdirect.co.uk which allows me to do far less wood preparation and thus far faster wood work. It comes beautifully packaged as well so I do recommend them especially if (like me) you don't have the space for the larger tools required to prepare the larger 2ft wide plans common at most wood mills.

As you will not this is a lot of wood but my intention is to have 4-6 strips of the Piece I and 26-32 blanks of the Piece II. Piece I requires 1 and Piece 2 17 so it gives a lot of reserves.

Cutting the Pieces

I started this time cutting the strips. The length was not fixed at this point (for reasons covered in the next section) and I cut about 0.5 mm over size in case of table saw mistakes.  One difference here was that I marked the ends prior to cut so I knew which board it came from, which cut it was, which way the board was facing and what the blank was for. 

cutting the primary strips 

Initial Sizing

Sizing started with the depth. Since the wood was already at a decent 18mm due to the supplier I did not need to do a lot so I just skimmed off 0.4 mm with my thicknesser. This was done to remove any slight marks caused by the original planing and to ensure everything was the same size.  One important note when thicknessing Is that you are likely to get extra bite at the ends where the drum/blades first 'bite' and leave and you can also get curves. This was not an issue with the 9x9 as the pieces were small enough that the thicknesser base plate forced it to cut level. With a much longer strip the wood can curve and not get corrected.

Either way may method was to set a cut. Push through the piece, trying to keep the feed steady. I then re-adjusted the cut to the same point and pushed it through again. The reason for this is that all planers/thicknessers can loosen their cut if the cut made is to large or the wood is touch (without this you would get a lot more damaging 'locks' where the tool stops working. This means you can sometimes get a larger cut/sand at the start of the piece than at the end. Once the cut point is re-set you then get any discrepancy ironed out. Don't flip the piece when doing this as otherwise the extra wood is taken of the bottom which can cause weirdness in shape.

Having a longer piece than required helps with this as the ends will get lopped off so any bite back will be dealt with when sizing to length later.

I then thickenssed the sides to the correct widths.

Grading

I now had a lot of blanks so I decided to grade the pieces. For this you need a flat surface. Fortunately for me I had an old (larger) table saw that is now broken but that I use as a surface. This has a planed, flat, steel top that doesn't warp and is a good base point. I placed each blank on the top and looked for gaps and shape issues. Each piece then had a double grade indicating firstly flatness/accuracy of the top and bottom and then another for the sides. I used four grades.

A - Lovely and Flat

B - slight imperfections

C - larger imperfections

D - Unuseable imperfections

So a piece might be A/A for perfect all over or B/D for slight imperfections on top/bottom but unusable for the sides.  This mark was made on the bottom of the piece (which is important for the vertical slot cut) and I aimed to ensure that any piece with imperfections had any convex curve on top and not on the bottom. This is simply because a piece that is higher than the surrounding wood just needs the sticking out section planing/sanding whereas if you have a concave piece then everything else needs reducing to the level of the single piece.

the strips estimated and marked on straightness and source

I was pleased with the result. The key indicator was the second (as the top can be planed post board generation)

My sorted list was

Piece I (1 required)

1 A/A

2 B/A

1 B/C

1 D/C

Piece II (17 required)

8 A/A

2 A/B

2 B/A

4 B/B

3 C/A

1 A/C

4 C/B

2 C/C

1 D/A

1 D/B

2 D/C

 so I had mostly A-Bs and only  a couple of C/As needed to round out the mix. This also provided me with a small group containing 'D's that I could use for test cuts and preparatory work.

checking the straight edges for flatness

Long Cut

This was a simple enough matter of setting the fence to only engage 1mm of the blade at a depth of around 5.6mm and running some test cuts to check depth and width. Once happy I ran all the piece through (piece I's both sides) making certain I could 'see' the A/B etc rating before cutting (as that was the bottom and we want the top to be face down and being cut into)/

the table saw prepared for the edge cuts on distance and height

cutting the edge cuts

First Sizing

Now I had my vertical cuts I could then play around with which piece might go next to which piece and building up my ideal board. In practice I may lose one or two due to bad cuts but if all goes well these will be the main basis for my 19x19. If this goes well some of the remaining pieces can be shortened and form another 9x9.

edge cuts all done on the selected 'best' pieces

Horizontal Cuts

For these I came up with a new approach. I printed off some half size strips which I then cut and glooped (a rather nice gel that sticks but is relatively easy to remove) onto one of the D pieces with a good side (so the D was for top/bottom). 

using two of the less decent pieces for the cross cuts

I then switched the saw blade from the wood cutting blade to a slitting blade. This is less decent at ripping stuff but is very good at cutting small slots. The larger the cut and the harder it makes the saw work so it is very much a 'horses for courses' type situation.

saw blade comparison. the default is on the left and the slitting blade is on the right.

I could then cut the slots in my test piece and then measure to ensure no miscuts had made a length mismatch between slots.

For the rest of the pieces I carried out a free cut at one end on all pieces and then attached this cut to the test piece with an appropriate plank. I could now 'set' the test piece on the blade and then cut the actual repeatedly. The primary concern is if the blade slowly loosens the test pieces cut widths instituting a marginally increasing unsteady cut.

My first attempt at a guide failed with one of the gaps to large at 23mm and the next to that to small so after trying to work around it I decided the chance of a process error was high and tried with a second guide piece. Here I ensured the strip was close to the bottom so the set square line was a better indicator. This was much more reasonable with all lines around 22mm. Note that though perfection would be wonderful it is not required. What I am after is enough accuracy that the human eye cannot look at the board and say - "hmm that column is wider than that one". When you get down to discrepancies of 0.2mm or less then that is not a problem. 1mm? Yes the human eye can pick that up.

Once I had a working guide I cut into one of the spare pieces to check the process. Here I used a thin plank to align then masking taped the top to secure the two pieces. Once done I checked the gaps on the newly cut piece and then flipped to check them in the opposite direction for any nasty visible mis-alignments and all seemed well. After the second cut through the guide all the tearing at the edges has now gone making using the guide much easier.

marking the accuracy of the cross cuts

The next step was to repeat on two of the important pieces and then to check alignment with each other. All good so far.  By the time I had cut four pieces I noticed that my guide was, very slightly, starting to show bigger gaps for some of the lines. Therefore I switched to a new policy of using my initial four as templates for two pieces each. The key is to ensure that each piece has the initial cut and then two guide cuts - so when I run out of pieces to guide then I will just re-cut to ensure the same depth etc. As a slight advantage this will also clean out the vertical cuts quite nicely.

cross cuts done

Human error (and to some extent incremental blade errors) will creep in so I now had to re-check my piece order and (in some cases) replace preferred strips with some of the spares. I had to replace around three in the end but bar two to three slight visible line changes was happy with the result. If I was after perfection, and to be frank in woodwork perfection is a journey you never get to but want to progress towards, then I would consider increasing the cutting list and having an extra ten or so strips which would have avoided this issue.

trying to sort out the optimal piece arrangement

Once I had my optimal arrangement I then had to trim some of the ends to make sure they were of matching size prior to gluing - after this I have no sanders that can cope with the lengths involved.

gluing - with central support to try and increase flatness

Gluing was organised in several stages. The pieces requiring star holes were done first and only after those holes are drilled will the full board be glued together.

side shot of initial glue showing end markings of piece order markings

Next up will be the generation of the star pieces and their drilling followed by completing gluing, surrounding structure and oiling. Phew...