Stephen

May 302020
 

It took another three weeks to finish off tweaking the design of the scorching machine, with many iterative changes as we test and adapt the machine (As usual things take much longer than you expect, but I have enjoyed the change of work).
We started by moving the machine into the house (after cleaning up a bit), we then made a extension to the input side to guide and support the plank entering the machine.

Input-rollers-on-the-machine

Input-rollers-on-the-machine


We did some test runs, and we can vary the amount of “burn” from a black highly charred finished to a gentle pale scorching. We have selected 5 levels of speed, all the way from a very slow 10mm per second (very highly burnt) to a much quicker 40mm per second to produce the pale finish. We also have the option of no scorching at all and this being our sixth level.
Our-six-scorch-levels

Our-six-scorch-levels



We put up a block of different burnt timbers on our house wall in the full sunshine to examine it and decide on what patterns we might like on different walls, taking into account their compass orientations, the windows sills and headers etc.
First-set-of-graduated-boards-on-a-wall

First-set-of-graduated-boards-on-a-wall

We-decided-to-flip-the-bottom-rows-

We-decided-to-flip-the-bottom-rows-



The next job was to build a spreadsheet of all our walls (fifteen of them), with their compass orientation, whether it is under a porch, then divide the wall up into three vertical sections (under the windows, beside the window itself and lastly above) to allocate what level of scorching is desired and then find solutions to the pattern of plank widths to fit into each section. This resulted in a table of quantity’s vs plank width and scorch level.

We started to process the wide planks first (they are nearest) and we had lightly scorched (#1) some planks when a plank got stuck. Shaun was running the machine whilst Stephen was wire brushing the finished planks so We did not notice for a while until Stephen looked over and saw the smoke poring out! We shut off the burners and when we examined the machine we found one of the motors had sheared it’s flexible connector…

Drive-adapter-sheared-off

Drive-adapter-sheared-off


We repaired the machine and carried on but a few planks later we had a jam again (the plank sometime wanders off or bends), we caught this one earlier and no damage was done. At this point we decided (being nerds and all) that we would add sensors to the machine to detect any stall and sound an alarm. In hind sight we should just have carried on whilst keeping a good eye on the machine as getting sensors to work reliably took 2 weeks (but that’s hindsight for you)!
The machine needed to know 3 things – When the plank enters, when it leaves and if its moving. The first two should be simple just put a switch near the plank to be pressed when the plank is in. The movement detector devised initially was a metal wheel on a shaft which had a sensor measuring the shaft rotation, the shaft being mounted on a pivot which kept the wheel in contact with the plank.
Input-plank-sensor-switch

Input-plank-sensor-switch

Rolling-encoder-mechanism

Rolling-encoder-mechanism



So having built the sensors and fitted them it took a day to write the software to run the machine and start testing. The metal wheel did not run very consistently on the plank even after filing some ‘teeth’ on on it so after a while a new system was devised to put encoder disk and sensors on the pinch rollers which force the plank against the drive rollers (an all together better way but you don’t always think of the best way first). This was fabricated, installed and worked reliably.
Encoder-disk-and-sensor-on-a-pinch-roller

Encoder-disk-and-sensor-on-a-pinch-roller

Protecting-the-plastic-encoder-from-the-heat

Protecting-the-plastic-encoder-from-the-heat



The next problem was that when we turned on the burners the output switch got too hot! Again a different system was needed, we needed to get the switch away from the heat so various pivoted arms where fabricated, tried and modified. But after a long time trying to get them to work reliably we gave up on this idea as well.
Wheel-lever-output-switch

Wheel-lever-output-switch


A non contact sensor was needed and a light sensor placed just below the plank would give different readings to show when the plank was present or not. This was fabricated and fitted and worked! We finally had a working machine and just a bit of fine tuning of the software was done. On the last Saturday a revelation was had! in that there had always been a lever to activate a switch built in to the machine from the beginning. The output flame guard was hinged on a rod which come out the side of the machine and this rod was turned when the flame guard was pushed up by the plank. This rod happened to have a bent end which was ideal for pushing a switch so we fitted a switch so now we have two output sensors which work…
Optical-output-plank-sensor

Optical-output-plank-sensor

Output-switch-using-the-flame-flap

Output-switch-using-the-flame-flap



So we finally have a machine for processing the planks, which will probably only take a week!
Final-Control-board-and-wiring

Final-Control-board-and-wiring

May 102020
 

You may remember that we have decided to lightly scorch all of our larch cladding to give the walls more texture both visual and physical.
We are using an ancient Japanese technique called “Shou-sugi-ban Yakisugi”, although we are not burning the wood as deeply.
As we have over 2300 m (7500 ft) of timber to scorch we decided to make a machine to automate the process (and of course because we’re nerds). We made a start on the machine back in March (Article), but we now have bought the motors and electronics to carry on.
So over the last two weeks Stephen has been working on this whilst Shaun does other things around the house.
First part of the job was to build the computerised controls for the motors so we can select the burn rate and pattern. This was completed by the end of the first week after lots of programming and learning about motor control feedback.

Control-Board-and-Motors-Day-1

Control-Board-and-Motors-Day-1


On Monday of the second week the motors were installed to the machine and testing of wood through the machine revealed some problems… the wood was moving a bit irregularly and had difficulties at slow speed. Lubricating all the rollers did not improve things enough and it was concluded we needed bigger motors! On Tuesday we remembered that we had some bigger motors bought for use in a giant CNC machine (which will be built eventually!).
Comparing-Motor-sizes

Comparing-Motor-sizes


So it was back to the electronics bench to rewire and reprogram the controller for the quite different motors. By the end of the next day everything was mounted and working well.
Control-board-and-larger-motors-mounted-to-machine-Day-3

Control-board-and-larger-motors-mounted-to-machine-Day-3


The next step was to improve the way the torches were mounted as we needed more control of the flames, so we made two rails with T slots in them then we clamped the torches by the tubes leading to the burners. The torches are mounted to pairs facing in opposite directions allowing maximum adjustments.
Torch-mounting-system

Torch-mounting-system

Gas-burners

Gas-burners



As expected the flames started to burn the plywood sides of the machine, so we made protective barriers comprising glass fibre insulation wrapped in foil behind steel plates, and added hinged steel plates across the wood to protect the drive rollers.
Burning-machine-Day-6-Inside-

Burning-machine-Day-6-Inside-


Then we made a stand for the machine to make it stable and allow the use of roller stands to support the long planks of wood.
Burning-machine-Day-6

Burning-machine-Day-6


It was time to burn the first plank! This worked quite well so we needed to wire brush the loose char off. So of course we needed another machine …
We bought an adaptor kit for a angle grinder but found it did not fit any of our angle grinders so we had to adapt it! So a quick bit of drilling and welding later we had a powered wire brush drum.
Drum-adapter-kit

Drum-adapter-kit

Gaint-wire-wheel-to-clean-burned-planks-1

Gaint-wire-wheel-to-clean-burned-planks-1

Gaint-wire-wheel-to-clean-burned-planks-2

Gaint-wire-wheel-to-clean-burned-planks-2



So we brushed the plank off and had a quite nice result.
First-burned-plank

First-burned-plank


Next week we will complete the machine with some extra rollers and guide to make it easier to use.

Oct 262018
 

We continued to apply the Resin and Fibreglass layers to the Filter Parts. We completed all of the insides with their two layers as well as the two layers on the insides of the lids and side panel. The outsides of the filter will have a single layer of resin and fibre-glass as they will not be submersed permanently in water so need less protection. The outside of the swimming lane return channel was coated last thing today.

Fibre-glassing the Filter Day 2

Filter-Fibreglassing-Day-2-1

Fibre-glassing the Filter Day 2

Filter-Fibreglassing-Day-2-2

Fibre-glassing the Filter Day 2

Filter-Fibreglassing-Day-2-3


Jul 112018
 

In better weather we completed the boarding of Sections D & E. We are getting better at cutting boards the correct size first time.

Roof-boarding-Day-7-D-E-Completed

Roof-boarding-Day-7-Inside-of-D

Roof-boarding-Day-7-Inside-of-E



We have now covered 140 square meters of roof which is 38% of the total in 7 days so we are on schedule to finish in about 20 days…

Stephen fails to do a ‘Shaun wave’ @ 4:37

Jun 262018
 

We had a delivery from UK-Mail today and the driver hit our fence post and snapped it off! He appeared to hit the post and break it off as well as flattening a few road cones placed by the fence to warn drivers that it was there! We have contacted UK Mail and will wait to see what they say about it.

Fence-post-snapped-off-by-delivery-driver

Fence-post-snapped-off-by-delivery-driver

Fence-torn by delivery driver

Fence-torn by delivery driver


Feb 272018
 

This morning, our 75 sheets of pretty plywood arrived. The lorry managed to get down our Loke, and we placed our flat bed trolley alongside. We then pulled off the plywood, sheet by sheet. This allowed the delivery bloke to get off to his next port of call.

Plywood-unloaded-onto-the-trolley

Plywood-unloaded-onto-the-trolley

We then moved half the pile by taking 3 sheets at a time and walked about 15 metres to our new storage tent. Eventually, we were able to move the trolley itself, with much slipping and sliding on the snow, and got the remaining sheets unloaded.

Plywood-stored-away

Plywood-stored-away

This plywood is much nicer than the last lot (well we bought it as pretty)

Plywood-has-a-decent-numer-of-plys

Plywood-has-a-decent-numer-of-plys

Nice-grain-on-Plywood

Nice-grain-on-Plywood


After lunch, we resumed in the lovely and warm workshop and created five more rafters.

N-O-Rafters-made-so-far

N-O-Rafters-made-so-far

That’s was enough for Shaun’s back so tomorrow we will finish off the last two rafters and get ready for doing the “K” section by measuring the final set of rafters! Hurray!

Oct 172017
 

This morning, we tackled the next rafter, the A4, to slice the bevel on the end of it, analysing the results. It was still inclining to curve outwards during the sawing and we surmised that the saw blade was ever so slightly off line. So we attached a thin 1mm plastic spacer on the sled at the bottom left corner to try and offset this skewness. And yes it did seems to do the trick. We then did the next rafter, A5, and repeated the bevel slice and this time we got a much better cut first time. We are a little bit more confident of the machine now. The proof came when we took these two rafters outside to get them up and installed. We decided to try and man handle them up ourselves as we had a great advantage with these rafters as the wall ends fits like a glove and actually anchors the whole rafter into place. We only needed to hold the top end to relieve the strain and make sure we don’t accidentally break the Bird’s mouth joints.

Roof-Section-A-Completed

Roof-Section-A-Completed

We fixed them the same by screwing a 8mm coach screws, one in the bottom and one in the top of each rafter and also glued of course too! It went very well and flat against the PA diagonal beam which further enhanced our confidence that the Bevel Slicing Machine is performing better.

Then after lunch, we went around measuring the other similar sections of the roof, namely the “M”, “I” and “E” sections by moving the tall scaffolding tower around, as all are having the same 32degrees slopes. During this task, we were concerned that we were getting spurious misleading measurements along the top flange distances because our template was not going flat against the Kerb of the Skylight and this was proven when we placed the spirit level on the template and the position of the Kerb was slightly proud of the steel webbing underneath. This caused us to reconsider and switch over to measuring the bottom flange distances instead as we realised that this aspect of the rafter element is much more “stable” and the main structural load points is on the bottom flange at the top and the shape of the wall end (the top-plate and legs) are much more accurate to where the Bird’s Mouth goes. So we took down a series of distances to then enter into our spreadsheet, ready for preparing the elements of the rafters in our workshop tomorrow. There are 10 more long rafters to make and a heap of the bevelled rafters to fill in those corners of the roof sections.

Aug 262017
 

The metal splice plates for the O ridge have been welded together.  More care was taken to weld them in short stages and swapping from one plate to the other resulting in almost no warping of the plate.

Ridge-O-Splice-plates-welded

Ridge-O-Splice-plates-welded

Now we have to make the wooden parts of these beams…

Aug 242017
 

The parts of the metal work for the K ridge were welded together, after they were finished it was noticed that the big metal plate had been distorted by the heat of the welding. We will have to investigate how to reduce this for the O ridge plates!

Ridge-K-Splice-plate-welded

Ridge-K-Splice-plate-welded

Tomorrow we will make the parts for the O ridge and maybe start welding them.

 

Aug 232017
 

After finishing the construction drawings, fabrication of the metal work started with drilling 64 holes in the flange reinforcement strips.

Flange-reinforcement-drilled

Flange-reinforcement-drilled

Then the two side plates were cut out and drilled for bolts.

Splice-plates-cut-and-drilled

Splice-plates-cut-and-drilled

Tomorrow the sockets for the adjoining  rafters will be welded on.