Sunday, November 29, 2015

Looking Around Pittsburgh (II)

I'd been meaning to share some photos of interesting houses I spotted in and around Pittsburgh. Not sure where all the time went, but better late than never. I found these houses more interesting than most, and thought readers here might find the same.

2nd empire homes were sprinkled in with some frequency. Love the neatly tiled Mansard roof:


Pseudo Tudor to be sure, but at least the simulated bracing was done in a realistic manner - and that's a rare thing to see:


I can't help but think that putting that short gable in between the main flanking gable roofs will tend to encourage roof leaking at some point. I wonder how often the owners make use of the (apparent) balcony above the main entrance?

I am often fascinated to see one material pressed into the service of simulating another - here, very tidily indeed:


The style of these dormers remind me of German designs I've seen in Rundgauben und Fledermausgauben - I wonder what impression my German readers might take?:


It's kind of amazing to take a totally inflexible material like slate and produce an effect like a thatched roof:


I wonder how weather-tight that roof has been over the years? Seems like it would be a challenge to make it durably water tight.

This portion of the house, er, raised more than a few eyebrows:


Some lovely timber detailing on this house:


Those knee braces lend an Art Nouveau touch it seems to me:


The diagonal mesh of beam work at the front is delightful. The low pitch and use of multiple horizontal elements there is an interesting counterpoint to the height of the building and the steeply-pitched main dormers which dominate otherwise.

Note the embossed copper panels fitted into the interstitial areas:


A view around the corner of the same house. The chimney crickets are about the same size as the adjacent roof dormer:


This house had some unusual grilles just below the eave:


I presume it is the attic space being provided with a modicum of diffuse light - I imagine that the effect seen from within would be very nice:


Here's a restrained Queen Anne style building, with an interesting ribbed chimney and a cylindrical tower seemingly extruded out at the side:


Some tidy slate work:


Another view - can't recall seeing too many places with this sort of golden decorative band below the eave - very nice indeed:


I like the 'dialogue' between rounded and flat surfaces, square- and round-end windows, eave brackets in some location and not others, the flared 'skirt' midway up the wall - a lot of interplay and I think it comes out very well:


Looks like a well made house.

Hope you enjoyed the tour as much as I did. Thanks for visiting the Carpentry Way.

Wednesday, November 25, 2015

A Ming-Inspired Cabinet (11)

Well, I've been getting a lot more familiar with the Zimmermann in the past couple of days, and it has been a pleasant learning experience for the most part.

Here's my Mark II (or is it III?) fixture for milling long stock:


Clean and flat, with no tear out - is it too much to ask? Not any more:


Here are the four inverted 'T' section beams - crisp arrises and square adjacent faces, and nice and straight - what's not to like?:


Today the three QS wide bubinga boards showed up via Fedex Freight and interrupted the milling:


Spreading these three boards out revealed a fairly awesome scene:


New to me altogether, in this material, were knots:


I'll be able to cut out around most of them, and the few that remain behind are not too critical as they will be in the upper and lower carcase boards, which reveal only their edges in the normal viewing positions.

Can't get much closer to 'perfect' stock at 23" wide, almost completely quartersawn:


I did a bunch of cross-cutting to obtain the remaining boards for the upper cabinet carcase, and then it was back to the milling:


The cutter has a positive rake, but it the cutter engagement is still quite a bit steeper than my 60˚ Funahiro plane. What you get, effectively, is rotary scraping with a HSS helical milling tool:


Beats grinding any day.

Gang-cutting was an option with the stock which will become the pillow blocks for the lower stand:


The mill made it relatively straightforward to straighten the stock, and I was able to get close to target dimensions. The pillow block width was set for 2.700":


Before surfacing with the FZ-5V, I had this sort of tear out on the faces after careful machine planing - the closest I could get with planing left tears propagating up to 1mm into the surface:


The rotary scraping however left behind a clean flat and tear out -free surface, which was indeed happy days for me:


Target dimension here was 1.1250" - - if I invested in a bit more metal, I could improve the fixturing yet, and I think achieving +/- of 0.001" to target dimensions is certainly within the realm of possibility:


What's this? More bacon slices crop up under the super surfacer:


Clean surfaces, free of tear out, arrises crisp, and parts are right at the desired dimension. Am I dreaming? I could get used to this:


Another view:


Round 216 of milling begins, with the stretchers (nuki) clamped into the fixture:


Getting the parts to dimension via the milling machine certainly isn't the fastest process compared to other approaches, but it is a rather certain one. I am so happy to get things down to size without tearing my hair out over tear out. More work remains on the stretchers, and then it will be time to form the legs into parallelogram sections. Feels very achievable and I look forward to seeing how these slight improvements in dimensional accuracy and parts straightness piece to piece translates into ease of assembly when the compound-splayed base comes together. I think the small differences will add up to something significant.

Happy Thanksgiving to my American friends, though in truth my Canadian friends are wondering what took you all so long :^) Thanksgiving is in October, after all. On to post 12.

Friday, November 20, 2015

A Ming-Inspired Cabinet (10)

This is the third project working with material from some big slabs of curly bubinga. Each time I delve back into this material I am brought face to face with one of the main challenges: the propensity of the wood to tear-out during processing.

Even with sharp knives, and four of them, in both jointer and planer, I have found I can only reliably dimension down to a certain point, beyond which I proceed at my peril. My shoebox Makita planer, given the high rpm, sharp knives and a shorter distance between the cutter head and the infeed/outfeed rollers, can obtain a slightly better finish, tighter to desired final dimension, than can the bigger planer, but I am still limited to dimensioning down to about 1/8" overside in a given dimension. Tear out, when it occurs, can pull fibers from nearly 1/16" into a surface, so if I want to be confident of obtaining a clean surface, I have to allow for that in the planing.

I've learned about this matter the hard way in the past, having to take some parts below intended dimension so as to produce a clean surface afterwards. I wanted to avoid that outcome this time around, and I have found that the tool that helps get me there is the FZ-5V milling machine with helical shell mill. Even longer pieces can be adequately supported and worked from each end to produce a flat surface.


Another view:
















A close -up shows the fine slices that can be removed, a layer at a time until a clean surface is left, with one round or two of final planing or super-surfacing remaining:


Above is a skimming pass taken closer to the dimension, though at other times I will mill off 1/16" or more. Generally speaking, it's not at all the type of milling, the heavier stock hogging, that one can undertake on the planer or the shaper. After all, we're talking about a 4 hp spindle, not an 11 hp one.

With these particular sticks, their final form was to be an inverted 'T' section, so in an effort to guard against unwarranted wood movement after the waste strips were ripped from the blanks to produce the 'T', I left the stock oversize in every critical dimension prior to doing the rip cuts:


I'll let the stock settle for a day or two before milling to final shape. The excess thickness in every dimension should allow me to realize the ideal stick hidden within - all being well also at dimension and straight and square. There is the ideal and then there is the reality. Just trying to shorten the gap between those two states.

Here are the short side pieces (top), with the pillow block stock (bottom) also being worked, step by step, down to the line:


It's certainly not a quick process to mill the stock in this manner, as the amounts taken per pass are rather modest, but it has allowed me to obtain clean surfaces without tear out, getting closer and closer to dimension, and that is really all that matters to me.

All for today- thanks for visiting! On to post 11.

Monday, November 16, 2015

A Ming-Inspired Cabinet (9)

While I'm letting most of the bubinga for this project sit for the next few days, and with more bubinga about to get shipped out to me I am holding off on any more rough stock breakdown for the time being.

I thought it would be worth exploring what sort of job the Zimmermann FZ-5V would do with its helical shell mill. A few people have been letting me know over the past couple of weeks that they are interested to see how I make use of the mill, so here goes 'part 1' in that regard.

I had planed the leg stock to a solid 1/4" over dimension, and could have taken them down further by planing them in the SCM S630 planer or the Makita shoebox planer. However, I've been finding that the planers are tending towards a little bit of tear out with some of the figured bubinga, even with fresh knives fitted, so taking the pieces right down to dimension with a planer seems decidedly unwise. I'll probably have to get the panels thickness sanded to remove the remaining 3/16" of material standing in the way of final dimension.

I thought therefore, that it would be worth a look-see with the milling machine, as I had been considering whether or not the Zimmermann would be a good way to process the leg shapes into their ultimate cross-section, which is rhomboidal. Getting these legs shapes done precisely is the most important part of cut out for the entire support stand, so I will do my utmost to get the results I want

This was mostly a test, and if it went awry there is plenty of material left on the legs to correct afterwards, not to mention that there is a spare leg to mill. The set up is pretty basic, with the bubinga leg placed across the vise and sandwiched between a couple of precision parallels to spread the clamping force:


The shell mill is helical with a positive rake:


Here's how this process looked, with a 0.3mm deep pass:



I realize that might not be the most exciting youtube clip ever, but I enjoyed watching the machine do the work. The machine takes shavings, and it can take them in thousandth of an inch increments. I did receive the cutter used along with the machine, and it seems sharp, however I am unsure as to how worn the edges on the cutter actually are, or what sort of materials it might have been used upon previously. As it is right now, it seems to cut well.

What I obtained from the shell mill was not what you would call a polished surface, like you can get off a hand plane or super surfacer, but it is very clean, with no significant traces left of the cutter and absolutely no tear out:


The lack of telltales from the cutter means the head is trammed in correctly, and the surface quality is better than what I would obtain out of my planer with 4-knife Tersa. The overarching idea here is to get the surfaces close to the mark with the mill, and finish them off with a hand plane and/or super surfacer. I'm glad to know now that is a fully-realizable plan.

The leg came out very square and within 0.1mm of dimension without me having to have been particularly fussy about getting to those results:


Whee!:


I enjoyed making some shavings in a whole new way today. That face mill worked very well I thought and should be the cat's meow for processing these leg sections into rhomboidal shapes, which will be the next step with these pieces once I have completed S4S with all the pieces.

That's it for installment 9 of what promises to be a lengthy build thread. Hope to see you again next time. On to post 10.

Saturday, November 14, 2015

A Ming-Inspired Cabinet (8)

While I started out with what was an enormous slab of bubinga, the sheer volume of wood does not correlate strictly with usability. I am preferentially using quartersawn or rift orientation throughout the cabinet(s), and that is where starting with a large bubinga slab afforded some options in grain selection. Once I had cut out the desired cants from the slab, I was left with a large remnant, namely the middle run of the slab. It is about 1/3 of the original width, and is a piece which is almost entirely flatsawn. A chunk like that is of far less general usefulness to me, though I do anticipate making use of it in this project. The flatsawn portion just does not convert in this case into wide pieces of stock however, just narrow and/or short pieces.

The two thick quartersawn cants I obtained from the slab were therefore precious, pretty much irreplaceable really, and it is from these two cants which I am intending to recover the panels for the doors on the cabinet. Those two doors are bifold in design, and therefore 4 panels are required. A 3" thick slab will re-saw into 4 pieces of about 0.75" in thickness, all being well. I say 'all being well', as this is a step carrying certain unknowns. I've had pieces of bubinga which, after having been re-sawn from a thicker cant, proceeded to become very unruly in their behavior and impossible to keep flat. It's a shame to take a board which is flat and perfectly usable, then re-saw it and end up with a bunch of scrap. It just can't be predicted what will happen, not with any certainty at least.

When re-sawing, stresses are released from the board. Some of these stresses are present as a result of the growing conditions the tree underwent, and some of these stresses may have been induced through a imperfect drying process. A piece of wood can be dead flat and look promising for re-saw, only to have boards go spriong! after cutting and you are then left with a whole bunch of nothing. Re-sawing is not a complete walk in the dark: the grain orientation of the board, along with its position in the slab, and therefore in the tree, are useful telltales as to how it might behave when sawn, and certain things can be anticipated.

In this case, having two quartersawn cants and needing to obtain 4 flat boards from each after re-sawing (given that I am making two cabinets in total) left me with no wiggle room whatsoever. No boards were permitted to move more than about 1/8" after cutting. It was verboten, and I made sure the cants were clear on this before undertaking any more work on them.

Well, in truth I did have an additional cant which was there as a back up for the panels, however I very much preferred avoiding having to cut into that spare piece since it had other vital uses on the project. I needed all the wood I can possibly convert out of the stock on hand, and as it is I had to drop another $1800 today to obtain 3 more wide quartersawn boards from Rare Earth Hardwoods in Missouri. These come on the heels of 3 long and wide bubinga boards purchased from the company last year, material which is also going into these cabinets.

These cabinets are positively going to devour material! I find it amazing how much can disappear into a piece.

In general, the safest re-saw cut is to divide a board exactly in half. This should distribute stress release equally to both pieces, and all being well, one will emerge after cutting with two flat boards. And, if the wood movement is severe after the initial division in half, then you will have a good idea of what lays ahead - i.e., more movement-  and can perhaps plan accordingly. That's a good time to head to the bar I suppose, or your wood therapist.

The initial slice and dice of the slab, detailed in the previous posting, was the first step in the process, and I let the recovered pieces sit a couple of days in my shop. No significant movement was noted, and so I proceeded to do an initial re-saw on each cant, dividing each into two boards about 1.5" thick. That also went successfully.

Today I decided, upon inspecting the re-sawn material and finding it flat, that I could tackle the next step, which was to slice each 1.5" thick piece into two boards. That process began with a pass or two on the jointer with each board:



I had meant to film the entire re-saw and the 'reveal' afterwards, however the camera ran out of battery charge.

The resulting boards, shown at the end of the above video, are worth another look:


Can you believe this stuff?! Nature is amazing!

The set of four sawn out of the other cant:


I was greatly relieved to have obtained full value from the re-sawing, and the material looks terrific. Think how it will look once there is a finish on there....

Currently, the re-sawn material is very flat and a full 1/4" over finished dimension, so I am feeling confident about having made it through this stage without incident. I was stressed but the boards were not. While in each set two of the boards appear to look really figured and two do not look quite as figured, this is simply an optical effect relating to camera position. If I shifted camera position a few inches left or right in each case, the figuring effect would shift to the other two boards in each set. It's fun to walk along and look at the boards and see the figure shimmering about as different portions of the grain reflect.

Onward and upward...thanks for visiting. Post 9 is next.

Tuesday, November 10, 2015

A Ming-Inspired Cabinet (7)

Last post in this thread took place back in September.

After acquiring some fabulous shedua stock for the drawers, I have been sidetracked with other projects. In the past couple of weeks, I have been able to turn my attention to this sideboard build, which will likely consume the next 4-6 months of shop time. Not just shop time, a fair amount of home time as well, as that is where I do my CAD work.

The past 10 days or so have been largely about working through the drawing with a fine-tooth comb, in preparation for rough cutting of stock. This is one of those initial stepping stones across the river I must traverse to realize a successful outcome in this design build process. After so many hours have already gone into a design, and having received the go-ahead from the client, one could be forgiven for being eager to get to work slicing up the material.

However, in this case, maybe more particularly than with other projects, I am dealing with material which is more or less irreplaceable. Take a look around for wide and thick slabs of bubinga and you aren't going to come up with much these days. The cessation of raw log exports from Cameroon, plus the economics of wood drying being what they are, have made the combination of wide and thick bubinga stock rather difficult to come by in what is otherwise a fairly readily available material.

I wanted to be totally certain about how I would slice and dice a slab like this:


About 600 lbs.  of material lying there, 50" wide, 9' long, and 3" thick. You only get one chance with cutting up a chunk like this.

While the design of the cabinet was largely resolved, as I worked on obtaining some take-offs of each part I began to have some second thoughts about one of the framing details.

Just for a refresher, here is that design as it stood in September:


The niggling detail that wasn't sitting right with me was the negative space created by the pillow block layer on the support stand. The pillow blocks act as spacers between a pair of beams, atop of which the cabinet carcase would be placed.

While I liked the aesthetics of that arrangement, and the practical consideration of the even air circulation around the beams that associated, there was a drawback. The drawback is that the pillow blocks are little more than spacers, and thus only modestly tie together the two beams in a structural sense. What you have there is essentially a horizontally oriented 'ladder'. And that ladder, when you get down to it, is not as optimally stiff as it might be be if an alternate structural arrangement were to be considered. A ladder is poor when it comes to shear loading, as would be experienced with weight bearing downwards.

Given that this is a sideboard, one with reasonable capaciousness in the volume of stuff that can be accommodated, it is not beyond the realm of consideration to see that the cabinet could become heavily loaded at some point in time. Added to this, there is time itself, a constant, as is gravity. What I fear is that over time the weight of the cabinet and its contents might tend towards deforming the middle of the cabinet slightly downward. I'm thinking the effects of this are the sort of thing that might start to accrue after 20, 30, or 40 years. Certainly, if you look at old furniture pieces in museums, one can find no shortage of examples where gravity and the proclivities of wood to move have made decidedly clear effects. And those effects are rarely if ever on the positive side.

The problem with some slight amount of deformation in the middle of the cabinet down the line is that it would have strong adverse knock-on effects in terms of the fit of the double bifold doors and the many drawers inside the cabinet. If the doors and drawers are to be fitted with reasonably close tolerances, then one also exposes the fitted parts to fitting problems if slight deformation in the surrounding casework were to occur down the line.

I concluded therefore that this area of the support stand framing was deserving of a second look.

While I had a certain attachment to the use of pillow blocks as a tie-in to the previously built side and coffee tables for the client, it was also obvious that from a normal standing position the pillow blocks were largely concealed from view. To get a look at the negative space framed by those pillow blocks would involve laying on the floor - after all, the support stand for this cabinet is all of 14" tall. In short then, the aesthetic aspect of the pillow blocks was a minor aspect.

From my understanding, for every increase in depth of a beam, stiffness is improves by a square of that increase. The current arrangement featured a pair of beams, one at 1.375" thick, and another at 1.125" thick, a total of 2.5" of material.

To wit:
For a beam with rectangular cross section, bending (flexural) strength is a function of the square (second power) of the depth (height). A similar, though more complicated, relation applies for I-beam shapes.
With all other factors the same, a rectangular beam with depth of 12 inches has a bending strength that is 4 times that of a beam with 6-inch depth.
Bending strength of the beam is dependent on the "section modulus".
From: http://www.structural101.com/Beam-Design---Basic.html

If you start looking at the most efficient arrangement of material to make a beam stiff, you end up at an 'I' -beam. So, with that ideal in mind, I reconfigured the lower beam into an inverted 'T' shape which tongues and grooves into the upper beam.

Here's the revised lower beam form:


This change more or less gives the lower beam a depth of 3", so as to make a total of 4.125" for total depth of the two beams connected together. The previous design gave me 2.5" and now there is 4.125", an increase of 65%. Without going into detailed calculations, I can expect, as a ballpark, an increase in stiffness with the new arrangement on the order of 2.72 times greater than previous.

Aesthetically, the change is slight, and, as noted above, most of the change will be out of view until one's eyes are below knee height:


The confluence of parts at the cabinet corner remains pleasing to my eye:


Feeling sufficiently confident about the design of the stand, and unlikely to make significant changes to the proportions of the cabinet carcase atop the stand, I felt I could start in on the cut out today of the parts comprising the lower stand.

The big slab was sufficiently heavy that I asked a couple of guys from the adjacent shop to help me lower it down to the floor. Once down, I marked some lines, snapped some ink down, and got busy with the larger Makita circular saw:


The board leaning against the posts is one of the wide Goncalo alves boards I am hoarding.

Able assistance to the cut out task was afforded by the larger re-saw:


Here are some of the rough-cut legs - keep in mind that I'm making two of these cabinets:


The rest lay piled up after cutting out - the stock for the front door panels, the stretchers, and what will be the inverted 'T' section lower beams:


What remained from the slab was still a chunk, still pretty heavy, but entirely flatsawn:


The standing piece to the left, also 10/4 stock, is a piece left over from a Ming-influenced table I built a few years back. That material will be sliced up further to become part of these two cabinets.

The sawing was sufficiently vigorous that the Makita 380mm saw began  to complain. Smoke is never a good sign. Stripping the machine down afterwards was straightforward however.

Best case scenario: it looks like I will need to replace the rear arbor bearing and the brushes:


Worst case scenario is that the motor will need rewinding, which is another way of saying that I will need to buy a new saw. Smoke generally means that the insulation around the windings has burned off.

The arbor windings don't look too bad, though there is some soot on the commutator surface:


Both brushes and bearing are fairly cheap and easily sourced, so hopefully this saw will be up and running in another week or so. Maybe that is just wishful thinking though. Fortunately it got me through some critical steps today, so I won't be needing it for a while. Those long rip cuts in a hard wood are tough on a saw, and I guess I must have pushed it a little too hard today.

All for now. Be prepared for a bunch more posts on this thread as the build gets underway over the next few months. On to the post 8 .