Lucky -?
Trial and error the arcs vs. thickness, but I'd say at 3mm thick most edge band will pull in nicely on gentle radii given good end-point clamping,?1.5mm might start feeling flimsy. Keep shelf?vertical, banding laying?across top, clamp one end secure and pull it down by hand to the other end then clamp final?end. If you have a frame press or assembly table with stop?dogs and?extending clamps, it's super easy and hands free.? If you take Mike's band clamp concept one step further, you could band clamp the arc across a bench top with the straps fastened at the ends of the bench and the board assembly secured vertically in the middle and get both hoop tension and downward compression spread across the entire surface simultaneously in one action as well.
Lipping the edgebanding on the widebelt sounds smart and easiest when you have one available. By scoring, I was referring to the excess length ends of the edgebanding. On softer woods I often just slide a razor edge to sever enough fibers to allow a snap-break. On harder woods, chisel/mallet until similar can be achieved, sometimes angling the?blade for a guillotine effect.? It's quick and true when I know I'm following up with sanding/finishing. If your bands are 3mm or less, probably 2-3 good blows with a chisel already wide enough for the board thickness, then snap, you're on to sanding and no tearback.
Kumiko? Oh yeah!
On Mon, Jul 15, 2024 at 5:06?PM David Luckensmeyer via <dhluckens= [email protected]> wrote:
Hi Brett:
?
Thanks so much for the encouragement. I appreciate your thoughts ¨C and I don¡¯t think you¡¯ve missed anything. The power of brainstorming is such that even if multiple ideas end up getting dismissed, there is
no telling what new ideas or solutions arise from those initial ones.
?
For example, I had never thought about the success of edge banding by fastening just the ends. I can picture how that would work but I¡¯d be so fearful that the glue line might not be ¡°perfect¡± along the arc.
But this raises another question: what thickness of edge lamination would you use? The face laminations are 3mm, but I find this a bit too stiff for the curve, although it is quite a gentle curve. I have prepared 31mm x 2mm thick strips for the lamination
project. But I can thin these down further in my wide belt sander, to 1.6mm which is the thinnest it will do without a carrier, although I¡¯m sure I could play with the manual ¡°cam-action¡± controls and trick the machine in going thinner. As a solid wood guy,
I don¡¯t have a lot of experience with edge laminations, and I¡¯m (irrationally) fearful of going ¡°too thin¡± ¡
?
I totally understand your nailing comment. I¡¯ve become very respectful of the use of nails, especially the plastic nails shot by Omer guns. I do have a 23g pinner and it is indeed nearly invisible. But I just
can¡¯t bring myself to use nails on this build!
?
I¡¯m not sure what you mean by the scoring comment? I¡¯m going to send the uprights through my widebelt sander one more time each side to clean up the width of the lamination and bring it flush to the Zebrano
faces. Is that what you mean? As for the lengths, Zebrano at 2mm is too thick/hard to score easily, hence the thought of using a hand saw, or going the router with backer method suggested by Imran and James. I may have missed your point entirely.
?
I¡¯m excited by this build. Wait until you see what I have planned in terms of ¡°kumiko windows¡± in the uprights, featuring solid copper strip (30mm x 3mm) for the kumiko details. I really need a shaper origin
but failing that I¡¯ll have to outsource to a CNC owner.
?
David B. I think you suggested a flatbed CNC router setup?!!
?
?
Lucky
?
Lucky -?
Forgive if my thoughts are misguided, I might have missed some details of context:
1. If you are cutting the curve in the boards, assuming concave, save the outside waste cut to make a clamping caul for the edgeband to sandwich until glue is set.
2. I have successfully?glued bows in solid edgebanding?by simply flexing across the arc and fastening the ends only.
3. If you can bury a nail or staple without destroying the sancitity?of the build...I feel you squinting at this remark ...? esp if a 23ga pin nailer shooting double toenails will hold, they are practically
invisible.
4. Solid edgeband end trimming success is so species dependent, but I generally prefer scoring outside edge in until bend/break and leaving proud enough to power sand or block sand to contour.
5. This set is going to be awesome finished, already loving the proportions and joints now - great work Lucky.
?
I would make the caul (one side matching the shelf curve and flat on the other side), use the regular clamps. You can make the caul the same length as the shelf's, so you can use the router to trim the lamination.
Power tools with the proper jigs are always my?preference.
?
On Mon, Jul 15, 2024 at 4:54?PM David Luckensmeyer via
<dhluckens=[email protected]> wrote:
Hi Brian:
?
Apologies, I should have provided more detailed photos.
?
The bookcase uprights in Zebrano are composed of a solid wood core lamination with 3mm thick shop sawn Zebrano veneers on both faces. This is the
material I¡¯ve been machining for the half laps.
?
?
?
If you look closely at the photo below, you can see that I have edge laminated the back side of the uprights with Zebrano veneers, prior to the half
lap machining process. This worked well.
?
?
The problem with the front edge is that it will be curved. I could not machine the curve until the half lap joints were done. This means that the
front edge laminations have to be applied after the machining is done.
?
The horizontal shelves sit out from the uprights by 20mm. In the photo below you can see how the shelves are proud, and you can see the laminated
nature of the uprights.
?
?
My dilemma is how best to apply/trim laminations to the front edge,
between each shelf, on the soon-to-be-curved front edges of the uprights. Gluing will be tedious. I¡¯m thinking of using dozens of clamps like this:
?
?
But how to trim the laminations to length after the glue dries? I guess the flush cut hand saw method is the best one I can think of.
?
Lucky
?
I'm not following the need to laminate the front edges of the uprights? The sides are solid wood, correct? Are you doing
it to cover the slots and make the front surface proud of the shelves by 2mm? Color me confused...
On Sunday, July 14, 2024 at 08:25:22 PM MST, David Luckensmeyer <dhluckens@...>
wrote:
Hi FOG:
?
By way of update, the aforementioned Leitz blade is phenomenal. It cuts as smoothly as any blade I¡¯ve owned which is saying something for such a large blade with relatively narrow
kerf width of 3.5mm. The details appear again here, just in case anyone else wants to consider it as a 400mm blade option: item number 133130, 400mm diameter, 3.5mm kerf. 2.5mm body, 30mm bore, 96 teeth, 10 degree irregular pitch, ATB.
?
This is the first time I¡¯ve used a blade of this size. The dust it throws off is horrendous; I assume it has something to do with making 125mm and 135mm deep cuts. My 3M Dustmaster
is in continuous use.
?
?
This is very fine dust from a couple of hours work at the saw:
?
In the end, the powered rip fence and associated programming has been gold. I¡¯ve used the special programs on the overhead display a lot, including rebating, trenching, etc., but I
must say that I have not really used the specific programming mode for the Kappa 400. It is wonderful.
?
There are 10 program positions, and each of those 10 has 100 spaces to enter specific parameters that can be called up at the press of an arrow key and the start button on the overhead
display. You can imagine its use as follows. For the first half lap joint, I needed a cut at 125mm deep, at 120mm from the end of the Zebrano board. So for P1 (where ¡°P¡± stands for ¡°Program¡±), number 1, I¡¯ve entered 125mm blade height, 0 degrees blade tilt,
and 120mm rip fence position. It works out that I need the fence (which is acting as a bump stop) to move over 16.1mm to give the half lap width I need, so I¡¯ve programmed the number 2 position with 125mm blade height, 0 degrees blade tilt, and 136.1mm rip
fence position for the second cut (i.e. both shoulders of the lap joint). And so forth and so on for all 7 half lap joint positions along the bookcase upright. After 4 cuts, my rip fence is out at about 1150mm. So at this point I have to flip the upright end
for end and start again on the final 3 shelf positions (working from the top of the uprights). This is made easy using the programming on the Kappa 400. Obviously my uprights all need to be dimensioned accurately which I have done.
?
I forgot to say that I could not use special programming function which the Kappa 400 offers, where the rip fence can be programmed for a number of grooves or trenches of a specific
width and depth, at fixed intervals. The reason? Because my shelves are deeper at the bottom and much shallower at the top, so the position of each half lap joint on the uprights is of decreasing distance (from the bottom up) and increasing distance (from
the top down). Programming mode made this process easy. Seems like cheating. Comparing uprights at random, and I¡¯d say the 0.1mm claimed tolerance is verified. I¡¯ve used lots of compressed air between cuts to make sure registrations are accurate, and I¡¯d say
the Kappa 400 has worked as advertised. I strongly recommend a Felder powered rip fence as a wonderful option.
?
Anyway, the 3.5mm blade width allowed for 6 passes to remove the waste for each half-lap joint. No drilling, no knocking out timber, no marking lines, and definitely no chiselling.
I¡¯ve decided I can live with the bumpy but regularly reliable surface in the middle of each half lap joint.
?
?
Last night I assembled the first and smallest of the bookcases. It went together very nicely but I think I¡¯m going to tweak the Walnut shelves (through the wide belt sander) so the
fit is less tight. In the picture below, you¡¯ll see the upper shelves don¡¯t seem the correct width. That¡¯s because the uprights have not been curved yet, to reflect the design where shelves at the bottom are full depth (260mm) while the shelves become less
deep by the top (210mm).
?
l¡¯ll use a template to pattern route the curve after I¡¯m completely satisfied with the joinery. And then another tricky part is headed my way: I have to edge laminate 2mm solid Zebrano
onto the front edges of all the curved uprights. I¡¯ve agonised over this because I could not cut the curves and laminate before machining the joints because I needed parallel edges. Adding the solid wood lamination to the edge after the machining is going
to be time consuming and I¡¯m not sure how best to do it. I could apply over long sections to each curved edge (between shelves) and then flush them, but how?
?
-
Router (risk of splinters)
-
Chisel (not looking forward to that)
-
Flush cut hand saw (might be the best option)
-
Cut the lamination exactly to length and glue it exactly in position¡
?
If I used the flush cut handsaw, I could use overlength Zebrano pieces (but maintaining continuous grain) and then not stress the clamping process. I guess after that many cuts (224
to be precise) I¡¯ll be pretty good with the flush cut saw? Thanks for your brainstorming thoughts!
?
?
?
Lucky
?
Hi David B and Michael:
?
Funny you should mention the flatbed CNC. Although not the same, I have been pondering how much use I¡¯d get out of a Shaper Origin. That tool would fit nicely somewhere in my shop!
?
Thanks David for your vote on the vertical method. I think the clean corners are swaying me strongly for that option. I¡¯ve wondered if I make a very accurate registration jig and use
a small pattern bit (I can get an 8mm dia. bit from across town) to establish the majority of the 20mm flat across the joint, then there would be very little chisel work at all:
?
¡¤????????
corners squared by the saw blade
¡¤????????
majority of waste drilled out
¡¤????????
shoulder established by the pattern bit
¡¤????????
minor chisel work to finish.
?
Your jig making has always been a source of inspiration for me. Your lock mitre jig is especially egregious! I absolutely love it. Of course, now I know I should have made a much larger
vertical jig that does not depend on the crosscut fence for registration. Something like what you¡¯ve done would have been much better. Cheers.
?
Lucky
?
I actually like your original concept of using the rip as a bump-stop, a 400mm blade cutting both edges, and a jig on the crosscut fence to hold the material vertically.? Here¡¯s a couple more jig I¡¯ve
made to stimulate your thinking.? I could quibble over the actual jig configuration (hence these photos of jigs I¡¯ve made to hold vertical materials on the slider as food for thought).
This is for cutting lock miter joints:?
 ??
This one is for finger joints - which is a shallower type of thing you are doing:
?
?
On Jul 9, 2024, at 3:18?PM, David Luckensmeyer via
<dhluckens=[email protected]> wrote:
?
Oh wow! The video came through. I love it. No one would even think the drawer is there! Thanks for sharing.
?
Lucky
?
Lucky, as I further thought about my suggestion I realized it was lunacy. I think you are correct about returning to the horizontal setup. I know we are talking about a lot of work for 280 repititions!?
I had a few do overs and re-designs as I progressed but ultimately was very happy with the result.?
To answer?your question, it is a drawer that holds DVD¡¯s on the bottom level, and Blu-rays on the top level as they are both different size cases. The drawer was incorporated into a raised panel staircase
I built 30 years earlier and had always planned to make a drawer similar.
There is hopefully an attached video below showing final install??
The CMT was excellent; best dado I ever had but I ran it on a cabinet saw, not on my K700S. I used a simple ¡°box joint jig¡± that we all learned in woodworking 101 but made a mistake thinking I could
do the sides separately and still have things line up. A couple of tries changing methodology were necessary. The drawer empty without the drawer front was over 75 lbs made of Maple. I used Blum slides (Blumotion) with ?¡°Tip-on¡± added for push to open tech
to eliminate visible exterior hardware.
?
On Tue, Jul 9, 2024 at 5:41?PM David Luckensmeyer via
<dhluckens=[email protected]> wrote:
Imran:
?
Thanks for your ideas. You¡¯re correct about the slotting blade. If I go the vertical machining route I¡¯ll need nearly every bit of the 400mm blade¡¯s depth of cut. If I were to go back
to the horizontal setup, I could remount the dado blade, set it to maximum height, and make a partial cut. This may well be a better solution because it gives me the exact width (to match the existing shallower trenches).
?
You have rightly identified the time-wasting point of needing to take two passes (minimum) and then deal with the waste by drilling a hole with a forstner bit to get rid of the septum.
I would then have to mark and chisel the waste which is fine for 1-10 joints, but not 280 trenches¡
?
Brian:
?
That looks awesome what you¡¯ve made. Is it a drawer insert for plates? I¡¯m very curious what you thought of the new CMT dado stack? What did you like? What didn¡¯t you like? Thanks
for sharing. I¡¯d rather not run two 400mm blades on the slider. So much to go wrong! EEK.
?
Joe:
?
Love those jigs. Pictures speak volumes. So many ideas to learn from. Thank you.
?
Methodology Quiz
How would you make 280 non-through cuts that have to be precisely machined at 125mm deep and 20mm wide?
?
-
Vertical method as discussed to date using the 400mm sawblade, to make two cuts to establish each shoulder, drilling out the waste, and then A) mark a line and use a chisel to square up the joint,
or B) use a router and small pattern bit to establish the shoulder and then chisel out the corners.
?
-
Horizontal method using the same dado stack that machined the shallow trenches, dado raised all the way up for the stopped cut, and then employ either option A or B above to clean up the joint.
?
-
Rely on the existing shoulders of the trenches (5mm deep on both sides), and use a router and pattern bit to obtain the 125mm ¡°half lap¡± joint, and then square up the joint using option A or B above.
?
-
Something else?
?
I personally can¡¯t see a way to avoid handwork of some kind. The reason why I favour the vertical method (as I¡¯m calling it) is because I can get precise and square internal corners
for the stopped cuts. I find it much easier to clean out the waste if the corners are already established. I have never liked chiselling corners square, even with a square chisel (e.g. Veritas). And I would
hate chiselling 560 corners!
?
Thanks in advance for your input,
Lucky
?
I have another more stable jig for demonstrating tenoning on the slider that could be adapted for this also.
?
--
Brett Wissel
Saint Louis Restoration
1831 S Kingshighway Blvd (at Shaw Blvd)
St Louis, MO 63110
314.772.2167
brett@...
-- Brett Wissel Saint Louis Restoration 1831 S Kingshighway Blvd (at Shaw Blvd) St Louis, MO 63110 314.772.2167 brett@...
|
