Project Wavy Bent

[krtwood]

I'm working on a limited edition set of 20 jewelry boxes.  I don't have a proper name for it yet and am referring to it as Project Wavy Bent because, well, I needed to name the file folder something and that was the first descriptive thing that came to mind.  Bent because the outside is a bent lamination and the wavy part we'll get to in the next post.

I started out in Sketchup and from there made a full size mockup out of scrap ply and paper all hot glued together.  The mockup showed me I was on the right track but had to make it a little smaller as it was too far around for one sheet of ply (5'x5' BB).  In the Sketchup drawing I have a basic layout of the front.  There will be a vertical slide out section on the right with hanging storage for necklaces and earrings and drawers on the left.  I haven't worked out exactly how that's all going to look.  I mainly just drew in the front to decide whether I wanted to go with walnut so I could buy the material.  I ended up with a couple nice 12" wide 8/4 boards that I can resaw for a bookmatch.  

 

Once the outer shape was finalized I used the cnc to cut out an mdf bending form.  My first attempts at doing the glue up I did it all on my assembly table and would put all of the clamps on one side of the form and then flip it over to get the other side clamped.  This was really awkward and tended to distort the thin plywood so I would end up with gaps on the second side.

After a few attempts at that I realized I could have the form supported by post in the middle and be able to get the clamps on both sides at the same time as I worked my way around.  I start on the table and get the first four clamps on.

Then I work around both sides and then use a ratchet strap to pull it around the other two curves.

Here it is with all the clamps.  Hard to see from this angle, but at the end closest to the camera I use a couple of track saw clamps as they have a low enough profile to fit under the overlap of the two ends.

I do each in three separate glue ups of 2 layers.  Once I got the process down pat I probably could have gotten away with doing 3 at a time, as the time went down from 30 minutes to about 15 minutes (not including spreading out the glue) and the glue (Ultra CAT) has a 40 minute working time.  But it took quite a while to get there and at that point I decided not to mess with what was working.  So I've been working on this since July.  Sometimes I get two glue ups done in one day, sometimes only one.  Instructions on the glue say 4-6 hours clamping time at 70 degrees.  Even in the middle of summer my basement shop rarely hits 70, unless I have the DC running for a while.  I've been putting an electric blanket on it which gets the squeeze out cured on the top in 8 hours but not on the bottom.  Even if the squeeze out is still mushy the lamination itself is okay to come out of the clamps.  The good aspects to this glue besides the working time is that it dries really hard, minimizing the spring back, and it doesn't stick to metal or plastic so it pops right off the clamps.  Speaking of clamps, there are a LOT more of them on there now than the first time I did it.  I kept modifying the form to find more places to get them on because it would come out a little wavy with high and low spots, and that is not the wavy that I am going for.  I still don't have it perfect all the time but it's pretty close.

Here is the amount of spring back after the first two layers, about an inch on each side.  Once all six are done there is effectively zero spring back.  Most of the time I have to pound the form out of it.

At this point I have 19 1/3 done.  So I'll be able to get doing on the next step very soon.  Very much looking forward to not having to share the shop with the clamp porcupine anymore.  I haven't been able to get to the planer for two months so it's been a stretch to keep myself busy with small projects.

 

 

[Coop]

That’s some serious clampage and strapping, but we’ll applied. Very cool  project. 

[wdwerker]

"Clamp porcupine "  wonderful term !  I've made several projects over the years that took every short clamp in the shop. Looks like a very interesting project to follow . 

[Chestnut]

This looks like it could be a really cool build. Look forward to seeing more.

[pkinneb]

Very cool!

[krtwood]

12 hours ago, wdwerker said:

"Clamp porcupine "  wonderful term !  I've made several projects over the years that took every short clamp in the shop. Looks like a very interesting project to follow . 

I ended up cutting down six of my 18" clamps to 8" so that I would at least have room to move around the thing.  That hurt.  You know I'm going to need more longer ones than I have now on the next one.

[wdwerker]

I have found that you can never have enough short bar clamps .

[krtwood]

The original idea for this was to use something like a contour map to create a 3d surface that was wrapped around the outside of a box.  This is the "wavy" part.  For a long time I was stuck on how to get started designing this until I came across a video on Suminagashi, Japanese water marbling.  I made a pan big enough to do it full size, though it probably would have been better to do it at half scale or so.  It kind of took too long to get all the ink everywhere and some areas kind of went past their prime look and became too chaotic.

I had some trouble getting the image transferred onto plywood (didn't have any paper large enough on hand) but got one that was good enough on the third try.

From there I got that into the computer and interpreted it into a 4 layer contour map.

Which I then used the cnc to cut into the plywood.  Lots of tabs to hold it together and I had it drill lots of holes in the areas to be removed.  I upgraded to this larger cnc pretty much with just this project in mind.

The height I made the tabs originally wasn't enough to prevent some breakage in the tight bends.  After the first attempt I both increased the height and started putting another sheet of ply with packing tape all over it on the outside during the glue up.

I built my tilting power carving bench with having to carve on a big triangle in mind

The first time around I just kind of dove into carving it.  It got really confusing as to what layer was what and thus hard to tell if I should be looking for another layer under a particular spot or not.  The second time around I started taking it down one layer at a time and removing almost all trace of the cnc'd dotted line on that layer before moving down to the next.  

This first one was one of my early glue ups and the carving exposed a lot of bad glue joints.  I didn't make the second attempt until I had done a great many of the regular version (There's 15 regular ones and 5 sculpted ones) and really had the process down.  There's one problem area that ends up on the bottom but other than that I am happy and relieved to say there's not a visible bad joint anywhere in all five of them.

I did as much as I could with the 4" turboplane, then almost all the rest with a 2" burr disc in the mini-grinder.  There's a lot more detail in the design than I can really do with that but most of it gets smoothed out.  There are however a few key areas where I went in with the flex shaft.  Anything I can do with the mini-grinder I can also power sand, but the smaller stuff is going to need hand sanding and destroy my will to live.  So I kept that to a minimum.

And here is the prototype 99% sanded.  Looking at it I decided that I did not like the way the design cut into the front edge of the case.  It looked fine from the side or even at an angle but head on it just looked bad that there wasn't an even frame all the way around, with it just being cut into randomly.  Also it cut too deeply into the back edge so I wouldn't be able to have a rabbet for the back panel.  So I reworked the design so it only slightly cuts into the front in one spot, which is right where a vertical divider is going to be that connects with the gap between the two ends at the bottom.  Then I got to do it again but this time shift the design two inches over so that spot would actually be where it was supposed to be...

The next post should catch us up with current state of affairs.  Not too much more progress beyond this.

[wtnhighlander]

This is really interesting. Hard to imagine what the final version will look like, but I look forward to seeing it!

[Chestnut]

I watched the vid you made on transferring the water marbling to the board and it was really interesting. I like the outcome on the curved case much better than the graphic makes it appear. I commented about how it'd be interesting to use contours from a geographic area that is sentimental to you, but realize that wouldn't really work bent around the outside of the case.

I can't help but notice that you have a lot invested in power carving. Have any advice for someone that wants to dive in sometime soon?

 

[Chet]

I have been watching this on Youtube.  Very fascinating project and techniques... which most of your stuff is.

[krtwood]

6 hours ago, Chestnut said:

I watched the vid you made on transferring the water marbling to the board and it was really interesting. I like the outcome on the curved case much better than the graphic makes it appear. I commented about how it'd be interesting to use contours from a geographic area that is sentimental to you, but realize that wouldn't really work bent around the outside of the case.

Ah, that's you   I once did a picture frame with contour map of a mountain range for my brother and his wife's certifcates for completing hiking all the 4000'ers in the state.  I could totally see something like that on this as it's kind of a mountain.  But then you're narrowing down the potential customer base.  They have to have a connection with that place and like the style of the piece.  And my style is well, not normal...

 

Quote

I can't help but notice that you have a lot invested in power carving. Have any advice for someone that wants to dive in sometime soon?

 

It depends on the scale you're working on, but I would go right for the turboplane.  I started with a kutzall burr disc and while it certainly works, it's really hard to clean.  It's a cheaper way to get started but it's kind of like buying a jobsite table saw.  You're going to outgrow it in a week.  I thought the burr disc would still be useful as a finer option, but I think the turboplane is actually better at making light passes.  You can hear it cutting and it doesn't really do much without a tiny bit of pressure, whereas the burr just removes material without you really knowing how much you're taking off unless you take off enough to make the grinder motor change sounds.  The turboplane throws chips everywhere whereas the burr kind of sends a stream of dust in one direction, so in certain circumstances the burr is easier to do dust collection.  But practically, it's going to make a mess either way and the turboplane makes less dust and doesn't spread the mess as far.  It's just utterly hopeless to try to do dust collection with it though. 

I have nothing but good things to say about the arbortech mini-grinder for smaller work.  Though I would recommend the opposite advice and get a 2" burr disc for it.  I haven't tried the mini-turboplane they have for it, but I have tried all the blades and you get way smoother results with the burr because of the curved edge.  The one I have is from saburrtooth, and it's much easier to clean than the larger kutzall.  I don't know if that's because of the different manufacturer or that it's the finest grit.  As far as grit goes, I would only get the finest grit.  There isn't really that much difference in how aggressive they are so you may as well just leave the best surface.

For sanding I use the Tim Skilton sanding pads with the wavy edge discs.  Klingspor sells them in a 50 pack for pretty cheap and you burn through them in no time.  I have used them in my cordless but it really doesn't like it and gets super hot.  So for this project I bought a milwaukee close quarters drill, but haven't used it yet.  I have tried my regular corded drill and it's just too awkward.  The chuck is too long and the balance is just wrong for the way you have to move it around.  So hopefully that's going to be worth it.

I have a Foredom flex shaft unit.  I have the lower end one, I think the SR, and it's a little underpowered.  I would get the bigger one if I had it to do over.  Might get it eventually if I end up using it more.  Also getting the burr caught up in your sweatshirt breaks the internal part of the flexshaft, so you should have an extra one or two of those lying around.  Or so I've heard

 

Full disclosure:  I did get the turboplane from a sponsorship but there were no strings attached.   I bought all the other arbortech stuff I have before that. 

[Chestnut]

11 minutes ago, krtwood said:

Ah, that's you   I once did a picture frame with contour map of a mountain range for my brother and his wife's certifcates for completing hiking all the 4000'ers in the state.  I could totally see something like that on this as it's kind of a mountain.  But then you're narrowing down the potential customer base.  They have to have a connection with that place and like the style of the piece.  And my style is well, not normal...

 

It depends on the scale you're working on, but I would go right for the turboplane.  I started with a kutzall burr disc and while it certainly works, it's really hard to clean.  It's a cheaper way to get started but it's kind of like buying a jobsite table saw.  You're going to outgrow it in a week.  I thought the burr disc would still be useful as a finer option, but I think the turboplane is actually better at making light passes.  You can hear it cutting and it doesn't really do much without a tiny bit of pressure, whereas the burr just removes material without you really knowing how much you're taking off unless you take off enough to make the grinder motor change sounds.  The turboplane throws chips everywhere whereas the burr kind of sends a stream of dust in one direction, so in certain circumstances the burr is easier to do dust collection.  But practically, it's going to make a mess either way and the turboplane makes less dust and doesn't spread the mess as far.  It's just utterly hopeless to try to do dust collection with it though. 

 

 

Full disclosure:  I did get the turboplane from a sponsorship but there were no strings attached.   I bought all the other arbortech stuff I have before that. 

Awesome information!

So I didn't mention I'll probably only be doing chairs with it ... to start. I at first was looking at the Holey Galahad because i was nervous the turbo plane might be too aggressive and cause me to do more damage than do good eg. the kick backs you had on your crayon vase (i realize that was an entirely different cutter just an example). I never thought about cleaning out the bur discs though that's a good point. The big benefit i can see to the arbor tech if i can tame it, is that i wouldn't need a coarse and a fine so the price is essentially a wash.

What is edge durability like does the turbo plane wear out? Can you sharpen it? I feel like i should just dive in and practice on some of the firewood that i have drying on the side of my house.

[krtwood]

4 hours ago, Chestnut said:

Awesome information!

So I didn't mention I'll probably only be doing chairs with it ... to start. I at first was looking at the Holey Galahad because i was nervous the turbo plane might be too aggressive and cause me to do more damage than do good eg. the kick backs you had on your crayon vase (i realize that was an entirely different cutter just an example). I never thought about cleaning out the bur discs though that's a good point. The big benefit i can see to the arbor tech if i can tame it, is that i wouldn't need a coarse and a fine so the price is essentially a wash.

What is edge durability like does the turbo plane wear out? Can you sharpen it? I feel like i should just dive in and practice on some of the firewood that i have drying on the side of my house.

I think you'll be surprised at how easy it is to use.  It can get away from you a little if you really push it but nothing like those kickbacks.   You have the most control pulling it toward you, so I do that when I'm being aggressive with it.

I've only had it a short time, but it's 3 carbide cutters.  I'd expect the durability to be similar to a router bit.  You should be able to sharpen it with a diamond file.  

[krtwood]

So the interesting thing I think at this point is the total lack of a reference surface.  The sides are all curved.  The faces/edges are not flat.  And the case has some flex to it, which meant trying to flatten a face at this point would probably lead to chasing my tail.  So I had to start somewhere.  So my first goal was to get rid of that flex by joining the two ends, so that I could then flatten a face.  Before I could do that I had to square up the ends, since I obviously can't be cutting through them after there's something joining them.  Well, I suppose I could have done some kind of temporary joining, flatten the faces, and then come back to squaring up the ends.  But really if they aren't absolutely square it doesn't make any difference.

Next I made a fixture to hold the case.  There are high stops at the ends to force them all to be the same width.  The ones with the sculpting had more springback after I removed a lot of material at the top curve.  There's a rail at the front and back which supports a template that I used to route a flat area on each end.  I had the bit extended as far as I dared to get to the lower side and it didn't completely flatten it, but the low spot is only 5 thou or so below, it just looks like it's more than that.

The walnut piece brings the two level with each other and extends out in both directions.  I used walnut here because you can just barely see it in the finished piece from the right viewpoint and I want it to disappear in the shadows.  

I used block with a cleat on the right side to reference off the edge of the walnut piece to locate the plate that joins the two ends in the same spot on each case, and to hopefully get the slot in it aligned parallel to the sides of the case.  The slot is going to receive a vertical divider, more on that at the end of the post.

With the two ends joined now I could flatten a face.  To do that I had to modify my shop built drum sander for a larger depth of cut.  Fortunately I never glued the cabinet together so I could remove the front and back of the cabinet and move the motor to the outside.  Then I just had to rig up a temporary support for the table.

If I just had a couple of these to do I would have done the other side with the drum sander as well, but with 20 to do and more than 1/8" to remove I did something I don't like to do and that is use the table saw to cut something that isn't flat.  They are big enough that my hands were nowhere near the blade and I had a lot of leverage.  It went fine.  The blue tape did it's job preventing splintering for the most part, just a couple spots I had to fix.

That brings us up almost current.  I've started working on the rabbet for the back panel.  Rather than use a rabbeting bit I cut a pattern on the cnc and am using a pattern bit so they all end up the same.  Then I can cut the panels on the cnc and not have to fit anything once I get it dialed in.  

That vertical divider is what is currently vexing me.  It attaches at the bottom into the groove in the joining plate.  It attaches to the top of the case... how?  I am thinking pocket screws, but on the sculpted version there is not enough material there for a screw.  I've got 1/4" at best thickness to the case there.  I was thinking about a "pocket dowel", where I would use the pocket hole jig with a regular 3/8" bit instead of the step bit to make a shallow hole in the case so I could insert dowels after sliding the divider in place.  But I'm open to other ideas.

On the right side of the case there will be a slide out unit for hanging storage on full extension slides.  So I have to also get another slide at the top.  I think that will be mainly attached to the divider rather than the case, but there's some opportunity to reinforce the divider/case joint there as well.

 

[Chestnut]

Is the divider going to be plywood? Honestly i don't see a lot of stresses in the dvider ply has decent edge gluing strength i'd just get the angle right and glue it no joinery. Should have a good amount of long grain to long grain glue surface.

[krtwood]

5 hours ago, Chestnut said:

Is the divider going to be plywood? Honestly i don't see a lot of stresses in the dvider ply has decent edge gluing strength i'd just get the angle right and glue it no joinery. Should have a good amount of long grain to long grain glue surface.

I question whether I can get the fit perfect on 20 of them.  Also, I should have mentioned it would be nice to be able to remove it for finishing.  Perhaps just a couple of blocks glued in on both sides to lock it down laterally. 

[Chestnut]

2 hours ago, krtwood said:

I question whether I can get the fit perfect on 20 of them.  Also, I should have mentioned it would be nice to be able to remove it for finishing.  Perhaps just a couple of blocks glued in on both sides to lock it down laterally. 

I guess i wasn't sure how the final product. I don't have the vision that you do. If there is going to be a back and a front so everything gets hidden the 2 blocks would work great. If the area is going to be seen at all is where things get iffy.

Absurd suggestion. You could find a way to clamp the divider in place and then do another ply of 1/4" lamination around the inside creating a slot at the top it would make it look like there was a perfectly cut groove.... jokes that sounds miserable.

[krtwood]

Yeah, there's a back panel and then walnut drawer fronts covering everything up.  The divider will have a walnut solid wood edge banding on the front and I figure I can apply that after fitting the divider so that if there's a tiny gap at the top I can cover it.  The edge banding will be in front of the plate at the bottom so I don't have to worry about fitting it between the two.

I cut all the rabbets for the back panel and cut the first back panel on the cnc.  I used a -0.007" tolerance as you usually need some tolerance to get things to fit.  It was way loose.  So I cut a second one with +0.002".  Still loose.  Super weird.  Then I went and looked at the template and sure enough the new Whiteside pattern bit i got was overcutting.  Try number three at +0.015" did the trick.  I would be more annoyed if I had only just enough plywood for all the back panels.  But some idiot had already used half a sheet for something else so I gotta go buy more anyway.  Of course it's supposed to rain for the next two days.

[Chestnut]

1 hour ago, krtwood said:

some idiot had already used half a sheet for something else

Awe man i get that jerk in my shop from time to time as well. I hate that guy.

[krtwood]

Been a while since an update.  This week I routed a few dovetails.

The layout of the tails is pretty odd because of the curved side.  Some of them have two tails on one side and one on the other because the top of that side of the drawer is going to get cut off.  The bottom drawer has one tail on each side but in different places.  This drawer is going to have three levels to it, to make use of the odd space at the bottom because of the way the case curves up there.  So the middle section will be deeper than the two sections on the sides.  It's not a big drawer so this may seem silly, but jewelry isn't big either. 

Nine drawers in each box.  I made one extra set of parts for each drawer in case I screw something up later, so twenty-one of each drawer.  So that's 189 drawers.  I don't think I have the stomach to start on the grooves for the drawer bottoms just yet tomorrow so perhaps I'll just work on my website redesign.  

[pkinneb]

Wow that’s a lot of drawers

[wtnhighlander]

I've been following this crazy build on YT. All I can say is, More power to you! I would have totally pulled my hair out, long before this.

[..Kev]

4 minutes ago, wtnhighlander said:

I've been following this crazy build on YT. All I can say is, More power to you! I would have totally pulled my hair out, long before this.

I've been following this on YT as well.. I'll admit that when I saw the post here, I ran to YT to see if there was a video...lol

Looking forward to seeing how you pull this one off!  

very nice work!

[krtwood]

7 minutes ago, ..Kev said:

I've been following this on YT as well.. I'll admit that when I saw the post here, I ran to YT to see if there was a video...lol

Looking forward to seeing how you pull this one off!  

very nice work!

Monday night!  It's a boring one though.  First half stock prep.  If you survive that, the long version of what I wrote above.  I imagine it's going to take a week to get the drawers assembled before things get interesting again.  I would just assemble drawers for five of them to get things moving as the sooner I have something done the sooner the chance to start selling them.  But being able to fit each drawer once and repeat the same thing for the rest with the same setup is kind of the whole point of making 20 of them so I've got to muddle through all of it.  Which means I'm going to have stacks of drawers everywhere, in addition to the boxes themselves stacked up on the table in the other room I normally stack that stuff on to get it out of the way.  I need to build the storage shelves I'm going to store these things on long term so I can free up the table.