Making "Dreamery"

[Mark J]

I was going to post a few tidbits of this project in the "What’s On Your Workbench” thread, but then I got to thinking that maybe I should take a chance and do a real journal. Unlike some of my previous journals this won’t be done after the fact, when I already know the project has succeeded, but rather it will be done prospectively. I tried this for my first journal some time back and unfortunately that project went to poop. Hopefully I have enough “street cred” on this forum to weather the embarrassment if this one goes to poop, as well. I am taking pains to say this as I think this will be the most difficult project I have undertaken so far. So no promises. There are a lot of points in the process where poop could just come to be. 

I haven't named the piece yet so the title, 4 Duet De v9.0, is just the file name in my computer.  If it all works out I'll come back and edit the title.  Here’s what I am proposing to make:

The basin is walnut and the base is butternut; I’m hoping, once finished, that combination will have just the right level of contrast and coordination.

The basin is similar to basins I have made before, so I will use a similar approach, including making the pointy part. But the deep overhanging rim will be a bit of a challenge.  I am thinking of it as a wide mouthed hollow form, but I’ve only done a couple of hollow forms, so….   At least I will be able to see what’s going on inside.

The base has some entirely new problems to solve. With my prior convolved pieces there was a connecting ring of wood at the very bottom of the piece that was available to create a mounting point, and even if the final form did not include an integral basin there was at least a temporary ring of wood at the top helping to hold everything together. Because the shape of the walls on this piece is externally concave rather than the convex shape of previous pieces, the single ring of wood that will hold it all together is in the mid section.

This project actually began early in October when I started to look back on some thumbnail sketches I had done earlier exploring ideas for another “duet” piece. The design I previewed above is the final choice after spending my hobby time designing and re-designing this piece over several weeks. It’s difficult to actually count up the number of different versions I have drawn and considered before settling on this form, but it’s easily over two dozen. These includes bases similar to my previous pieces:

And also, at the behest of some artists I have spoken with, at least one base with three pillars instead of four, which as you can see suffers from “plumber’s butt”.

Well, the good news is that with all the designing, I’m getting better with Fusion 360.

After the design phase I needed to figure out how I would make the piece, so I’ve spent my time the last couple of weeks engineering. The good news is I now see a pathway to make the project. Since, unlike a chair or dresser build, this is an unfamiliar process to most of you (even if you’ve seen my previous journals), I thought I would layout the broad strokes of my plan to make it easier if you choose to follow this journal. And it gives me a chance to show off some Fusion 360 mojo.

Starting with the squared up blank mounted via a screw chuck in the top face, the first step will be to turn the outside of the form.

Then begin hollowing by removing the bottom half of the inside. I will use the outer wall contour as a guide to the shape of the inner walls. My target thickness is usually ¼”, but I will leave the wall thicker at the level of the connecting ring.

Then I will glue the bottom of the base to a sacrificial block of poplar which is in turn mounted to a screw chuck. This allows the piece to be flipped on the lathe so that the base is now mounted to the spindle. The first chuck, which had been attached to the top face, can then be removed exposing the top face for hollowing out. Again the outer contour is the guide to shape and wall thickness is determined directly.

With the inside and outside contours completed it will be time to cutback the sides to reveal the form.  The piece is removed from the screw chuck and a template is attached to the bottom of the poplar block using the previous screw hole.

Then through the magic of Fusion 360 the waste wood is removed and it will look like this (in theory).

How that magic is actually going to happen is a matter of some debate in my head. I have two plans in mind. One is to use a bandsaw follower and the template, but that has had mixed performance in the past. The second idea is to carve away the waste wood with a Dremel.  I may end up with some hybrid of the two; the jury is still out.

After the cutback the edges get chamfered. And then the sacrificial block has to be removed. I plan to saw this off with a flush cut saw, and hope that the legs lengths are close enough that only a little sanding will be required.

 

So that’s the plan as I’ve drawn it. Of course it remains to be seen if it survives first contact with the wood.

[wtnhighlander]

Whether the project turns to poo or not, someone will learn something valuable from this. I'm certain that I will!

[Chestnut]

I've already learned a ton about your process from this post. It's helped fill in some gaps from previous explanations. I'm excited to watch this through.

If all things go to poo that's not a bad thing necessarily. I learn a lot from failed experiments

[Coop]

I’ve seen your projects before and if there is poop, your’s don’t stink!  Looking forward to this!

[Mark J]

I should back track a bit briefly. I showed 3D renderings of the final goal, but I have a process to get there.  It starts with some small scale thumbnail sketches.  

Followed by full scale sketches making assumptions about the final size of the wood blank.

When I have some designs I would like to see in 3D I bring them into Fusion 360 again using an approximate size for the blank. I will then tweak the curves in F. 360 making small changes.  When I have settled on an approximate design I choose a block of wood to become the blank. This then gets squared up resulting in four flat and perpendicular sides and two flat and perpendicular faces.  (I have showed that process before, but I can do it again if needed).

Then based on the actual size of this trimmed blank I go back to F. 360 and tweak the design to make sure that it fits.  That’s how I got the images I showed earlier. Nice pictures to judge the form, but I can’t really use them to make the project. For that I need accurate life size 2D cross-sections. So back to the graph paper.

I use these drawings to help me engineer how I’m going make the piece. I look at how I am going to hold the work at each step and confirm that tools are long enough and can reach into the recesses without impediment. Then I mark out the guide holes which are 1/8” diameter holes drilled to about ¼” shy of the intended contour line.

Next I mark up the blank to show every possibly important landmark. These include the center of each face, the location and depth of each guide hole, and any diameters that may be of interest (e.g. rims).

Lastly, one more trick, and this is hard to explain. Although I’m carving away a lot of each face, there is some portion of that face that is going to remain as part of the final surface. I make a drawing to depict what’s going to remain, photocopy this and transfer the lines to each of the four faces of the block with carbon paper. It is very helpful during the actual turning. It’s not so much a guide as to how to remove wood as it is a warning that you’re getting close to removing wood that’s not supposed to come off.

In this pic you see the photocopy, and the relevant lines are the two broad curves at the top and the bottom. (the two curves at left and right just show me for the sides relate to the top and bottom curves).

I know that is clear as mud, and I can go into this in more detail if that’s where someone’s interest lies, but I’m figuring you guys would rather see some actual sawdust.  I'm just trying to say here that nothing is left to chance if I can help it.  And maybe explain why the design and planning phases would go so many weeks.  

[Chet]

I would be a pretty basic woodworker if I hadn't pushed projects to new levels.  I have learned how to make adjustments and repairs by having the poop show up.  AND I have learned just as much or maybe more from other peoples poop incidents.

So lets go Mark, I am ready to watch this.

[Mark J]

With the blanks marked up It’s time to start drilling the various holes preparatory to turning. Since I’m going to turn the butternut base first I’ll start with that. Given the open nature of the base shape I don’t really need guide holes for the inside curve, that will just parallel the outside curve. But given the concave outside curves I can’t place guide holes from the top or the bottom, and placing them from the side means drilling the holes from the extreme corner of the block.

First step to do this was to make a 45* cradle to hold the blank. Mine is nothing special, I made it on the table saw with the Incra Express/1000 HD. It’s two identical pieces clamped together with a spacer block in between. The blank is held in place on the cradle with double stick tape and the cradle is then clamped to the drill press table.

Now the tricky part. These holes have to be placed somewhat meticulously if the end of the hole is going to indicate the desired contour line deep inside the blank.  Most bits will slide off the corner to one side or the other.  I have a couple of times in the past extolled the virtue of a drill point countersink bit for starting holes. First, although at first glance the tip does not appear to be that fine, the entry point is well defined, and easy to see, so the bit can be very precisely lined up with your target mark in two projections. Second the bit itself is short and fat, so very stiff. It does not wander. My bit has a 1/8” drill point and a 5/16” diameter countersink, matching the two sizes of holes I commonly need in a blank. Here is the DPC bit lined up on corner of the blank.

Here is what the starter holes looks like.

After the starter hole is drilled nothing is moved as I want to maintain the alignment. The DPC bit is carefully swapped out for 1/8” twist bit, which for most holes I drill is a 6” long “aircraft” style bit. So having a lot of quill travel is essential.

I carefully adjust the mechanical zero stop until the tip of the twist bit is exactly level with the top of the starter hole and then set the electronic zero. Next the hole depth is set electronically. Note that each target mark on the blank also shows its hole depth in decimals. Now I drill the guide hole. Once the first hole is done it’s rinse and repeat.

One thing I have learned to do is to drill the shallowest hole first, then the next shallowest, etc. It is possible to forget to change the electronic depth control and drilling too shallow requires a lot less crying than does drilling too deep. Here are all the holes drilled, all in a row and all on the tip of the corner. If you are unsure of any hole depth you can probe with a dipstick.

Next step is drilling the clearance hole for the screw chuck. Back to the DPC bit (a keyless chuck is worth it’s weight here). Again, precision centering matters. I have lights to the left and right of the drill and find that watching the shadows converge as the bit descends can be very helpful in getting the bit to pinpoint the target.

This time when I drill the starter hole I go deep enough to leave a shoulder from the 5/16” countersink part of the bit.

Swap the DPC for a 5/16” twist, it seats neatly in the starter hole and I drill a bit more than ¾” deep for the screw chuck. Then mount the blank.

 

The project is running ahead of the journal, so I should have some more for you later today or tomorrow.  

 

 

 

 

[Mark J]

When I begin the turning it’s helpful to know what parts of the waste wood can just come off without a lot of caution. In this case the diameter of the bottom “rim” is smaller than the diameter of the top rim and everything between the rims is waste, so that defines the first chunk that can just go.

If you have a square blank with the grain running perpendicular to the lathe, as in this case and as in most bowls, it is better to start the rounding from the flat face rather than try to attack the work form the sides directly. First it is a lot less pounding and there is much less risk of splintering the wood off the back side of the corners.

Here I’m set up with my radius square turning tool ready to engage the bottom face. The periphery of the bottom rim is marked and you will also see a straight line about an inch down from the top face demarcating the end of the “good to go zone”.

Likewise I know that the middle is just going to come out, so I can just start removing it. And when the inevitable “are we there yet’s” arise I can probe the guide holes with a dipstick.

It is possible to see the curved line that demarcates the top rim when the work is spinning. Cameras don’t have the same persistence of vision that humans have, but you can just barely see that line in this photo.

As I remove more material I start to pay attention to the valley curve. I still have a good bit of wood to remove, but I want the surface I am working on to start paralleling the final form.

As the valley gets deeper I need to get creative with the tool rest for good tool support.

Seems like it will take forever, but I keep working the contour deeper, and suddenly I’m seeing the bottom of the guide holes.

 

The guide holes were positioned approximately ¼” short of the target, so there is still more material to remove. Leaving a quarter inch means there’s room to refine the shape. As I do this it’s still necessary to keep track of the diameter of the waist and this can be measured with calipers. The calipers can then be directly compared with the full size drawing.

Up until now I have been using a standard (single bevel) carbide scraper to remove wood. These cutters are efficient at removing wood and they can be had from various sources keeping the replacement cost down. On the down side the surface it leaves behind has a lot of tear out.

So for closer in work I switch to a negative rake carbide scraper (double bevel), which is less efficient at removing wood, but leaves noticeably less tear out. These cutters are proprietary so they’re a bit more money to replace, hence I save them for when it matters more.

I continue to refine the shape while removing some of the remaining excess material. In assessing a curve it can be helpful to view it against a black background. The wood is over exposed, but you can see the curve in stark contrast. There’s a flat spot visible on the upper wing. If you close your eyes and feel the curve, that flat spot stands out, too.

Now I am about 1/16” from the target.

It’s time to change tools, again. For the final contour work and to get the smoothest tooled surface I am going to switch to high speed steel negative rake scrapers. Like the double bevel carbide scrapers, these NRS tools have two bevels, which makes them less aggressive and smoother cutting than a single bevel scraper. Where the carbide tools hold their edge a long time and have no burr, it is the burr that allows an NRS scraper to cut so smoothly. The action is much like a card scraper in flatwork. So a NRS needs to be touched up every few minutes of use. Here are the two that proved most useful. I was a little concerned at the outset that tools might get trapped in the valley, but things worked out.

I wish I could take still shots of the tools in action, but that just can’t be done solo and I don’t have a photographer. As I continue to refine the curve there are a couple of other tricks I can use. If you recall I have a photocopy of the side view of the base, I took that and cut out the curve. Now I can use both the negative and positive curve cutouts to assess my progress.

Confession: I created a curve in a 3D program then by a hand to eye process transferred that curve to a piece of paper. The curve on the piece of paper does not exactly match the curve on the computer. That piece of paper is now suspended on an easel next to the lathe and I am transferring the curve to the wood using another hand to eye operation and the curve in the wood is not exactly the same as the curve on the paper. This isn’t CNC, but just the same the goal is to get the curves to be very very similar, and of course, still graceful.

I know from previously exploring design variations that the curvature of the upper wing is more important than the rest of the curve. The upper wing determines how high the basin will sit, and that has a big effect on the overall aesthetic. So I am concentrating my efforts on getting that part of the curve right. And I think I’m there. A quick check of the waist and a lean back look at it overall.

By jove, I think I’ve got it. Time to start sanding. Despite the NRS I still have a bit of tear out to deal with on the underside of the upper wings. The wings are thin and so were vibrating a bit too much to get a really smooth cut with the NRS. No matter, a bit of p80 grit paper just in the areas affected solved the problem. I’ll continue from here sanding through the grits to between p600 and p1200. During the planning stages I thought I would have to sand the outside contour by hand, due to the tips of the wings being sharp, but it looks like there’s enough room to sand with the lathe running slow (90 rpm).

[wtnhighlander]

Excellent explanation! Some of the magic used to create these forms is becoming a little less clouded.

Thanks.

[gee-dub]

Thanks for the great explanation and show-n-tell.  The different depth holes acting as witness markers is brilliant.  Really fun stuff and I love the results.

[Coop]

If NASA had you and the $$, they wouldn’t need Russia’s support! 

[Mark J]

Most of yesterday's shop time and half of today's spent sanding.  I thought I would spare you the swipe by swipe demonstration, but there a few point I'll make, particularly as no one asked . 

Why sand now, after all there's more turning to do and much of what I'm sanding is going to get cut off later?  First it's much easier to sand a piece that is still mounted on the lathe rather than handheld.  Even if the lath motor can't be used it still serves as a handy holder.  Second it's much easier to sand something that is a continuous round form than one that is interrupted.  And third there's a lot of sanding that's going to happen, much of it by hand, and I prefer to break up the task. There's a risk that the finely sanded surface will get a blemish, but it's easier to repair a p1200 surface than create one.  

Why so fine a grit?  Because these pieces will inevitably be held in the hand and examined closely (hopefully by someone who has just paid money to own it).  

I like to use a 3" hook and loop Abranet knockoff that I get from Woodturner's Wonders.  It is very comfortable to hold and gets into spaces, and I like that I can blow it out from the back (into the DC).  

For this project I had to do some spot sanding with p80 on some rough spots on the wings that did not clean up with the NRS, then p120 all over followed by p150, p180, p240, p320, p400, p600, p800, p1200.  I will confess that when I got to p400 the first time I realized that a had a deep scratch that was not going to come out with high grits.  So I did some spot sanding with p180 to get rid of it, then worked my way up again.  Scratches like that are almost always the result pressing too hard with a coarse grit of sandpaper, but I find that they are often difficult to see until the rest of the surface gets smoothed out.  This is not the first time I have had to go back grits.  I keep re-learning this lesson that sandpaper is to be brushed or even wiped, but not scrubbed.  

Here's what it looks like, and with a MS wash.

 

[Mark J]

Yesterday I began working on removing the bottom half of the inside. First step was to rearrange the turning set up. The tailstock was now in the way of the work at hand and the articulated arm was better positioned directly on the lathe bed than on the standoff arm.

I didn’t mention earlier that, as many of you already know, I prefer using handle-less tools, and when I can, I mount them in my articulated arm hollowing rig. Most of the tools that I commonly use need to be positioned on the center line of rotation (zero altitude) and with zero pitch and zero roll. The articulated arm does that perfectly, way better than a human. It frees me up to concentrate on the tool's contact point and movement.

In this photo you can see the bottom face of the blank marked with two circles. The outer circle indicates the inside margin of the foot. The inner circle marks the approximate diameter of the narrowest point of the base (waist).

Once I carve away these pencil lines I will have no reference mark, so it’s important not to remove them too soon. In the case of inner circle I am going to excavate out a cylinder of that diameter to the depth of the waist of the base form. I start out with the radius square tool which does make a crisp margin, but is a bit slow at bulk removal.

Then I remember I have this tool.

The cutter on this tool is also carbide, like the scrapers, but in addition to being small it differs by having a sharp up turned rim, something like a pie pan or cupped shape. Positioned at an angel like this it is acts as a shear cutter rather than a scraper.

It is capable of delivering a smoother surface than the carbide scrapers, but here I am more interested in speedier removal of material.

The goal is to make the cylinder a bit deeper than than the waist, without actually running into the woodworm screw. A quick check of the depth shows two inches which you can see from the diagram is deep without being too deep.

Using calipers (careful not to scratch) I can find the distance of the waist from the foot. This distance can then be transferred to inside surface.

Looking back at the previous posts I may not have made it clear that I am making the waist a bit thicker than the remainder of the base in order to increase the strength at this naturally weak point. Here is the cross-section of the base wall for illustration

 

t will be slimmed down later, but at this stage I am looking for the wall thickness at the waist to be on the plump side of ¾”. What I have is morbidly obese. You can see that I haven’t turned down to the pencil line in addition my cylinder has ended up a little cone shaped which accounts for the increased dimension.

So break out a radius square tool and expand and straighten out the cylinder.

That’s better

Replace the internal mark

Now I’m ready to start removing wood from the remainder of the bottom creating the inside curve of the base.

Eventually I found that in order to define the waist I needed to remove some material from above it.

Ultimately the time comes to reduce the thickness of the waist and shape it further, which removes my pencil line. But I have a good sense of where it should be, and I can also pinch the outer and inner sides of the waist to get a good feel if it is staying centered.

The goal is a basic wall thickness of ¼” so we’re still a bit heavy. And the surface needs further fairing.

Time to bring back the NRS tool.

At first I used the NRS where the articulated arm was set up. But it’s a different tool and needs more “elbow” room as it were. So I had to take the 15 minutes to remount the articulated arm on the standoff.

But with that done it was only a short bit of work to get the curve fair and the wall thickness down to .28” With sanding yet to go, that will do nicely.

For sanding I will probably start with p120 to tweak the curve a little bit more, but I am very happy with the form. I won’t regale you with the sanding saga, but doing it will take time and I won’t be able to start on it tomorrow, so it may be a few days before our next installment.

Things are going well, but the ground I have been traveling thus far has not been particularly favorable for an ambush by the forces of poop. I have a few question marks to deal with coming up so it’s not bad time to have a day off and a day or two of sanding to think.

One issue is that the foot has come out a little smaller than I had planned. Despite meticulous planning and precise measurement, reality always deviates. As an example look at these two opposing sides of the bottom of the base. There is a minimal but real difference in the width of the rim.

They should be equal, but they’re not. Error just creeps in (I think it may be one of those pesky laws of thermodynamics). Hence, I try to be as precise as I can so my mistakes will be more accurate.

Back to the foot being smaller. The wider parts of the foot were supposed to come out to a width of 7/8”, but the reality is 5/8”. This carries no aesthetic consequence, but remember I will be gluing a block of poplar to the foot as a temporary mounting for the next phase. My original calculation was that I’d have about 10 sq in of glue surface. I think the reality is closer to 7. That’s a goodly amount of glue surface, but the wood of the foot is quite thin, and if the glue joint fails it will certainly necessitate a lot of crying. The real risk to the joint is the sudden jolt of a big catch. I should be able to avoid a catch with my techniques, but I could reduce the risk further by excavating out more of the top half of the base now while it is securely on the woodworm screw. But, there’s a “but” on that, too. Removing that wood now will reduce the mass of the top half and I’ll get more vibration when I go to remove the rest and thin out the upper walls and rim. Gotta think on that a spell.

There are some decisions to make with the sacrificial glue block, too. Do I go with round or square? Rounding off the corners makes it safer to work around (I already nicked my knuckle when sanding the base, twice in the same place), but if I make it square, the same size as the original butternut blank and line it up accurately when I glue it on, then its corners will be a guide as to where to line up the corners of the cutback template. I think square is clearly the way to go, but I’m just doing an imaginary walk through of how I would do it.

[Coop]

Your shop and tools remind me why I’m not a turner but, admire you and those that are! 

[Chestnut]

Before this thread I was pretty happy with my turning set up and tools, I thought it was pretty good .... boy how naive I was.

The articulated arm leaves me a bit confused. I get that it doesn't have a handle and the reason for it, but how do you engage with it to make the tool cut. It needs some sort of handle for you to push the tool to engage the cut?

Also what RPM are you running? Not cause I want to try and replicate just trying to understand order of magnitude. If it's in the tens, hundreds, or thousands.

[Mark J]

9 hours ago, Chestnut said:

Before this thread I was pretty happy with my turning set up and tools, I thought it was pretty good .... boy how naive I was.

The articulated arm leaves me a bit confused. I get that it doesn't have a handle and the reason for it, but how do you engage with it to make the tool cut. It needs some sort of handle for you to push the tool to engage the cut?

Also what RPM are you running? Not cause I want to try and replicate just trying to understand order of magnitude. If it's in the tens, hundreds, or thousands.

As to how I hold the articulated arm, I usually have one hand on the "wrist" and the other on the tool   It only takes finger tip control.  Both of my hands end up controlling the motion, but it is possible to do one handed.

As to rpm, I have the presets set at 100, 300, 600, 900, 1200.... I was doing most of the turning, inside and outside, at 900; some 1200 for the interrupted work on the outside.  I only occasionally go more than 900,  and do a lot at 600.  But it depends on the work being done, the materials, the tools.  

[pkinneb]

Mark thank you so much for taking us through some of what goes into this! Like @Chestnut blown away by how you do these! 

[Mark J]

I'm glad you're enjoying it.  I wish there was an easy way to take some action video vignettes while working solo, but using the still camera one handed with the lathe off is hard enough.  

[Mark J]

I have a quandary which y’all may be able to help me with. First let me get you up to date.

I was busy yesterday with distractions, i.e. any life task which is not woodworking related. But this morning I finished sanding the inside of the bottom. Here it is all purdy and smooth and then after a refreshing mineral spirits wipe down.

The surface looks good, but I realize that, with sanding, the inside rim line has migrated every so little towards the periphery, narrowing my feet further. I should have seen this coming; it was entirely predictable.

What was once to be 7/8” is now 4/8”, and the sides of the rim are now contributing next to nothing to the available glue surface area.

Still with the feet measuring about 3” in length that comes out to about 6 sq in of glue surface, which I think will be enough. But it does reduce my margin for error.

I still have the opportunity to remove material from the top half while it’s on the screw chuck, but I think it’s worth keeping the mass to dampen vibration, so my plan is glue on the sacrificial poplar block and go easy on shaping the top.

That leaves me with one quandary that I also should have seen coming. In the photo below you can see a clear MS margin identifying the boundary of the curving inner wall with the flat foot.

I’m short on glue surface so I can’t afford to starve the joint, which means I’ll get squeeze out along this boundary. I won’t have access to this side of the glue joint until long after the glue has cured. Eventually, after the cutback operation, I intend to use a small flush cut saw to saw the basin’s feet off the sacrificial block. Undoubtedly I will have some residual cured glue on the inner wall which I will have to remove. And by the way I plan to use either TB I or TB II.

So what to do here? How to keep the glue from bonding outside of the joint area, or make it easy to remove after the joint is sawn apart? All without changing the  appearance of the finish.

A couple of thoughts have passed my mind. I could use hide glue, which can be loosened with heat. But that doesn’t mean there won’t be residual glue in the pores that will change the appearance of the finish coats. And I don’t think that it has the same strength as TB?

Maybe shellac would keep the glue from bonding? (no idea, you tell me). If so, once cut free, I could set the base in a dish of DNA and hopefully dissolve and wash away all the shellac? (never heard of doing this, I just made it up).

On the plus side, the wall thickness came down a smidge.

[wtnhighlander]

This is totally a WAG, but I would try marking the boundary of the curve, and masking it with tape before gluing. That should leave little more than a hairline of glue the should easily shave off with a sharp chisel.

But bear in mind, most of my work is thick and meaty enough to clean up with an angle grinder.   I have no experience with anything this delicate.

[Mark J]

According to Bob the Glue Guru at Titebond, TB does not stick to wax (but wax is then hard to remove), so we thought it was possible that TB wouldn't stick to the Osmo.  I'm doing the experiment now, but the OPXO will have to cure over night, and then after I apply a bead of glue that will have to dry.  So there's going to be a "rain delay".

The problem I was seeing with just using masking tape is that it would be difficult to tape right up to the line since it's a compound curve.  But not only could I use a razor knife to trim the excess tape, but any little bit that is left over the line might come off when I sand the feet flat.  The last step before gluing.

If the Osmo experiment works I may end up trying both.

 

[Chestnut]

I'm a bit late I'd bet but my advise is to finish the inside with what ever your final finish is going to be. It'll give you some barrier to keeping the glue from absorbing into the wood and should make it easier to remove. Just assume you'll have to touch the finish up. I like the idea of both finish and tape though.

[Mark J]

I did the experiment. Prepped a scrap bit of butternut to p1200 and applied some beads of TB Original and TB Hide Glue. I applied some of the glue when the Osmo PolyX-Oil was freshly applied and the rest after the finish had cured for 12 hours. I let the second dose of glues cure overnight.

And the results of this experiment is:    Nahgud .  All of the glue applications are tightly adherent to the wood. I was able to get some of the hide glue to pry off with a chisel, but this left obvious damage wood under neath.

I also tried using a bit of water to dissolve the hide glue. A little water did nothing, A lot of water dissolved very little, but just enough to spread a dilute solution of glue around the bead, that later left a film.

Oh well, now I know.  On the good news side, the masking tape idea worked extremely well—at least as far as applying the tape. I won’t know how well the masking tape protects against squeeze out till later.

First, while my glue and Osmo experiment was curing I made up the sacrificial block. I decided to go with BB instead of poplar, because the plywood was near perfectly flat and my poplar has a shallow cup. Poplar is much easier to turn than BB, and you can flatten the cup by turning the piece on the lathe, but I don’t plan on turning any of the sacrificial block and so I was able to avoid this task.

I carefully measured the width of the base piece and cut a square of the BB the same size using the Incra jig. Then I penciled in the diagonal lines, front and back, as they may be necessary later for aligning the cutback template. Last, I drilled a 5/16” clearance hole for a screw chuck.

Now for taping off. I realized that I did not have to tape off the entire circumference of the base. Much of this wood will be removed with the cutback, so I only need to worry about the areas where the four feet are going to go. This is good, because it obviates the need for any piece of masking tape to be applied in overlapping manner with any other piece of masking tape.

To make the demarcation line more obvious I made a red pencil line about 1/32” peripheral to the margin. Here I also marked out in red what I think will be the approximate size of the foot, so I’m taping off way enough. Some scribbling with regular pencil so I know when I’ve sanded enough.

I positioned a single strip of tape such that I’m just at the top of the red line in the center of the arc. Put a nick in the bottom edge of the tape as a strain reliever, and stuck it down. Then carefully trimmed back the excess tape using the razor knife and the red line as a guide. Note that all the cutting is occurring over the foot and away from the sanded surfaces.

I took another bit of flat plywood, and using spray adhesive put on a sheet of p120 sandpaper. This is the sanding surface I will use to flatten the bottom of the base preparatory to gluing on the sacrificial block. Hopefully this will also remove the remaining excess masking tape. For this operation I use a large face plate as the flat supporting surface for sanding board. The face plate is mounted to a live center (i.e. it rotates freely) and this is mounted in the tail stock.

The sanding board is then sandwiched between the target and the face plate. Note also that the sanding board is sitting on a block of scrap wood to give it adequate height. I used to attach the sanding board to the face plate with screws, but this is simpler and faster. Moreover, this allows me to shift the sanding board around (left & right and up & down) to expose fresh areas of sandpaper.

Sanding is by hand rotation; this is a motor off operation. The loose sanding board can decide it wants to move, particularly if you are too aggressive when advancing the tail stock quill.

I was concerned that the sandpaper would tear off the masking tape, but you can see here that the that the masking tape is abrading nicely showing a clearly defined line of demarcation. And ground up masking tape on the sandpaper.

At this point I re-trimmed the tape with the razor knife, then finished off the sanding. The results were great.

Last step for tonight was to glue the sacrificial block to the base. I want to give this at least an overnight cure before I try to turn the piece again. Mount the sacrificial block and chuck.

Trial fit is good. I’d say I’m within a 1/64”. I need the sacrificial block to be both a perfect square and perfectly lined up with the base, as the sacrificial piece will be what I reference off of when I eventually attach the template.

Then the glue up.

Sorry no action shots there, you know once the glue is out you want to keep moving. Alignment was carefully done, and exterior squeeze out was wiped off with a damp rag. Although I thought I was reasonably generous with the glue, I did not get much squeeze out on the outside. No going back at this point, but I think it will be good.

Next up is turning the upper half of the inside.

[wtnhighlander]

Glad the tape held up to sanding! I was worried it would peel, when you first mentioned that.

[Mark J]

Having let the glue cure overnight, yesterday morning I backed off the tail stock. The glue seems to be holding .

After removing the piece from the first chuck, and taking off the live center, the second chuck can be put onto the spindle.

Then set up for turning. For this bit I don’t need to use the standoff and can mount the hollowing rig directly to the lathe bed.

Since the standoff is a double layer of ¾” ply this has always necessitated a significant vertical readjustment of the hollowing rig. Which is something of a pain. While cutting the plywood for the sacrificial block I realized that all I had to do was make a “stand up” pad for the hollowing rig to obviate this adjustment. I hate to tell you that it has only taken me three years to come up with this “genius” idea. But hear you can see see the “stand up” and the standoff compared.

Getting started

Progress.

Starting to get curious as to where the bottom is

Very curious

It can’t be much deeper

It wasn’t. I just touched it with the tip of the diamond tool and pop out came the middle.

A little more bulk removal, then some smoothing with the double bevel carbide scraper.

Then pull out the negative rake scraper for the final shaping and smoothing. I have bought at least half a dozen different tool rests to cover a variety of approaches. I haven’t used the clunky OEM rest in a couple of years. But wouldn’t you know that was the only one that would work here, and in fact worked well.

Work the contour just a little bit more and I’m right on the mark.

And not a moment too soon; here comes the inspector

I’m ready to start sanding, but this includes the last opportunity to adjust the contour. The wall is thickest at the waist and the inner contour has a bit of a point to it. I’d like this to be approximately centered on the waist. This is a little difficult to determine by feel, so I came up with this. Using the tool rest I hold a bit of plywood scrap against the top rim and measure to the middle of the waist (being careful not to scratch).

Then use the calipers and the same board to measure down to the point. As far as I can see no adjustment of the position of the point is needed. Sure hope I’m right.

Next step for the top is sanding. When then story continues I’ll be working out how to make and attach the template for the cutback. I have a plan. After that we’ll actually do the cutback, then some carving, followed by more sanding. Leading up to sawing the sacrificial board free. And then of course there’s the basin. My what a lot of work there is to do.