Woodworking Talk banner

Hard Problem: Seeking Guidance

4628 Views 26 Replies 6 Participants Last post by  kiwi_outdoors
Nothing like a challenge. Sorry for the long post. There's a lot to this.

Mine seeks to recreate this table:

My attention focuses for now on the table's circular skirt. Most of the other stuff is done, or I know can be with my skills.

It needs to be 3/4” thick, 5” high, and with an outside radius of 43.” Anyone familiar with the formula for a circle's circumference (2 * r * pi ) would find requiring just short of 12' of board length, more like every so slightly North of 11 ¼.'

I've come to the conclusion that there are basically three ways to do this. Perhaps you know of others.


The first involves building my own steam bending unit and form, and bending the wood into shape; attaching it with brackets at the two loose ends.


The second involves cutting circular rig segments out of plywood and lying them out on flat surface in a circle, and then cutting the next level's circular segments, placing them on top of the first level, staggering them from the first level's segments while gluing the two levels together to form the first two levels of this ring. An example, even if for another table design lies here:


Additional levels of the ring can be created this way, each slightly staggered from the prior until the 5” height is achieved. Routers and circular jigs are indicated to smooth out the inner and outer curvatures of the ring.

This involves a lot of time and waste; waste being the byproduct of many circular wood projects people take on, that unlike rectangular ones, grows exponentially in size with radius, as opposed to rectangular ones which grow as multiples.


The third method I see involves kerfing the wood at proper depth and distance between kerfs for the bend, and forming it temporarily around a jig of the small table top so it doesn't break under its own reduced kerfed strength, and then gluing the kerfs with expanding glue to build their strength up, only released it from its form thereafter.


One more thing. The inner surface of the ring has an appearance like the inside of a hex nut as its threaded. This allows it to spin and go up and down like the video.

In prior rings that I made too small and had to discard (using method two) I was otherwise able to successfully make this threading using this stuff:


This go around I'm interested, while the skirt is still a flat board to router out a straight line (with a home made fencing jig) on what will become the inside surface of the ring, that gradually travels up the 5” height of the board for its near 12' duration, then bend the wood using techniques one or three, to make the screw threads.


As this skirt holds up the extendable table leaves and neither side of the ring will lean up against any support structure (think unlike the outside of a wood guitar up against an interior form that lies inside the instrument) it needs to be strong when circular and parallel with the ground. I'm hoping once glued, particularly with expanding glue like Gorilla Glue this is will be the case.

Has anyone thoughts on this, from another approach to wood bending services North of NYC?

See less See more
1 - 20 of 27 Posts
You need to read this!


See the article on "Expanding Round Table." it's a link in the "Large Round Table on the left side column, Projects.

Large Round Table
A very complex diningroom table

See less See more
  • Like
Reactions: 1
The numbers bother me. A radius of 42 inches gives a diameter of seven feet. Below assumes a diameter of 42 inches.

In effect you have to make the inner circumference of the circle about 5 inches smaller than the outer of the ¾ inch board.
If you kerf bend and make your cuts ½ inch deep and about every two inches you should be able to create the circle that you want. That is what the math says.
Be sure to draw a line every 2 inches as a guide to cutting. Use a SCMS or a RAS with the depth of cut restricted to ½ inch.
I'm not sure how to tell you to join the ends of the apron. It is possible that pocket screws might be the best and mitered ends of the ¾ at 7.5°.

BTW - If you don't have a sliding compound miter saw, this project would be the perfect time to get one.
Thank you, "NoThankyou" for taking the time to reply.

I misspoke about the radius of the table being 43" (you having quoted 42"...but not a big deal at your end.)

I meant to say that its diameter (i.e. twice its radius) is 43".

According to my math, which perhaps is wrong, the circumference of a circle is d * pi, where "d" is diameter. That would put the circumference just North of 135", or just North of 11 1/4. ' Certainly the inner circumference, with a diameter of 42 1/4" would be smaller.

Were the table actually 43" in radius like I said, I think--but may be wrong--e'd be looking at a table whose skirt circumference would lie just North of 22.5.'

I would join the skirt's free ends with a mending plate, likely bent slightly from its flat purchase dimensions and attach it to both loose ends, outside, inside or both sides of the circumference.

I actually watched a 4th technique on Youtube yesterday of gluing together strips.
. It seems promising in that skewing the strips, which in my case may encompass the entire 360 degrees, not the 180 of the video, each level skewed from its inner level's break point between the loose ends, and glued together, to both maintain curvature and integrity of the ring.
See less See more
Because your skirt is floating and unsupported by not being connected to the main field of the table, your best method is to kerf-and-fill the curve. This will be more rigid and hold shape better than a bent lamination.

For the fill, do not use an expanding glue. Get some resin used for fiberglass construction.

For the kerf cuts, you want to use the thinnest blade you can. I was able to do it with a bandsaw and a vertical backstop fence to control depth. With your skirt being taller, you might consider getting a circular trim saw with a high-speed-steel blade, so the kerf is nearly as thin as a bandsaw.

Joining the open ends:
The best method for joining the open ends to form a complete circle would be to use Dominos (or similar). But the problem is that your unfilled kerfs wouldn't be strong enough yet. So for this, form your complete circle in your jig, and fill only the first few inches worth of kerfs to either side of that joint. Clean up the resin to near-finish, and then plunge and join the boards with Dominos.

Because your skirt is fairly tall, make sure your circle forming jig used for filling the kerfs supports both the top and bottom of the skirt. Otherwise it will be very easy for the top and bottom to not be identical bends. (Although less of a concern because you have a closed circle.)


See less See more
  • Like
Reactions: 1
Thank you Rick for your sagely advice.

I think I'm going to go the kerfing route to achieve this circular shape, and if I do I will definitely incorporate as many of your suggestions as possible.

Under the notion that "brilliant minds think alike" ; - ) I envisioned making my circle forming jig as two 43" diameter disks, held apart by a bunch of well placed nuts and bolts, that insure this distance remains constant and the same across all measurement points between the disks, somewhere about 3" in distance.

If a visual is clearer, I would adopt Jimmy's (DiResta) approach here, only with the aforementioned two disks, not one, and (of course) not permanently attached to the ring:

Then again, approaching Jimmy's 2nd method of making a ring at 3:15 seems interesting too, even if the height were built up in stages.

Given that I would hope the resin would bear most of the responsibility for maintaining the shape, I'd probably be looking to cover as much as the ring's kerfs (outside my jig) with this liquid before detaching the jig and doing the rest of each curf within the jig. Once complete I could apply your great suggestion about joining the two loose ends (invisibly) with dominos, so these dominos bear far more responsibility for keeping each loose end vertically aligned with the other, than "in the business" of resisting ring expansion (the resin's job, along with strength/rigidity.)

The thin blade idea is strong too. I have access to some kerfing calculators.

The last time I tried this I ended up with a hot mess. Wood, unlike MDF, need I tell you, is less consistent in strength along its length, and a 12' board, pushing the limits of its own strength to remain intact when cantilevered during the build, is that much more susceptible to breaking once kerfed. I was thinking of kerfing on ground, above a rachet strap, and then placing the jig on top and rolling/securing the kerfed board to the jig. I'm fine to resin portions of the ring each day (outside the jig) until complete, that are closest to the ground as I roll the whole contraption. I think I'll need a lot of mixing cups as I can't do this as a one shot deal.

I'd guess more evenly spaced kerfs from thin blades that are less deep is the approach most resistant to breakage(???)

My first successful ring, (although too small) was done with the plywood method of approach #2 above.

Now to create a jig to router out a track from the inside of the assembled ring that rises a fixed distance with each equal partial turn of the ring, so as to make "threads" for the ring to rise and lower on with its spin in each direction.
See less See more
I was thinking of kerfing on ground, above a rachet strap, and then placing the jig on top and rolling/securing the kerfed board to the jig.
There are 2 positions where your kerfed board will be most susceptible to breaking, and this method includes one of them.(a standing circle) So I don't recommend it, even with your jig partially supporting it.

The second position is when it's standing on-edge, but not round yet. In this position, the board will want to lay flat again, so the unsupported end will twist and break. However, this is easier to overcome. With your kerf laying flat on the floor, either slide another board under it (preferred), or tape another board to the top of it in several places. This will stop the kerfed board from twisting while you tip it from flat to upright. Once upright, you just need to bend it enough to stand upright on its own. So cut the tape and get it curved into a horseshoe.

Once it is upright and can't twist, it is very stable to work with, even without any type of form controlling it. From here, get it onto a board large enough to use as your forming jig. Like 48x48.

I envisioned making my circle forming jig as two 43" diameter disks....
Instead of making a disk-form on the inside, make your form on the outside. The beauty of this is that it can be made from a bunch of smaller pieces that you tack down to the 48x48 jig board I mentioned above. (Also, if you use screws to hold them down, you can remove parts of the jig while working, such as when Dominoing the ends.)

By having the form on the outside, you can epoxy resin the inside without interference. just use some wax paper or masking tape to prevent the resin from bonding the kerfed board to your jig.

Once your non-filled kerf board is secured to your jig with a couple of clamps or even tape just to hold it down to the 48x48 board, now you can tilt the whole thing near-vertical (say 75 degrees), and apply resin to the first few inches to either side of the main joint. OH! don't forget to put some wax paper between the ends. You don't want to accidentally glue them together yet.

After the resin sets, Domino the joint, and glue it together.

Get it back into the jig, and resin the rest of it in a similar fashion. Tilt the whole jig while you work in each area.

P.S. Contrary to what I said before, you won't have much risk of the cylinder being skewed. That's because you are first closing the loop with your domino joint.

I have access to some kerfing calculators.
Because you are filling the kerfs with resin, don't bother with a kerfing calculator, except to confirm that you have more than the minimum. But you're going to test your kerfs long before that anyways. (see below)

I'd guess more evenly spaced kerfs from thin blades that are less deep is the approach most resistant to breakage(???)
Yes. Here's how I would start out. Take a scrap board of the same material and same width as your actual board, and begin making shallow kerfs about 3/4" apart (or 1/2" if you feel more comfortable). Then see if the board bends smoothly at the desired radius. If it is too stiff, increase the depth of the kerfs. You want the kerfs as shallow as possible but deep enough to bend without too much stress. At 1/2 to 3/4" apart, you already know you have enough kerfs for the curve, so that isn't a variable you need to worry about.

I think I'll need a lot of mixing cups as I can't do this as a one shot deal.
Resin doesn't stick to polyethylene (HDPE), so the rectangular cup on the top of the resin can is reusable. You just break the hardened resin out of the cup....it's a rather satisfying thing to do. :vs_laugh:
Edit: P.S. Standard plastic sheet or bags are made from polyethylene. So you can use that instead of wax paper.
See less See more
By the way, I always just bought my resin from automotive stores, but I was surprised to see that HD and Lowes apparently carry it too.

See less See more
So that everyone understands the math but based upon 42 inch diameter.

Using ¾ inch wood, the inner diameter is 1½ less than 42 or 40½. Circumference is then just a hair more than 130-3/8.
A saw blade kerf, 1/8 and cut ½ deep and bending the wood creates a isosceles triangle of space with the included angle about 15°. There is room for 60 cuts in the apron. But only 51 or so are needed of the spaces are fully closed to form the isosceles triangle. Put the 61 cuts in with 2 inches between cuts.

The 7.5° mitered ends comes from half the included of the isosceles triangle.
From someone who has done it ......

That kerf is way too wide. You'd end up with a choppy form.

Rick has the experience you need for this project..... :smile2:

The result will have small facets between the kerfs unless you use a fine kerf blade or a bandsaw such as Rick used.
Rick, woodnthings and NoThankyou, please accept my thanks for taking the time to add color commentary here, especially you Rick.

I must say, I am a member of several forums concerning various interests I have, from cars to music performance and far and away this one has the nicest AND most knowledgeable people by far compared to the others.
  • Like
Reactions: 1
Rick, others:

...another question if I may.

As I indicated originally, and as can be seen by in the initial product video above of the table whose skirt I seek to copy (BTW: this table has no IP rights associated with its duplication-I'm sensitive to that) the skirt is designed to rise when spun (unlike a standard hex nut) in a counterclockwise direction, and lower when spun in the opposite direction.

(Threading with standard righty-tighty threads, raising the skirt might risk closing the leaves, which open with clockwise expansion.)

In the past I have created the "threads" for the skirt to behave in the reverse threaded hex nut like behavior I describe above by gluing a flexible dentil molding product (linked above) to the inner circumference. The idea has worked, but such protrusions from the inner wall of the skirt, towards the table's center, take away, if so ever little, from the space under the table top's base to store the leaves.

(To make a long and tangential story short, the optimal hinge point for these leaves to open from a concealed 8 node star under the base table top is a mathematical optimization that endeavors to maximize both the leaves' size and ability to store under the table much like a camera lens--but I digress.)

This time around I am interested in creating the "threads" by routing them into the inner wall of the ring; so rather than having the threads as protrusions from the inner wall until the aforementioned dentil molding , the are depressions in it; I'm thinking about 5/16" deep.

I'm also thinking that prior to even kerfing the flat board that will become the skirt once kerfed and curved, that I create this threading in what will become the inner wall of the ring. I can lie the board on the ground with its longest dimension facing North-South, with the skirt's top (when compete and properly oriented) on the East side, and create a straight line indentation in the board for its length, with my router and a quick fencing setup, that starts the router close to the East side of the board's 5" width on its South end, and works its way to the West side for the approximately 11 1/4' duration, in a Northernly direction.

(If this is conceptually difficult to follow and/or poorly explained on my part, imagine placing the board right above the X axis of graph paper in an orientation that it never exceeds 5" of height (the skirt's height) on the "Y" axis. My cutting slope, y = mx + b where "m" is the fractional rise over run that begins at the graph paper's origin (i.e. b = 0 ), would be around y = 5"/11.5'x , or y = 5"/138"x, or y = 0.036x (all the same mathematically))

Maybe I should wait to create this depression based threading until after the ring is made, but I've got to think this a harder routing job than that described above, working with the board in its straight form prior to curvature. Perhaps I should still to use of the dentil molding so as to not further compromise the board by removing even more "meat" (and strength prior to epoxying) than kerfing it will already do.

If you have any thoughts or wisdom here it would be appreciated.

One more thing, I own a band saw and I think you suggested its use, perhaps in lieu of a circular trim saw, along with perhaps some method of fencing that would control the cut depths in my work piece.

As my work piece would be on its side, and would require transfer to the jig once cut, and in its highly fragile state, I'm I doing these cuts with an uncut board that backs my work piece: perhaps the working piece taped to it?

See less See more
Why don't you show me how your skirt interacts with the rest of the table. I haven't looked into this type of expanding table before, but I can probably come up with a simple but elegant solution for you.
Kerfing a board 11 ft + long

If I needed to do this, I'm using a 7 1/4" thin kerf circ saw blade in my radial arm saw with an extended table on either side of the blade. This would make 1/16" wide kerfs from the top or visible side of the board. An indexing metal "blade" could be used for uniform spacing away from the blade, much like a finger jig on the table saw. As you make each kerf, you drop the thin lever from the fence into the previous kerf and create the next one. It would be set one kerf space away from the blade. It could be a loose metal blade or hinged for convenience ....
A bandsaw obviously won't work because of the throat depth limitation. A table saw wouldn't be very workable, working upside down and no long support on either side. The RAS seems to be the best choice, at least in my opinion. A support board could be used underneath the workpiece on the RAS with the blade raised a suitable dimension to accomodate the additional thickness.

This is quite the project. :surprise2:
See less See more
Fair request sir. I'll take some pictures and post them. Thanks.
There was a lot of trail and error "woodnthings" but most of the table is done.

IMO, this below is quite the project, and on my bucket list. At least this Jupe table spinoff comes with instructions:

https://mechanicallumber.com/plans-for-sale/ : - )
Are there other videos of the table you're trying to build, beside the one you originally posted? In other words, are you designing from scratch, or have you gleaned ideas from other sources?

IMO, this below is quite the project, and on my bucket list. At least this Jupe table spinoff comes with instructions:

https://mechanicallumber.com/plans-for-sale/ : - )
I watched a couple of his videos this morning when I noticed your link to them in your first post. He said he spent years developing his design.

I didn't have that luxury. My first Jupe table was designed in 2 weeks using Win/DOS-based 2D CADD. The more famous one was done in a week because I upgraded to SolidWorks.

The current owner found me a few months ago, after years of searching, to troubleshoot why the table wasn't working. After weeks of preparing to upgrade the electronics from the other side of the country, some connector got reconnected, and the table started working again.

(I hate this video because the auto-volume went too high. You really can't hear the gear box, like it seems. That was a $500 gear box...it was definitely quieter than the video makes it seem.)
See less See more
Hey Rick:

Thank you for taking an interest in my wood working journey.

To the best that I have been able to acquire on the internet there are only two more videos of the table whose skirt is the subject of this thread.

While I would be glad to link them, I can assure you, having starred at them in slow motion many times, they glean little to no additional information about how one might go about constructing this table and its inner mechanics. Fortunately I think I've figured out most of the parts, and when time finds me able (hopefully later today,) I will take pictures of what I've done so far and you will see much of my mechanics revealed from "behind the curtain," and develop a better understanding I believe that words alone make it difficult for me to convey.

I made this table years ago in its non-expanding form with a skirt too small in diameter to accommodate reasonable sized expansion leaves. Time has me now revisiting this project in the hopes of completeing it. The skirt was done with plywood, or "option 2" in my initial post.

As it regards the separate Jupe like table of Scott Rumshlag, yeah, I think he's quite brilliant. Then again, that you were able to formulize your own version within a matter of weeks is pretty outstanding--especially given the relatively limited software tools available at the time you made it compared to today's 3D design software. I enjoyed watching your video, which seems to incorporate a much simpler to understand (and implement) design, but by no means "simple"...very impressive.

The current Fletcher-Capstan implementation of the original Jupe table, although virtually impossible to DIY, is quite impressive, featuring, as you may know, optional electrically mechanized movement and aluminum honeycomb made table leaves with wood veneer so as to be lighter and more dimensionally stable than wood.

Thanks for sharing.
See less See more
Here is the full table in perspective view, in its smaller size configuration:

It is 43" in diameter.

The skirt is not as large in diameter as the table top, and needs to be redone and made bigger to lay flush with the table top's outer lip so as to accommodate the maximum size extension table leaves folded up inside the base table when not in use. This will become clearer in a moment.

This is the unfinished and uncut ring that will form the eight extension leaves:

Circle Architecture Wood Mirror Window

This piece's inner ring has a diameter of 43" and an outer diameter of 61" for a total of 18" larger, or 9" of radius larger. It is only 1/4" thick because such thickness affects the ability for the leaves to store under the table, as they sit in large part on top of one another, overlapping, like a camera lens.

The next picture is with the table top and skirt removed:

Table Wood Wood stain Plywood Furniture

The inner circle is fixed to the table base. The outer ring pointed to, not to be confused with the extension leaves ring shown prior (that has yet to be sliced into 8 equal pieces) sits on top of 2 boards in an "X" configuration, and can spin freely in either direction. One of those 4 prongs of the "X" is above my finger. I point to a dowel that was cut down on the ring because it was misplaced. In a working form of the table, 8 equidistant dowels will emerge from this ring and point to the sky about 2" or so in height. These dowels help open the leaf extensions when this ring is spun clockwise as looked at from above. This becomes clearer in pictures below.

Here is a picture with this ring removed so the "X" pattern can be seen.


The circular table skirt rests on top of this "X" pattern, rising with a counterclockwise (as seen from above) spin of the table skirt, and lowering with a clockwise spin.

In the next picture the too small circular table skirt and its "threads"--made of the aforementioned glued on dentil molding ride on top of this "X." This molding allows the skirt to raise and lower with its being spun. This dentil molding, in protruding from the inside surface of the skirt, takes up precious space to store the expanding leaves. I would rather effect this threading by routing into the skirt, either while a still a straight board before kerfing, or after. This is the essence of my last post that inspired request for these pictures.

Equally important, the threads resist downward movement of the skirt, but less so, albeit less important, upwards movement of the skirt towards the sky. Threads routed into the inside of the skirt would resists up or down movement of the skirt.

Here is the table skirt spun counter-clockwise and as a result is lower to the ground. This lower position is necessary to be able to extract the leaves from within inside the table, and then raise those leaves to the height of the base table top by spinning the skirt clockwise once those leaves have been extracted.

Wood Barrel Arch Circle

Let's now take a look at the internal hardware that makes this work. The following is an 8 point star whose center lies in the center of the table.

Wood Plywood Vehicle Wheel

Near the end of each star point is a bolted hinge point with a folding up arm with its middle section cut out. This middle cut out of each arm sits on its own dowel (of 8) mentioned above that have yet to be installed on the internal ring. As the skirt is dropped and this ring spun by the owner clockwise, the dowels turning clockwise open the arms.

The position of the bolted hinge point is a complex optimization exercise. While all 8 bolts are equi-distant from the center, there is a distance window where the largest leaves can still fit under the table. Make the hinge point further out or closer to the center, and the leaves that will fit under the table when stored must become smaller to fit.

When these arms are folded up they are roughly 20" from center as shown:

Wood Plywood Tool

This size fits under the table top, with a radius of 21.5" but not the current skirt, which needs to be widened, and why this thread first began, and how best to make this circular skirt.

The last picture shows how 1 of the total of 8 ring segments fit under the table, featuring a temporary cardboard ring segment for measurement purposes:

Product Wood Wheel Architecture Plywood

These ring segments fold somewhat overlapping on top of one another, again, like a camera lens. The leaves can only close as far is the small disk in the center of the table that keeps the base table top elevated, even with the skirt is dropped, permits them to.

The basic principle to expanding the table is to drop the skirt by spining it clockwise, exposing the leaves by spinning the ring clockwise, and then spinning the skirt counter clockwise to raise it and the leaves, (along with the arms and star inside the table) to the base table top height.

The process is reversed to fold away the expanding leaves.

I hope these pictures help for a better understanding and my desire to router the new larger skirt's threads.


See less See more
1 - 20 of 27 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.