End Grain in Segmented Bowls

by A.J. Golichowski

End grain is the bugbear in turning but if it is featured in can produce an interesting and good-looking bowl. I learned to look at and for nice end grain while operating a radial arm cutoff saw in a cabinet shop. 

While browsing through the American Indian Basketry issue of Arizona Highways Magazine (July 1975), it occurred to me that end grain could be used to suggest basketry materials.  As a project I chose the polychrome olla on page 19 of the magazine.  It was only after starting the planning that I realized that it would require 1392 small pieces of wood to create the olla!

To manage so many segments I developed a process that makes it feasible to handle such a quantity of small cross section pieces.  This method is also a great way to use up those odd pieces of wood that are too nice to throw away. 

For instance the dark wood group of three, Photo #1, is made of walnut Korina and mahogany three quarter inch ripping waste.

The oak pair, Photo #2, is made from cut-offs of three different colored boards of random red oak selected during cutting operations.

This technique works equally well in projects with fewer and larger segments.  Bowl #3 was adapted from an African vessel to use up some zebrawood and mahogany scraps.  The end grain panels were designed to show off the interesting end grain found in some Japanese shipping crate lumber. 

Because of the complexity and the care and precision required, it became a long and tedious operation.  The process in the case of Bowls #1 and #2 begins by imagining the pattern as being on the surface of a cylinder as in Photo A.  It is plotted on ¼ inch cross-section paper and broken down into units two squares wide and one square high representing the layers in the basket.  There are 29 units in height and 48 units in the circumference for a total of 1392 units.

Photo A

I will describe the process sequence without regard to the pattern.  The finished size of a facet in the pattern is taken as .250 inch in height and .500 inch in width.  This represents the end of a segment.  The diameter of the finished bowl is taken as .500 inch x 48 or 7.64 inches. Half of that or the radius is 3.82 inches, the nominal length of the segment.  Make it 3.75 inches.  The segment angle is 360 degrees divided by 48 = 7.5 degrees.  The segment is imagined as a right-angled wedge.

Diagram #1

 

Two wedges are matched to form a rectangle.  Then spread apart to a width of .750 inch.

Diagram #2

This will provide a space for the kerf of the band saw.  The basic building block then becomes a rectangle .250 inch x .750 inch x 3.75 inches. 

Photo B

Since each block will become two segments, only 696 or half of the 1392 will actually be needed.

However, since the thickness must be held to .002 inches or less, and the surfaces must be flawless (some of final interfaces will be very small), an allowance of 15-20% must be made to permit rigorous culling for blemishes like off-color, snipe and tear outs.  Otherwise the flaw will surely show up as the outstanding feature of your bowl.  The allowance will also cover stupid mistakes.  Twenty-nine of the blanks plus a guard blank on each end are liberally spread with glue to be sure that there will be no glue-starved surfaces, some of which will become very small. 

Photo C

Two slabs can be glued at one time as in Photo C.  Care must be taken to insure all the slabs are clamped with equal pressure so that all will be of the same length.  Additional clamping is needed to keep the work flat.

Note: All the glued joints are long to long, with no cross stresses. Any movement caused by humidity is radial.

The glued up slabs are then surfaced on both sides to a uniform thickness of 5/8 inch plus.  The router set-up in Photo D worked well for me.  Sanding is not advised.  Next, they are trimmed to a width 3.5 inches.  This will create a flat for mounting on the band saw carrier.

Photo D

Diagram #3

 

It is essential to make a half dozen or so of dummy slabs of common lumber and machined along with the bowl slabs.  These are needed to try the band saw set-up and as gluing cauls later.

Photo E

A dressed slab is mounted on the carrier, Photo E, which rides a guide clamped to the table, which is tilted to 7.5 degrees. The blade must be new, aggressive and well tuned (3/8 inch 6 point skiptooth, in my case), so that the 7.25 inch long slice can be made in one smooth pass with no hesitation or vibrations.  This action is harrowing and should be well rehearsed! The dummy slabs are used in trial runs to try the angle and adjust the carrier so that the slab is sliced down the exact middle producing a pair of major segments or staves and in effect cutting 58 minor segments in one pass and all 1392 in only 24 passes to make 48 staves. A pair of staves will be mirror images and will be 180 degrees apart in the assembly.  A strict numbering system must be devised to keep all this in order.

The angle of 7.5 degrees is too acute to measure with a protractor, so four or six staves are tried together as 30 degrees or 45 degrees.  The band saw surfaces need only a touch with the low angle block plane to make a good gluing surface.          

The rip fence of the bench saw is set at 3.5 inches.  The stave is passed with the sawn face down cutting off a corner to make two equal faces and easier clamping.

Diagram #4

Now, all the staves are assembled dry to see how nearly the cylinder will close.  If it is reasonably close, the next step is to band saw away the surplus stock on the inside.  The profile of the inside of the bowl is scribed on the stave and with the table set at 7.5 degrees; the excess is carefully cut away.  This waste is numbered and coded and gently put away as bonus material for one or two smaller bowls.  Photos B and F.

Now, the surfaces to be joined are greatly reduced and should be dry assembled again and tried.  With the block plane the staves are carefully dressed, not sanded, as needed to close. I tried for high spots by rubbing on carbon paper.  This is the most tedious operation but the one that makes or breaks the whole thing.  The assembly must fit together by itself with only a light clamping.  Excessive force will create stresses that will lead to failure later.

Photo F

Now the gluing begins.  Photo F.  First, two staves are matched and glued and clamped using the dummy staves as cauls.  Two pairs are glued together to make a block of 4 staves.  Two sets of four are glued and matched to make a block of eight, which is about as many as can be clamped successfully using the practice staves and C clamps.  The rest of the segments are made up similarly and the six blocks are dry assembled again and tried for closing and adjusted.  It is disassembled again and the final gluing starts.  I used yellow glue.

Two stave groups are glued and all six are assembled again, always checking and correcting the fit.  Continue one joint at a time.  Don’t be tempted into doing it all at one time!

 When the cylinder is finally closed and glued and clamped, it will be dripping with glue.  Put it aside for a couple of days to dry.  You don’t want to look at it for a while, anyway!

The ends are scraped and dressed and a disc of MDO paper faced plywood is glued to the mouth end.  The disc is slightly smaller than the diameter of the cylinder so that diameters can be scribed from several joints to find the center.

Attach to a 6-inch faceplate.  The other end is fitted with a shouldered plug to accept a live tail stock center so that the heavy assembly is supported well. 

Photo G

After the cylinder is balanced and rounded with a scraper or gouge by working from end to middle, the plug is removed and the inside is smoothed about half way in.    I chickened out at that point.  Then a shoulder is formed in the bottom opening.  A disc will be fitted to make the bottom after the bowl is thoroughly dry.  Make it a loose fit, with hot melt glue as an expansion joint. A mounting disc of MDO plywood is now glued to the bottom over the opening and a center is scribed.  The faceplate is transferred to the bottom end and mounted on the spindle and the work is supported between centers.  Photos G H

Photo H

Photo I

Photo J

The contour is shaped with a scraper or gouge.  The end grain cuts easily “uphill” from the ends to the middle.  Starting with 60 grit cloth the bowl is smoothed to high finish.  The pores are filled with numerous coats of gloss lacquer each sanded down until the pores are full.  Then a finish coat of whatever you like best.  Photos I and J show how I handled some of the bonus material.

I was most satisfied in ending up with everything in place and all the joints good, since the olla patterns are fixed with no opportunity for correction.

I am nearly 90 years old and the projects were done 25 years ago with some pretty primitive equipment.  I’m sure that with today’s technology my method can be refined to produce some outstanding projects.  Let me know if you want to try this or if you have questions.  You can contact me at albertjohn0606@sbcglobal.net

A.J. Golichowski with some of his carousel horse and Beatrix Potter creations

Read the recent article about Albert in the South Bend Tribune newsletter by clicking here.