Segmented Fail (?) -- 20% rant, 80% cry for help

  • Advertise with us
Blog entry by dspahn posted 11-26-2011 09:22 AM 3291 reads 0 times favorited 14 comments Add to Favorites Watch

So after watching some videos on segmented bowl construction and turning, I decided to give it a shot while I wait for my cole jaws to arrive in the mail. I actually went to a lumber place and bought a few pieces: some walnut, alder, cherry, and something with the word “african” in it… maybe mahogany, but I’m not 100% sure.

And then, after I got home, whinging internally about spending $40 on an amount of wood you could maybe build a desk for a barbie doll out of, I realized that I didn’t have any good way of cutting strips of wood with 22.5 degree angles.

Now, I like math. I always have. I understand that 22.5 degrees is half of 45, which is half of 90, which is a quarter of 360, and why 8 pieces need 22.5 degree cuts in them. I could give you the measurements in radians if you’d like. But tonight I learned the hard difference between theory and practice. More on that later.

So to rectify my lack, I decided to create an angle jig for my table saw. First I created a sled. Pictures below:

And here’s the underside of the sled and what it rides on:

Then I used a piece of MDF (I think), and some small pieces of dowel to hold it in place at 0, 22.5, and 45 degrees as needed. The nice thing about this method is that while it’s not infinitely adjustable, to add a new angle, all I need to do is drill a new hole. You can see the jig in the first picture above.

I used these tools to find my 22.5 and 45 degree angles:

The 0 degree angle was found by placing a framing square against the blade of the table saw, and then drawing a line.

The tools in the picture are: a couple of framing squares, a protractor doohickey, and what I think is called a compound square. To drill the holes for the dowels, I used my shopsmith in drill press mode.

In retrospect, rather than create the sled and try to find the angles, I might should have used my radial arm saw, but the table on that needs to be completely replaced, and it’s harder to get at than my table saw. You see, I’m trying to do a woodshop in a one car garage that also holds my motorcycle, several shelving units, a good sized tool cabinet and a bandsaw.

So, rather than dive right into cutting up my expensive wood, I decided to give the new jig a test run on some scrap wood I had left over from something else. The cutting went swimmingly. Here’s what I ended up with:

Which all looked well and good until I actually tried to assemble them into an octagon. And here’s where the frustration set in:

If I put 7 pieces flush together, the 8th piece just doesn’t fit very well. Which means my 22.5 degree angle was something other than 22.5 degrees. And here’s the gaps if I place the pieces together not quite flush, but very close:

And now for the questions. Am I close enough? And by that I mean, if I decide to chop up my “nice” wood, and glue those segments together, will I end up with a nice looking ring, and thereby a nice looking bowl? Or do I need to get closer to 22.5 before ruining the “nice” wood? And if I need to get closer to 22.5, how do I do it? I have a table saw, a radial arm saw, a band saw, a shopsmith, and a router. And as far as angle type tools, I have several framing squares, two compound squares, that protractor doohickey, and that’s about it. Can I even do what I’m trying to do with the equipment at hand? Or do I need to obtain something else? And if so, what do I need?

On the bright side, even though I’m not ecstatic about my results, I still had fun killing a few hours in the garage.

Thanks for reading, and thanks in advance for any tips or advice you might have to offer.

14 comments so far

View David Craig's profile

David Craig

2137 posts in 4606 days

#1 posted 11-26-2011 10:52 AM

You did good by not moving straight to the exotics and putting together test pieces. I personally feel the gaps are a little too much to make a good segment and I would test again. Just a few observations. I don’t see where you have a stop block on your cutting sled. I would check the length of your pieces and double check that each one is exactly the same length. You might have the correct angle but if pieces are say 1/32nd off each other and you have 8 pieces those slight differences will add up. I would get a small protractor and check the angle of the test pieces. The biggest caveat with your angle jig is that if the dowel holes are not exact it is difficult to fill and redrill to correct any deviancy from the true angle. I think you would be better off with a block that could rotate a full 360 and use tightening pressure from a wingnut or something along that line when setting the angle. Then, if you needed to make minute compensations, you could adjust accordingly. Hang in there, you are off to a good start.


-- There is little that is simple when it comes to making a simple box.

View dspahn's profile


85 posts in 3877 days

#2 posted 11-26-2011 11:09 AM

There was a stop block. It was clamped to the fence of the table saw. All of the pieces are very very close to the same size. There are a few that are slightly larger than the others, but only by half a millimeter or so. Maybe I’ll try sanding them all to the exact same size tomorrow to see if that fixes the problem. Good thought. :)

I initially wanted to make a block that could rotate, with wingnuts to lock to desired position as you described. The problem with that, and I realize there’s no excuse for lazy, is that I would have had to first make a circle jig for my router. Tomorrow, I’ll probably try again and do it the right way.

Is it considered “cheating” if I just take the pieces to the disk sander to get them to fit? Do you think it would be noticeable to anyone if one piece of the circle is 1/16th shorter than all the others?

Also, instead of putting 22.5 degree angles on each side of each segment, can you just put a 45 degree angle on one side of each segment? Because 45 is a LOT easier for me to measure out accurately given the tools I have on hand.

View dspahn's profile


85 posts in 3877 days

#3 posted 11-26-2011 11:14 AM

Hmmm. I just used my protractor doohickey to compare all the angles on all of the pieces. It seems like the angles are slightly different on each side of a piece. Which got me to thinking that the original strip of board ought to have parallel sides…. and that is not something I checked for. I suppose it’s possible that the jig is right but the board was wrong. ;)

Tomorrow, I shall conquer this octagon! crazed laughter

View Tootles's profile


808 posts in 3999 days

#4 posted 11-26-2011 12:59 PM

What occurred to me is that you said:

The 0 degree angle was found by placing a framing square against the blade of the table saw, and then drawing a line

A better way is to make the sled base too big so that the first time through the saw cuts the sled. That gives the line parallel to the outermost spinning tooth.

You didn’t say how you measured the 22.5° angle, but probably the best way is by bisection of a 45° with a compass.

-- I may have lost my marbles, but I still have my love of woodworking

View HamS's profile


1844 posts in 3886 days

#5 posted 11-26-2011 02:56 PM

probably the easiest way to get a 22.5 degree andl is to takea square piece of paper, fold it corner to opposite corner and then fold the 45 deg corner again. that will give you a real 22.5 that you can reference. I am not a turner so I cannot say this with authority, but I do not think those gaps are close enough for a good looking glue line and I think the bigger the gap the weaker the joint. I think that a segmented bowl is dangerous enough to turn without the chance offailing glue joints.

-- Haming it up in the 'bash.

View dspahn's profile


85 posts in 3877 days

#6 posted 11-26-2011 05:30 PM

@Tootles: That is an awesome idea! I will incorporate it into today’s attempt. Thanks!

@HamS: That’s a good idea, but that, too, has sources of error: Is your paper truly square? Have the folds been made 100% completely in half? I’m not rejecting the idea, more just thinking out loud.

The first thing I’m going to try though, is gluing up 4 of the segments into a half circle, then glue the other 4 segments together, and see how much I have to take off with a disk sander before they meet up right. If it’s less than 1/16th, I’ll consider this a success.

View KayBee's profile


1083 posts in 4743 days

#7 posted 11-26-2011 08:05 PM

I don’t do segmented turnings, so take with a grain, or pound, of salt. My understanding is that you cut it close, then sand to get a perfect fit. Even with jigs and no human error, it’s hard to get something that complex to fit straight off the saw. Try making a holder/platform for a belt sander if you don’t have a stationary sander.

-- Karen - a little bit of stupid goes a long way

View Bearpie's profile


2601 posts in 4515 days

#8 posted 11-26-2011 09:42 PM

From my limited experience on segmentation I have found that it is critical to have tight glue joints or your turnings will have thick glue lines here and there and that is definitely not attractive! It is also imperative that all your pieces are precisely the same length! It also helps that your board sides are parallel. Sometimes you can cheat a bit here and there but not every time. Good luck!

-- Erwin, Jacksonville, FL

View sras's profile


6722 posts in 4626 days

#9 posted 11-27-2011 09:47 PM

Tight glue lines are important.

For me, I would glue up two sets of 4 segments. This would result in two half rings with 3 tight glue joints each.

There is a good chance that the half rings do not have parallel edges. Gluing these two halves together would result in two bad joints.

Instead, true up these edges – either with a sanding belt or disc on on the table saw. Use a sled or jig to hold the piece. Remove only enough material to get a tight joint.

Now you have eight tight glue joints.

-- Steve - Impatience is Expensive

View able339's profile


47 posts in 3872 days

#10 posted 11-27-2011 10:10 PM

A stop on your cutting sled along with one or two good hold downs (clamps) would be helpful also… Try a sandpaper face on your Jig board – that will help to keep the workpiece from slipping too. It is also possible that your carpenter’s square is not truely square! You need to check that out too.

-- TNJames

View dspahn's profile


85 posts in 3877 days

#11 posted 11-27-2011 11:40 PM

@sras: that is exactly what i did. i meant to take and post a picture, but i accidentally left my camera on while hooked up to the computer during the last uploading, and thus the battery was dead. ;)

@able339: all very good ideas! thanks!

View sras's profile


6722 posts in 4626 days

#12 posted 11-28-2011 12:14 AM

Funny! I looked through the thread to see if the idea had been brought up – and I missed it!!

I have a book on segmented turnings that uses the same method so you’re in good company with the technique.

-- Steve - Impatience is Expensive

View Grandpa's profile


3265 posts in 4172 days

#13 posted 11-28-2011 06:56 PM

I don’t think you would want to start sanding the ends of the pieces. If you do that the cut area is going to get either shorter or longer unless you can maintain the 22.5 deg accuracy. things will just get all messed up in a hurry. Same thing with the 45 deg cut. The cut area would not be the same size as the mating part. You need an accurate jug and you are on the way to getting it. One thing you haven’t talked about is the flexing of the drill while making the holes. I am always amazed at how much things move that I don’t see. what about moving away from the plywood for the sled and going to a large plastic cutting board from the $ store. There is no grain to cause the drill to flex and all that. What bout an adjustable surface. Make the jig just like the one you currently have and then make an adjustable face by using a good hinge and a machine screw through a tapped hole for getting things exact? I use a Starret protractor on a machinists square for checking and setting up. I have found that to be a very valuable tool in my shop. The sandpaper idea is always good for holding pieces in place. that could make the jig off if the paper didn’t have the exact amount of sand in all areas so the adjustment would be needed again. Stay on this site and you will get enough ideas that some of them will have to work for you, LOL. lets us know how this works out, please.

View dspahn's profile


85 posts in 3877 days

#14 posted 11-30-2011 08:12 AM

Pictures of the same pieces above after doing what sras suggested. I’m pretty sure this could be used in a bowl. Thanks for all the friendly advice. Lots of really good tips here that I will probably use this weekend when I actually start my real bowl. Tonight, however, I did something else. New blog upcoming. ;)

Have your say...

You must be signed in to post the comments.

DISCLAIMER: Any posts on LJ are posted by individuals acting in their own right and do not necessarily reflect the views of LJ. LJ will not be held liable for the actions of any user.

Latest Projects | Latest Blog Entries | Latest Forum Topics