How do you measure your segment lengths (segmented turning), to the 32nd of an inch?

[dirk martin]

I want to try my hand at some segmented rings.
I've got the Segmented Projet Planer software.
I draw my vessel in the software, it calculates segment length, and on one ring, the length of my segment is to be 1-25/32" long.
20 segment ring.
First of all, how important that each segment be 1-25/32" long...on the dot?
If I need to be dead on, within 1/32, how are you guys measuring your segment lengths, before you cut them, to achieve this?

[dirk martin]

Never mind....
Perfect video found.
He's even using the same Incra fence that I am.

http://woodtreks.com/how-to-cut-segmented-pieces-for-turning-projects/42/

[Quinn McCarthy]

Dirk,

I use a micrometer now. I used to use a ruler that went to 1/64ths but you can hardly read it. I was over by just a hair on a 72 segment ring. When you multiply a hair X 72 you have a considerable amount of error. I was still able to use the ring but it was close to too big. I use a jig that I saw in a Curt Theoblald DVD that I have based on a 12" disk sander. There are a few different segment cutting jigs out there. Make sure that whatever one you choose you are very careful in the contruction because it is the cummulative error that is a problem in segmented turning. In CUrts DVD he shows step by step the construction on the segment sizing jig. Also Malcolm Tibbets has 5 DVD's out now. They are a great resourse.

Hope that helps.

Quinn

[Ryan Baker]

1/32 inch is a pretty huge amount for segments. Make that much error on a 32 segment ring and you are a full inch off. It adds up fast. The angles matter more than the length though.

[dirk martin]

I'm not so sure sure about that, Ryan. The angle is pretty easy to get dead-on, but cutting each segment, to a perfect length, accurate to .001, or .0001, is the hard part, and that's what get magnifies around the circle.

[Michelle Rich]

Yes, but if you are consistant & each row is oversized all the way ..then you still have a perfect vessel. I've been segmenting for 25 yrs..I use a micrometer..if I'm off equally I'm just fine. it won't be the exact vessel you drew, but it will still be a fine vessel. And remember you made your rows pieces longer than they really neeed to be & oversize anyway & you will turn a lot off..you may just end up close to the vessel you drew. Go make some & I bet you see what it's all about..But get a micrometer..you will need it. 1/32 is not close enough

[dirk martin]

Ok, let me see if I have this right....I'm learning a lot here....
Let's assume, my miter angle, is perfect. Dead on to .00001.
But let's say my edge length is a tiny bit over, on each segment.
Let's also say that I'm gluing up the whole cirlce at once, rather than half and half.

Are you saying that if my angle is dead-on, then that last segment, will silde into the circle perfectly, even if my edge length is a bit long? Are you saying that having an edge that's too long will affect my diamter, but not my segment, fit? Am I interpreting that correctly?

And, if that's true, then are you teaching me, that if we swtiched this around; and my segment edge length was perfect for an eight inch diameter circle, that even though that will guarantee me of a perfect 8" circle, it won't guarantee me of perfect segment connections?

[Bill Bulloch]

If your angle is dead on and you cut all your segments to the exact length, then you should have a perfect circle. It'll never happen...if one of your segments is a minuscule off then you will not have a perfect ring period. A perfect ring is something you should strive for when cutting, then you glue up into half rings and sand as necessary. I have had a lot of rings that looked dead on when dry fitted, no light shining through at all, but when I glue them up in half rings there is always a gap of some size.

You are expecting too much. Concentrate on good tight joinery.

[Robert McGowen]

Let's assume, my miter angle, is perfect.
But let's say my edge length is a tiny bit over, on each segment.

Are you saying that having an edge that's too long will affect my diamter, but not my segment, fit? Am I interpreting that correctly?

......my segment edge length was perfect for an eight inch diameter circle, that even though that will guarantee me of a perfect 8" circle, it won't guarantee me of perfect segment connections?

The fit of the segments has absolutely nothing to do with the length of the segments, given your assumption that ALL of the segment lengths are exactly the same.

For simplicity sake, say you cut a four-sided picture frame with 8" sides, but use a 60 degree angle for the cuts instead of 45 degrees. You will get a perfect 8" square on the outside, but the joints themselves won't be anywhere close.

The problem, as Bill states, is that you realistically will never get 16 segments in a ring that are cut exactly the same length, and probably not even cut the exact same angle.

[dirk martin]

Assume all my segment angles were perfect 9 degrees.
Are you saying that 20 of those will make a perfect circle, regardless if some segment lenghts are longer, or shorter than others?

I just want to be sure, that a perfect angle on every segment, will get me a perfect circle, even if the segment lengths are mixed.....

[Robert McGowen]

Assume all my segment angles were perfect 9 degrees.
Are you saying that 20 of those will make a perfect circle, regardless if some segment lenghts are longer, or shorter than others?

I just want to be sure, that a perfect angle on every segment, will get me a perfect circle, even if the segment lengths are mixed.....

I don't think that anyone has said that and it certainly is not true...... ?????

[Michelle Rich]

I think there is confusion with terms. Your angles can be precise and the ring will go together IF the width (the measurement of the widest part of your triangle shaped piece) is pretty consistent. THey will never be perfect, but you can take your time & be very careful cutting and they will be within 100ths. If you are 1/32 off on either side of the true measurement, you will come up with an ovally shaped piece. If you are always 1/32 over the width you will just have a bigger row when glued up. When I say seg length I am referring to the measurement from the widest to the narrowest of the seg, or the measurement from outside the bowl to the inside. I think going & doing it will help you see what it is all about and you will have hands on & it will become clear.
If I were you I'd get a few books on segmenting & get a better grasp of what it's all about. Have fun!! And lets see what you come up with!

[Bob Coates]

Ok, I gotta add some pictures to the confusion and my .02. I did an experiment this morning and took some pictures. First I cut most of my segments in 15 sections in each ring on a dedicated chop saw. The block segment length is gotten from my excel sheet for the closest ring diameter and a stop block set on against the fence as shown. For this case all white blocks were cut to the same size, I then cut 5 pieces of dark wood and added them to the segment and dry fit. If my reasoning is correct the block segment length is not critical? Then took two out and dry fit as shown in both pictures. This seems to me that the angle is most important for creating segments and not the length, did not try it with just one to see if it would totally upset. fit Granted you may not be able to predict the size of the ring.

[Adam Alloway]

I am trying to segmented turning but almost every one i do has a gap of some sort or the angle is off. the best angles i get are off my bandsaw but the cuts are too rough to glue with out sanding. the cheap o craftsman table saw i got and http://www.rockler.com/product.cfm?page=21487&filter=sled I still get messed up rings. not sure it it is moving when i cut, the saw blade is crap, or the sled is broken. I did get and angle finder and see that sometimes the angles are off or maybe cus its a plastic angle finder from lowes.

If anyone be kind enough to email me. nedrix007@gmail.com for some help, I will also try my local turning club, I hope i can make this months meeting.

[Bill Bulloch]

Bob, your experiment only adds to the confussion. The Angle is determined by the number of Segments in a ring: ((360/# Seg)/2)). The Length of the segments determine the diamenter of the ring.

Your experiment is confussing because you have lost prospective of what you have done. In your first example you use 15 segments which resulted in an 12 degree angle: (((360/15)= 24) / 2)= 12)). In your second example you have 5 segments made up of 4 pieces (3 white blocks and 1 dark block). In this case the resulting angle is 36 degrees: (((360/5 = 72) / 2= 36)). In your last example you have 3 Segments made up of 6 pieces (5 white and 1 dark block). In this case the resulting angle is 60 degrees: (((360/3=120) / 2 = 60)).

Now here is a though to pounder: You can use different lengths, diffferent number of segments and different angles all in the same ring --- as long as the total of your angles come to 360 degrees.

[Robert McGowen]

Bob C., Your experiment is totally flawed. In the example with 5 spacers, you essentially created 5 larger segments that are all the same length and therefore fit together. Your are correct in that if you just put in one spacer, which you did not do, they would not form a ring, because you would have 14 equal segments and 1 segment with the spacer that is longer. What about if you put all the spacers together? You certainly would not form a ring that fit together. How about 2 spacers across from each other? Then you would have an oval, but all of the segments would fit together. The fact that you set the spacers up the way you did, just showed what you wanted it to show and excluded the facts.

[dirk martin]

Mr. Coates, please realize that the comments others are giving, are meant for good, common, education....not to ridicule.
Depending on how you read some of those comments, they can be taken harshly, and I hope they are not taken that way.

I'd like to make some statements, that I believe are true. Guys...please let me know if any of these are not true.

1) In order for the ring to fit together perfectly, all the angles used in that ring need to add up to 360.

2) It is not possible to get a perfect angle, but it's important to get as close as you can.

3) Segment angle determines the # of segments in the ring.

4) Segment length determines the diameter of the ring.

5) All of the segments, do not have to have the same edge length, however this formula needs to result in a whole number: (total ring segment count) / (segment count with different length edge). In other words, a 20 segment ring, can have 4 segments that are of a different length than the rest, because 20/4=whole number. Those 4 segments, though, need to be equal in length to each other.

6) Titebond glues work great. CA glue is too brittle.

7) Pressing two freshly glued segments together, and holding them with good hand pressure for 30 seconds, and then setting aside, should yield a good, solid bond, as long as both faces are flat. Clamping is ok, but not neccessary.

8) Freshly glued pieces should be left to cure for at least 30 minutes.

9) After 8 hours, segments can be stressed in terms of sanding them.

10) After 24 hours, segments can be turned.

11) If glued rings, are clamped, the purpose of the clamp is to hold the segments in place, not to force the removal of a gap.

12) Having dead-on angles, is not enough to get an easily joined ring. Segment length also needs to be consistant.

13) Dead-on segment lengh alone will not get an easily joned ring. Segment angles need to be consistant.


I believe all of the above are true.
Correct me if I'm wrong.

[Robert McGowen]

Mr. Coates, please realize that the comments others are giving, are meant for good, common, education....not to ridicule.
Depending on how you read some of those comments, they can be taken harshly, and I hope they are not taken that way.

Dirk is correct if he is referring to my comment. I am just trying to help people understand what they are doing.



5) All of the segments, do not have to have the same edge length, however this formula needs to result in a whole number: (total ring segment count) / (segment count with different length edge). In other words, a 20 segment ring, can have 4 segments that are of a different length than the rest, because 20/4=whole number. Those 4 segments, though, need to be equal in length to each other.


I am guessing that you are correct, but the critical factor is that the different length segments need to be equi-distant from each other. You won't end up with a circle either, so you will have to take that into account. You would probably only want to do this on some part of a feature ring. Here are 2 examples of different length segments on several rings to make a pattern:

187598187599




7) Pressing two freshly glued segments together, and holding them with good hand pressure for 30 seconds, and then setting aside, should yield a good, solid bond, as long as both faces are flat. Clamping is ok, but not neccessary.


Pressing them together won't do much. You need to rub the ends back and forth for a few seconds and then hold them together for a few seconds. (It is, after all, called a "rub joint") :D Maybe 10 seconds total time per joint. This technique might be used for a feature ring, but in my opinion is way too slow and not even the best way to make a regular ring.



9) After 8 hours, segments can be stressed in terms of sanding them.


I sand the rings after 30 minutes of drying. YMMV



10) After 24 hours, segments can be turned.


I turn the rings 30 minutes after glueing them together. YMMV

I would say that everything else you posted is correct. :)

If all you do is achieve #12 and #13, you will probably end up doing a fairly good job! :eek: :D ;) :cool: :rolleyes: :eek:

This is just my way of doing things. Hope this helps. :cool:

[Bob Coates]

No Bill both segments contain 15 segments it just different placement of the dark wood. All white segments are the same size in both rings.

Bob

[dirk martin]

Robert, you wrote:

"Pressing them together won't do much. You need to rub the ends back and forth for a few seconds and then hold them together for a few seconds. (It is, after all, called a "rub joint") :D Maybe 10 seconds total time per joint. This technique might be used for a feature ring, but in my opinion is way too slow and not even the best way to make a regular ring."

What do you feel would be a faster technique?

[Robert McGowen]

What do you feel would be a faster technique?

Check post #14 on the thread you started a few days ago. :rolleyes:

http://www.sawmillcreek.org/showthread.php?161966-CA-glue-for-segmented-blanks&highlight=

[Thom Sturgill]

TECHNICALLY if the angles are correct you could glue into a perfect circle with the pieces being such that some are recessed and some protrude. REALISTICALLY we tend to glue the pieces such that the outside edges mate so BOTH angle and length need to be consistent. If you allow a little extra in width (inside to outside) you can increase the segment length to a easier to measure value and simply turn off a bit more.

[Malcolm Tibbetts]

With perfect angles and very consistent segment lengths, it is possible to glue entire rings together in one step. I sometimes do this, using rubber bands as my clamp, when gluing small (2” dia) rings. The risk is while a dry fit might show perfect fitting joinery, during the actual glue job, one (or more) of the outside “point to point” alignments might not get perfectly aligned as it was during the dry fit. If just one segment is not perfectly located, then somewhere in the ring a slightly thicker glue joint will be created. This is why most experienced segmenters use the half-ring technique. It’s like an insurance policy against finding a sub-standard joint later in the project.