[steamingbill]

Hello,

Following a post from another member regarding cutting and fitting 24 pieces into a polygon I thought I'd ask a more general question.

There are various easy tests to see if you have cut a perfect 180, 90, 45 or 22.5 degree angle. Thinking about it 30 and 60 and 15 shouldn't be too difficult for most people. I guess 7.5 is ok as well 12 x 7.5 = 90

How do people set up their saws / routers to accurately cut other angles and how do you check that it is done succesfully ?

Its a bit late when you discover that your 33 piece polygon has a gap in it

I guess it depends upon the type and magnitude of misclosure for example shouldn't be too hard to fix a 33 piece polygon that had the first and last pieces overlapping - shave a bit here and there.

The way I see it for polygons with heaps of sides 0.1 degree error could be critical.

ie 33 x 0.1 = 3.3 degrees so there is either a gap or an overlap of 3.3 degrees on the last join. Is that the right way to think ?

My maximum is only an octagon.

Regards

Bill

 

[AndyL]

Interesting question. I would say if your design calls for making a 33-sided polygon, change the design!

 

[cagenuts]

Measure them with a digital angle gauge like the Wixey.

 

[jsears]

A lot of things to consider when your are trying to cut exact angles. Is your saw capable of it? Is it set up properly? Does your blade flex when cutting (even a little bit)? Is your miter bar on your miter gauge tight in the slot. Is the miter slot perpendicular to the blade?How accurate is your miter gauge, and what are you using to set the angle of the miter gauge? Is YOUR sawing technique correct? Are you holding your stock tight to the miter gauge, so that it does not move (again, even a little bit) when you are cutting the pieces?

If the answer to any of these is "no", then the angles will not be perfect.

 

[rwbaker]

Mr

Sorry you are having these issues - have you tried going back and shaving a hair off the other pieces to close the gap. What you are doing is akin to barrel making, where you have many staves of the same width and angle, and one of the staves that is less wide and many times of different angles, held by the hoops. Jack Daniels was one of my customers for many years.

I use 1/2/3/ blocks or setup angles and a digital gauge. The setup angles are a geater tolerance for metal and wood. These can be had on the cheap from WT Tool or Harbor freight 30? or Starrett from 200 up - mine were inherited but I use the WT for wood (drill tables, tablesaw, shaper angles, etc.)

Good Luck - Baker

 

[woodworker352]

Having done a small amount of segmented work with bowls I have found that there are two basic ways of dealing with getting tight joints.

One is to refine your angle until there is no gap between the pieces. There is no such think as a perfect joint so do the best you can. You can put the pieces on masking tape. You can apply glue, pull the pieces together with any minor errors distributed across all of the parts. Small errors distributed across all the parts is less noticeable than if all of the gap is between two segments.

The other way is to glue up all the segments except the ones on opposite sides. Put a small section of dowel rod in the center between the opposite sides and pull the circle together with clamps. After the glue is dried you can sand or shave the remaining two joints on opposite sides to perfectly flat. Then glue up the remaining two joints. Sorry but this does not work for odd number pieces.

I hope this makes sense. If you look up Malcolm Tibbets I think he has some of this on his web site. He is an expert at making tight joint for segmented work.

 

[vallejotex]

I use this (my scientific approach), 360 degrees divided by number of pieces of wood to be used to make your design = to the corner degrees divided by 2. For example if you've measured the circumfrence with the measurement of 60 inches and want 24 pieces to make it happen then: 60/24=*2.5 inches each board. for the width of each board. 360/24=15/2=7 1/2 degrees.
*NOTE: This is the measurement on the back side (inside radius).When cutting the board measure the back side.
I hope this helps you. i use this alot for figuring the crown moulding cuts when not using flexible moulding, this is an option for some jobs only.

 

[PinchPoint]

Cut them oversized angle wise and make a shooting board using a wood plane. Make light passes and fit this all together. If you can...

 

[friendly1too]

I recall ann episode of Router Workship in which they used a jig to cut angles on the edge of a board, using a straight cutter. The jig consisted of a surface held at an angle to the cutter held up with wedges cut to the complimentary angle. It seemed very accurate and I'm sure you could create one to the angle you need.

 

[steamingbill]

Interesting question. I would say if your design calls for making a 33-sided polygon, change the design!

Andy,

That was simply a random example - I am not building a 33 sided figure.

However there is a chap in another thread who has cut a 24 sided drum - I will ask him how he did it. Will also google how people make barrels - coopering.

Was hoping to start a discussion regarding how different people ensure accuracy when cutting unusual angles. Looking for some shop tips.

I guess there are 4 main parts to the question

1. Get the hardware set up right to produce repeatable results - as discussed in one of the posts above

2. Get the angles right - for me this would mean making a very accurate jig, and being able to confidently test the cuts produced. As I type this am visualising an a relatively large measuring circle somewhere in the cutting machine or jig.

3. Do some fine tuning if necessary after the initial cuts are done

4. Get the joints glued together as tight as possible

As a further example its fairly easy to cut and glue up a square is there anybody out there who would be equally comfortable cutting and building a 13 sided regular polygon ?

For me it involve a lot of mucking around and fine tuning of the cuts as discussed in some of the other posts above.

Regards

Bill

 

[AndyL]

Andy,

That was simply a random example - I am not building a 33 sided figure.

I know, sorry, I was being a bit facetious. What I really meant was I was interested to know the answer to your question, but I was pretty sure whatever it was would be too hard for me

 

[steamingbill]

Found these on the internet

http://dls-website.com/documents/WoodworkingNotes/Polygon_Notes.pdf

Woodworkers Guild of Georgia, Miter Calculator

Some people might find them useful.

Have posted a question on the other thread asking the drum making people how they manage to accurately cut their multi sided polygons.

Googling Compound Mitre Calculator yields a heap of sites that might be interesting

Off to play with Google - too hot to work outside

 

[johnjory]

Hello,

Following a post from another member regarding cutting and fitting 24 pieces into a polygon I thought I'd ask a more general question.

There are various easy tests to see if you have cut a perfect 180, 90, 45 or 22.5 degree angle. Thinking about it 30 and 60 and 15 shouldn't be too difficult for most people. I guess 7.5 is ok as well 12 x 7.5 = 90

How do people set up their saws / routers to accurately cut other angles and how do you check that it is done succesfully ?

Its a bit late when you discover that your 33 piece polygon has a gap in it

I guess it depends upon the type and magnitude of misclosure for example shouldn't be too hard to fix a 33 piece polygon that had the first and last pieces overlapping - shave a bit here and there.

The way I see it for polygons with heaps of sides 0.1 degree error could be critical.

ie 33 x 0.1 = 3.3 degrees so there is either a gap or an overlap of 3.3 degrees on the last join. Is that the right way to think ?

My maximum is only an octagon.

Regards

Bill

I used to make a tool for cutting accurate angles. It was called Anglewright. It had a scale that showed 2/10 degree lines that were about 1/32 of an inch apart. This meant that 1/10 degree (6 minutes) accuracy could be fairly easily be achieved. I closed the company closed and sold the rights to make it so you may be able to find the tool somewhere.
However, the accuracy you are asking about is much finer than that.
If you make a polygon with 33 sides you will make 66 cuts. This means that if the saw is 0.1 degrees off and you merrily connect the pieces together the last joint will have sides 6.6 degrees off. There is no wood working tool that would provide the accuracy ( about 1 second of a degree) you are looking for.
The way to do it is to cut the angles as accurately as possible using something like the AngleWright, a sine bar, the Bridge City Toolworks, or a universal bevel protractor and make enough sides to create an assemblies that are 90 or 60 degrees. Then true the assemblies to that angle. then you only have 4 or 6 joints to worry about.

 

[neville9999]

Interesting question, I will follow all the suggestions. One thing you must do, whenever you work on a new shape and you are satisfied with your angle, then keep a sample of it numbered correctly, so if you do have the, say 13 piece polygon worked out, save the angle and write on it a 13, at least then you dont have to solve that angle again, one by one you can put a set together. N

 

[gmercer_48083]

I generally glue the segments to make two parts 1/2 the diameter... once the two halves are dry then sand each half on a flat surface... then glue the two halves together to form the completed diameter.

 

[DesertRatTom]

This is a very old string. But a good question. My solution has been to use the Rockler table saw sled, picture below. It has a swing arm with extremely precise markings for the angle of the fence. It's all in the setup. I set it up so Zero is perfectly 90 to the full kerf blade, and that the blade is perfectly parallel to the miter slots. On this once you set the fence 90 to the blade, you then move the indicator so it is on the zero line. After that, you can cut any angle you wish, and it's markings are sufficiently far apart so you can do half and even a quarter of a degree cuts with astonishing accuracy.

If you are going for ridiculous accuracy, you can clamp the workpiece to the fence because it can slip slightly. Add a strip of self adhesive sandpaper to the front of the swinging fence. Use a stop block to get all pieces to the same length.

At the bottom is a pdf of the Rockler file on calculating to determine length of each piece to get a particular diameter. I would precut to a little over the desired lengthy, then cut one end to half the angle required, then place the stop block in position and then cut the other end. Thank Rockler for the pdf by buying stuff from them. I don't get a toaster from them, but they are the only woodworking store withing driving range, so I have a lot of their stuff.



Here is a picture of the Freud Glue Line blade. It is a rip blade but does an equally great job on end grain. It also has an ample amount of carbide, so it can be resharpened many times. It also has every fourth carbide cut flat so it is excellent for 1/8th spline cutting.

I tried for some time to get perfect cuts, but found the thin kerf blades would deflect slightly so the angled faces didn't quite meet. I also suggest using a Glue Line blade for ultra smooth gluing surfaces. Put one coat of glue on the ends, let it dry, then do the finished glue up. This will improve the joint strength. Use the dries clear glue for this application.

Hope this is helpful. By the way, Rockler makes a great, thick draftsman's triangle, clear plastic with a 90 and 45 angle that are right on. I use it all the time. Really handy.