PDA

View Full Version : Turning spirals



paul48
29th May 2016, 09:25 PM
I am having some trouble with the maths involved in turning a segmented bowl with a continuous spiral.
I cut, thicknessed and glued alternating cherry and jarrah timber strips 18mm wide and 900mm long.
I then worked out the diagonal angles to cut 9 x 75mm blocks each bevelled at 20 degrees and allowing for a 3mm saw kerf so that, when assembled into a circle, the alternating colours met perfectly at each join.
My problem is that, when turned on the lathe into a circular bowl, the loss of timber from the peaks of the joins meant the each segment was out of line by about 2mm and the spiral is no longer continuous all around the bowl. There is probably a mathematical formula I should be using but I confess to not paying much attention in geometry class way back when.
Anyone got any ideas?

Paul

powderpost
29th May 2016, 10:00 PM
Paul, I did a W.I.P., how to make a segmented vase. Have a look at http://www.woodworkforums.com/f8/segmented-vase-174358. There is a better way to refer you to a thread, but I don't know how to do it. The vase doesn't feature a spiral, but by introducing a second or third coloured segment, you can induce one or several spirals.

Jim

paul48
29th May 2016, 11:00 PM
thanks Jim
Have done several segmented bowls in different configurations but really want to get this continuous spiral done. Might be it's not even possible but I like the challenge anyway.

Paul

joe greiner
30th May 2016, 02:18 PM
I am having some trouble with the maths involved in turning a segmented bowl with a continuous spiral.
I cut, thicknessed and glued alternating cherry and jarrah timber strips 18mm wide and 900mm long.
I then worked out the diagonal angles to cut 9 x 75mm blocks each bevelled at 20 degrees and allowing for a 3mm saw kerf so that, when assembled into a circle, the alternating colours met perfectly at each join.
My problem is that, when turned on the lathe into a circular bowl, the loss of timber from the peaks of the joins meant the each segment was out of line by about 2mm and the spiral is no longer continuous all around the bowl. There is probably a mathematical formula I should be using but I confess to not paying much attention in geometry class way back when.
Anyone got any ideas?

Paul

I'm having trouble visualizing which dimensions are where. Do you have a sketch elevation of your design? Even if only MSPaint? Words alone are inadequate.

Cheers,
Joe

paul48
30th May 2016, 09:22 PM
I'm having trouble visualizing which dimensions are where. Do you have a sketch elevation of your design? Even if only MSPaint? Words alone are inadequate.

Cheers,
Joe


381583381582
Thanks Joe
Photo 1 above shows the 9 sides ready for glue up.
Each side is 75mm wide and, as you can see, the pattern match is spot on.
Photo 2 shows how the pattern "slips" when I turn the 9 sides into one circular bowl.
Frustrating but I'm not sure how to arrange the initial cutting pattern so that the pattern match is correct at the end of the turning phase.

Regards

Paul

Glenn.Visca
30th May 2016, 09:34 PM
I can't answer your question, nor even begin to consider the geometry, but isn't the pattern slippage evidenced in the 1st photo on the inside ?

paul48
30th May 2016, 10:17 PM
I can't answer your question, nor even begin to consider the geometry, but isn't the pattern slippage evidenced in the 1st photo on the inside ?

Glenn

It is indeed. I thought that was a result of the timber I removed to cut the segments - each at 20 degrees
Seems that slippage carries through after I removed the peaks on the sides. I thought about gluing the segments with that slippage built in but that only works if I round the bowl to exactly a specific depth and I don't know how to work that out. The rough sketch below shows how I worked out the original cutting pattern.

Paul381597

Paul39
31st May 2016, 04:37 AM
Paul48

It appears that the individual pieces are leaning out when the slabs are put together.

I think if each individual piece were parallel to the base on an axis from the center to the outside, they would meet evenly. In order to make the piece taper from top to bottom, you would need to make the wall thicker so as to have more to cut off.

You might be able to calculate thinner sections and step out as you came up, but that calculation makes my head hurt.

This book might help: http://ecx.images-amazon.com/images/I/51DIaxQMVPL._SX373_BO1,204,203,200_.jpg

Ray Allen with turning:

https://s-media-cache-ak0.pinimg.com/736x/a6/86/31/a68631cd6be3805f6a3ff7e30603cea8.jpg

Short video: https://www.youtube.com/watch?v=owT-XMNiBAY

joe greiner
31st May 2016, 03:31 PM
Glenn

It is indeed. I thought that was a result of the timber I removed to cut the segments - each at 20 degrees
Seems that slippage carries through after I removed the peaks on the sides. I thought about gluing the segments with that slippage built in but that only works if I round the bowl to exactly a specific depth and I don't know how to work that out. The rough sketch below shows how I worked out the original cutting pattern.

Paul381597

This looks like it should be a 3-lead screw, but it isn't. You show a sloping joint beginning at the lower left. The next cycle should start at the 3rd vertical joint, and the next at the 6th, and finally at the 9th (back at 0). This might have the effect of making the sloping joints slightly steeper, without steps.

Joe

Hlokk
31st May 2016, 04:32 PM
While there is a way to make this with the spiral all lining up no matter how you cut it, it will be rather difficult as the segments will need to have non flat surfaces.

As you create a helix, you basically need to "torsionally twist" your segments. You could have it line up the whole way through the blank if you make each segment piece out of two pieces but that will cause "ripple" in the helical pattern even though it will be continuous lines.

Perhaps ps the easier way to do it would be to "extend" the pieces such that the alignment line is inside the blank instead (you'll need to turn down to exactly that line though). Or perhaps offset more and make it a feature?

paul48
31st May 2016, 09:21 PM
Thanks to all who replied.

Paul39 - It does look like the pieces are leaning in the photo -that bowl exploded on the lathe with almost all vertical glue joints failing. They are in fact plumb vertical. The photo on the lathe is my second attempt using a better quality glue which held together well.
Joe - I didn't plan it a a 3 lead. I plotted the lines so that the timber pattern flowed from the bottom left to the top right. The spiral was dictated by the 701mm length, 130mm height of the cut out block and 18mm width of the individual timbers. The sketch shows this but isn't to scale or even the right number of alternating timbers. I plotted this to 701mm to allow for the 9 x 75mm segments plus all the saw kerfs. It worked well on the paper model and then worked really well in the actual build. I did look at various leads in a similar manner to my hollow twist candlesticks but this method just seemed a lot easier.
Hlokk - Not sure I understand the first part of your reply but I am also thinking the line will have to be somewhere in the middle of the block. If I take the step measurement of the inside of the unturned bowl and halve it, that should be close to the centre error. I then need to glue the segments with that error margin built in to the front surface instead of lining the spiral exactly at the front as I did in the bowl in the photo. Have a few other projects to complete first but will come back to this issue in a few weeks. Hate it when I can't figure out an answer to a problem so I really appreciate you guys taking the time to help me along.

Paul

joe greiner
1st June 2016, 01:35 AM
The 3-lead screw was a false interpretation of the problem. The real culprit is the angle between the staves. I think if you increase the number of staves, to 18 or 36, the step will be reduced.

The attached picture shows how the surfaces of the stripes project to the final surface.

As a last resort, you could add a thin vertical veneer between the staves to fool the eye.

Cheers,
Joe