another geometry question

[MikeG.]

.......Jacob you are exact in what you have pointed out regarding stave width for 15 staves located by circumscribing circle of 170 mm diameter. The arc length for each stave is 35.6 mm as you mention........

Circumscribing a circle was also a mistake. If you need a finished cylindrical shape of diameter X, X must be the dimension to the face of the staves, not the corners. Because your circle is correct but its relationship to the staves is wrong then necessarily so is your calculation for the width of each stave.

 

[sunnybob]

hey, I got a 4 page discussion going without using the "sharpening" word.
Is that a new record?

 

[Jacob]

........
However to get 2 semi cricles you need an even number of staves - half in each semecircle. If you glue up three lots of five staves you will have to adjust SIX edges by eye without any reference plane.

"Reference plane" again! This is not a woodwork technique. Woodworkers use their eyes.

 

[Brandlin]

........
However to get 2 semi cricles you need an even number of staves - half in each semecircle. If you glue up three lots of five staves you will have to adjust SIX edges by eye without any reference plane.

"Reference plane" again! This is not a woodwork technique. Woodworkers use their eyes.

You speak for ALL woodworkers do you Jacob?

And who are you to judge what is and is NOT a woodworking technique? If it works it's a technique for working wood.

The fact is, its FAR easier and more accurate to mate 2 semicircles by sanding on a flat surface than it is to shoot or sand several individual edges and align them.

Every woodworker I have worked with that makes stave based cylinders does it this way. YOU may chose to do it another way.

 

[dizjasta]

.......Jacob you are exact in what you have pointed out regarding stave width for 15 staves located by circumscribing circle of 170 mm diameter. The arc length for each stave is 35.6 mm as you mention........

Circumscribing a circle was also a mistake. If you need a finished cylindrical shape of diameter X, X must be the dimension to the face of the staves, not the corners. Because your circle is correct but its relationship to the staves is wrong then necessarily so is your calculation for the width of each stave.

Hi Mike regarding a circumscribing circle which aligns with joints at 170 mm diameter. What you say is correct and finishing the exterior of the tub to a circular shape would not be possible if 170 mm diameter is required. I seem to recall that when Bob first brought this issue to everyone's attention he made made no mention of making the tub's exterior cylindrical. His main concern was to produce the interior to the shape and dimension he needs using his drum sander.

 

[Jacob]

........
However to get 2 semi cricles you need an even number of staves - half in each semecircle. If you glue up three lots of five staves you will have to adjust SIX edges by eye without any reference plane.

"Reference plane" again! This is not a woodwork technique. Woodworkers use their eyes.

You speak for ALL woodworkers do you Jacob?

And who are you to judge what is and is NOT a woodworking technique? If it works it's a technique for working wood.

The fact is, its FAR easier and more accurate to mate 2 semicircles by sanding on a flat surface than it is to shoot or sand several individual edges and align them.

Every woodworker I have worked with that makes stave based cylinders does it this way. YOU may chose to do it another way.

The usual woodwork way of matching two pieces against each other is to put them together and see how they fit and make adjustments accordingly. It cuts out the middleman - the two pieces need only fit each other - not a third 'reference surface'.
Obviously your sanding process will work, in fact I used it a lot for finishing boxes with sawn off lids. But it's easy to dub them over a bit and defeat the object.
It's the 'reference surface' notion that I'm sceptical about - it seems to crop up a lot lately - even as the method for verifying the flatness of planed boards on workbench, with a perfectly planed 'reference surface' top.

 

[xy mosian]

If the inside dimension is critical. Then produce a template, cut disc, to the required diameter of the inscribed circle. Then each joint can be checked easily and adjusted if needed.
If it is the outside dimension which is critical do the opposite. Make a template, cut a hole, to work with the joints on the outer side of the drum. Again each joint can be checked and fettled as needed. Working with the outside will be easier as there will be more contact, two points, with each piece. The other way around will be difficult as there will only be one point of contact with each piece.
xy

 

[Brandlin]

The usual woodwork way of matching two pieces against each other is to put them together and see how they fit and make adjustments accordingly. It cuts out the middleman - the two pieces need only fit each other - not a third 'reference surface'.
Obviously your sanding process will work, in fact I used it a lot for finishing boxes with sawn off lids. But it's easy to dub them over a bit and defeat the object.
It's the 'reference surface' notion that I'm sceptical about - it seems to crop up a lot lately - even as the method for verifying the flatness of planed boards on workbench, with a perfectly planed 'reference surface' top.

You use a reference plane every single time you pick up a plane, run a board through a jointer or thicknesser, and every time you use a router table or table saw. Your sharpening stones are a reference plane, the inside edge of your square is a reference plane... you use them ALL the time.

Many people have made comments here about which technique might be better, or even whether the answers people have provided will produce the required result. That kind of debate is useful, positive and educational.

There are many ways of doing most things. Some require more skill, knowledge or different tools than others. Some might give better results than others. None of them are 'wrong'.

It's your pejorative attitude to others on here and your holier than thou attitude about telling people they are "doing it wrong" that Zilch-Wedlock people off. Me included. You are not the "woodworking police".

 

[Jacob]

The usual woodwork way of matching two pieces against each other is to put them together and see how they fit and make adjustments accordingly. It cuts out the middleman - the two pieces need only fit each other - not a third 'reference surface'.
Obviously your sanding process will work, in fact I used it a lot for finishing boxes with sawn off lids. But it's easy to dub them over a bit and defeat the object.
It's the 'reference surface' notion that I'm sceptical about - it seems to crop up a lot lately - even as the method for verifying the flatness of planed boards on workbench, with a perfectly planed 'reference surface' top.

You use a reference plane every single time you pick up a plane, run a board through a jointer or thicknesser, and every time you use a router table or table saw. Your sharpening stones are a reference plane, the inside edge of your square is a reference plane... you use them ALL the time........

Not in the particular sense I'm trying to describe above. I don't check the flatness of a board by resting it on a "reference plane" etc etc.
Non of my sharpening stones are flat either.
Feel free to disagree with anything I say - but no need to get shirty about it! :lol:

 

[dizjasta]

Sunny Bob I hope you are in good fettle . In an earlier post I got the impression that you would use your router to machine the angled staves using a dovetail bit. As has been hammered home by others this will create problems. If your preferred power tool is a router could you consider cutting annular rings from from sheet material to the dimensions you need and gluing the rings together to produce a stacked shape that you require? I feel I should go now back to my shelter in case another "Ear Raid" is imminent!!!! Regards dizjasta.

 

[sunnybob]

Has a cease fire been called?
i've been hunkered down with a bottle of vodka. mmmmm.

Ok, so far,

Plan A was to make a tub (for want of a better word) of staves that I could then use the bobbin sander on the inside to create an inner circle. After that I was going to make the outside also circular.

I'm fairly sure now that my cutter is 12.5, not 12. So the first attempt was useless. i redrew the circle following a utube video (much more down to my level than all the complicated stuff).
I got as far today as cutting the 16 (yup, I got sixteen sections now) pieces oversized. Then i used the belt sander and a tilting table to sand one side of each piece to 12 degrees, to fit the paper template.
I then cut off excess on the bandsaw table (tilted to 12 degrees), leaving enough spare to sand down to exactly fit the template.

Then I had to make a stop on the sander table so i didnt sand too much off any one piece.
Then,,,, aw carp, then the bluddy NVR switch on the sander broke.
can you believe that? on a saturday afternoon in a country where every shop closes for sunday.

I know my nearest tool shop has them for the eye watering sum of 75 euros. My other option is axminster (14 quid) but a 2 to 3 week delivery time.
I have one possible souece at an electrical wholesaler who knows me and is kind, but he doesnt open till 8 am monday.
So thats it, A wasted weekend.
Its so bad that even my shotgunning was cancelled today due to heavy rain.
Now where did i put the remains of that vodka bottle>

 

[niagra]

https://www.scribd.com/doc/64153983/Sta ... -Updated-2

 

[Yojevol]

Or , you could do this way:-

Use 16 staves. Set a jig to hold each at 22.5deg. Use a straight cutter. Glue up in pairs, then quadrants, semi- circles then full circle. Trim off by what ever means you have.
By this means you can get the jig pretty accurate and you can true up at each glue stage.
Brian

 

[sunnybob]

Its done.
Ended up with 16 and a 1/4 staves to get the correct diameter.
Lots of sanding. No. I mean LOTS of sanding!!

a couple jigs to make a turntable for the belt sander to get it pretty close to circular.

I hope I've kept you all amused on this brain teaser. I wish I was smart enough to understand some of the answers.

 

[galleywood]

Sunnybob
Could you post some pics please.

 

[sunnybob]

I dont take pics while I'm working, I dont have a video camera or mount and if i start faffing about with the SLR I lose the thread of what I'm on because everything I do is as i think of it and things change by the minute.

At the moment, its nowhere near finished and doesnt warrant a pic, but if the thing does finally get completely finished and not end in the scrap bin, i promise i will post a couple.

 

[MikeG.]

Or , you could do this way:-

Use 16 staves. Set a jig to hold each at 22.5deg. Use a straight cutter. Glue up in pairs, then quadrants, semi- circles then full circle. Trim off by what ever means you have.
By this means you can get the jig pretty accurate and you can true up at each glue stage.
Brian

The problem with this is locating each stave. How do you know when you are gluing up that you've got the stave in the right position? You aren't lining anything up corner to corner, and if they start creeping out of line you'll end up with a bigger or smaller circle. Pulling this tight with a band will deform it. I think this may be one of those ideas that's worth thinking about, but in reality it won't actually work.