Dr.Al
Old Oak
Thanks Scott
-
Hi all and welcome to TheWoodHaven2 brought into the 21st Century, kicking and screaming! We all have Alasdair to thank for the vast bulk of the heavy lifting to get us here, no more so than me because he's taken away a huge burden of responsibility from my shoulders and brought us to this new shiny home, with all your previous content (hopefully) still intact! Please peruse and feed back. There is still plenty to do, like changing the colour scheme, adding the banner graphic, tweaking the odd setting here and there so I have added a new thread in the 'Technical Issues, Bugs and Feature Requests' forum for you to add any issues you find, any missing settings or just anything you'd like to see added/removed from the feature set that Xenforo offers. We will get to everything over the coming weeks so please be patient, but add anything at all to the thread I mention above and we promise to get to them over the next few days/weeks/months. In the meantime, please enjoy!
Garden Table
- Thread starter Dr.Al
- Start date
The accuracy and precision of those joints is something to behold but, sorry there is a but, there is no mechanical strength, only the glue holding them in place. If/when the wood moves and the glue detoriates what stops them from just falling out?
Dr.Al
Old Oak
It's the same point that Richard made earlier in the thread & it's a fair comment. Too late to do anything about it now though.
Add a screw up through the dovetail, maybe angled towards the table's perimeter to maximise screw length and downward force resistance, preferably stainless steel. Slainte.
Dr.Al
Old Oak
Yeah, that's probably the best option; that or a steeply angled dowel. However, I can't think of a reason why I need to do it until it becomes a problem so I'll probably just leave it as-is and do a screw repair if/when required.
Dr.Al
Old Oak
My time in the workshop has been a bit intermittent again, but I've been gradually working on the remaining slats. I started by picking out a few of the planks I'd prepared and marking up suitable lengths ready for chopping up:
The process from there was fairly similar to before. On the first one, I marked the shoulders (using the 3D-printed spacers) as before:
I then added 20 mm onto each end and then sawed the ends off:
On the later ones, I shot one end square, then marked a square shoulder line 20 mm in, then lined that up with the table to mark the angled shoulder line. That seemed a slightly quicker way of doing it, but it didn't make much difference.
For the angled ends, I decided to stick with a planed (on the shooting board) end, but as it was about 45° I screwed a little plywood angle block to my shooting board. The angles weren't guaranteed to be exactly 45° (as they're marked off the table rather than with a gauge), so I fixed the angle block a millimetre or so away from the main fence and then I could adjust the angle by inserting a steel rule somewhere along the back edge:
Strictly speaking, the ends don't have to be parallel with the shoulder line, so I could have just shot the ends at 45°, but it didn't add much time to make them parallel.
The next step was to mark the depths and saw / chisel / router plane the half laps:
After the first one, I realised that router planing the angled lap would be much easier if the grain direction was such that it would give a smooth finish when planed towards the outer point (i.e. if the grain ran from left-to-right in the photo above). In that direction, there's a much bigger surface for the router plane to sit on. All the remaining pieces were oriented that way round.
The dovetails are very slightly offset from the centre (as that seems to make them look a bit better, although no-one will ever see them as they'll be under the table!). As I had eight of them to do, I thought it was worth printing out a couple of little gauges to make marking them up very quick and easy:
The gauges sit on the end of the plank and can be easily centred (as the little "wings" are about the same width as the planks) and then a pencil can be used to draw the shape of the cut-out:
They can then be chopped out as before. After transferring the marks to the body, I then needed to chop out the sockets. The angled ones were quite a lot more faff than the simple sockets, but the process was made a bit easier by bringing out my skewed chisels and a 3 mm dovetail chisel. I also used the pointy tip on the router plane to help get as far into the corners as possible:
Even with those, I still ended up using a knife to cut the very last little bit in the acute corner (on both the socket and the acute corner on the tail board).
The first one in place:
Still a long way to go:
I didn't take much in the way of photos of the remaining pieces as it was just more of the same (with the addition of another, home-shaped, skew chisel with a more extreme skew). There are some little gaps here and there, but given I'd never cut a dovetail half-lap before embarking on this project, I'm quite pleased with how they all came out. Time will tell whether they remain in place or if the boards fall out if/when the glue fails.
I haven't glued them in place yet, but it's definitely looking much more like a table top now:
The next job is to get started on the base. I've got an idea for the design of this, but I have no idea whether I'm going to be able to do it with the tools I have available. There's also a lot more work needed on the details of the design, but I'll worry about then when I've worked out whether it's even possible!
The process from there was fairly similar to before. On the first one, I marked the shoulders (using the 3D-printed spacers) as before:
I then added 20 mm onto each end and then sawed the ends off:
On the later ones, I shot one end square, then marked a square shoulder line 20 mm in, then lined that up with the table to mark the angled shoulder line. That seemed a slightly quicker way of doing it, but it didn't make much difference.
For the angled ends, I decided to stick with a planed (on the shooting board) end, but as it was about 45° I screwed a little plywood angle block to my shooting board. The angles weren't guaranteed to be exactly 45° (as they're marked off the table rather than with a gauge), so I fixed the angle block a millimetre or so away from the main fence and then I could adjust the angle by inserting a steel rule somewhere along the back edge:
Strictly speaking, the ends don't have to be parallel with the shoulder line, so I could have just shot the ends at 45°, but it didn't add much time to make them parallel.
The next step was to mark the depths and saw / chisel / router plane the half laps:
After the first one, I realised that router planing the angled lap would be much easier if the grain direction was such that it would give a smooth finish when planed towards the outer point (i.e. if the grain ran from left-to-right in the photo above). In that direction, there's a much bigger surface for the router plane to sit on. All the remaining pieces were oriented that way round.
The dovetails are very slightly offset from the centre (as that seems to make them look a bit better, although no-one will ever see them as they'll be under the table!). As I had eight of them to do, I thought it was worth printing out a couple of little gauges to make marking them up very quick and easy:
The gauges sit on the end of the plank and can be easily centred (as the little "wings" are about the same width as the planks) and then a pencil can be used to draw the shape of the cut-out:
They can then be chopped out as before. After transferring the marks to the body, I then needed to chop out the sockets. The angled ones were quite a lot more faff than the simple sockets, but the process was made a bit easier by bringing out my skewed chisels and a 3 mm dovetail chisel. I also used the pointy tip on the router plane to help get as far into the corners as possible:
Even with those, I still ended up using a knife to cut the very last little bit in the acute corner (on both the socket and the acute corner on the tail board).
The first one in place:
Still a long way to go:
I didn't take much in the way of photos of the remaining pieces as it was just more of the same (with the addition of another, home-shaped, skew chisel with a more extreme skew). There are some little gaps here and there, but given I'd never cut a dovetail half-lap before embarking on this project, I'm quite pleased with how they all came out. Time will tell whether they remain in place or if the boards fall out if/when the glue fails.
I haven't glued them in place yet, but it's definitely looking much more like a table top now:
The next job is to get started on the base. I've got an idea for the design of this, but I have no idea whether I'm going to be able to do it with the tools I have available. There's also a lot more work needed on the details of the design, but I'll worry about then when I've worked out whether it's even possible!
Dr.Al
Old Oak
I've been playing with more sweet chestnut since finishing (apart from glue and final planing) the table top and I'm now starting to feel a little bit more confident that my plan for the base might be achievable. This seemed like a good time to write about it in the hope that any major flaws will be pointed out before I do anything irreversible.
The plan is to have a column (probably circular, possibly octagonal) that goes up the middle. There will then be four legs going diagonally down to the ground and a cross thing at the top to support the table. This is what I'm hoping it'll look like:
I've drawn that with a mostly circular column with an octagonal bit at the bottom, but more likely I'll make it either entirely circular or entirely octagonal. Octagonal would make the mortice and tenon joints for the legs a lot easier, but I think circular might look better. Another compromise option would be to make it circular and then plane some flats near the bottom, just wide enough for the legs.
Thoughts/opinions welcome as always.
Making that central column would involve eight planks glued together and then (if I go for a circular column) turned into a cylinder on the home-made lathe (which I think has a bed that is **just** long enough!). I did a bit of research into making segmented turning blanks and all the advice said "make the blanks by setting your table saw's blade over at an angle". I don't have a table saw any more and I really don't want there to be a reason why I need one!
As an alternative, I tried doing a web search for how to make segments without a table saw. The general advice seemed to be "find someone with a table saw who can do it for you". I saw one post where someone asked if there was any way to do it with just a bandsaw and there were eight or nine answers, all of which could be summarised as "no". Not quite what I was hoping for!
Nevertheless, a plan gradually formed in my head while I was working on the table top and with it out of the way I decided to have a go. I started by preparing quite a lot more sweet chestnut plans. I had one (planed on four sides) left over from when I prepared the material for the slats and the size seemed reasonable: according to my Regular Polygon Calculator, it should allow me to make a cylinder about 115 mm diameter, with an octagonal hole through the middle that's about 55 mm across-flats. Therefore, I decided to plane all the others to the same size as that one.
All done and ready for the next stage:
With the exception of the one I'd previously prepared, I labelled one side of each piece as "IN": this was chosen based on blemishes/knots mainly: if one side had more imperfections, I picked that as the side that would become the inside of the "tube":
While I could possibly attack this problem purely with hand tools, I think it would stretch my skills a bit beyond what I'm capable of. It would also take a very long time, which would be frustrating if all the wood ends up as scrap at the end (a very real possibility!). Instead, I decided to use the bandsaw. First of all, I tilted the table over to 22.5° (measured with one of those little digital angle gauges that come in handy for all sorts of things) and moved the fence over to the down-hill side of the table. The latter job involved flipping the support bar over as it's longer on one end than the other and the longer end was pointing up-hill.
The fence support rail is only just long enough for what I need to do, but it is long enough.
The first cut was done by running a flat side of the narrowest piece (which was the one I'd prepared as part of the slat preparation) against the fence and feeding the plank through the saw:
That left a slightly rough face (as you'd expect from a bandsaw) but at roughly the right angle.
During the week, I'd 3D-printed some little supports, each made in three pieces - a main support, a 20 mm diameter cylinder to fit the dog holes on the bench and a little wedge that fits into the notch on the main support:
Those supports drop into the dog holes on the bench and with the wedge inserts removed and a handy bit of scrap clamped in place as an end-stop...
... I was ready to to plane that rough surface away:
The planing job was quite ergonomic as I'm planing parallel with the bench top. I set a sliding bevel to 67.5° and used it, in combination with the slightly cambered blade on the #7, to prepare the edge in much the same was as I would normally prepare an edge square to a face:
Before starting on the second side of that plank, I fed the other seven pieces through the bandsaw to do the rough cut on the first side of each of them:
The reason for that is that I could now set the bandsaw's fence up for the second cut and (give or take any planing discrepancies) they should all end up the same width.
When I came to start the second cut on the first piece, I found that the little cut-out (for mounting accessories) in the fence was at almost exactly the wrong height:
I didn't think it was a good idea to try sawing like that, so I had a rummage around in my various drawers and found a piece of 8 mm × 40 mm × 500 mm gauge plate. That got placed against the fence and then I used a clamp to hold an off-cut of plywood onto the end of the fence to stop the piece of gauge plate sliding forward with the wood:
I could then feed the plank through with the corner of the first prepared side pushed up against the steel bar:
The little wedge pieces were then inserted into the support blocks and the support blocks spun round (to suit the grain direction) and I could repeat the process of preparing an edge at the right angle, with #7 and sliding bevel to check:
That brings us up-to-date. Tomorrow, I'll prepare the other seven blanks (which, so far have only had one side rough-cut). Then they'll need to be joined together. At the moment, I'm thinking I'll just do that with glue (and lots and lots of masking tape!). Alternatively, I could put some sort of spline in each joint (perhaps by ploughing a groove in the edges I've just planed and then inserting a long and relatively thin bit of wood in the groove). Do you think this is necessary? I think of long edge joints as being strong enough with just glue, but I'd like to hear other people's thoughts before it's too late!
Going back to the model, this is what it will (probably, hopefully?) look like with the top removed:
I haven't really figured how to attach those cross pieces to the central column. At the moment, the CAD model has some half-lap joints where the cross-pieces intersect and some sort of (probably impossibly to cut how I've drawn it) mortice and tenon joining the column to the cross pieces:
As well as the thoughts on cylinder vs octagon and on splines vs simple glue joints, I'd also welcome any thoughts on this support design and the associated joinery.
The plan is to have a column (probably circular, possibly octagonal) that goes up the middle. There will then be four legs going diagonally down to the ground and a cross thing at the top to support the table. This is what I'm hoping it'll look like:
I've drawn that with a mostly circular column with an octagonal bit at the bottom, but more likely I'll make it either entirely circular or entirely octagonal. Octagonal would make the mortice and tenon joints for the legs a lot easier, but I think circular might look better. Another compromise option would be to make it circular and then plane some flats near the bottom, just wide enough for the legs.
Thoughts/opinions welcome as always.
Making that central column would involve eight planks glued together and then (if I go for a circular column) turned into a cylinder on the home-made lathe (which I think has a bed that is **just** long enough!). I did a bit of research into making segmented turning blanks and all the advice said "make the blanks by setting your table saw's blade over at an angle". I don't have a table saw any more and I really don't want there to be a reason why I need one!
As an alternative, I tried doing a web search for how to make segments without a table saw. The general advice seemed to be "find someone with a table saw who can do it for you". I saw one post where someone asked if there was any way to do it with just a bandsaw and there were eight or nine answers, all of which could be summarised as "no". Not quite what I was hoping for!
Nevertheless, a plan gradually formed in my head while I was working on the table top and with it out of the way I decided to have a go. I started by preparing quite a lot more sweet chestnut plans. I had one (planed on four sides) left over from when I prepared the material for the slats and the size seemed reasonable: according to my Regular Polygon Calculator, it should allow me to make a cylinder about 115 mm diameter, with an octagonal hole through the middle that's about 55 mm across-flats. Therefore, I decided to plane all the others to the same size as that one.
All done and ready for the next stage:
With the exception of the one I'd previously prepared, I labelled one side of each piece as "IN": this was chosen based on blemishes/knots mainly: if one side had more imperfections, I picked that as the side that would become the inside of the "tube":
While I could possibly attack this problem purely with hand tools, I think it would stretch my skills a bit beyond what I'm capable of. It would also take a very long time, which would be frustrating if all the wood ends up as scrap at the end (a very real possibility!). Instead, I decided to use the bandsaw. First of all, I tilted the table over to 22.5° (measured with one of those little digital angle gauges that come in handy for all sorts of things) and moved the fence over to the down-hill side of the table. The latter job involved flipping the support bar over as it's longer on one end than the other and the longer end was pointing up-hill.
The fence support rail is only just long enough for what I need to do, but it is long enough.
The first cut was done by running a flat side of the narrowest piece (which was the one I'd prepared as part of the slat preparation) against the fence and feeding the plank through the saw:
That left a slightly rough face (as you'd expect from a bandsaw) but at roughly the right angle.
During the week, I'd 3D-printed some little supports, each made in three pieces - a main support, a 20 mm diameter cylinder to fit the dog holes on the bench and a little wedge that fits into the notch on the main support:
Those supports drop into the dog holes on the bench and with the wedge inserts removed and a handy bit of scrap clamped in place as an end-stop...
... I was ready to to plane that rough surface away:
The planing job was quite ergonomic as I'm planing parallel with the bench top. I set a sliding bevel to 67.5° and used it, in combination with the slightly cambered blade on the #7, to prepare the edge in much the same was as I would normally prepare an edge square to a face:
Before starting on the second side of that plank, I fed the other seven pieces through the bandsaw to do the rough cut on the first side of each of them:
The reason for that is that I could now set the bandsaw's fence up for the second cut and (give or take any planing discrepancies) they should all end up the same width.
When I came to start the second cut on the first piece, I found that the little cut-out (for mounting accessories) in the fence was at almost exactly the wrong height:
I didn't think it was a good idea to try sawing like that, so I had a rummage around in my various drawers and found a piece of 8 mm × 40 mm × 500 mm gauge plate. That got placed against the fence and then I used a clamp to hold an off-cut of plywood onto the end of the fence to stop the piece of gauge plate sliding forward with the wood:
I could then feed the plank through with the corner of the first prepared side pushed up against the steel bar:
The little wedge pieces were then inserted into the support blocks and the support blocks spun round (to suit the grain direction) and I could repeat the process of preparing an edge at the right angle, with #7 and sliding bevel to check:
That brings us up-to-date. Tomorrow, I'll prepare the other seven blanks (which, so far have only had one side rough-cut). Then they'll need to be joined together. At the moment, I'm thinking I'll just do that with glue (and lots and lots of masking tape!). Alternatively, I could put some sort of spline in each joint (perhaps by ploughing a groove in the edges I've just planed and then inserting a long and relatively thin bit of wood in the groove). Do you think this is necessary? I think of long edge joints as being strong enough with just glue, but I'd like to hear other people's thoughts before it's too late!
Going back to the model, this is what it will (probably, hopefully?) look like with the top removed:
I haven't really figured how to attach those cross pieces to the central column. At the moment, the CAD model has some half-lap joints where the cross-pieces intersect and some sort of (probably impossibly to cut how I've drawn it) mortice and tenon joining the column to the cross pieces:
As well as the thoughts on cylinder vs octagon and on splines vs simple glue joints, I'd also welcome any thoughts on this support design and the associated joinery.
Dr.Al
Old Oak
Well, I was hoping for more of a response to all my questions / requests for help than I got, but I guess I'll just have to carry on and hope for the best. This morning I carried on planing the angles onto the various planks until they were all the same width and at what appeared to be the correct angle as far as the sliding bevel showed:
I arranged all of them on the bench with the grain direction pointing the same way (just because it seemed like a good idea) and with what seemed to be the best pattern match from one plank to the next. I then labelled them A-H:
As a trial, I added some masking tape and pulled them round into an octagon:
This shows what they look like from one end:
There are some small gaps, suggesting the angles weren't all absolutely perfect, but it should be fairly easy to sort out: I think I can pick any plank and plane it (with the #7 over to one side or the other as appropriate) and it'll close up the gap. The result won't be a perfect octagon (as I'll potentially be planing one plank more than the others), but the effect should be small enough not to be noticeable.
If you look at the size of it versus the rule, you might notice it's a bit bigger than I said in yesterday's post (where I said it would make a 115 mm diameter cylinder with a 55 mm hole through the middle). That was just a stupid mistake as I'd loaded the polygon calculator but left it on the default of a hexagon rather than changing it to an octagon. With the calculator set-up properly, it shows the outer octagon to be 162 mm and hence the inner one 102 mm across-flats, which matches the actual size of the column.
To try to help make the decision on octagonal vs cylindrical column, I thought I'd do the model properly for each:
In the meantime, the next job is to tweak the angles slightly to minimise gaps in the joints, trim the planks to length and then try to make the decision on whether to try to add splines of some sort. Adding splines will be far from easy, so I'll only do it if I (or someone else) can convince myself that it's necessary.
I arranged all of them on the bench with the grain direction pointing the same way (just because it seemed like a good idea) and with what seemed to be the best pattern match from one plank to the next. I then labelled them A-H:
As a trial, I added some masking tape and pulled them round into an octagon:
This shows what they look like from one end:
There are some small gaps, suggesting the angles weren't all absolutely perfect, but it should be fairly easy to sort out: I think I can pick any plank and plane it (with the #7 over to one side or the other as appropriate) and it'll close up the gap. The result won't be a perfect octagon (as I'll potentially be planing one plank more than the others), but the effect should be small enough not to be noticeable.
If you look at the size of it versus the rule, you might notice it's a bit bigger than I said in yesterday's post (where I said it would make a 115 mm diameter cylinder with a 55 mm hole through the middle). That was just a stupid mistake as I'd loaded the polygon calculator but left it on the default of a hexagon rather than changing it to an octagon. With the calculator set-up properly, it shows the outer octagon to be 162 mm and hence the inner one 102 mm across-flats, which matches the actual size of the column.
To try to help make the decision on octagonal vs cylindrical column, I thought I'd do the model properly for each:
In the meantime, the next job is to tweak the angles slightly to minimise gaps in the joints, trim the planks to length and then try to make the decision on whether to try to add splines of some sort. Adding splines will be far from easy, so I'll only do it if I (or someone else) can convince myself that it's necessary.
Mike G
Petrified Pine
Sorry, Al, I've been busy.....
Personally, I'd go for the octagon. The rectilinear feet look a bit incongrous with a cylindrical stem. If you were to shape the legs/ feet in a more traditional manner, then I think the round would be OK......although, again, it might look better shaped/ tapered. My next question would be why the need for double cross-support-thingies at the top. It's that complication which is making the joint design difficult. You're clearly seeking to align them with the slats in the table-top, but how about just rotating the top 45 degrees? That way you'd be giving support to those weaker parts of the table top, and helping hold the dovetails in their housing.....and of course, you'd only need a single cross piece in each direction, which could just drop into a housing at the top of the octagonal column.
I'd glue up the octagon in 2 halves, and then do your final adjustment planing when mating those two together. There's absolutely no need for any splines, dowels, dominoes, or anything else, for those joins.
Personally, I'd go for the octagon. The rectilinear feet look a bit incongrous with a cylindrical stem. If you were to shape the legs/ feet in a more traditional manner, then I think the round would be OK......although, again, it might look better shaped/ tapered. My next question would be why the need for double cross-support-thingies at the top. It's that complication which is making the joint design difficult. You're clearly seeking to align them with the slats in the table-top, but how about just rotating the top 45 degrees? That way you'd be giving support to those weaker parts of the table top, and helping hold the dovetails in their housing.....and of course, you'd only need a single cross piece in each direction, which could just drop into a housing at the top of the octagonal column.
I'd glue up the octagon in 2 halves, and then do your final adjustment planing when mating those two together. There's absolutely no need for any splines, dowels, dominoes, or anything else, for those joins.
Dr.Al
Old Oak
No problem Mike; thanks for the reply: I really appreciate it. The design side of woodworking still feels very alien to me at the moment.
Sorry for asking stupid questions, but what do you mean by "a more traditional manner"? I'm not averse to changing the foot design at all; the feet were designed purely as the first way I thought of doing them. Octagonal would certainly be a lot easier though and shaping/tapering a cylinder is starting to sound rather challenging to me!
Hmmm, that's a very good point. I'd drawn it that way as I thought it best to hide the supports underneath the cross bars. I'll draw it at 45° and see how it looks. As you say, it'll make the joinery far simpler.
That's a great idea, why didn't I think of that?! I might still do a small tweak before gluing the halves together (as I can see one place where it will definitely help), but doing it in two stages will make everything more straightforward). Thank you.
Thanks again
Mike G
Petrified Pine
spb
Nordic Pine
To my mind, a cylindrical turned pedestal goes with curved, rounded feet with optional carved details, but that's also something I wouldn't expect to see outdoors. For a garden table, I'd expect the aesthetic to be more solid than slender and graceful, and the octagon works much better.
Dr.Al
Old Oak
Thanks both.
Gotcha, thanks. I don't particularly like that style, so I think I'll stick with octagonal.
That makes a lot of sense, thanks.
Here's what it looks like with an octagonal base and a single cross:
From above looking straight down (ignore the slightly weird patterning on the top - something went weird with the grain pattern and I couldn't be bothered to fix it):
Construction detail came out a bit small, but is hopefully clear enough if you click the full screen icon thing in the bottom right:
Gotcha, thanks. I don't particularly like that style, so I think I'll stick with octagonal.
That makes a lot of sense, thanks.
Here's what it looks like with an octagonal base and a single cross:
From above looking straight down (ignore the slightly weird patterning on the top - something went weird with the grain pattern and I couldn't be bothered to fix it):
Construction detail came out a bit small, but is hopefully clear enough if you click the full screen icon thing in the bottom right:
duke
Old Oak
I also like the octagonal column , here is a thought. Feel free to discard it.
Dr.Al
Old Oak
I've just realised what should have been more obvious at first glance: that cross design won't work as-is. It covers up the central parasol hole.
Back to the drawing board I think. I guess I could just have four arms with M&T joints at the top.
Thanks Duke. I do like the shaping to the legs: it certainly adds a bit of interest over my very simplistic design. I'll see if I can incorporate something like that.
I don't think I can do the lower bar though, for two reasons. Firstly, the column I've made wouldn't be long enough to go that far down (I've made it with the gap at the bottom in mind). Secondly, I want to leave space for our cast iron parasol stand at the bottom.
Dr.Al
Old Oak
Now I really don't think I would have thought of that! I'll model it & see how it looks; thanks!
I did suggest the placing the parasol in the middle of the table was a bad idea.
.
A thought has just occurred to me, probably daft. The parasol will go down through the centre of the column. Is there room between the legs for an additional stand or weight to stop the whole thing toppling over in the wind? Or will you just retreat indoors if a breeze picks up?
.A thought has just occurred to me, probably daft. The parasol will go down through the centre of the column. Is there room between the legs for an additional stand or weight to stop the whole thing toppling over in the wind? Or will you just retreat indoors if a breeze picks up?
Dr.Al
Old Oak
I did suggest the placing the parasol in the middle of the table was a bad idea.
You certainly did
For the parasol? If so, yes. Part of the reason for the gap at the bottom of the column is to allow space for a cast iron parasol stand underneath. We've got one that looks a little bit like this random one I found on the web:
The intention is that it'll sit underneath the table, a bit like this (I haven't measured the dimensions of the parasol base yet, so this is just a rough idea - I will measure it before making the legs so it doesn't clash like the CAD model!):
Dr.Al
Old Oak
This is what it would look like with an off-set cross:
Top view:
(the double cross thing is because you can see the legs underneath the offset support arms)
and with a mortice/tenon cross:
Top view:
Top view:
(the double cross thing is because you can see the legs underneath the offset support arms)
and with a mortice/tenon cross:
Top view:
Mike G
Petrified Pine
I see some sliding dovetails in your future, Al:
They'll need to be quite deep at the octagon to be strong enough.....and the complication is that the female side of the joint is with the grain.
They'll need to be quite deep at the octagon to be strong enough.....and the complication is that the female side of the joint is with the grain.
Steve Maskery
Old Oak
I think you are asking a great deal of those feet-to-column joints, TBH. Looks fab, I agree, but I'm not convinced it will stand up to prolonged use, especially outside.
Sorry
S
Sorry
S
Dr.Al
Old Oak
That's definitely not out of the question (and it wouldn't be the first time I've cut sliding dovetails). I would be quite nervous about the grain issue: a tapered dovetail is going to be trying quite hard to split the wood apart.
I guess the advantage is that they're stronger than a mortice and tenon would be?
I think you are asking a great deal of those feet-to-column joints, TBH. Looks fab, I agree, but I'm not convinced it will stand up to prolonged use, especially outside.
Sorry
S
Don't be sorry: I'm much happier with people raising concerns when I share the CAD model rather than waiting until I later share the photos of the finished joints!
What would you suggest instead for attaching the feet?
Mike G
Petrified Pine
Well, you could do wedged through M&Ts....multiple per bracket/ support......but you'd have to do them before gluing the octagon together, obviously, to be able to get the wedges in. And ordinary (un-wedged) M&Ts would be relying solely on the glue.
Last edited:
Dr.Al
Old Oak
Dr.Al
Old Oak
Me neither. I'm sure I'll try them in some scrap at some point, but never in something that's had this much work put in
Pete Maddex
Old Oak
Weld up some stainless for the base, a socket to hold the column with some nice curved legs.
Pete
Pete
Steve Maskery
Old Oak
Well personally I'd have four straight legs and rails. Strong as an ox and no complex joinery. The one I made a gazillion years ago (still in use, I understand) was made like a gate-leg table so that it could be folded away for winter storage.
There was still a square hole down the centre to accommodate the umbrella.
I'd do the same again, but maybe not bother about making it foldable. I don't think we ever took it down, we just covered it over for the winter.
S
Are you sure you really need that cross arrangement at the top of the column? The table top looks like it will be a pretty strong item.
Why not just fit a short horizontal block say 300mm long and the width of the 2 middle cross bars to the column and screw the top on through that. have a matching pole hole in the middle of it.
I'm thinking those arms you have will be a nuisance when sitting at the table if you cross your legs and bang your knee on them.
Oh and how do you secure the parasol to the cast base with its clamp screw inside the column - do you need a base? just infil the bottom of the column with a piece with a hole in it.
My table took ages using machines - couldn't imagine doing it with hand tools. keep going
Why not just fit a short horizontal block say 300mm long and the width of the 2 middle cross bars to the column and screw the top on through that. have a matching pole hole in the middle of it.
I'm thinking those arms you have will be a nuisance when sitting at the table if you cross your legs and bang your knee on them.
Oh and how do you secure the parasol to the cast base with its clamp screw inside the column - do you need a base? just infil the bottom of the column with a piece with a hole in it.
My table took ages using machines - couldn't imagine doing it with hand tools. keep going
Al, if you do decide to go for splayed legs jointed into your very tidy column, have a look at Roy Underhill's demonstration of how to cut suitable joints using hand tools here
video.pbsnc.org
He uses a template (cut from tin as he seems not to have a 3D printer!) and his table is on a smaller scale than yours, but I'm sure you could beef it up suitably and make a neat job of it.
The Woodwright's Shop | Hancock Pedestal Table | Season 29 | Episode 10
With its turned top and dovetailed legs, this walnut tripod table is a American design.
He uses a template (cut from tin as he seems not to have a 3D printer!) and his table is on a smaller scale than yours, but I'm sure you could beef it up suitably and make a neat job of it.
Dr.Al
Old Oak
I guess that would be one perfectly valid way of doing it, but I'd rather make it out of wood (also stainless steel is obscenely expensive!)
Dr.Al
Old Oak
The original plan had been to have four straight legs and up until not very long ago I was still planning that. I agree that it would be stronger, but I also think that on a small table like this (from your photos, it looks like yours is a lot bigger), the legs (especially with stretchers) would get in the way of my lanky legs a lot I think. Also, it's nice to experiment. If it proves not to last very long then I'll just have to write it down to experience.
To be honest, I'll be glad to see the back of this project now!
Dr.Al
Old Oak
It's a fair point, but I don't think it'll be any worse than any other rail arrangement. I'm also planning to make the table quite a bit taller than a typical garden table (we're both very tall) so there should be leg clearance. Maybe I could get rid of the crosses, but it feels a bit risky to me.
Getting to the parasol screw shouldn't be a problem: there are two of them and one is fairly low down and should remain accessible. I never tighten both (especially since I 3D-printed an insert for the parasol base to make it about the right diameter hole for the parasol, so it arguably doesn't need the screw at all).
I possibly could get rid of it, but I'd rather design with it in mind and then if it turns out it's needed then all's fine. If it turns out the table feels robust enough to support the parasol without it, then I can perhaps get rid of the base.
My table took ages using machines - couldn't imagine doing it with hand tools. keep going
This hasn't been entirely hand tools (I've used the bandsaw a fair amount). I have to admit that I'm getting a bit fed up of this project now, so I'm not ruling out shoving some of the remaining planks through the thicknesser to speed things up a bit. I don't have a surface planer and I hate sanding, so there will still be plenty of hand planing even if I do that.
Dr.Al
Old Oak
Thanks Andy, I'll give it a watch. At the moment, I'm leaning towards @Mike G's suggestion of sliding dovetails for the rails at the top, but it'll be interesting to see what he used for the legs.Al, if you do decide to go for splayed legs jointed into your very tidy column, have a look at Roy Underhill's demonstration of how to cut suitable joints using hand tools here
The Woodwright's Shop | Hancock Pedestal Table | Season 29 | Episode 10
With its turned top and dovetailed legs, this walnut tripod table is a American design.
He uses a template (cut from tin as he seems not to have a 3D printer!) and his table is on a smaller scale than yours, but I'm sure you could beef it up suitably and make a neat job of it.
Dr.Al
Old Oak
Following an excellent suggestion from @Mike G, I decided to glue the shaft together in two halves and then (when the glue was dry) tweak the fit before gluing the two halves together. I'd wondered about cutting each piece to length before gluing them together, but in the end I decided to leave that until later (partly because I still haven't decided exactly how long the column needs to be!)
I applied some masking tape along the long edges:
Then used a recently acquired little silicone glue roller to apply glue along the edges:
The pieces then got taped together:
After applying the first three bits of tape to hold it in place, I used some plane shavings to scrub out the bulk of the squeeze-out:
The two halves were then left to dry out overnight.
While I was in a gluing frame of mind, I also dismantled the table top, applied glue and then reassembled it, but I didn't take any photos of that as they would look very much like the ones I posted earlier in this thread.
The next morning, I removed all the masking tape:
After offering the two halves up to each other, I could see that they needed a tiny bit of adjustment to get good mating faces. For a while I pondered finding a way to support the halves on a corner (so that I could plane the edges horizontally), but in the end I just clamped the parts to the bench and planed at an angle:
The two halves could then be glued and taped together:
After leaving them overnight again, I stripped off all the masking tape:
There was a bit of squeeze-out and a few blemishes, so it seemed a good time to pass a plane over the outside surfaces. I'd very carefully arranged all the planks so that the planing direction was the same (i.e. towards the top or towards the bottom, but not both), but hadn't made a note of which way that was. Having looked at the assembled column I realised that the edges were of course no longer visible (having been glued together), so I couldn't work out the grain direction. Not a big deal: I just had to guess and see whether there was any tear out. I got it right first time, but if I hadn't I just would have had to flip it round and then I would have known the correct direction for all the remaining sides.
I'd 3D-printed a little marking gauge to help cutting the ends of the column. For now I've only cut one end:
After marking it up, I noticed a bit of a nasty blemish on one of the corners, so I moved the line further along a bit and marked it out again. I then clamped the column to the bench and sawed all the way round, working on two edges at a time and sawing down to this point:
After I'd gone all the way round, I sawed down until I'd got to the half-way point, keeping an eye on four edges at a time:
Finally, I shoved a scrap bit of plywood into the hole in the column (to stop the saw from falling onto the F-clamp) and then finished the cut:
The vice jaws open **just** enough to loosely grip the column across the flats, so I held it in the vice (with an arm for support) and used my home-made block plane to clean up the saw cuts a bit.
With a fairly clean end surface, I couldn't resist balancing the table top on top just to get an idea of what it'll look like:
The next job was to start preparing the arms that will attach to the column and support the table. I pulled out what was left of the big plank of sweet chestnut and marked it up for a couple of arms:
After chopping that bit off, I did the same again for another pair. I could then get on with planing them:
I've planed both faces of all four pieces. I've also planed the square edge (the one that will support the table) and used the shooting board to square up the long end grain face. For now I've left the short end grain face (so I can decide later how long it needs to be) and also the angled face (which I'm considering making slightly curved).
I picked a fairly arbitrary 12 mm for the depth of the tapered sliding dovetail I'm considering using to attach the arms and marked a shoulder line all round on one piece only:
I think I'll have a go at cutting one dovetail with this piece (with the mating socket in a bit of scrap) and, if it feels like it'll work well, I can continue with the real socket and the rest of the pieces. That should minimise the chance of scrap.
I've also marked (with pencil) an initial go at a rather arbitrary taper angle (which, at least for the next hour or so, can easily be changed, so comments are welcome!).
That tapers in about 4 mm over the 115 mm width of the arm (an angle of about 2°).
I applied some masking tape along the long edges:
Then used a recently acquired little silicone glue roller to apply glue along the edges:
The pieces then got taped together:
After applying the first three bits of tape to hold it in place, I used some plane shavings to scrub out the bulk of the squeeze-out:
The two halves were then left to dry out overnight.
While I was in a gluing frame of mind, I also dismantled the table top, applied glue and then reassembled it, but I didn't take any photos of that as they would look very much like the ones I posted earlier in this thread.
The next morning, I removed all the masking tape:
After offering the two halves up to each other, I could see that they needed a tiny bit of adjustment to get good mating faces. For a while I pondered finding a way to support the halves on a corner (so that I could plane the edges horizontally), but in the end I just clamped the parts to the bench and planed at an angle:
The two halves could then be glued and taped together:
After leaving them overnight again, I stripped off all the masking tape:
There was a bit of squeeze-out and a few blemishes, so it seemed a good time to pass a plane over the outside surfaces. I'd very carefully arranged all the planks so that the planing direction was the same (i.e. towards the top or towards the bottom, but not both), but hadn't made a note of which way that was. Having looked at the assembled column I realised that the edges were of course no longer visible (having been glued together), so I couldn't work out the grain direction. Not a big deal: I just had to guess and see whether there was any tear out. I got it right first time, but if I hadn't I just would have had to flip it round and then I would have known the correct direction for all the remaining sides.
I'd 3D-printed a little marking gauge to help cutting the ends of the column. For now I've only cut one end:
After marking it up, I noticed a bit of a nasty blemish on one of the corners, so I moved the line further along a bit and marked it out again. I then clamped the column to the bench and sawed all the way round, working on two edges at a time and sawing down to this point:
After I'd gone all the way round, I sawed down until I'd got to the half-way point, keeping an eye on four edges at a time:
Finally, I shoved a scrap bit of plywood into the hole in the column (to stop the saw from falling onto the F-clamp) and then finished the cut:
The vice jaws open **just** enough to loosely grip the column across the flats, so I held it in the vice (with an arm for support) and used my home-made block plane to clean up the saw cuts a bit.
With a fairly clean end surface, I couldn't resist balancing the table top on top just to get an idea of what it'll look like:
The next job was to start preparing the arms that will attach to the column and support the table. I pulled out what was left of the big plank of sweet chestnut and marked it up for a couple of arms:
After chopping that bit off, I did the same again for another pair. I could then get on with planing them:
I've planed both faces of all four pieces. I've also planed the square edge (the one that will support the table) and used the shooting board to square up the long end grain face. For now I've left the short end grain face (so I can decide later how long it needs to be) and also the angled face (which I'm considering making slightly curved).
I picked a fairly arbitrary 12 mm for the depth of the tapered sliding dovetail I'm considering using to attach the arms and marked a shoulder line all round on one piece only:
I think I'll have a go at cutting one dovetail with this piece (with the mating socket in a bit of scrap) and, if it feels like it'll work well, I can continue with the real socket and the rest of the pieces. That should minimise the chance of scrap.
I've also marked (with pencil) an initial go at a rather arbitrary taper angle (which, at least for the next hour or so, can easily be changed, so comments are welcome!).
That tapers in about 4 mm over the 115 mm width of the arm (an angle of about 2°).
Dr.Al
Old Oak
After lunch I got going on the first joint. I started by using a Western rip saw to cut (working directly on the line):
I found that really hard work; perhaps my rip saw needs sharpening (not that I've used it much since I bought it). For the shoulder line I did my usual thing of sawing slightly away from the line...
... and then having a pleasant few minutes chopping with a chisel (one of my favourite jobs):
I also cut a shoulder at the bottom end of the tail, again sawing slightly clear of the shoulder line and tidying up with a chisel:
As I was a little concerned about splitting wood (and, for that matter, just generally making a mess of the joinery!), I decided to start by cutting a socket in an off-cut. This was one of the bits that I was going to use for a slat, but messed up marking it out:
I started by chopping a clearance pocket using an 8 mm mortice chisel:
To cut the taper angle, I plonked my dovetail gauge upside down on the off-cut and eye-balled the angle:
For the first side of the socket I used a western saw:
From then on, I switched to a Dozuki, which I generally prefer (and I found it much easier for this joint):
I also added a middle saw cut to help with removing the waste:
The first bit of waste was removed with an 8 mm chisel and then I switched to the router plane to get down to the right depth. The grain direction on this scrap was the wrong way round and the grain fairly steep, so the surface at the bottom of the socket left a bit to be desired but I guess that's not the end of the world.
The first attempt at fitting wasn't too bad:
For this one I decided to tweak the socket. In general I'd tweak the tail instead, but since this is a scrap off-cut, it made more sense to tweak the socket and then I wouldn't remove too much from the tail (that will be tweaked later to fit the column). After a few goes at tweaking I got a good fit:
However, the taper angle wasn't quite right, so there's a bit of play if you try to rock the arm from side to side. With lots of glue in the pocket, that probably wouldn't matter too much, but it was a bit disappointing.
When I cut the sliding dovetail housings on the shelf unit I made for my other half, I used a jig (basically a plank of wood with a 1:6 angle planed on one edge). I'd considered using the jig again here but (a) I couldn't find it and (b) I thought it would be good to have a go at doing the joint freehand. What it actually meant was that I spent much longer tweaking the fit (including, rather unsuccessfully, trying to tweak the taper angle to get a match between the two parts). Overall, I think it took a lot longer and also made for a poorer joint. No doubt if I was better at eye-balling angles it wouldn't have been such a problem!
I found that really hard work; perhaps my rip saw needs sharpening (not that I've used it much since I bought it). For the shoulder line I did my usual thing of sawing slightly away from the line...
... and then having a pleasant few minutes chopping with a chisel (one of my favourite jobs):
I also cut a shoulder at the bottom end of the tail, again sawing slightly clear of the shoulder line and tidying up with a chisel:
As I was a little concerned about splitting wood (and, for that matter, just generally making a mess of the joinery!), I decided to start by cutting a socket in an off-cut. This was one of the bits that I was going to use for a slat, but messed up marking it out:
I started by chopping a clearance pocket using an 8 mm mortice chisel:
To cut the taper angle, I plonked my dovetail gauge upside down on the off-cut and eye-balled the angle:
For the first side of the socket I used a western saw:
From then on, I switched to a Dozuki, which I generally prefer (and I found it much easier for this joint):
I also added a middle saw cut to help with removing the waste:
The first bit of waste was removed with an 8 mm chisel and then I switched to the router plane to get down to the right depth. The grain direction on this scrap was the wrong way round and the grain fairly steep, so the surface at the bottom of the socket left a bit to be desired but I guess that's not the end of the world.
The first attempt at fitting wasn't too bad:
For this one I decided to tweak the socket. In general I'd tweak the tail instead, but since this is a scrap off-cut, it made more sense to tweak the socket and then I wouldn't remove too much from the tail (that will be tweaked later to fit the column). After a few goes at tweaking I got a good fit:
However, the taper angle wasn't quite right, so there's a bit of play if you try to rock the arm from side to side. With lots of glue in the pocket, that probably wouldn't matter too much, but it was a bit disappointing.
When I cut the sliding dovetail housings on the shelf unit I made for my other half, I used a jig (basically a plank of wood with a 1:6 angle planed on one edge). I'd considered using the jig again here but (a) I couldn't find it and (b) I thought it would be good to have a go at doing the joint freehand. What it actually meant was that I spent much longer tweaking the fit (including, rather unsuccessfully, trying to tweak the taper angle to get a match between the two parts). Overall, I think it took a lot longer and also made for a poorer joint. No doubt if I was better at eye-balling angles it wouldn't have been such a problem!
Dr.Al
Old Oak
Given the gappiness of the test joint, I decided to quickly knock up another jig, much like the one I used before. There's a photo of it later, but as I said, it's really just a board with an angle planed on one edge. I'd tried "persuading" the arm into the dovetail socket quite enthusiastically (after taking the photo of the fit!) and it hadn't shown any signs of splitting, so that gave me some confidence to start cutting the socket for real.
One of the challenges of cutting the sockets in the column is the process of chopping out the clearance hole at the bottom of the socket. As the column is essentially hollow, chopping down into it isn't a great thing to do. A few years ago I made a little Machinist's Jack for use on the milling machine's table. I've since lost one of the bits and not got round to replacing it, but here's what's left:
In combination with a couple of scraps of wood, that served as a nice way to add some support in the middle of the column, right under where I'd be chopping:
Chopping was done with the 8 mm mortice chisel as before, with a bit of masking tape to show depth:
The edges (and a relief cut in the middle) were cut with the Dozuki again:
After chiselling and router planing, I had a nice finish inside:
The grain direction was in the kinder direction this time (but of course that means that if I use similar joints for the legs, then the grain direction will be the other way for them). The first fit was a fair way off:
However, adjustment was nice and quick as I could just clamp the jig thing in the vice at the taper angle and just slide the plank along a bit, pare with the chisel, test fit, slide it a bit further along the jig, pare a bit etc and in very little time I had a good fit:
I didn't take any photos of the jig in the vice, but there are some photos on page 7 of the shelf unit build log.
The result was a much better angle fit between the two dovetail parts than I got on the previous attempt.
The jig isn't quite as easy to use as the one I used for the shelf unit as, when paring the socket, there really isn't any way of clamping it in place, so it just has to be held roughly in the right place with the chisel held against the angled face:
That's basically as far as I've got so far, but I did mark up the tails on the other three arm pieces. For my birthday recently, I got given a second one of my favourite marking tools, so I set one for the straight side and also the left-hand end of the tapered side and set the other for the right-hand end of the tapered side. It was then very quick to mark the ends and join with a rule and pencil:
I made a point of arranging all the pieces with the tapered side on the same edge in the hope of minimising the chances of me doing something daft when I'm cutting the sockets:
That's it for now; the next post will probably be when I've had time to finish off those remaining three sliding dovetails and am starting to think about legs.
One of the challenges of cutting the sockets in the column is the process of chopping out the clearance hole at the bottom of the socket. As the column is essentially hollow, chopping down into it isn't a great thing to do. A few years ago I made a little Machinist's Jack for use on the milling machine's table. I've since lost one of the bits and not got round to replacing it, but here's what's left:
In combination with a couple of scraps of wood, that served as a nice way to add some support in the middle of the column, right under where I'd be chopping:
Chopping was done with the 8 mm mortice chisel as before, with a bit of masking tape to show depth:
The edges (and a relief cut in the middle) were cut with the Dozuki again:
After chiselling and router planing, I had a nice finish inside:
The grain direction was in the kinder direction this time (but of course that means that if I use similar joints for the legs, then the grain direction will be the other way for them). The first fit was a fair way off:
However, adjustment was nice and quick as I could just clamp the jig thing in the vice at the taper angle and just slide the plank along a bit, pare with the chisel, test fit, slide it a bit further along the jig, pare a bit etc and in very little time I had a good fit:
I didn't take any photos of the jig in the vice, but there are some photos on page 7 of the shelf unit build log.
The result was a much better angle fit between the two dovetail parts than I got on the previous attempt.
The jig isn't quite as easy to use as the one I used for the shelf unit as, when paring the socket, there really isn't any way of clamping it in place, so it just has to be held roughly in the right place with the chisel held against the angled face:
That's basically as far as I've got so far, but I did mark up the tails on the other three arm pieces. For my birthday recently, I got given a second one of my favourite marking tools, so I set one for the straight side and also the left-hand end of the tapered side and set the other for the right-hand end of the tapered side. It was then very quick to mark the ends and join with a rule and pencil:
I made a point of arranging all the pieces with the tapered side on the same edge in the hope of minimising the chances of me doing something daft when I'm cutting the sockets:
That's it for now; the next post will probably be when I've had time to finish off those remaining three sliding dovetails and am starting to think about legs.
