Marri Dining Table and Chairs

[Superbunny]

I really like this, I think your mother-in-law is right. If I produced a set like this I'd be very pleased. Well done.

SB

[skywarka]

Morning all

The back slats for the prototype chair. My design required 2 of these per chair so I needed a system to make 16 mostly the same.

For those designing chairs I read somewhere that the number of vertical back slats should be even. If it is odd the middle slat lines up with our spines and can be uncomfortable ( if you have a bony spine ). I did notice that several designs in well known furniture stores have an odd number ..mmmm.

I had read and watched from the 'net that there are basically two ways to do this i.e. bend wood (thinnish) - steam or laminate .

I tried steaming first - seemed less involved / costly and didn't require much that I couldn't throw together. So I made a steam box out of old rectangular metal water down pipe -off cuts from a house being built down the street. Covered the outside with polystyrene foam (10mm I had lying around from an new fridge we bought - never throw stuff out it may come in useful ha ha).

Where to get the steam, hmmm lots of home made stuff on the 'net but being an engineer I understand how hazardous it can be playing with steam and vessels etc. So by chance I saw a clothes jacket steamer at Kmart for $50.

Took the nozzle off the end and plugged it into a wooden stop at one end of the drainpipe steamer, a small vent / leak at the other end.

Put the 90x12mm wide pine in for about 30 mins. I had read that steaming is about 1 hour per inch of thickness. The steamer on a full fill (a bit over a litre) lasts for 45-50 minutes. I used distilled water to stop any scale build up in the boiler.

Took the wood out (with gloves its hot) and tried to bend it to see if the steamer had done its job. Well it was certainly not plasticine but was noticeably more pliable than cold wood.

So I needed a press to keep it to shape whilst it cooled. I made a temporary one from the offcuts of the legs pieces. Heated up the pine again (not ideal but this is experimenting) and then put it in the press. What did I find...

Well pressing takes lots of clamps and some muscles and speed - the wood cools quick - less than a minute, the press should be made to a decent standard. My first one was just three rows of timber in parallel nailed together with cross braces - rough as guts. So the timber came out a bit bumpy cross wise as well as for the curves

It worked though but... springback which I had read about but can only be determined in practice. Although the slat had been pressed to the same curve as the leg when it was taken out of the press (I left it in for 24 hours) it relaxed to a less bent curve.

So to make the required curves to match the legs, I would require a press with smaller radius curves to allow for springback. Also I read that with a hardwood like Marri this could vary and I also found that the steaming time was now 45 minutes.

All this was getting too hard (and costly - water, losses due to variability in springback, playing around with moulds / presses until I found the right curvature. The variability though was the real killer.

So laminating was looking easier. This required a couple of decent bits of gear though. Oh no expenditure! which is why I tried steaming first,

Next post laminating.

Moderators - hopefully I am not doing the wrong thing here breaking up the story into bite size pieces please let me know if you have any suggestions for improvement.

[skywarka]

To make the laminated backslats I need three plys of say about 3mm each to give about 10mm. My experiments with steaming showed that I need more clamps and a decent press. But springback was likely to be minimal.

I luckily found that Bunnies had some 300mm F clamps for $5 bucks each - a bargain really so bought 12. I know more expenditure but I was getting hooked in trying to crack this and clamps are always handy. They were also going to be just big enough to go across the back of the chair (narrow backs).

I also bought a piece of rough Marri 100x30x1000mm to experiment with.

I honed the jointer blades and tuned up the settings beds and blades - very important as it was well out when I got it. Got the manual off the net. Remembering that all this was new to me I read / watched a bit about setting up a jointer after I had sourced it. Plugged in the the dust extraction, the cyclone really came into its element for this work. Nothing but really fine dust hit the vacuum bag.

So made a true face and edge on the trial piece (again use the push pads and goggles etc). So now I had to band saw say 4mm plus off the face about a metre long.

I learnt about tuning a band saw - guess what when cutting fine and accurately in hardwood, drift rears its head. Also tension on the blade and size of blade.

You can see I was learning a lot during this project which is nice of course.

Remember this is the bottom of the range carbatec unit (cheaper than Bunnies Ryobi ones at the time). So guide bearings are not top quality and the fence is well pretty basic. I was trying something a bit up there - thin accurate plys in hardwood 100mm deep 1 meter long.

The first laminates were awful - they looked like the sea - full of waves (not blade marks). I was a bit dejected at this point thinking I might have to buy the chairs.

But I studied the net more about bandsaw tuning and then did the following. Found the drift, added a vertical extension to the fence, tensioned the blade to the max I could, bought a wider blade - 10mm instead of 6mm which apparently won't fit (so why do they sell one then?). Paid real attention to setting up the guide bearings. Then saw really slow and concentrate. Got really tedious for ~50 plys during the main project.

Looking back for accurate work like this a 14" bandsaw would be the minimum - but they cost a lot more new. If I could find a second hand one....

Anyway - success and the crowd went wild.

I repeatedly (4 off) got 4-5mm laminates. So all I had to do now was thickness them to 3mm. One side was smooth after each ply the main stock is run over the jointer again to remove the saw marks and provide an accurate face for the next ply cut.

Well thicknessers don't go below say 6mm do they grrr for safety reasons i.e. the blade doesn't munch the base plate. During all this I was reading and watching about thicknesser tuning.

Yes it was second hand and not set up evenly. So hone the blades, set the blades square to the base, and line up the feed / out tables to minimise snipe.

To deal with the 6mm limitation I put in a false bed made out of two pieces of white melamine fixed together 30mm thick then.

Learnings - blades need to be really sharp for hardwood especially if there is any figure / curl. The grain is all over the place. 3mm is the bottom limit as the wood will disintegrate (especially if the blades are not sharp) and you try to take too much off at once. As with the bandsawing - be patient go slow.

Snipe - its almost impossible to eliminate with the cheaper thicknessers with no spindle lock. The play in the support post bearings is such that snipe will occur no matter what. The disadvantage with cheaper tools. But allow extra at the end of the timber and take very small amount off for the last passes and its a lot cheaper than a $1000+ unit. I also used some hose clamps (jubilee clips) around the posts above the spindle head to try to minimise the movement. A pain in the but to undo and do up 4 every time you lower the cutter.

But again success - I now had three smooth 3mm plys - bendy - nice. All I had to do was glue them and put into a press.

Glue - In the end for the real deal I chose Titebond Extend II (longest open time I think) I found the selleys stuff started to go off before I had laminated in the trial. I was brushing at this stage 4 faces 90x1000mm.

In the end I got an artists screen printing rubber roller to spread the glue. I then put it into the decent press I had made at this stage and clamped up. I found that I had to put end stop inside the press as the laminates slide around / move as you clamp them. Let it dry for 24hrs. Worked like a charm - no springback to speak of. See attached pics.

As I said lots of clamps and a decent press/ mould. In the second pic you can see the jointer in the background and the bandsaw to the side in half the garage - where's my shed!?

[skywarka]

Oops forgot to attach a pic of the thicknesser setup

[skywarka]

The third key step (for those following the ramble) in making the chairs was joining the bits using floating tenons. What I like about this techniques is that the mortices can be router cut and the tenons can be individually sized to deal with any variations in the mortices (i.e. errors).

So again lots of reading and in this case watching the famous Tom make templates to route anything (thanks Tom).

Sizing the mortices is standard stuff and the tenons can be made in long strips on the thicknesser and router table. (Initially oversize of course). In the main project I used the waste bits / offcuts from the legs etc to make the tenons).

I made a basic mortice jig for use with my small B&D router (see attached pic). The straight mortices are no problem but the angle ones require an angle block to present the work piece at an angle to the bit. You can see the angle spacer stuck (carpet tape and then screw from the back) at the bottom of the jig.

You will note this is real basic all the stops are temporary and everyone is held with F clamps. Again for a one off job the 20/80 rules applies. Lots more fiddly to set up but not as much to make the jig. I used pocket hole screws to attach the vertical and horizontal piece of the main jig structure.

In reality I should have spent a bit more time on the jig - like make it long enough to be clamped onto the work bench (router table) without interfering with the router stop when cutting the mortices.

Again dust extraction was key here to see what was going on. The beauty about the little B&D router was its dust extraction housing around the bit. A lot of the cheap new ones don't have this.

For this work the stops and marking out are key for consistency but at the trial stage in the pine I made a few errors but no big loss. I had to hone my techniques for the main game.

This part proved to be the least problematic though (once I figured out the angles from the drawing and put in the correct spacer). I would have lots of mortices to cut for 8 chairs I think 32 per chair. I always set up the jig for the same joint and did all of the same type together.

Oh yes another tip watch, turning pieces over end to end with angled mortices, you can put it the wrong way around. So think about it first and mark faces etc. I made a couple of errors when not concentrating. Again though the beauty of this technique is that the tenon can then be cut and sanded (rough paper) to suit.

A good fence is essential for this work. I used the one from the GMC router as the original B&D one did not have micro screw adjust to locate the lateral position of the bit (centre line of the mortice longways) and screw on extenders. Despite the GMC being cheap it has all the features.

I also ended up buying a sliding mitre saw to cut the pieces. (just a small cheap one as mentioned early on). At the prototype stage I hand cut the pine. This made me realise what I would need for hardwood and 8 chairs and the accuracy / repeatability required.

Not much else to say for this just repetitive work for 8 chairs.

I've also attached a pic from earlier - the extended bandsaw fence and off to the side on the back of the router fence one of the yellow push pad safety devices. Oh yes that pic reminds me, for the longer band sawing I got a cheap $20 support roller to support the pieces as they left the small table - set it up using a long straight edge.

Let me know if I am going on a bit (yes) but just trying to convey my thought process at the time to assist others.

[skywarka]

Hopefully the posts will be a bit smaller from here, less complex stages.

So now I had all the bits. I measured each mortice and cut / sized each tenon to suit by hand. Used 40 grit sand block. The tip here is to have snug tenons hand fit and pull apart (not too tight, not too loose), we want good joints for the chairs to be structurally sound. If a tenon is no good just make another only cost is time. A lot of the modern chairs are screwed bolted together - thats why the chairs creak after a period of time as the fixings work loose under the stress of use.

For the prototype I wasn't too fussy but again it gave me an insight into what I would be up for. I also filled in a couple of mortices that I had cut in the wrong place - one of those errors I mentioned but not a deal breaker just be more careful when doing the real thing.

I glued the back and fronts first, used the 300mm F clamps for the back top and middle rail. The front was wider so I had some 900mm sash clamps just one clamp. I used a small cut down brush the $2 ones with metal pen like handle from Bunnies. A tip from fine woodworking.

Just check that all is square in all directions - this was easy. Leave to dry 24 hours.

Gluing the sides in i.e. the front to the back well with this design "Its not easy".

I had the sash clamps but the shape is a trapezoid front 450mm back 300mm so if you apply pressure unevenly the shape skews - if that makes sense . Also it can skew so that all four feet are not on the ground. Real tricky tightening this up for the prototype.

Remember you have a limited time for the glue in this case top and middle rails, 8 joints, 16 mortices to push glue into with the brush, then assemble, then clamp, check for square and symmetry in this case. A real problem for time before the glue goes off. Oh yes and wipe off excess glue.

So for the main production units I made two aids or jigs. A frame for the front and back to allow the sash clamps to pull the small back in square to the front and a seat insert inside the frame to assure symmetry

And as they say here is a pic. This was critical to success. I am sure experienced hands must know better way to glue backs to front but this is what I figured out. It also not as much of a problem if the front and back are the same width.

Well at this point I had my prototype chair (as per earlier pics) and I figured I could do this, make the frame that is. What about upholstery next post.

[skywarka]

For the trial this was the easiest part. Again watched some example / guides via you tube. Went to Clark Rubber got a square of foam (high density) cut it to shape on a base of some 17mm mdf I had lying around

Then cut out some of the white fluffy stuff from spotlight - Dacron? in a 30mm sheet from Tontine - the pillow guys I think.

Turned it over and used my existing small staple gun to tack the $2 piece of cloth from spotlight into place. Start in the middle of each side and work to the corners which I had rounded of to say 30mm radius.

Two things with upholstery getting the finish smooth and folding corners. Following the advice from you tube worked fine ( I think I searched for upholstery dining or slip seat covers.)

I might add here that the leather for the final seat covers proved to be more difficult due to its thickness and stretchiness. I'll cover that later.

The only fault with the prototype (when we sat in it - no it didn't collapse) was I had judged by eye for my drawing made the back, slope back too much. Its a dining chair not a lounge chair. Again no problem thats why make a prototype just remake template for the final design.

So I had my prototype (photos earlier - blue seat cover) which gave me enough confidence and I could now consider buying my Marri.

[skywarka]

Well as some of you will know Marri is an interesting timber. It used to be wood chip fodder a few years ago because of all the defects. Scratch that now called features. The ladies in particular seemed to like features.

So I need

• a slab or two or three for the table top (going to be approx 2.4x1.2m)
• Something for the table legs (slabbish refractory design we had in mind)
• and then some sticks for the chairs.

After hunting around Perth for suppliers I eventually found the Timber Bloke in Naval base (good man - Derek - you may have seen him at the Wood show). I liked the idea of using salvaged trees. Right up my alley really.

I wanted some slabs with some features but not huge cracks. Too much black resin detracts in my opinion. We talked price but the cost of a one piece slab that size was just too much ( I had set a budget of $3000 overall).

So I chose 3 slabs from the same tree dried of course. Derek finished one surface (the top) nicely for me and I took them home and stuck them in the house.

I was going to make the chairs first so I could let the slabs acclimatise for a few months. Apparently wood moves of its own accord and I was conscious of this. "Her indoors" was vey accommodating as I started to make the chairs I had to store the slabs in the dining room. See pic and we'll get to the table again later.

After I made the prototype chair I went back to Derek and was able to choose the timber for the chairs. I selected raw timber that had as little as possible features / gum veins. The legs in particular were going to be structural.

The back legs cost 75% of the chair frame the finished blanks need to be 1100x110x30mm so raw planks were a decent size 150x40+ I think But the jointer could manage this.

Well I had outlaid significant cash and was past the point of no return. I had all the bits I needed tools, timber, a bit of an understanding of whats involved I could set sail for the promised land. (Dramatic music sounds)

Next post tomorrow I think making the real chairs.

[Sawdust Maker]

That is a brilliant dining setting

anything that the approval of the MIL has to be beyond the pale

and thanks for the belated wip and description of the steps you went through

[skywarka]

The best thing about a prototype is you can figure out most (not all) of the issues without to much loss if it goes wrong.

Obviously I jointed and thicknessed all the timber stock to size first - nothing new to comment here other than thicknesser snipe again - but I knew what to do by this stage to deal with it.

So for the main chairs the key issues were 8 of them! So lots of repetitive work. I decided my workflow would be to prepare most of the small bits first breaking the flow up into parts where a big step occurred.

So made all the small parts first and left the big back legs until last. I also decided to do the chairs one at a time for the back leg stage.

I also found during the prototype stage that routing the mortices is easier in square rather than curved wood especially in a simple jig I had made. So I just had to remember extra depth in routing where the legs curved.

The front leg tapers I did on the jointer - just put a small block under one end (the top) (sticky carport tape again) This presented the square leg to the blade at an angle and then just shaved off small amounts until the desired taper had been achieved

Then for the back slats. I only had a couple of losses making the plys for the laminates - lack of concentration at bandsawing and thicknessing. I morticed them directly into the top and middle back rail. Not too tricky just used the curved stilt as a measure and skimmed a little off to make a small tenon. There was an error here that I didn't correct. The back slats enter the mortice at an angle so the tenon shoulder need to be different heights. I noticed this at the firs chair but chose not to fix it as the half mil gap cannot be seen unless you look up under the top rail at the back.

This error occurred because I changed the design from the prototype on the run (resist this) In the prototype I put the back slat full thickness into the mortice so the problem did not occur.

The attached pics show all the small parts morticed, the back legs stock and a part assembly of one of the chairs before the back legs were shaped to see if I had calculated the trapezoid angles in the joints correctly. I had of course because the prototype had worked but you know what they say measure and check a lot before cutting.

Also the back legs with the mortices. You may notice what looks like extra mortices here an there. No these were not mistakes - Marri can be full of gum veins so when the gum crystals are dug out we have little tunnels. Even though I had selected the chair stock to have minimum features - some were hidden inside and only revealed during prep.

[skywarka]

Well before you know it (well a few months really as I was doing this after work and on weekends) the chairs were done.

As I mentioned earlier the biggest headache was gluing the backs to the fronts but the rig I made worked OK. See pics

Now I had to upholster. Again I had never done this before but had made the prototype. Originally I had in mind that false suede material for the chairs seats. But Her Indoors decided black leather - she was right of course.

I sourced the leather from Direct Leather in Osborne Park. It was funny when I first called and said I need leather for 8 dining chairs about such and such square metres. The chap said that will be about a hide. I thought that was a strange term to use.

Little did I know leather comes in hides. When I went to the store there was this flat cow spread on the table, nice black leather though - cost for the one hide was about $200 - so $25 per chair - not bad.

As mentioned got the foam from Clark rubber - I took a chair base 12mm chipboard) and they even cut the bits to size for me. The white fluffy stuff from Spotlight.

The foam had to be chamfered around the top edge to provide a better lead in to the edge of the seat. So again the net came to the fore - how to cut foam rubber neatly without buying a foam cutter. A sharp knife is not good for this - creates an uneven cut.

Well it turns out the electric carving knives does the job - so a Sunmbeam unit for $30 was the go. - worked like a charm.

Then the leather - well this was trickier than I had thought to get a nice smooth finish that is. I nearly gave up a couple of times (lots of brow sweat here).

The issue is that the leather is thick 1mm or so and stretchy. I also didn't want to have a fold on the corner as is traditional in upholstery. I had seen one chair without any fold. The trick seemed to be to have a radius on the corner say 30mm and then pull the leather tight over the corner.

Well this gave me a headache as whenever I pulled the corners in I got wrinkly sides and front on the chairs. So I would have a go get frustrated and then go back to it again. Took me a while but it turned out I just need more tension overall on the stretchy leather than I had in the initial temporary tack I had done as proof of concept. Then trim off excess underneath.

Hopefully the pics show what I mean. They show the back of a finished seat, the back of the temporary tack (thin staples and no trim of waste) and the the wrinkles in the unfinished one at the bottom of the stack of two.

[skywarka]

Well now the chairs were made, I could do the table- relatively simple but with a couple of challenges

Trimming and edge jointing the three slabs
Drilling 4x6" deep holes in the end grain of the cross beam
Clamping the slabs whilst the glue dried.

Compared to the chairs this was simple but heavy work. Her indoors provided some of the lifting here.

I decided to make the base first - simple refractory arrangement - two end slabs and a cross beam.

For moving house purposes I wanted to make the table deconstructable so I decide on half inch whitworth threaded rod into the cross beam two on each end. Then glued in with resin - the same stuff for filling the Marri holes. I knew this would be tricky - to keep the hole that deep square to the beam into end grain.

For those that don't know long bits like to wander. I did not have a drill press, any way one which could get on top end of a 1500x50mm beam. So how to drill horizontally and accurately into end grain.

Firstly the correct bit - a ships auger or something close to it is the go. They tend not to wander as much. This was probably one of those jobs that would have been better with and old hand bit and brace. But I saw some long bits 3 for $10 in Bunnies (their crafty watsit brand). I was lucky 2 of them were true, one required a little filing on one of the faces that sits in the chuck.

The next trick was to make sure I presented the long bit square to the end grain.

I made a guide block (with holes for each size of bit) and used a complicated arrangement of F clamps to hold it onto the end of the beam (see pic). Then started with a small bit and took it real slow with the hand held drill. I checked for true very frequently for the first couple of inches in ( A ruler edge on the top for left / right and a square for top / bottom. Lots of brow sweat in this one. Then repositioned the guide block, upped the bit size went again, and then a third time up to 12mm which due to the slight wobble in the bit a smidge over.

Now this is when I had some real luck (or a cunning design ha ha) I had intended to make the hole larger than the rod and glue it in (hence 6" deep - no way would this pull). But I knew the OD of the thread was just around 13mm and the hole was a frac over 12mm so I tried to start the rod directly into the wood.

Now this was hardwood - with no defects (that I could see - chosen for that reason). So I put two nuts locked on the end of the rod and gave it a go - well with a bit of furniture wax for lubrication it worked like a charm. The rod cut a lovely thread into the wood - with a bit of to-ing and fro-ing. I couldn't have planned it better.

The end pieces were then as per the pics - nothing really to say - I just tennoned the top cross members in and screwed from the top down - again from dismantling.

[skywarka]

Now I had my base I could get the slabs from the acclimatisation area and eye up the arrangement. See pic

O yes before I forget - To make a faux through beam appearance I used a tip I found from one of the other members here (sorry I forget who - it was a Jarah table - but thanks for the great tip). I had the off cuts from the main beam and used powerful magnets to hold the end pieces to the nuts you could see in the previous pics.(from old hard drives which are odd shaped but can't be seen - or you could buy proper round ones - neodymium).

Now I had to cut off the edges of the slabs that were to be jointed, then joint them, then glue and clamp. This is heavy work and some of you will cringe at the next bits but what can I say it worked.

I needed a +2.4m non-bendy straight edge to guide my circular saw (< 20mm thick or my saw blade would not go through the slab). Proper straight edges cost so I bought a rectangular hollow section from Bunnies (30mmx15mm I think)- I checked it was true before leaving the shop by trying the edges against each other - they all seemed true. - Worked well enough with F clamps and middle backing to stop flex half way along. A table saw is the go I expect but I don't have one or the room.

Now I could have tried to joint 2.4m by hand with a smoothing plane ha ha or... I could try to run the 2.4m 40mm wide slab over the 6" jointer small bed which is what I did.

How do you man handle a big slab over a little jointer. As I have said before - It isn't easy - but with two end stand rollers to extend the bed - need to be set up carefully true to the main bed - it worked.

The weight of the slab acted in my favour as no pushing down was required and again I made several passes of small amounts - see pic.

Then we were in the home straight - dig out the gum crystal and dust then fill with West Systems Resin with a bit of black brickies oxide to fill the holes. The only issue here is that some of the holes were big so I did several small mixes to fill the files in layers then sand off the excess at the end. Used masking tape underneath and on edges to stop the resin running where its shouldn't.

Then we just had to sand the edges - wire brush to get loose stuff off and then the edge of ROS to smooth.

Now how to glue 2 big bits 2.4m long and clamp. I do not have clamps 1500mm wide (1050-1200mm table) and the edges are natural or waney I believe the term is.

So I tried hold down tape straps - you know the things for the trailer - see pic. I used spacer blocks to keep them off the glue strips and just like sash clamps two over and two under as they pull just like sash clamps i.e. they can cup the boards. I checked for true as I increased tension and adjusted the necessary clamp. Now those that may think there is not enough force, each strap has 450kg max pulling force so approx 2 tonnes would be plenty - which it was. - and here are the pics

If you look at the grain you can tell where the joints are - not too shabby if I saw so myself.

Oh yes I used biscuits for vertical alignment - It was a two person job gluing up say 15 biscuits and two edges 2.4m long twice before the glue goes off (titebond extend II). I did one joint at a time


...and then it was made the crowd went wild - now just the finishing

[skywarka]

This was the simplest bit - just takes time.

I finished off the main slab joint tops with a belt sander with sanding frame. Interestingly my ancient +15 years GMC belt sander (had to replace carbon motor brushes $6 half way through thats how old) has the exact same mounts for the sanding frame as the current light green ryobi one - so I got one as a spare part from Bunnies $10 or so. The GMC didn't come with one but had the mounts slots. Amazing really - not sure what that says about what gets manufactured where by whom.

I sanded to 320 - 400g using ROS or hand as required. The only error if you like that occurred was I forgot to round off the edges of the chair pieces in my haste to assemble. So it was more time consuming than it should have been to do once the chairs were assembled.

Then just used the famous brand Wipe on Poly satin - I did 7 coats for all with light sanding in between the early coats. I will do two or three more layers on the table to provide the maximum protection against accidental damage - which is why I chose WOP. I usually just use F&W fine buffing oil for my projects - not good for hot and wet.

So thats it - if you have been following the saga. You've seen the end result at the start. Hopefully in amongst the text there are some tips others might find useful. I tried to cover as much as I could.

Thanks again to all those who helped me find inspiration to solve my problems and who have given kind comments. I'll keep any eye on the thread so if you have any questions feel free to ask.

I have just heard I have an order for a matching buffet so guess what I'm doing for the next few months

[artme]

Absolutely stunning setting Tony!!