Upholstery WIP - chair from scratch

[Arron]

Here's a few more things I've learnt on the way.

Today I ordered a cording/piping foot for my wife's sewing machine DZ1141 Universal Metal Double Welting Piping Cord Foot FOR Singer Brother Janome | eBay. Almost all good upholstery needs piping, mainly just to disguise the messy bits where surfaces join. Its usual to make your own piping from the same fabric as the chair covering and a bit of twine or whatever. Last time I used a zipper foot on the machine, which was OK but cant really get in as close as I would have liked and thus there was some puckered-up piping. The cording/piping foot actually has two grooves in the foot, one on either side of a tiny needle cut-out. As the sewing progresses, the twine is guided through one of the grooves and the needle punches through the cutout right beside it. I think the second groove is for doing double piping - though not sure about that. Its worth getting one of these things apparently.

Be careful to buy a cording/piping foot, not a cording foot, which I did previously and is something quite different.

Also, dont forget to cut your piping fabric on the bias.

I also made a staple remover. Its just an old screwdriver with the end ground into two sharp points, and bent after heating with MAPP gas. If you keep the width across the points less then 9mm then you can drive either one or both prongs in below standard (T50) staples. I have found about 25 degrees being a good angle for the bend, enough leverage without feeling misaligned when you drive it in. On one Youtube video the guy said he uses it in his left hand and drives it with a pair of pliers in his right, rather then a hammer. That way he doesnt have to go searching for his pliers to pull recalcitrant staples. It works very well. Another tip from the same guy was to keep a magnet or two on the benchtop when de-stapling, sooner or later all the staples end up sticking to it, making cleanup easy. I guess when you have thousands to do every bit counts.

remover.JPG



I also found that this site is the best way to learn about chair/sofa woodwork - Clark Chairframes cover. You can see everything here, naked of upholstery.

Chairmaking is not the sort of careful woodwork we are used to - like screws driven straight into endgrain and joints clamped up with nailguns. I was a bit taken aback at first then I decided that clearly it works so when in Rome ...



And finally, here's an interesting picture of a chair frame with the components named. About the only thing which is not intuitive to a woodworker is probably the stretcher rails. These are the components to which the upholstery is stretched down and then tacked, so they need to be robust and set in such as way that they dont interfere with the alignment of the fabric.

chairNamed.GIF

cheers
Arron

[Arron]

I glued up the next batch of components yesterday. Obviously there isn't much interest in the woodwork so I'll not dwell on it.

It sure is a robust looking beastie. It wont matter, all that is hidden.

moreadded.JPG


cheers
Arron

[Xanthorrhoeas]

Of course you could always use teased horsehair instead. You can still buy it in Australia (or at least could a couple of years ago when I bought some cartons of it) from Leffler in Melbourne.

Hi Arron, I realise that you were not interested in using it, but, for completeness and for anyone else reading this, I have checked with Lefflers and, unfortunately, they no longer import the teased, curled horsehair, so that option no longer exists for new furniture.

If anyone is reupholstering a piece that has horsehair stuffing they should wash and dry the stuffing - then reuse it. That works well for me.

[Arron]

Hi Arron, I realise that you were not interested in using it, but, for completeness and for anyone else reading this, I have checked with Lefflers and, unfortunately, they no longer import the teased, curled horsehair, so that option no longer exists for new furniture.

If anyone is reupholstering a piece that has horsehair stuffing they should wash and dry the stuffing - then reuse it. That works well for me.

I appreciate that. Now I know that if I come across any in a wrecked chair I will wash it and put it away just in case.

[Arron]

Again, skipping over the woodwork because its pretty basic.

I've done the long swooping top rail in the last few days. It took several attempts to get the method right.

I started thinking I'd make a form out of mdf and laminate thin strips. I made the form and cut a stack of 2mm thick veneers on the bandsaw. It only took a few minutes of trying to bend the veneers around the form to realise that wasnt going to work. The problem is the chair has and needs a flat top rail, while trying to bend timber in 2 directions leads to a cambered curve. Even 2mm was way too thick to get the curve, and using thinner veneers would make it possible but too time consuming.

So I figured I would make a 'wall' of 20mm thick scrap timbers laminated together until I got sufficient height and then cut a 50mm deep curved rail out with a bandsaw.

This is the 'wall'

toprailraw.JPG

I actually had to cut it in two parts at the apex and then join them back together to get the right curve. I didnt realise this, I really am no good at visualising curves.

Once fixed to the chair it was a simple matter of cleaning up the top of the rail with plane and spokeshave, and bevelling the sides so they dont obstruct the line of the fabrics.

At this point the top rail is just structural - solid but ugly.

Then I laminated on 4 layers of 2mm thick high quality veneers onto the top surface. This is the visible bit, so they are just to tidy up the appearance. The combined veneers form a 8mm capping which overhangs the lower, structural part of the top rail by about 5mm each side. The upholstery will come right up to the bottom of this overhang, with a row of piping immediately underneath and kind of partially overlapping the capping to give a neat finish.

Unfortunately I had to use three pieces of veneer per layer (butt joined), with the two joins located over the points where the back legs meet the top rail. I will applique on a carving of some sort to hide these, I think.

Here is the chair, woodwork finished, with the top rail all shaped and ready to go.

chairwooddone.JPG

This was quite a lot of work. I have found that before when mimicing high quality furniture - you start out wondering how they can charge so much for their furniture and end up wondering how they can do it for the money.

Even before I finished I found this photo and realised this would have been a much quicker way to do it, silly me :

clarkchair.jpg

Just start with a base of flat timber, then build up from bits of timber with the curved bits bandsawn where appropriate. It would be ugly and bulky but its all hidden anyway. The problem is I still havent adopted the right way of thinking - I still think like a solid-timber woodworker who feels required to make every bit of the item well finished because every bit may perhaps be seen.

Anyway, next is carving some appliques.

cheers
Arron

[ian]

Arron

can I thank you for sharing this journey

[Arron]

Today I mentioned the appliqued carving to my wife and she insisted the chair should have no carving. Mixing old and new styles, apparently. I guess she's right so I'll skip it.

I just carved my makers mark on the rear of the front leg, up top where it's visible but you have to search for it. On the other leg I carved "2016" just to balance it out. I have found that whether you are selling things, giving gifts to friends, or simply making for yourself, the personalisation is always a hit with people.

makers mark.JPG


Then I did the finishing.

What I was aiming for was a very dark brown finish - almost black but definitely not black. I wanted the grain to be visible and the wood to be 'streaky', but only when you looked closely. I wanted it to look like very old and mellowed Makassar ebony.

Like this but with even less contrast, in fact much less.

ebony.JPG

I knew it wouldnt be a very natural look, but then I like faux finishes - as long as they are very good faux finishes. I like them much more then I like a mediocre timber and the visible parts of this chair are Tassie Oak and you cant get much more mediocre-looking then that.

So this was the finishing schedule:

1. sand to 240 grit
2. wipe with white spirits to reveal any unnoticed glue smears or scratches
3. wipe with water to raise the grain
4. when dry, sand with 240 grit again
5. apply several base coats of pigment in a red/brown colour using a rag. Being Tassie oak, its still not very dark at this point.
6. apply a coat of darker brown using an airbrush, making it artificially streaky (actually the airbrush does this naturally)
7. apply a very light coat of black/brown across the lot by airbrush, to bring it up to the desired darkness
8. spray on two coats of nitrocellulose sealer
9. sand with 400 - very, very lightly so as not to cut through the stain
10. blow off all the dust
11. spray on 3 coats of precatalysed lacquer
12. after it has set hard (several days), sand with 1500 grit, very lightly, just to knock off the dust nibs
13. buff with car cleaning compounds

The trick with the white spirits is something I learnt a few years ago and is well worth doing. Glue-smears especially stick out like dog's. It has to be real white spirits though, turps or any other substitute I know of do not work so well.

I dont normally do the grain-raising step because I dont normally work with stained timber and thus am able to do a very thorough sanding after applying a sealer. Given that this is stained timber, the risk of sanding through the stain is too high to sand thoroughly after staining, so its better to do the grain raising first. In this case it worked very well and I think I will do it in all cases from now on. In fact the preliminary grain raising worked so well that there really wasn't any need to sand from there on down right through to the final coat, I just sanded after the sealer out of habit.

So a few days ago I diligently got some offcuts and put them through this process till I was happy with the look.

Today I set about doing the chair.

It did not go well.

Everything was good till I finished step 7, which looked OK so I left it for a while. When I came back it was not dark brown, but black - 100% black.

Thinking about it, I had mixed up a new lot of colour and simply plunged in and sprayed it on. Had I tested it on some scrap first I would have realised the colour was all wrong. Stupid me.

Panicing, I mixed up some dark brown and sprayed that on, hoping it would tone the black down. I didnt test that either, still hadnt learnt the lesson it seems. It didnt tone the black down, it simply obliterated the grain and the streakyness.

So I ended up with what I tried so hard to avoid, plain opaque black lacquer (a la 1980).

Its actually a very nice spray job - in fact its brilliant considering it was done outside on a windy day. Its just that I've taken a wooden chair and made it look like a plastic one.

I get annoyed thinking about it but then I console myself by thinking that this chair is meant to be a learning experience, a mistake magnet, so I should be glad I'm making the mistakes now and not on something that counts.

Heres the look.

chairlacquered.JPG

Next I start the actual upholstery.

cheers
Arron

[Arron]

Into the upholstery.

Starting with the front surface of the chair back/sides.

Quality upholstery seems to require a quite a few layers and in this case the layers to be done are:

1. Webbing
2. Hessian
3. Foam
4. Calico
5. Batting
6. Upholstery fabric

I'll do a different post for each one.



And just to get a few things out of the way:

Yesterday I went to Home Upholsterer (Wholesale Upholstery Fabric & Supplies | Home Upholsterer) for some basics. They have web sales but if you can visit in person all the better. You are right, Shane, they are helpful people and very DIY friendly. I have previously bought from professional upholstery suppliers and they are not helpful and not even cheap when they break their stock down into DIY-level quantities.

The lacquer on the chair is iron-hard but regardless I still took the time to wrap the legs in newspaper and tape. This will stay on till the end of the project. I find with upholstery that you are flipping the chair around a lot and its easy to do damage. Likewise I put some tape and padding on the back rail and then pinned some fabric over the benchtop.

Finally, if there is an upholsterer on this forum please step in and give some guidance. Even if just confirming that I'm not off-track and not saying stupid things it would be very helpful.

cheers
Arron

[Arron]

As I understand it, basically there are three ways that I could provide a firm foundation for the chair back and sides.

Plywood
Springs
Webbing

Most upholsterers dont like plywood because one stray knee can crack the plywood and leave you with a mess to fix. Webbing and springs resist damage because they part and let the appendage through.

Springs are the deluxe approach and may be good where the chair is going to be sat in a lot and comfort and longevity are an issue. In this case they arent, so they would simply be an expensive overkill.

So that leaves webbing.

The choices with webbing are:

linen webbing (or linen/flax)
jute webbing
rubber webbing (ie pirelli webbing)
elastic webbing

I dont see either linen or rubber webbing on any of our local suppliers catalogues/websites so that makes it easy, just jute and elastic webbing to choose from. I discussed this with Home Upholsterer and they confirmed that both would be work in this situation, but if there was no need for an elastic effect then jute webbing would be a safe and economical choice. Jute webbing is about $0.50 per meter and elastic webbing is $2-3, so its best to learn to work with jute webbing wherever it will suffice.

So I bought a 33 metre roll of jute webbing for $15. There is a 10lb and a downgrade 9lb version, which is apparently no good for seats but OK for backs, but the difference is only $2 per roll so go for the 10lb.

I have noticed that you can buy webbing at Spotlite and Clarke rubber. At Spotlite they seem to only have the downgrade ones.

Then you need a webbing stretcher. You can make one easily. There are several different types but I prefer this type.

stretcher.JPG

Its a bit more fiddly to work with then the other types but it doesnt involve penetrating the webbing with sharp spikes which doesnt seem smart with elastic webbing. I doubt its an issue with jute as the weave is fairly open anyway.

Here it is in action

stretcherInUse.JPG

So I went ahead and checked that all arises were rounded and sanded smooth to minimize wear on the webbing. Then I stapled all the webbing in. Not much to say about it. Here's a picture.

webbingdone.JPG

There were some lessons learnt.

1. The jute webbing frays easily so you cannot put a row of staples near the end and expect it to hold. Eventually it will pull through. You need a fairly long (say 50mm) length of webbing on the loose side of the staples, preferably with that loose end tacked down with another row of staples. If you dont have a lot of space, it seems the usual practice is to staple the webbing down, then turn it over and staple it down again with the second row of staples more or less over the first. I found all this out the hard way and needed to disassemble and redo.

2. You need good staples. I bought two packets of these from Masters. Do not use these or anything similar to them for upholstery. They are fragile and one or both legs will often break if you need to remove them. When they break you get little sharp spikes in your woodwork which you cannot remove and will surely abrade anything they come into contact with. I had some good Arrow staples before this which were much better. I use 12mm for anything 'structural', and 8mm for anything temporary or not under stress.

staples.JPG

3. I've been using an electric staple gun until now and its really not up to the task. Put 10,000 staples though them and they are hopeless. Tomorrow I will buy a air stapler. I think if you buy an air stapler try first to buy a long nose one. They arent essential in upholstery but they do some interesting tricks. Some people even grind the noses down further to minimum thickness.

4. Tension of the webbing is an issue with me. Basically, I tensioned till it made a drumming sound. Seems OK.

Arron

[Arron]

OK, I've done the next step, in this case hessian (burlap) foundation layer.

I really dont know how important it is to use hessian, or what the arguments for using it are. I find that typical of upholstery, plenty of instruction but no information on the reasons behind things. Maybe the preference for hessian was largely just cost saving. I guess once it was cheap, though now its actually more expensive then many basic fabrics like cotton drill, which is much sounder. Or maybe its used because it has a (small) ability to mould to the shape of what it covers. Or maybe its favoured because the open weave allows you to see how stuffing underneath lies, and poke a regulator (big needle) through and move the stuffing around. Or maybe its just tradition.

Anyway, I do know its traditionally used as a foundation fabric over springs or webbing, below the fill.

I once disassembled some dining chairs that didnt have a hessian layer and the webbing edges had started to cut into the foam.

I have read that there are different 'rub-counts' or grades and the stuff used for upholstery should be of a higher count, but I didnt think the difference would be big enough to warrant half a morning's drive so I got some from Spotlite ($20 pm at 2400mm wide).

So here it is, stapled all around. You can see from the photo that its not stretched particularly tight.

hessian2.JPG

If anyone can enlighten us on the importance of using hessian and what else could substitute then please do.

cheers
Arron

[q9]

The hessian is just there to ensure the next layer doesn't fall through the webbing/springs/etc. It's primary quality is that it is fairly hard wearing and cheap. Some dudes use offcuts instead.

Was there a reason not to weave the back?

[Arron]

Was there a reason not to weave the back?

can you clarify what you are asking ?


********* edit *************

rereading it I guess you mean was there a reason I didn't weave horizontal webbing through the vertical webbing. The answer is because the back is rounded so the horizontal webbing would, if it had any tightness at all, simply shortcut between its two anchor points, thus dragging the entire back webbing forward.

cheers
Arron

[Arron]

The hessian is just there to ensure the next layer doesn't fall through the webbing/springs/etc. It's primary quality is that it is fairly hard wearing and cheap. Some dudes use offcuts instead

Sounds right. I hadn't thought of it as hard-wearing but it's been used for sacks for centuries so I guess there must be a reason for that.

[q9]

Ok. From the photo it is a bit hard to tell how it lines up. I wonder if angling it wouldn't have worked? Could also be completely unnecessary

[Arron]

Ok. From the photo it is a bit hard to tell how it lines up. I wonder if angling it wouldn't have worked? Could also be completely unnecessary

Angling it would also have compromised my nice curve. If you look at the original design, it's all about the curves.

Anyway, I was motivated to watch some YouTube videos again and it seems vertical-only is the norm for tub chairs. Some apparently competent upholsterers, like this one https://m.youtube.com/watch?v=m90eTpARot4 , even use just a single sheet of fabric instead of both webbing and hessian.

Cheers
Arron