Rhys Jones

STUDIO LOG var _gaq = _gaq || []; _gaq.push(['_setAccount', 'UA-18629025-2']); _gaq.push(['_trackPageview']); (function() { var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s); })();

[rhyscjones]

Chair brackets have been SLS prototyped and silicon moulds have been set. Just need to create some wax copies for casting. The chair Conformation is a step closer to realisation...

[rhyscjones]

Rendering of front chair bracket in Maya. 

[rhyscjones]

Curved Forms

Following the preparation of the European Beech laminations, they are arranged into 8 members (4 backs and 4 legs). This is for the making of 2 'production chairs', which i have decided to name Conformation (I will talk about this at a later stage). 

The glue-up was done with epoxy, without the addition of fibreglass sheet - I have decided that it is not necessary. The laminations did not need to be steamed as the Beech is incredibly flexible, despite being a tougher, harder timber than the pine used in mock-ups. Subsequently, Beech is highly suited to steam bending - the fibres will gladly conform to a new shape. This is one of the many reasons why i have selected this timber. It is also quite uniform (reducing wastage from slabs) and easy to work. It is highly suited to ebonizing, which is something i may explore.

The formers utilize shaped pine blocks at key points around the form to help spread the clamping load/pressure. To get these shapes i did a dry-run without the former base and traced the outer-edge of the laminations onto the blocks. They were then shaped via disc sander and bobbin sander. For the glue-up, 4mm plywood (cut short-grain to allow flex) is covered in tape and placed on the outside edge of the laminations. Next to this, a strip of eco panel (scraps from a fellow students project) covered in tape is used to help even the clamping pressure and allow for discrepancies in the form and blocks. This set-up has proven to work very well, producing very little gaps when compared to a male-female former set-up. It also requires much less work.

After all glue-ups are complete the forms are trimmed (using the formers as reference) and dimensioned using the jointer and thickness planer (we are lucky to have wide beds here).

[rhyscjones]

Formers and Laminations

After recieving the cnc templates, the formers are cut made for the 'production chair', and the laminations are cut.

The formers are constructed from 25mm chipboard, using the mdf template to 1:1 route each layer (in this case 2 layers) double-sided tape holds the template to the chipboard in each case. The layers are glued together, and screwed onto a base. All surfaces that will be in contact with glue are covered (in tape or plastic film). 

The European Beech is broken down and dimensioned for cutting laminations on the re-saw (jointing between each cut). The laminations are thickness-planed to 1.55mm. Their original order is maintained (as to give the 'illusion' of solid timber). This is one of the things considered when arranging them for glue-up.

There are a lot of laminations here, the overall wastage was probably around 60%. The next part will be gluing the curved forms.

[rhyscjones]

Woven prototype 4

In response to some of the issues raised on woven P3, I have taken a new direction, with the under-structure; throwing myself into new, bolder forms. I have produced a 'ribbon' or 'tendon-like' member which spans the front and back rails, shaped to look paper-thin. The under-structure of woven P3 was recycled in order to achieve this is a short space of time.

The rails themselves have been radically shaped from solid blocks of pine (particularly the back). This is probably too much, however it does tell me a lot, rather than creeping slowing towards an outcome. I have successfully broken the static nature of woven P3 - this base is alive with energy - there is incredible tension between the back and front legs - so much so that it appears the front legs could pull-in from under us. This energy needs to be pared back. To achieve this, the front legs can be angled out a little. 

The structure is much more rigid. I will take this form to the final piece: Interwoven. There will of course be alterations. The back rail will not twist so radically, instead the ribbon will travel flush into its surface and be pinned. The intersection into the front rail is quite good and will remain, though will be made less busy. Although the structure looks very busy through the seat, when utilizing the nylon mini-tape, the seat will appear much denser and vision through it will be impaired (though the surface will still give the impression of lightness, with the surface seemingly penetrable).

As this prototype borrows the legs from woven P2, i have included an image of the latter, helping to compare the two. I believe the structure has been solved without compromising the aesthetic i am aiming towards. The final image in the set (a side-on) illustrates the addition of a 'fillet' between the seat and leg structure. This will help reduce flex and visually marry the two.

[rhyscjones]

Woven P3 complete

Assembly and shaping

After completing joinery and shaping the members of the chair, the upper and lower halves are (separately) assembled.

The seat/backrest have an additional step in this process which involves gluing the mitres with epoxy, then inserting splines. Note: a straw can work well to clear off excess glue (see images).

The chair is then shaped and refined. Holes are drilled for weaving, and a groove is cut into the underside of the seat to allow attachment to the base (without interrupting the weave). Once again i have used 80lb fishing line to form the seating surface, only this time the weave is not as close together. This is because it's only function at this point is for seating, not visual feedback.

Analysis:

I find the upper half to be a great improvement. I see potentional in the direction i am taking it, even if it requires more hand work. I think at this stage the nature of the chair has changed drastically from where i started. It is now very much a no-compromise piece which says a lot of about craft-design practice. 

The base is an interesting development. I think that this new form does indeed better match the upper half. However, i am sceptical about whether the additional member at the back is necessary for this. It seems to have more of a visual function than a physical one. Perhaps there is a different way of going about this?

Strangely enough, this chair seems to communicate a kind of traditional sensibility; like it is reaching into the past, referring to forms and logic which is inherent in classical design. I'm not sure why at this point... Perhaps it is the clear language of form - that each piece responds to a need based on visually coherency and structure - (the stresses and forces). It is a chair that no longer tries to defy our senses, instead it prescribes to a sensible logic that is somehow comforting...

But this still does not seem right to me. It was my original intention to challenge the senses, not produce a piece which is simply static. There needs to be energy, just enough to bring the piece alive. Furthermore, the under structure still flexes despite the use of epoxy and the addition of fibreglass. You can only bend the material to a certain point - you cannot transform it - at least if you are to honour it... A new approach is needed, and i think i know what it needs to be... 

[rhyscjones]

woven nylon thread

For weaving my chair, i have been looking at using nylon. In particular, woven nylon, for its sense of 'warmth' over single cord (i.e. fishing line). Inspired by the crafted quality and warmth of traditional single rattan-cane weaving; i am hoping to create a surface which reflects these qualities (generated through a similar act of weaving), but exceeds them by the use of modern materials which are infinitely stronger and long-lasting.

In a search for a thread which is reasonably flexible/forgiving and will fulfill my needs, i have found bonded nylon cord utilized in the manufacture and repair of parachutes. This particular cord has a tensile strength of 50 lbs (23kg) and is used for sewing harnesses and heavy tacking. 

Furthering my investigation, i discovered 'super tack cord',  a waxed, flat-braided nylon mini-tape with a tensile strength of 80-90 lbs (37-40kg). Both products are black in colour (which is what i am after), however, the width of the super tack cord seems to be more appropriate in referencing traditional methods of weaving. It would also produce a tighter, more uniform surface - thus more comfortable.

These products are their respective information was found at paragear.

I am yet to do tests, so the second cord may prove inappropriate...

[rhyscjones]

Woven prototype 3 : laminations, formers and glue-ups

seat/backrest

Pictured here is the side profile of the backrest/seat. This member has changed it's role a lot in the new design. What was originally a relatively simple construction, is now quite a complex form which requires a large measure of handwork.

The laminations are glued in much the same manner as the legs/under-frame (without the addition of fibreglass sheet). They are cut at quite a large width to accommodate the amount of curvature and shaping that will later take place. This curve (which can be described as giving the chair a 'waist', is intended to reduce the overbearing presence of the upper-half of the chair.

The innermost lamination is broken up into three parts with the centre piece covered so it can be removed. This is to allow the profile (as pictured) to be cut into the members. This profile is intended to help define the seating surfaces, emphasising their separate function amongst the chair's busy form. This, if successful, would be a good example of a complication helping to simplify.

[rhyscjones]

Woven prototype 3 : laminations, formers and glue-ups

legs/under-frame

To construct the new under-frame, i have opted to use an epoxy adhesive in combination with fibreglass sheet (super fine weight), placed between each lamination. The need for rigidity in these members is to negate the need for stretchers which span the sides. I am using a 2-part epoxy adhesive called Techniglue CA made by ATL Composites - more about it in this post.  

The new form requires that the laminations are bent around a very tight radius. To do this i am steaming the section on each lamination which will be subject to the most amount of stress. This was done simply over a boiling kettle. Once 'elasticised', the laminations are clamped around the form and left for a day to dry out. They are then removed and left for another couple of days to completely dry out before gluing. Any excess moisture in the timber appears to affect the curing of the epoxy and thus the glue bond. This i found out with a previous test - disastrous. 

During the glue-up, the fibreglass sheet (cut into strips) is placed between each glue-line. Additional glue is applied directly onto it until the fibres are saturated. Once dry, the stock is cleaned-up and machined. The new seat-supporting member is made in a similar fashion, without the addition of fibreglass. Only one leg/side contains fibreglass. This is to allow me to the test the difference in rigidity.

The two forms are matched (by cutting into the leg) and glued. Shaping then takes place. Note: additional material is attached to the supporting member to bend it inwards (this is to line it up with the seat).

[rhyscjones]

Updated information for reference on:

Epoxy Adhesive

This is a 2-part epoxy resin adhesive called Techniglue CA made by ATL Composites. It consists of a resin and hardener (or curing agent) which is mixed 2:1 respectively. The reaction is chemical based, however, the curing can be influenced by excess moisture, effecting the bond. On average this adhesive needs about 12 hours to completely harden, however this figure can be effected by the amount of hardener added, additionally, a higher surrounding temperature may speed up the chemical reaction. On average, the epoxy has a 20 minute ‘open time’ (this is the time in which it may be applied or ‘worked’). This particular product has a gel-like consistency which makes it easy to spread with little mess and wastage.

In use, it was found that the epoxy will not adhere to ‘non-polar surfaces’ such as plastics like polyethylene. This means formers and jigs can be designed to utilize plastics as ‘releasers’.

[rhyscjones]

Birth of woven prototype 3

By tracing the form (in this case the actual chair) of my previous prototype, I am able to rework the profiles, informed by the physical feedback my prototype chair can offer.

This process tends to involve a lot intermittent sitting then drawing. This a relatively free and intuitive way of developing the design, so all drawing is done by hand. In these sketches, have inked the new shape in green. Note the addition of a new member which cradles (dry joinery) the back of the seat. It's purpose is to triangulate the junction - offering support. However, it originally came about from needing to visually link the top and bottom halves of the chair - creating a kind of pathway and granting a shared purpose between the seat and legs.

[rhyscjones]

Upholstered production prototype 2

The seat and back, now upholstered in vinyl, are inspected and attached to the frame of P2. 

I'm glad i was able to do a test run, as several changes have come to mind. For starters, the seat is wayyyy to thick. It looks like a bubble - it needs much less foam. I would be happy simply sitting on the timber - the leather is more of a surface treatment for me... Perhaps a thin, high-density foam which tapers at the edges. The problem is, the leather needs something to cushion the harsh corner, or it will get cut and damaged when hit.

The fixture underneath (a hidden metal ribbing) is miuch too obvious and irregular. The upholster recommended a seat pan that can be fixed to the underside - i agree. It should be relatively easy enough with the 3D ply.

As for the back, because of the extreme shape and taper, almost half of it needs to be hand stitched after being pulled over. This will only get worse when i increase the cut-away. I don't mind paying for the labour, especially if it's going to be high quality. My only concern is the seam - on a piece which has such a defined edge, which side does the seam sit on? If one at all?

Just a few things i need to discuss with the upholsterer.

Just a note: proper, robust hardware is needed for attaching these upholstered components. I need something i can really trust. I need a chair i can really trust...

[rhyscjones]

Considering form

This chair has yeilded some possitive results, but has also brought a whole lot of problems to light. This process of unpicking the design took place before the weaving had even been finished. Essentially, it looks good sitting on my bench. On paper, the side profile is gorgeous! Alas it is not meant to be... When the chair is placed on the ground, the under frame slinks away, leaving only the seating half. This woulnt be so bad if it didnt upset the visual balance so much. 

Those stretchers are still bothering me... They look directionless - floundering in space. Their shape has logic, but it just isn't apparent. I tried spraying them black (to pretend they weren't there) and mocking-up some stretcher variations, however, they didn't seem to work.

Using similar methods of mock-ups, (blue and white masking tape and odd bits of veneer) i have decided to give the chair's upper half 'hips'. Also, to help link the two halves, i will try adding a curved member which reaches up to cradle the back of the seat on either side (just dry joinery at the point of contact, though linked via laminations to the legs). These i will implement on the 3rd prototype soon to come...

[rhyscjones]

The details - woven prototype 2

Shown are some of the techniques used in constructing this organic form. Mitred edges are glued with epoxy, then splined before shaping. The majority of shaping must be done by hand using a combination of files. The seat is held to the base via 3 bolts on either side. I was thinking of routing a groove for the nuts in the final piece, then patching it with the nuts embedded.

For the first time i am having a glimpse at the finished form during a dry-run. I find this prototype to be much more refined than the previous, however, it appears as though the legs/base has slunk away in the process with the seat being much too dominant. I will have to weave it to find out what it feels like. I have attempted to make it more of a 'dining chair' by raising the backrest and making it more upright (this has reduced the seat depth).

[rhyscjones]

Construction of woven prototype 2

The construction of this prototype (which has taken place alongside the other chair over the last 2 weeks) is a direct progression of the last woven chair (P1). Woven prototype 1 was a response to the need for contrast between my craft practice and the chairs which utilize outsourcing and a simplified making process. The difference being that these woven chairs are incredibly light and generous in form (though not in dimension or material), where i have complete control over these elements ('in-house').

This prototype attempts to rectify some of the issues i had with the last design which primarily involved profiles and form (or curvature of the members). I have attempted to simplify the structure and visually define the seating planes.

The curved side members were glued using polyurethane in the same manner as the previous prototype. Blocks are glued onto the members to facilitate the new curved joinery and profiles.

Note: The Radiata Pine laminations proved slightly too thick for one of the seat sides during glue-up, and snapped. This i fixed using epoxy and some splines (as shown in the images). I will have to either reduce the lamination thickness or steam them in the problem areas prior to gluing (or both). 

[rhyscjones]

Profiles for CNC

One thing i can take away from the 3D model of my chair are the profiles for forming the laminated curves. Using these in conjunction with measurements taken from the CAD file, i will be able to produce an accurate frame - on which i can sculpt the brackets.

The files will be sent to Integrity Signs for CNC routing out of 12mm mdf, providing an accurate template from which to 1:1 route the formers from chipboard.

Note: In the above image, the form on the left (for the backrest/stretchers) has been drawn on the wrong side of the line (something i need to change...). For the formers that I will be making, it is important that the majority of the curves are 'male' (convex). In this case the shape is 'female' (concave), aside from the two, smaller outer-curves. This will become clear once i post images of a glue-up.

If the contrast between the two different curves is large, then a former must be made in two parts with a pivoting joint. This is something i will not have to do for these chairs (but has been done for inter se).

[rhyscjones]

3D modelling part 2

After meticulously recording information and measurements using production prototype 2 as a guide, I set about modelling the final piece in Form Z. (Not really a program i would recommend for this, however, it is all i have to choose from at this time).

The aim is to achieve an accurate model from which i can derive formers (to be CNC cut), and model the front and back brackets; mirror them, and have them rapid prototyped. This process ensures accuracy throughout the piece, enabling it to come together at the end.

But alas, if only it were that easy. After a week of modelling, problem solving and number crunching... the brackets cannot be resolved in a 3-dimensional space. Their form, though so simple to me, is too complicated for the program. Or rather i should say; I have designed these forms as though 'for craft', when in fact, to access these technologies is to design for the technology. This would mean considering the processes needed to arrive at a form - one is limited by the medium, in this case, the parameters of the program. In craft it is that of the material itself. 

I have worked at length with those i know who can help me. However it is the same problem. I am sure a professional in this field could achieve what i want quite readily. However, this is a luxury that i can not afford. 

And so, without throwing these forms away, i have opted for something a little more tradition - pattern making. I will attempt to sculpt a master model by hand of both left and right front and rear brackets. These i will send away for casting. My only hope at this point is that i have enough time to do so.

The separation of an idea from the reality is well illustrated in the development of the Vitra chair Vegetal, designed by the Bouroullec brothers. While both the concept and the reality are valid - they are two very different things. The development of this chair was very costly and drawn out for this reason. Perhaps there is a parallel to my chair here? There is, however, no comparison in cost and resources...