#4 Shaping Rib Stock

The wood that has been seasoned for making ribs now needs to be straightened and shaped to a distinctive ‘tear drop’ shape. This is narrow at the top so as not to obstruct the flow of water while forming sheets yet wide enough at the bottom to form pegs at the ends. These rest in holes to connect the ribs to the frame. Later posts will discuss shaping pegs, drilling for the stays and sewing, and otherwise finishing the ribs.

This drawing was made to calculate and keep track of the angles and dimensions of two sizes of ribs. The ribs that are being made in the following photos are the smaller size; 5/8″ deep with 3/16″ diameter ‘pegs’ centered 1/2″ from the top. Single faced laid moulds have ribs that are 1/16″ wide at the top. Double faced laid and wove moulds use ribs with 3/32″ wide top edges. The width at the bottom is not critical but the ribs need to be thick enough there to form the pegs that fit into holes in the mould frame. In addition to the sizes above I have used ribs that are 15/16″ deep for large moulds and even 1-1/8″ deep for very large moulds. And I have used 1/2″ deep ribs for very small ‘test’ moulds.

The rough stock which has been previously seasoned. (See previous post)

One side of each rib is flattened on the jointer.

Then the top edge is also made flat and straight, at 90 degrees to the side.

The ribs are sawn slightly oversized at the proper angle. The angle depends on the size and type of mould. For these moulds the angle is 7 degrees off of perpendicular. (Since the wide sides will form an angle of 14 degrees subtract half of that from 90 degrees) This makes a slightly acute angle of 83 degrees.

Then the rough side of each rib is sawn to the right angle on this suction fixture. Here the saw is set to create an angle of 14 degrees but leaving the rib a bit thick for one more trim.

This shows how the ribs are held to the special attachment with suction supplied by a shop vac. as they are fed past the rip saw blade.

The rib on the right has been jointed, straightened and had its bottom sawn to 83 degrees. The arrow points to the juncture of the two jointed sides; this angle is now 90 degrees. This corresponds to step 2 in the photo below.

Stage 1 is the rough stock. Stage 2 has the left and top faces jointed and the other two faces sawn to the correct angles. (So far in this post stages 1 and 2 have been completed.) In stage 3 the stock will be straightened once again and reduced a little more, but with the 90 degree angle on top corrected to 97 degrees. ( When the ribs are installed in the mould ‘standing up straight’ the top surfaces will be aligned with the top of the mould). Step 4 shows the finished stock. The reason for reducing the dimensions of the ribs in two steps is to make them straighter. They will distort slightly each time material is removed; this way they are likely to distort less the second (last) time.

The wide face of each rib is once again flattened and the top edge jointed, but not at 90 degrees this time. This shows the jointer fence being set to 97 degrees to joint the tops of the ribs.

The ribs are now trimmed to their final depth. It is best for the fence to have an angled edge to match the top angle of the rib. This is easily done by setting the angle for the rib bottom on the saw and feeding a thin strip of wood through. Then it is clamped in place as a temporary fence to saw the wood accurately to width.

The ribs are ready for the second and last run through the vacuum fence. At the upper right you can see the 5 suction holes that hold the rib against the fence. The clear plastic piece at the top can be adjusted to help hold the ribs in place. The vacuum hose plugs into the large hole on top.

The ribs are pushed through one after another, each pushing the previous one, and they fall off onto the floor.

The rough side of each rib is now planed smooth. A block of wood spans the cutter head and presses the rib down against the in-feed and out-feed tables. First the rib is pushed and then pulled to keep the right hand at a safe distance. The same method of adjusting the tablesaw fence is used here as was previously used in thicknessing the mould frame. First a test piece is sawn on the vacuum fixture and planed smooth. Then it is measured to see how much more needs to be removed to end up at the right thickness. The fence is adjusted by that amount using the dial indicator.

The bottom edges are rounded using a rounding router bit and an angled fence. A narrow band of the rough (but accurate) surface is left. After the ribs are complete and ready to install in the mould frame this will mostly sand off with a few swipes of a sanding sponge.

If the rib stock was correctly made the pieces will slide easily into this jig which will be used later to shape dowel-shaped pegs on the ends.

Thus 5/32″ radius round over bit was used to shape the bottoms of these ribs.

Angled edges on the fence allow the rib to be rounded in one pass each side.

Another view of the vacuum/suction fixture. The saw blade can be tilted in the saw to prepare ribs of different angles.

The rib is sucked tightly against the angled fence while being pushed through.

#3 Prepare Frame Stock for Paper Moulds

The wood for making Paper Mould frames has been prepared to make pieces that are relatively stress free and stable. Now the stock will be shaped to its final dimensions prior to cutting to length and joining at the corners with dovetails.

Shown above is a block of poplar that I shaped to demonstrate the method I use to make the frame pieces of even thickness. I used exaggerated angles to make it easier to see.

To begin one side of each frame piece is jointed perfectly flat on the freshly sharpened jointer. I am experimenting with better guards for machines to make them safer to use.

Next one edge is also jointed perfectly straight and at exactly 90 degrees to its previously jointed side.

The wood rests on its newly flattened side with the jointed edge against the tablesaw fence. It is sawn (with the rip saw of the previous post) at precisely 90 degrees to the jointed face. The fence is set so the pieces end up about .015″ (1/64″) wider than the intended final width. They will all end up exactly the same width and the slightly rough edge will be left that way until after the joints are cut. This makes it easy to identify the bottom edge of the mould frame. It also helps avoid mistakes as the pieces are put through all of the steps to create the dovetail joints. Even though the sawn edge is rough it leaves a precise, square, straight edge to measure against. The wooden ‘featherboard’ pressing on the piece being sawn means my fingers don’t get close; a push stick is used to finish the cut.

This photo shows how the home-made dial indicator fixture can be used to help adjust the fence very precisely. As the fence is moved laterally, the dial on the left will measure the change.

The first step to thickness the wood uses the rip saw raised to just over half of the width of the frame pieces. It is set at a slight angle; about 1/3 of a degree. Each piece is sawn in two steps to create a very shallow ‘V’ shaped depression. This piece was withdrawn partway through to show the process.

You can see the shallow V-shaped space when a straight edge is placed there.

If the saw is sharp and the fence is straight this will leave edges that will measure exactly the same width all along the length. When placed rough side down on the jointer only the outer edges will touch the table. When pushed through the jointer the pieces will be planed to a precise thickness. It works very well! Here a test piece is used; first sawn and then jointed at the end. This is measured with the vernier calipers to see how much the table saw fence needs to be adjusted, with the help of the dial indicator, to achieve the desired thickness.

The mould frame stock is finished and ready to cut length. Each piece has one rough side and three smoothly jointed sides and all are virtually identical in thickness and width. And, as you can see, the pieces are very straight.

More detail

To test to see how well a jointer is working see how a freshly jointed piece acts by gently moving one end as shown in these two photos.

If the piece pivots on the middle as above the piece has a ‘belly’. The star indicates the approximate pivot point.

These two photos show a better outcome…

One end will move and the other stays in the same place. It pivots near the star at the right. This shows that the piece is ‘hollow’ (microscopically) but this is better than having a belly. Woodworking planes work this way. If you are preparing to join two pieces of wood, set the plane for a fine cut and try to make the edges hollow. The tool will not allow that to happen (except to an insignificant degree). The next 4 photos will attempt to show a machine process of slightly hollowing a piece first, then jointing it flat. It ends up sounding kind of insane but what I’m trying to describe becomes pretty intuitive once you get the hang of it.

The guard is opened so the piece to be worked can be dropped in. The jointer is set to a very fine cut; about .005″-.010″. This is possible because the wood is already very flat.

The workpiece in place with about an inch hanging over the out-feed table.

The guard is pushed in and with light pressure at both ends the piece is started here, moved fairly rapidly from right….

…to left and ending in this position. The ends aren’t trimmed at all. This is repeated 2 or three times and has the effect of slightly hollowing out the middle, while removing any ‘hump’ or ‘belly’. Don’t move the guard, just pull the piece back through to start again. You can hear and feel when the knives can no longer reach the wood to cut any more. Then the piece is pulled completely out to the right, lifting slightly to clear the knives. This sounds harder than it is. Being slightly hollow the ‘board’ now rests on its ends so it can move into the cutters without any tendency to rock. As the left end of the piece slowly passes the knives a true plane is established on the bottom edge of the wood which continues to move along to rest on and be supported by the out-feed table. As the board is slowly fed past the knives, that plane is extended until the right hand end leaves the in-feed table and passes the knives, finishing the cut. This is repeated one more time to remove a slight ‘divot’ that will likely occur at the left end; this last pass especially being nice and slow.

The thicknessing method.

Here is the poplar ‘demonstration block’ being sawn at a slight slant. I set the saw at 3 degrees (instead of the usual 1/3 degree) to make it easier to see. The usual angle is somewhat arbitrary, just enough to create two clean edges for the wood to ride on without having to joint too heavily to remove the saw cuts.

Partially cut.

The hollow is visible beneath the straight edge. When the piece is turned over only the outer edges will ‘ride’ on the jointer table as the saw cuts are planed off.

Measuring the thickness at numerous places along both edges with this vernier caliper showed variation of less than .001″. (1/40mm) The measurements hardly strayed off the line possibly showing variation closer to +/- .0005″!

After jointing a smooth face replaces the rough sawn one.

Measuring wood in thousandths of an inch might seem ridiculous. But being able to control dimension with basic low-tech tools is kind of satisfying. And it makes the later steps easier. Wood will always move; these ‘perfect’ pieces of wood will not stay that way but they will likely be closer. I am convinced that the care taken now pays off in the long run as the pieces of a mould end up being more ‘relaxed’, each doing its part and not working at cross purposes. At least not as much.

The question comes to mind: “Why not just use a thickness planer?” Early on I did. Even for ribs, though that was soon abandoned. The rollers mashed one side of the wood and distorted it. Probably the need to find a way to make the delicate ribs without damaging them led me to this way of making the frame stock. I can’t remember the entire process of discovery.

In my shop a thickness planer is a valuable tool but I have not sharpened mine in years! It is a lot of trouble to change the knives. I use it to save time when shaping larger pieces of wood. If wood is jointed or planed with dull knives the edges are actually ‘pounding’ on the wood as well as cutting. This smashes and bruises the surface. When the wood gets wet smashed areas will swell and ‘rise up’ making the surface rough. Wood cut with very sharp tools won’t do this as much. I hardly do anything to smooth the wood of a paper mould after the wood work is done. Edges are slightly rounded with a plane and the outsides are smoothed just a little with a fine sanding sponge.

More on the thicknessing method.

In the drawing above the rectangle on the right represents the desired final width dimension (B) of the mould frame in cross-section. On the left the rough, DOTTED area represents the part to be sawn off. This in preparation for planing off (on the jointer) the DASHED area. If you try to saw the frame stock to width in one pass some problems may occur. To make the bottom dimension (A) the same as the top dimension (A2) will take very careful adjusting of the tablesaw. No matter how careful you are the two are not likely to end up the same. For one thing at the top edge of the wood (along C2) the circular saw blade will be cutting 1-1/4″ to 1-1/2″ above the tablesaw surface while along the bottom (at C) the wood will rest on that surface. Dimension “A” is likely to remain very consistent while dimension “B” may vary a bit with slight differences in pressure as the wood is fed through the saw. Circular saws can ‘flutter’ which is likely to cause a slightly more irregular cut at the top than at the bottom. This may leave a surface that is not functionally flat; it may rock slightly on the jointer. If it rocks at all, the thickness will not be consistent after the rough saw marks are planed off.

Trimming the frame piece from both sides with the saw set at a very slight angle reduces or eliminates these problems and it’s so easy! First there is no need to carefully adjust the tilt of the saw to try to make the top and bottom edges equal in one pass. The tilt of the saw is not at all critical; just enough to create the shallow depression along the center but not so deep that an excessive amount of wood needs to be planed off. (About 1/3 degree for moulds) The saw cut only needs to be half as deep so there is less opportunity for a ‘wobbly’ or ‘fluttery’ cut. Dimension “A” at BOTH top and bottom is ‘defined’ by the distance between the fence and the saw blade at the level of the table. This is as accurate as you are likely to get on a tablesaw. Since the jointer knives are set to very close tolerances the two sides will end up virtually parallel. When being pushed over the jointer the piece rides on the edges only (C) so it can’t rock.

For yet one more attempt to explain this in words check out my reply to the comment posted below.

#2 Sharpen Tools

The wood for moulds, ribs and deckles has been seasoned, straightened and trimmed to near final dimensions. At this point I pause to get the necessary tools tuned up and ready to go. (I’ll post general information first, and include more detail at the end for those who are interested). These are current sharpening methods for the tools I depend on the most.

This 8″ Hollow Ground Planer saw blade plays a major role in my method for making paper moulds. Many sharpenings have reduced the diameter of the saw which started out exactly the same as the new one behind it.

I found a few of these on Ebay. I don’t know if they are still made.

When the saw is new the teeth are all filed nearly straight across like this and it doesn’t cut well.

The teeth should be filed at 15 degrees like this. Every fifth tooth is a raker which is filed straight across and at a lower level to cut slightly below the tips of the cross cut teeth. This blade cuts very smoothly and I use it a lot for the finer work making moulds. Steel isn’t as hard as carbide but is less brittle so the the teeth can be shaped to sharper points. It seems to cut more freely and with less chip-out on the relatively soft woods used for moulds.

This is a rip saw. Its teeth are all shaped like little chisels and it is used for cutting wood parallel to the fibers (‘ripping’). I’ve learned how to use it in combination with a sharp, well-tuned jointer to create perfectly straight pieces of perfectly even and consistent thickness. This and the hollow ground planer are the only two saw blades needed for making moulds.

This is my home-made saw filing vise which is used to keep both types of saw sharp.

This is a home-made stake and anvil that I use occasionally to put ‘set’ in the teeth of the rip saw. Used with a hammer and punch this tool is used to bend the tips of the teeth to create ‘set’. The cut made by a saw must be wider than the saw’s body to prevent binding. The ‘anvil’ is just a hardened bolt with its outer edges eased to provide an angled striking surface.

A very sharp, well tuned jointer is essential for this method of making moulds. On top is a very simple tool that holds a dial indicator to check for proper alignment of the in-feed and out-feed tables of the jointer.

There are countless ways to sharpen. I have recently discovered a very effective way to hone jointer knives. I purchased an old surface grinder some years ago to help me make bookbinding tools. It came with a magnetic chuck, shown here dismounted from the machine. The top surface is perfectly flat and can be turned into a powerful magnet with the turn of a lever. This can hold the jointer knife as well as this roller assembly firmly in place to create a controlled honing angle. A more detailed description is given in the second part of this post.

Another simple shop-made device holds a dial indicator for setting the jointer knives. This tool also gets used a lot for making small adjustments when using the tablesaw, jointer and router table.

This machine works very well for sharpening chisels. I’ll use the four chisels shown for paring the dovetail joints of the mould frames and for detail work on the deckle joints.

A Deeper Dive…

The information given above just scratches the surface. Below is more detail for those who are interested. There is still much more to know and to discover.

Marks on the saw vise show the correct angle to hold the file. The saw must first be ‘jointed’ on the tablesaw. The saw is set for exactly 90 degrees and lowered so the teeth are just below the table surface. A carborundum sharpening stone is held directly over the saw blade. With the machine running the saw is raised EXTREMELY slowly until a few sparks are visible. (Dim the lights). This will dull the teeth slightly so that all are at the same height. The teeth should be altered the smallest possible amount. Only enough to show a tiny shiny spot on each tooth (minus the rakers).

Teeth are filed alternately on both faces until the shiny spots just disappear, leaving the teeth sharp and all at the same height. The raker teeth must be jointed in a different way with each tooth held at ‘top dead center’ in the vise. Careful adjustment and a file used with a block of wood joint the cutting edges about 1/64″ lower than the other teeth. Then they too are filed sharp until the jointed spots disappear. The angled teeth slice the wood fibers like little knives while the straight teeth rake out the waste between like little chisels. Since the higher, angled teeth contact the wood first the the fibers are sliced, not ‘torn’, leaving a clean surface.

The saw gets reversed in the vise to file alternating teeth.

I rub the file on a file card frequently. I spray a little WD40 on the card which helps the file cut smoother without ‘chattering’.

This rip saw is much easier to sharpen. Here, it has been jointed on the tablesaw. You can see the shiny spot that now needs to be filed away.

First the front faces of the teeth are dressed smooth. Here the saw is positioned with a tooth at top dead center. This way you can angle the file to ‘feel’ the correct relief angle while filing. The narrow margin visible along the edge of this tooth is where it was jointed. This saw has very little ‘set’ remaining and that will soon need to be addressed. Less often a saw also needs to be ‘gummed’ which means the gullets between the teeth are filed deeper and the teeth re-shaped as needed.

The tops of the teeth are filed until the shiny part just disappears.

The saw vise can be adjusted for different diameter saws.

The stake and anvil are adjusted with shims so the blade sits flat on the ‘anvil’ (a large hardened bolt head that has been ground to a slight bevel at its edges). A tooth is positioned to hang over the edge just enough to be bent slightly with a punch driven by a hammer blow.

The center bolt can be moved for different diameter saw blades.

This tool is just a sturdy piece of wood as long as the longest table with three screws on the bottom. A hole at the far end holds a dial indicator which is adjusted to agree with the other three points to define a plane. You can then move the beam around to measure misalignment of the jointer tables. They should be exactly parallel to each other (even when set at different levels). Once you get the machine aligned this step doesn’t need to be repeated. I mention it here because it is important for the machine to be set up well for the precise work needed to make moulds this way.

The dial indicator ‘zeroed out’. Assuming the table is flat the tips of the three screws and the tip of the dial indicator together agree to define a plane. The beam can be moved around to check alignment since the spring loaded tip of the indicator will move up or down, measuring differences in thousandths of an inch.

Moving the beam along the in-feed table shows variation from parallel of the out-feed table. Here it registers zero at the front edge of the out-feed table.

Here the indicator registers zero at about the middle of the out-feed table. This is as far as it can go without having a screw drop off the edge. (The pencil mark indicates the location of the two middle ‘feet’). The weight keeps the three screw feet contacting the surface of the in-feed table. By referencing the flatness of the longest table the dial indicator can measure deviation over about half that length.

The jointer knife honing jig that relies on my ‘mag-chuck’. The roller is the core of a defunct printmaking brayer that I found in my scrap box. I carefully drilled the two pieces of angle iron so the roller is held parallel with the flat surface of the magnetic chuck. It is also important that both ends of the roller are set back exactly the same distance from the front ends of the angle irons. These bump up against a piece of rod which in turn rests against the back of the jointer knife being honed. Everything is locked into place when the lever actuates the strong magnets in the chuck. This establishes a honing angle which can later be altered slightly by inserting different sizes of rod. In place above and below is a piece of 3/16″ diameter threaded rod which defines the first (shallowest) honing angle.

The knife is first honed with a diamond plate.

The results of using the diamond plate.

Swinging the lever from right to left de-magnetizes the chuck. This has enabled the first rod to be replaced with a smaller (5/32″diameter) rod. This moves the roller closer to the sharp edge of the knife to establish a slightly steeper angle so this India stone can remove the ‘scratches’ left by the diamond plate, but only at the very edge.

Now an 1/8″ rod makes the angle a little steeper still so this ‘Washita’ Arkansas stone can polish the edge further.

To remove the burr the knife is placed bevel side down. The magnetic base holds the knife, a 1″ wide rule and the 5/32″ rod firmly in place so this Arkansas slip stone can polish a very shallow micro bevel on the flat side of the knife to take off the burr. Few woodworkers will have access to a magnetic chuck. But I thought it deserved mention since it works so well.

A simple home-made dial indicator fixture which I use all the time. Three screws on the bottom establish a plane. A dial indicator mounted at one end measures vertically and one mounted at the other end measures horizontally in thousandths of inches. A magnet helps hold the fixture in place. I often place a weight on the top of the tool for this purpose, too.

The dial indicator is zeroed out on the out-feed table.

Then it can be moved over to set the knives at the same level.

#1 Season and Prepare Wood for Moulds, Deckles and Ribs

Mould Frame and Deckle Wood

Wood usually contains internal stresses. If the trunk of a tree grows at an angle ‘reaction wood’ will develop on the bottom or top to strengthen the tree. A windstorm may damage a tree causing different growth patterns, introducing internal tensions. Improperly drying wood can also introduce stresses. None of these are good for woodworkers. Most are not apparent until you begin to work with the wood. When it is sawed into narrower widths the release of internal tensions will usually cause wood to change shape. Sometimes this is very slight as shown above. If you are very lucky there is no discernible stress and the wood will hardly change at all.

This wood, originally from the same large plank, shows greater tensions being released. The wood I use is marketed as “Genuine Mahogany”. I don’t know the exact species. It is good for mould frames and deckles. It has little difference between tangential and radial shrinkage and swells relatively little when wet. And it is soft enough to receive brass and copper nails which is important for mould making.

Boiling wood for mould parts helps to release stresses. Lignin, the substance that binds wood fibers together, softens with heat, allowing the fibers to shift slightly. This boiling tank is made from a 4″ diameter piece of iron pipe. The bottom end is threaded and bushings are screwed in place to reduce it so that a 110 volt water heater element can be screwed in. I bring the water to a boil with the wooden parts submerged then unplug it to let it rest for an hour or so. Then I bring it back to a boil before removing the wood.

This also tests the wood. If the wood warps or twists badly it is not good for making moulds. If the wood barely changes you can be more confident that the wood will not warp or twist during the life of the paper mould. Many, many hours of work go into a mould, making it prudent to test and prepare the wood ahead of time. I prepare the frame and deckle wood in several steps. It gets boiled, allowed to dry, straightened (jointed, in woodworking parlance) and trimmed smaller. Later it is boiled and dried again and jointed and trimmed a second time. I sometimes add an extra boil for the “L” shaped deckle stock. In the process you will get a good idea how the wood will behave and with luck gain confidence in it.

This wood has been boiled and allowed to dry. Two sides of the wood are now jointed (made straight) using a power jointer. This removes any warp or twist.

A tablesaw has been used to trim the other two sides parallel with the jointed sides. Thus the wood gets a little smaller with each step. The internal tensions are gradually reduced as the wooden parts get closer to their final dimensions. The time and work actually spent preparing wood this way is not great. However you must plan ahead a few weeks to allow for drying between steps.

Deckle stock is prepared in the same way but it is shaped into a rough “L” cross section. In this photo the small scrap pieces that were sawn out remain very straight. This is a good indication that this wood is relatively stress-free.

Here, the shape of the deckle has been further refined; very close to its final shape but still slightly oversized.

This frame stock has been boiled and straightened twice. After drying out fully it will be ready for final shaping. Now the wood is still only ‘rough shaped’! To become a paper mould it will be further shaped very precisely and with very sharp tools. This will be described in future posts.

Rib Wood

This is a plank of vertical grain Western Larch, a deciduous conifer often called Tamarack. I use this species for making ribs for moulds. I have Jim Croft to thank for this. He lives in Idaho where larch grows. He had it specially sawn for me with the growth rings in the ‘vertical’ orientation (approximately 90 degrees to the wide face). Larches can die and remain standing for many years without decaying. With the aid of my Optivisor and a pin I counted the rings. This tree was well over 300 years old when it died (the heart of the tree is not in this plank). It must have grown in a dense stand; the growth rings are miniscule. This makes great rib wood. Even with this old growth wood I use only the outer, finest grained wood for ribs. (I am spoiled). I feel a little guilty cutting it up into little pieces. Before Jim provided larch I would comb through stacks of clear white pine at my local lumber yard looking for perfect boards.

This tree only gained about 1/32″ to 1/16″ in diameter per year! As a result of growing so slowly and in a dense stand this wood has practically no internal stress.

It is easy to see if a vertical grain plank has been sawn parallel to the growth rings. It is harder to tell if it has been sawn parallel to the trunk in the other orientation. Here I have split a scrap of the plank and found grain ‘run-out’. This is best avoided for delicate paper mould ribs. I will re-saw narrow sections of this plank to eliminate the run-out before sawing out the tapered ribs.

Here the wood has been re-sawn to eliminate the run-out. This piece had a crack that shows the orientation of the grain. I drew a line to accentuate this. The cracked part won’t be used for ribs; there is solid wood on the other side and the other blanks had no cracks.

The prepared blanks are sawn into tapered rib shaped pieces. I cut these pretty close to final dimension since the wood is so stable.

Rib stock for a couple of moulds. In Britain these are called bars but I have always called them ribs.

Rib stock only gets boiled once, just to test for warp and twist. This wood turned out great. After boiling I stack the pieces so that air can circulate all around.

#20 Making Laid Facings

For a detailed look at how this loom functions go this site: paper.lib.uiowa.edu/epm*. Here you will find an extensive list of related internet links. Near the bottom of the list select “Z3”. It will take you to a video that I made of the process of making a laid facing on this loom. The quality of the video is not great but it was the best way I could think of to explain the workings of the loom.

*This is a companion website to the book “European Hand Papermaking” by Timothy D. Barrett. It is published by The Legacy Press (www.thelegacypress.com). The book includes an appendix on mould making that I contributed.

Photo by James Kleiner

In the foreground is my current loom which I made in 2011. In the back is the original loom which I used throughout the main part of my career making moulds; 1982-2012. The new loom incorporates many improvements though both work on the same principles.

Chain wire is measured out for setting up the loom.

Each length of chain wire is draped over the steel rod visible along the top of the photo. The wire’s ends pass through a slot in the wooden wire trough and down through a pair of holes in a plastic spindle. Then they pass through holes in a ‘wire slide’ and are secured by twisting the ends tightly together. A one pound weight is attached at the bottom to hold each wire taut. (Later each wire will become a ‘pair’ as the two halves are twisted. They resemble tiny chains as they twist around the laid wires, thus the name).

Setting up the loom involves adding a weighted wire for each of the spindles. The length of the wire and number and spacing of spindles varies depending on the size of mould being made.

The wooden ‘wire trough’ has been elevated while the loom is being strung up. In use the trough covers the spindles. The spindles are driven clockwise as the weights are reeled in by a crank. The weighted cords slip as they are unreeled so the spindles do not reverse direction.

Foam board spacers have been added to separate the wires below the twisting mechanism and a couple dozen laid wires have been added, partially completing this facing. Each time a laid wire is added a crank turns all of the weighted spindles simultaneously, twisting the chain wire pairs 1/2 turn. This action incorporates one laid wire into the facing. Next, a different crank is turned to lower the entire twisting mechanism (spindles, weights, wire trough etc.) a specific amount, the measure of one laid wire and one space. These steps are repeated over and over until the laid facing is complete. The video gives a clearer idea of how this works.

This facing is complete. A bundle of laid wires is shown which will be made into another facing.

This facing has been cut off the loom. It will become part of a mould being documented for this sequence of posts. Laid backing wire will be covered in a later post. More later!

#19 Straightening Paper Mould Laid Wires

Wire can be purchased in coils or on spools. In either case the wire comes off curved. In order to make a laid paper mould facing or backing it is necessary to make it much straighter.

Above, a piece of .0226″ diameter 1/2 hard phosphor bronze wire right off the spool resting on some of the same wire that has been run through a straightener and cut into lengths. While not perfectly straight, the processed wires can be made into a laid facing.

The major curve of the wire is called ‘cast’. This corresponds roughly to the diameter of the spool or coil. Another way that the wire is curved is called ‘helix’. The helix can vary, from very slight as shown above, or very ‘steep’ as shown below.

The same two pieces laying flat. The smaller circle of wire is the 1/2 hard wire mentioned above, while the larger circle is spring tempered wire of the same diameter. This wire has a more pronounced helix shape and won’t lay as flat.

A traditional way to straighten wire. The wire straightening blocks above and below are from Ron Macdonald’s collection of tools. These photos were given to me by Serge Pirard who studied with Ron and now makes moulds using many of Ron’s tools.

I tried to straighten wire this way but soon gave up!

Above and below are photos sent to me by Sergey Nasaev of Russia.

This is the wire straightener that I made using the information sent to me by Sergey. Ball bearings are stacked in pairs to create grooves which guide the wire back and forth to remove the cast. I expected that the wires should come out almost perfectly straight, like the pre-straightened wire I had been using for years. So, I was a little frustrated at first when the wires weren’t as straight. In retrospect this turns out not to be necessary or even desirable. If you try to make perfectly straight wires this way you may go crazy!

I made a second straightener out of steel with bronze rollers. The base plates of the first design were made of plastic. Steel, being much harder, can be more accurately calibrated. Now I can record the best settings for various sizes of wire so this unit can be used for all sizes of paper mould laid wire. Wire straighteners can be purchased but I chose to try to improve on what I had instead. (Besides, I like to make things and this looked like a fun challenge.)

A closer view of the steel and bronze wire straightener. The first set of seven rollers is used to remove the cast. The second set is used to remove any remaining curve caused by the helix.

The horizontal bank viewed from above. You can see that the rollers are set to bend the wire back and forth and less aggressively as the wire moves from left to right. Adjusting screws are located at either end of the slot between the rollers and are accessed through deep holes in the front plate. The two tightening screws are visible at the bottom of the photo.

The photo above shows the use of a built up pad to orient the cast of the wire as it enters the first rollers. This photo has been added to show an improvement in the process. Please refer to post # 58 for more details. And see “An Addition” at the bottom of this post.

The vertical bank. I have the best luck using five rollers here. Wire this light requires only 7 or 5 rollers per bank; heavier or harder wire needs more rollers. The middle five rollers have been selected to remove a helix spiraling upward. If the helix curved downward the upper right roller would be removed and one added at the blank spot at the lower left.

The wire enters the horizontal bank of rollers. The cast is removed and slightly reversed by the first three rollers. Each trio of rollers creates a bending action as the wire is forced around the middle one. The remaining rollers form overlapping groups of three that apply a gradually diminishing amount of force. The wire is forced first one way, and then the other, to gradually change its shape from very curved to nearly straight.

After the first set of rollers are adjusted the wire will look fairly straight when viewed from directly above. But it may still have a noticeable curve when viewed form the side. This curve is mostly the result of the helix and the second set of rollers have the job of removing it. Since this is much slighter they don’t have to push as hard. The wire lengths seldom turn out completely straight. If the straightener is calibrated right the wires repeatedly cycle from nearly straight to less straight and then back again; but fortunately staying within a useful range. I don’t fully understand all that is happening but think the wire has acquired some slight variation in the spooling or coiling process. Also wire twists slightly as it is unspooled. Either may cause the wire to shift slightly in the rollers, affecting the final shape. (This is my current theory.)

I pull the wire with this pair of wire cutters.

One jaw rides along the fence to keep the pulling angle constant.

When the outside handle hits the back stop the handles are squeezed to nip off a length of straightened wire. Then the cutter is moved back to the front stop to grab another length to begin again.

The top three bundles are .0226″ diameter 1/2 hard wire and will be made into laid facings for three moulds, two 12″ x 18″ moulds and one 10-1/2″ x 15-3/4″. The lower three bundles are .0254″ diameter wire and will be used to make the backing wires for these moulds. They are all to be “double-faced laid” (also known as “modern laid”) moulds.

Wire straightening is fairly new to me. I made moulds for about 25 years using wire that I purchased pre-straightened in lengths. After seeing a video of Ron Macdonald pulling wires from a coil I was inspired to try again. I like the idea of not needing a source of pre-straightened wire. I think that the hand pulled wires give a more interesting surface which carries over to the paper made on the mould (though this is pretty subtle).

I have tried to convey this information clearly. If something doesn’t make sense or if you would like further clarification let me know. I often have extra photos that can be posted and can also try to state things more clearly if I am aware of your questions.

An Addition:

I made this device to add to the front of the wire straightener. It helps to orient the wire as it enters the rollers. Ideally the first bank of rollers removes the cast only but without some help the wire has a tendency to twist and wander a little as it passes between them. If the cast enters the rollers a little slanted the rollers won’t just be working at removing the cast but will also be altering the helix. I don’t think you can completely isolate the two with this simple technology but this does seem to help a little.

The wire is pulled between two sheets of polycarbonate. The bottom sheet is screwed down to the wooden base. The top sheet floats, held somewhat loosely in place by the two brass pins and by the white plastic (acetal) disk. A couple layers of tape hold the sheets apart so, when weighted, the wire is free to move. The wire is pulled past the disk while being held flat between the plastic sheets to hold the cast of the wire in line with the grooves in the first bank of rollers. This helps the rollers to work on the wire in a more uniform fashion.

Another view with the weight removed.

The weight has been added. The wire can move freely enough that the cast isn’t pulled straight. The wire isn’t pulled around the white disk hard enough to change the curve (cast) any. The disk just helps keep the wire oriented.

Order of Operations

Paper Mould and Deckle Construction

Following is a list of the topics I plan to cover as I document some European style moulds and deckles that I am making. As you can see this is a big topic. Some fairly simple steps are covered but others will be rather vast. I won’t follow a strictly traditional method of making moulds; instead I intend to extensively document the method I have arrived at after almost 40 years in the trade. If all goes well over the next few months pages will be added to cover all of these steps in detail.

Season Wood

Rough Wood to just over final dimensions

Sharpen Tools

Prepare Frame pieces

Cut Frame Joints

Waterbar Mortises

Prepare Ribs

Drill Sewing Holes

Fit Ribs to Frame

Fitting Stays

Assemble and Glue Mould

Clean up Mould Frame and fit Brackets and Rub Strips

Pin Ribs and flatten Ribs

Straighten Laid Wires

Make Backing and Facing on Loom

Fit Wire Facing and Backing to Frame

Sew down Facing

Make Deckle Parts

Deckle Joints

Deckle Camber

Deckle Glue Up

Deckle Final Fit to Mould

Repairing a broken paper mould

A lot of gorgeous paper was made on this mould before it broke. I undertook to return it to a useable condition and learned a few things in the process.

The bottom of the broken side with rub strips removed. The insertion of the brass brace rod may have helped cause the break.
This frame was put together to hold the mould flat and to enable a router to cut a true surface on which to glue strips of new wood.
The first layer glued and screwed with epoxy and then machined (with the router) and ready to receive the next layer.
Beginning to repair the top of the mould. The (blue) blocks were clamped in place to guide the router to protect the chain wires where they overlap the frame. Notice that the wood that has been exposed is ‘just like new’. This mould has been in and out of water for years and has no remaining finish to keep water out. Yet beneath the darkened surface the wood is perfectly sound.
The frame has been cut away to just above the rib pins. You can see that the broken part of the mould has been almost entirely replaced, leaving only the narrow part in which the rib pins are inserted.
A longer area is routed away for the final top layer which will end up flush with the original top of the mould.

The mould had problems with sides that warped outwards. Braces had been added to restrain this but weakened the mould frame, resulting in a break which made the mould unusable. I decided to repair the mould by replacing parts of the mould frame with staggered strips of new wood which would be screwed down and glued with epoxy. The first step was to rough out a long area on each side to prepare a space to add the first layer.

The mould was then clamped into a specially made temporary frame. This provided a reference surface to enable a router to machine a true surface for glueing the new wood strips to.

New braces were installed along with the first layer. These are of a different design and should be much stronger than the ones they replaced.

After two layers of wood were added to the bottom, the second longer than the first, the mould was flipped over and clamped in the frame with the wire side up to repeat the same process of cutting away old wood and adding new strips but working from the top of the mould.

Great care was taken not to let the router bit contact the laid or chain wires. The remaining parts were carefully chiseled way to leave a good surface for gluing the next strip in.

A final layer was then added to bring the ‘patch’ flush with the top of the mould. The copper edging was tacked back into place, finishing the top repair.

One more layer of wood was added to the bottom and given a rounded shape to match the remaining original parts of the mould frame.

The first cut has been roughed out freehand with a router fastened to a long base to span the sides of the mould. The little ‘pedestals’ were left to give the router a surface to ride on. They were chiseled off before machining a smooth surface with a router in the temporary frame.
Two new braces were installed along with the first layer of wood. Threads were cut on both ends of 1/8″ brass rods. Each end was then threaded into a tapped delrin block. This gives a very strong connection.
A second layer has been added to the bottom of the mould.
The router is attached to a long base that spans the width of the mould.
The first top layer glued and screwed
The broken side of the mould showing the repair. One more layer will be added to the bottom to finish the repair.
Gluing on the last strip on one side of the mould.

What I learned.

The sight of the ‘like new’ wood that was exposed during the repairs reinforced my suspicions that any kind of finish which attempts to ‘protect’ a paper mould from water is most likely pointless. I think if the wood is allowed to dry thoroughly between uses the mould will be fine. I think wet wood (wood is ‘hydrophilic’) allows water to flow better (along the surfaces of the mould while sheet forming) than wood that has been covered with a water resistant (‘hydrophobic’) coating.

Preparing and testing wood before using it is important to prevent the mould from distorting when wet. Unproven wood can lead to multiple problems due to warping; deckles that don’t fit well, difficulty couching, and uneven stresses on parts of the mould which can shorten its life.


Rub strips made of boxwood or hornbeam (I can’t tell the difference) are nailed to the bottom edges to protect the mould from wear. On this mould (and others I’ve seen) these are attached in short sections that have gaps between them. The gaps seem intentional and it’s hard to see what purpose they serve. You can see here that a single piece of wood was nailed in place (the grain is continuous) and THEN sawn in two! The saw cut goes down into the mahogany frame. Was this to keep the strong, hard, possibly ‘ornery’ boxwood from distorting the frame? There may be another reason but I can’t think what it would be. The nails used to attach it are ring shank nails; possibly bronze boat nails.


This my very first blog post. The choice may seem arbitrary and, in fact, it was! I hope to share a lot of stuff I’ve learned over the years but the order of topics will likely follow no discernible pattern. I hope some of these posts will be interesting to you.