Paper Mould Making


Directory to Mould Making Posts

With a few exceptions these posts appear in reverse order in the blog. Click on the category link “Paper Mould Making” for links to these posts.

Construction of Moulds and Deckles

  • #1 Season and Prepare Wood for Moulds, Deckles and Ribs
  • #2 Sharpen Tools
  • #3 Prepare Frame Stock for Paper Moulds
  • #4 Shaping Rib Stock
  • #5 Mould Frame Joinery
  • #6 Rib Spacing Layout
  • #7 Shaping the Rib Pegs
  • #8 Drilling Holes in the Frame to Support the Ribs
  • #9 Finishing the Ribs
  • #10 Assembling the Mould Frame and Ribs
  • #11 Fitting Waterbars and Brace Rods
  • #12 Gluing Moulds
  • #13 Making Corner Braces
  • #14 Fitting Corner Braces
  • #15 Pinning Ribs, Corners and Brace Rods
  • #16 Leveling the Ribs
  • #17 Finishing the Mould Frame
  • #18 Definitions of some Unfamiliar Terms
  • #19 Straightening Paper Mould Laid Wires
  • #20 Making Laid Facings
  • #21 Making a Backing on the Loom
  • #22 Fitting Backing and Facing to a Double Faced Laid Mould
  • #22-1/2 Fitting the Laid Facing to a Single Faced Mould
  • #23 Sewing a Double Faced Laid Mould
  • #24 The Functions of Backing Wires
  • #25 Wove Backing
  • #26 Making the Grid for Wove Backing
  • #27 Mould Brass Sheathing
  • #28 Fit Wove Facing and Sew
  • #29 Copper Edge Strips and a few more details
  • #30 Getting Ready to make Deckles
  • #31 Tool Kit for Deckle Joinery
  • #32 Deckle Joinery: the Mortise Side
  • #33 Deckle Joinery: Tenon End Part 1
  • #34 Deckle Joinery: Tenon End Part 2
  • #35 Deckle Joinery: Tenon End Part 3
  • #36 Deckle Joint Dimensions
  • #37 Fitting the Deckle to the Mould
  • #38 Deckle Rim Curvature
  • #39 Gluing the Deckle
  • #40 More Deckle Details
  • #41 Shaping the Deckle Part 1
  • #42 Final Shaping of the Deckle
  • #43 Adjusting a Deckle: (another advantage of the ‘pinwheel’ joint)
  • #44 Preparing a Deckle for Sheathing
  • #45 Sheathing a Deckle

Miscellaneous Paper Mould Posts

  • Further Thoughts on Mould Construction
  • Repairing a Broken Paper Mould

#46 Investigating an Amies Wove Mould

In which some interesting variations (and a couple of mysteries) are discovered.

This pair of wove moulds was given to Cathleen Baker by the late Larry Lou Foster of Tuscaloosa, Alabama who acquired them in 1968 while traveling in England. Cathy loaned me this pair and another laid pair to examine.

The moulds are small, 11″ x 12″, and of fairly standard construction but with a few puzzling features. The bottoms are fitted with boxwood rub strips and these curious, roughly shaped ‘minimal’ cast brass corners. These moulds, being so small, are naturally very stout. In this case at least, brass corners must have been added to take wear rather than to brace the structure; that there are only two screws each and that the corners are so small supports this idea.

The moulds have been altered and/or re-conditioned. They seem to have started out (as many small moulds do) plain, without rub strips and brass corners. This is indicated by the remains of a miter partly hidden beneath each of the eight brass corners. A miter (combined with the regular joint) is a traditional way to neatly finish the bottom corner of a plain, unbraced mould. Corner braces and rub strips were added later, possibly because the mould was wearing badly on its bottom edges. The narrow copper strips that protect the edges of the wire facing were originally folded down over the ends and nailed. The current edge strips stop just shy of the corners. You can see the two holes left when the nails were removed and also an area which was carved out of the wood to accept the thickness of the original strips. The joints seem to be nailed and countersunk; filled holes appear on both faces of each corner. Also interesting is this variation of the double dovetail joint in which the top ‘pin’ (this refers to the part that overlaps the ‘tails’ in dovetail joints) is offset by about 3/32″. My guess is that this was intended to add strength to the end of the frame. Gripping the mould to the deckle during formation might tend to pull the bottom edge out, thus torquing the top edge inward; in this case the inward thrust would be stopped by the narrow ‘ledge’ cut into the adjoining (long) side of the frame. (This is illustrated more clearly at the bottom of this post).

One of the moulds has a decided twist, though the mahogany sides are all very straight (as is the grain of the wood). Both moulds are slightly concave; the metal screen sags by about 1/16″ in the middle. This may be a result of the long term stresses of couching. The photo shows the back side of the mould where the pegs of the ribs are visible. These moulds have ‘through holes’ and the pegs formed on the rib ends pass all the way through the frame. The entire front of each mould is sheathed. The sheathing ends right at the two front corners and is not folded around the sides. You can see this in the previous photo.

Mystery #1: Why are all of the holes that the rib pegs rest in uniformly ‘lemon shaped’? And why are they oriented the same way? I can imagine that a file or a sort of broach must have been used to shape the holes. If anyone has any insights please share!

This is truly weird for me to see. The holes have little points at 10 o’clock and 4 o’clock. This uniform orientation seems to argue that a jig or mechanical device was used to make the holes.

The brace rods are hammered flat at the ends and set into oblong mortises. The ribs are made of wood with a very uniform grain; not resembling any coniferous wood that I have worked with. They may be made of a species called Parana Pine. I have a couple of samples given me by Serge Pirard and I fancy there is a resemblance. (I will at some point make a mould using this wood for ribs.) The wire structure that supports the wove facing is a variation that I haven’t seen before. There are two ‘grid’ wires between each pair of ribs; this is not unusual but what IS unusual is that only one of them is stitched in place. (See post #26) The other ‘floats’, giving support to the facing but not held in place by stitching. ‘Floating’ grid wires are pretty standard but usually they are flanked by a two sewn grid wires, one on each side.

In each space between ribs you can see the two grid wires; in the usual repeating pattern a sewing wire passes under three (backing) laid wires and passes up and over two wires of the mesh and back down, taking a spiral path and binding the layers together. But this is true only for the grid wires on the right in this photo; the ones on the left are not sewn at all.

I was gratified to see that the stitches are carefully placed to ‘lie low’ by crossing the mesh wires at the low points as they spiral around the backing wires. (Gratified because I sometimes wonder if I imagined this detail; it seems such a fine point and may or may not be a common feature). The sewing wire is unusually small at .006″ diameter and the mesh wire is relatively coarse at .011″ diameter. The mesh is fairly coarse at 36 wires per inch by 40 wires per inch. The 4 stitches visible in this close-up are indicated by blue arrows. The path of a floating grid wire is shown by the yellow arrows. Ideally this grid wire is held in a sort of ‘trough’ formed by a pair of zig-zagging mesh wires; here you can see that this wire wanders off to the right a bit.

When I first studied moulds in the early 1980s I noticed an inward slant on their sides but with time came to doubt my first impression. This feature would make it easier to place and remove the deckle. I started out making moulds this way but soon switched to making them with no slant at all. These moulds show an inward slant (actually more of a slight curve) that was probably shaped with a plane after the mould frame was assembled. This slant is on the short end of the mould, parallel to the ribs.

This shows the slant on the long side, at right angles to the ribs.

Mystery #2: Is it possible that grooves were actually carved out for the backing laid wires to rest in? It sure looks like that here, along with a transverse groove to hold the chain wire. Usually there is what I call a ‘ledge’ here but there’s no sign of that on this mould. It seems unlikely that the laid wires could have been pressed into the wood. When wet the wood would tend to return to its original shape, pushing the wires back up. The copper nails in old paper moulds are often remarkably loose; I was able to pull these out with my fingernail so I could peek under the facing. I would have loved to examine this more thoroughly but couldn’t ‘dig any deeper’ without damaging the mould.

These deckle joints were made in the fashion that I prefer with joints fitting together in a rotating ‘pinwheel’ pattern. This eliminates the necessity of cutting opposite forms of the complicated joint. Notice the single inset wear plate of brass at lower right.

The left front corner of the deckle has a brass plate inset into the vertical inner edge. It is bent at a right angle and nailed into a shallow mortise. There was also an “L” shaped flat brass plate that was set down into the groove behind the deckle rim. Now it is missing. Both front corners also have bent brass plates along the very narrow vertical face of the rim. These ‘wear plates’ are standard features on most British moulds I’ve seen.

The left front corner viewed from a different angle. The extra reinforcement (I believe) reduces wear on the part of the deckle that first hits the edge of the mould as the deckle is quickly placed on it (over and over and over…).

This is a pretty deckle. It is very ‘curvy’ and the relatively thin sheathing (.008″) is nicely fitted.

The front and back edges are shaped to a very narrow edge; less than 1/8″.

I have not carefully examined a lot of moulds and deckles. I thought I should get organized to make it easier to note features and variations. The survey that I devised ended up describing about 75 features that I am interested in recording for comparison. I first print blank versions with only the questions to complete by hand. Then I enter the information for a digital record. Pictured is a completed survey of this pair of moulds.

More about the offset in the corner joint

I’ve altered and annotated the above photo to show how the offset part of the joint adds an extra strengthening feature to the completed joint.

The red lines show what the edges would look like if the facing and edge strips were removed. The yellow arrow shows the offset and the black arrows point to the extra joining surface that is created on the right hand part of the mould frame. This little ledge would keep the mould frame (on the left) from being pushed in at the top. That’s the theory I’ve come up with to explain the function of the extra feature. There must be some reason to do it this way since it makes cutting the joints a little more difficult.

#45 Sheathing a Deckle

I start with paper patterns. First I determine the width of brass needed to wrap over the sides of the deckle. Then I lay out with pencil and rule an “L” shaped pattern for each side.

These are cut out and fixed in place with a few pieces of tape.

A sharp pencil is used to mark the corner cuts.

The corners look like this.

Each pattern is fixed to a piece of soft brass sheet. This brass is .017″ thick.

A sharp utility knife cuts through the paper, marking the brass in the process.

The result looks like this.

The sheet is carefully cut to shape. The edges are smoothed with an abrasive pad.

Getting this part out is a trick. If the two cuts travel a short distance along the same line the waste piece will come out cleanly. I think I try too hard to make the edges meet exactly when the brass is formed to the deckle. It might be better to cut both sides back just a little, showing a bit of wood along the edge.

A propane torch softens the brass. When annealed it droops of its own weight.

Vinegar, salt and an abrasive pad restore the brightness.

I try to pre-form the sheathing so it doesn’t work against the nails to spring back.

A few nails (#19 brass escutcheon pins here) tack the top edge in place.

More nails along the outside edges. I space them every 5/8″.

After nailing I burnish the edges down. I file off the domed nail heads a little and polish everything with an abrasive pad.

This is how it looks from the bottom.

At the front the brass traditionally wraps around each corner a bit. This probably indicated the front of the deckle as well as adding a little strength.

The wood of the back corners is usually carved out more abruptly because excess pulp was ‘tossed off’ the back. Running the brass sheathing straight over the back leaves room for these tight curves.

It looks nice but is it really necessary? Probably not most of the time. With very heavy use wood becomes fuzzy with rubbling against the vat person’s hands. And I have had one report of some amazing callouses developed from that rubbing. Perhaps the brass made it easier on the hands and also ‘washed off’ better, preventing stray fibers being rubbed into ‘knots’, unwelcome flaws in most paper.

#44 Preparing a Deckle for Sheathing

One of the moulds being made has some experimental features. It and its deckle are made of larch; an early effort of mine to start looking for alternatives to tropical hardwoods. I sheathed the mould in brass, mainly to demonstrate sheathing on the blog. The larch deckle will also be sheathed. This gives me an opportunity to try, for the first time, using traditional copper staples at the corners.

Here the deckle is being drilled for the staples. It is easiest to do this while the deckle still has square edges. One could easily do this free-hand. This is the same drill press set-up I used to drill for the acetal pins that were used in place of staples for the other deckles.

Pictured is another method of shaping the inward sloping bevels of a deckle using a rasp and file. This deckle is receiving small inward curves along the short sides. A rasp makes quick work roughing these areas. It will be followed by a file to remove the rough rasped surfaces prior to scraping and sanding. Traditionally flat areas were left where the deckle is gripped to the mould during sheet formation. You can see this in the middle of both short sides above. Perhaps the gently tapered areas were designed to encourage fibers to wash off the deckle during use.

Here a file is being used to smooth the rasped surfaces. Note the ‘stair step’ device at the back. It enables the deckle to be tilted at various angles so the file and rasp handles can ride on the bench surface to guide them in shaping the inner edges of the deckle.

One of the long sides set up for shaping with rasp and file.

A wire-width groove has been chiseled between the holes for the staple. If this step is skipped splinters of wood are likely to pop up as the staple is hammered down. This wire is .050″ diameter.

The soft copper wire readily conforms as it is tapped into place.

Small recesses are chiseled for the staple ends. I copied the orientation of the recesses from the old deckle I took apart.

The way the staple ends are clinched makes sense. They end up parallel to the contours of the deckle so they can lie flat, leaving a smooth surface to be covered with brass sheathing. I hesitate to leave these ends exposed, fearing that they would catch fibers (or flesh) in use if the deckle was left unsheathed. Perhaps it would be OK after all as they are fairly unobtrusive.

The stapled deckle atop a drawing I made of the old deckle as I dismantled it.

#43 Adjusting a Deckle

(Another advantage of the ‘pinwheel’ joint)

This mould and deckle were intended to be an A4 size. In converting from metric I made an error and ended up with a deckle that was too loose in one direction. (It fit fine the other way). Since the deckle was made with joints in a ‘pinwheel’ orientation it was possible to correct this. If the joints had been ‘opposing’ it could be adjusted only one way. (About half the time you’d be out of luck).

Adjusting the deckle requires re-working the joints of the tenon side; making the tenons longer by cutting back some wood at their bases. Often, only two of the joints (at opposite corners of the deckle, diagonally) will need to be altered. That was the case here since the deckle fit fine the other way.

The joints are glued in the usual fashion and the overhanging wood is cut away.

I’ve only had to adjust a deckle in this way twice. It is extra work and best avoided. The main advantage of pinwheel joints is saving labor when cutting them, but here is another small advantage.

#22-1/2 Fitting the Laid Facing to a Single Faced Mould

This fills a gap in this sequence of posts. These few photos should provide enough information since the differences between this and the other two types of mould are covered elsewhere in the blog.

The facing is taped in place just like the double faced laid but directly above the ribs without any backing wire below.

The top edge of the ribs have been made a little narrower than those of the double faced and wove moulds.

This mould started out as my first and only attempt at making a laid facing the original way; on the mould, one laid wire at a time. This is the explanation for the extra holes, scars and divots along the top edge. This attempt was not successful and the unfinished hand made laid facing was replaced with this one made on my loom.

The chain wire still needs to be moved back to fit into the ‘trap groove’. More tape will be added at the ends to cover the wire edges before the mould is sewn.

#42 Final Shaping of the Deckle

If the deckle belongs to a laid mould small grooves are filed into the rim where the chain wires stick up.

Placing the deckle on the mould and rubbing it around will cause the chain wires to leave visible dents in the wood. The deckle is turned 180 degrees and rubbed on the mould that way, too. (Or both ways on two moulds if it belongs to a ‘production style’ pair of two moulds and one deckle). If the chain wires were carefully aligned when making the mould they will match up both ways (or all four ways for a pair of moulds) and there won’t be a need to file extra grooves.

The dented marks are enlarged just a little at first with a file. Then the deckle is returned to the mould to check the alignment of grooves with chain wires. Some of the marks will be hard to see at first; deepening the ones that show first will allow the others to show up.

A ‘slim’ or ‘extra slim’ triangular file works well.

The notches should be deep enough that the rim rests on the laid wires to help make a neat edge on the paper.

Shaping the Chamfers

The ends of the chamfers can be carved with a chisel.

The middle parts can be shaped with a spokeshave.

It is traditional to include small chamfers on the short ends of the deckle too. The purpose of these is not clear to me. (This may be important in some way for production use in commercial mills). The short sides have been left flat on top (not chamfered) for the moulds I’ve been making recently. I usually narrow the inner edges of the front and back to about 1/4″. Sometimes the deckles are left full depth on all four sides (with no chamfers at all) if thick paper is to be made. Over the years I’ve offered to make the inner deckle rims to any dimension but have received little feedback. 1/4″ seems to work well for most contemporary sheet forming methods.

This single facet chamfer would probably work fine but I broaden the chamfer up over the top surface in a gentle curve.

A cabinet scraper can be used as part of this process.

Sandpaper finishes the job.

The outside curves are also sanded.

The entire deckle is sanded with the exception of the four inside (vertical) edges which retain their original ‘jointed’ surfaces. Along their bottom edges (where they meet the deckle rim) the corners are left fairly sharp and eased only slightly with very fine abrasive paper. (320 or 400 wet/dry over a small block works well).

When the shaping is done I soak the deckle thoroughly to raise the grain as shown here. After it dries I give it a once-over with a fine sanding sponge before applying any finish. As with the wooden parts of the mould the finish can be boiled linseed oil, tung oil or Watco oil. The finish won’t last very long (and that’s OK). The deckle and mould are ready to use!

#36 Deckle Joint Dimensions

Following are some dimensions for deckle joints. I make two ‘weights’ of deckles; ‘light’ and ‘standard’ (which correspond to ‘light’ and ‘standard’ moulds). I have long felt that small moulds tend to be overbuilt and have recently ‘slimmed down’ the wooden parts of those smaller than about 16″ x 20″.

These joints are based on the old deckle joint that I examined in post #32. Do not rely on these figures as cutting dimensions; use them only as a guide to understanding the proportions when designing your own version.

An Important Consideration

It is important that the channel behind the deckle rim is not so deep as to interfere with the joint. The photo above shows a narrow margin of ‘extra wood’ between the ‘bottom’ of the channel and the overhanging part of the joint. The black lines have been added to describe the shape of the deckle part prior to a joint being cut there. This shows how the shape of the part relates to the joint cut on its end.

There would be a problem if this same deckle part had a channel cut this deep. A gap would be created where the arrows point. It is good to keep this in mind when designing the joint.

Dimensions of Deckle Stock

Both light and standard weights of deckle stock are described in inches and millimeters.

Light Deckle/inchLight Deckle/mmRegular Deckle/inchRegular Deckle/mm
F = 3/4″F = 19mmF = 7/8″F = 22.2mm
G = 7/16″G = 11.1mmG = 1/2″G = 12.7mm
1/2F = 3/8″ *1/2F = 9.5mm *1/2F = 7/16″ *1/2F = 11.1mm *

*This measurement is approximate and is left oversize until after the joints are cut in order to facilitate creating the special deckle rim curvature. (See post #38)

Light Deckle/inchLight Deckle/mmRegular Deckle/inchRegular Deckle/mm
1 = 1-1/8″1 = 28.6mm1 = 1-1/4″1 = 31.8mm
2 = 7/16″ *2 = 11mm *2 = 1/2″ *2 = 12.7mm *
6 = 3/16″6 = 4.8mm6 = 1/4″6 = 6.4mm

*The dimension given is the final dimension; it is used when planning the mould and deckle fit (posts #5 and #30). This is left oversize until after the joints are finished (see post #37).

Sample Dimensions of Deckle Joint Parts

The charts give dimensions for two sizes of deckles in inches and millimeters.

Use this table of vertical dimensions for the photos immediately above and below.

Light Deckle/inchLight Deckle/mmRegular Deckle/inchRegular Deckle/mm
A = 9/32″A = 7.1mmA = 5/16″A = 8mm
B = 3/16″B = 4.8mmB = 1/4″B = 6.4mm
C = 9/32″C = 7.1mmC = 5/16″C = 8mm
D = 5/32″D = 4mmD = 3/16″D = 4.8mm
E = 1/8″*E = 3.2mm*E = 1/8″*E = 3.2mm*
F = 3/4″F = 19mmF = 7/8″F = 22.2mm

*This dimension is the same in both deckles. It is impractical to make the slot any narrower than this.

Use this table of horizontal measurements for the photos immediately above and below.

Light Deckle/inchLight Deckle/mmRegular Deckle/inchRegular Deckle/mm
1 = 1-1/8″1 = 28.6mm1 = 1-1/4″1 = 31.8mm
2 = 7/16″2 = 11.1mm2 = 1/2″2 = 12.7mm
3 = 1″3 = 25.4mm3 = 1-1/8″3 = 28.6mm
4 = 11/16″4 = 17.5mm4 = 3/4″4 = 19mm
5 = 7/16″5 = 11mm5 = 1/2″5 = 12.7mm
6 = 3/16″6 = 4.8mm6 = 1/4″6 = 6.4mm

Some Dimensions for the Mortise End

I make this end of the joint first. I ‘shoot’ for the dimensions shown and then ‘make’ the other side fit. Slight dimensional errors made when cutting this side of the joint can be corrected with corresponding adjustments to the other side. Not all dimensions are given here; some must be deduced from the previous photos).

Light Deckle/inchLight Deckle/mmStandard Deckle/inchStandard Deckle/mm
1 = 1-1/8″1 = 28.6mm1 = 1-1/4″1 = 31.8mm
2 = 7/16″2 = 11.1mm2 = 1/2″2 = 12.7mm
3 = 1″3 = 25.4mm3 = 1-1/8″3 = 28.6mm
4 = 11/16″4 = 17.5mm4 = 3/4″4 = 19mm
5 = 7/16″ 5 = 11.1mm5 = 1/2″5 = 12.7mm
C = 9/32″C = 7.1mmC = 5/16″C = 8mm

When making his joint the part labeled with the red “X” would not yet be trimmed away.

#41 Shaping the Deckle Part 1

Rounding the Outside Edges

First the corners are sawed off at 45 degree angles.

This can be done on the table saw using this overhanging fence.

This roughing stage is finished.

The lower deckle has been partially rounded over with a router. One more fine cut will take off the last rough saw marks.

This is how the edges are rounded.

After the outside edges are rounded the front and back need to be chamfered to narrow the inner edge.

Roughing the Chamfer on the Inside Edges

Most deckles have chamfered areas along the front and back (long) sides. The inside edges of the deckle are narrowed to roughly correspond to the desired thickness of pulp to be collected on the mould. One way to rough out the wood is with this panel cutting bit. A spokeshave can be used later to smooth the angular surfaces into a gentle curve.

Rounding the Deckle Rim

The inner edges of the deckle rim need to be rounded off so they don’t catch on the mould’s copper edge strip as the deckle is taken off the mould.

I do this by hand using this burred scraper. It’s the blade of the scraper plane that was used earlier to level the ribs.

First an even bevel is scraped the full length of each side.

The bevel is refined into a curve with the scraper then smoothed with sandpaper.

These last two photos may show the process more clearly.

#40 More Deckle Details

The clamps can be removed after a few hours, leaving hardened globs of epoxy here and there. But the epoxy stays sticky for a while so is best to let it cure overnight before removing the excess.

This can be cleaned up with a coarse file or with a chisel.

A chisel works best on the inside surfaces. This one has an offset handle to allow paring in hard to reach areas.

A little part of the deckle rim needs to be cut away at each corner. Some of the waste has been routed away at this corner and another pass will remove the rest.

Since the rim of the deckle is slanted inward a little triangular area (A) will stick up just a little where it laps the adjoining rim. A sharp chisel laid flat on the adjoining deckle rim slices this off easily.

“Stapling” the deckle Joint

Above, a pair of 1/8″ diameter holes are drilled through the deckle at each corner.

Traditionally the deckle joints would be stapled with copper wire. The clinched ends of the staple would be hidden beneath brass sheathing that covers the deckle. This deckle is not to be sheathed (and the staples’ sharp ends wouldn’t be covered) so I substitute a pair of acetal pegs for each staple. These will be machined flush with the surface of the deckle.

Lengths of the acetal rod are inserted and trimmed to length. The holes may need to be drilled larger to make the pegs fit. (The rods are supplied a little bigger than their nominal size).

I turn a shallow groove on the pegs to lock them in place when they are epoxied in the holes.

The pegs are started into their holes.

When they are tapped in place glue squirts out the other side.

The pegs and cured epoxy will be filed flush before the deckle is shaped.