Shipshape 7

How to build an Anglo-Saxon ship

Paul Constantine.

Planks or Strakes

Plank Shapes

In Shipshape 3 the importance of Plank Runs was described.

In a nutshell, the ship has 9 planks on either side, but what shape are they? Each one is different from the others and there are variations in shape that occur in many places along the length of each plank. They are narrow at the ends and wider in the middle. They are curved and they are bevelled. To determine the shapes in detail is a huge task which had not been seriously attempted up to the time of the build. The lead people tasked with providing some answers to plank-run questions, were John Cannon and Clive Cartmel. Their one-fifth scale model (see right) was their primary working tool, but John also made other models (see Shipshape 3)

       It is usual to work from 1:10 scale models, but John and Clive worked from 1:5 as the thin plywood that they were obliged to use to represent the planks had a scale stiffness more closely related to the real planks that would be used in the ship. This was particularly relevant towards the ends of the ship where the planks have to be twisted and secured. The 1:5 scale planks gave a more accurate prediction of the force that might have to be applied to the solid wood planks, to make them conform to the shapes required

         It has been explained that the Anglo-Saxon builders seemed to have used narrow garboard planks next to the keel which is not the usual practice found today and which had been recommended on some ship drawings that copied modern systems. John, through intense study of the original records returned to the planking arrangements that the builder’s used and found that they offered more structural support to the keel/cutwater transition whilst providing sufficient plank width for attachment to keel, underlout and cutwater. This is possibly one of the most significant discoveries stemming from the construction of the ship. To confirm their findings, John and Clive planked the 1:5 scale model differently on either side of their hull to check their findings. The picture left shows the garboard strake to the left of the keel as being narrower than the garboard to the right. The difference may seem small, but any difference will be five times greater in the ship. The significance of the difference relates to the support of the joints in the keel structure closer to the ends of the ship. It indicated that the ‘narrow planking’ close to the keel, as used by the Anglo-Saxons was probably the superior supporting system. This is the system that has been used in the ship’s construction.

Making Planks

The way that planks are made has been mentioned at various stages in this series of articles, but it is worth adding to that information here.

        The tree is riven using wedges (see left). This is not a violent process, but one which involves careful placing of the wedges and concentrating on the way that a split may be progressing to ensure that it doesn’t need re-routing. Sometimes the axe is laid on the timber and tapped in with a maul to direct the splits.

There is no guarantee that the interior of what appears to be a perfect section of tree on the outside will also be perfect on the inside. Much time and effort can be expended only to find that a section of wood has twisted grain, internal knots not visible on the outside, or a serious fault around the pith, which is the very centre of the tree. Such a fault was found in the stern underlout (see right), which meant losing a huge amount of time having to replace it.

A riven section will be quite thick initially. To bring it closer to the required size the plank thickness can be marked on its edge and a series of ‘V’ shaped cuts (see below) made to indicate how much wood now needs to be removed. The series of hills between cuts can be driven off with the axe to approximate the required board size.

 A roughly hewn board can be about 3ins thick on its outer edge and more axe work will be needed to bring the plank to an even thickness. A thickness gauge like the one cetrally placed on this plank, left, can be used to check for the uniformity of the plank as it is worked down to size.

Grain Pattern

Every board must have the correct grain pattern, which can be seen by looking at the end of the board. The annual rings should be as short as possible and of uniform length. This pattern is automatically produced by riving the planks instead of cutting through the tree using saws, as we might do today. Most people would not understand this now, but the Anglo-Saxons knew all about it. Riving  makes the boards stronger and reduces the chance of the boards splitting to allow the ship to leak, as described here.

Left.                                                                      A view of the end of the tree.

The annual rings go around the trunk – BUT – there are other cells for food storage that run from the centre outwards. They are called the Medullary Rays. All trees have them, but they are not easy to see. They are best seen in Oak, Beech and in some cases Sycamore. They are very numerous, but the diagrams below only show a few, where the wood is starting to split. As the sap dries from the tree and the wood shrinks, the rays cause a weakness that splits the wood. The splits cross the annual rings.

Centre.

The tree has been cut into planks. The top half has been sawn through, as is done today. The bottom half has been riven with axes and wedges along the weak lines of the rays. The two systems produce different grain patterns. The important thing to note is the direction of the medullary rays in the planks. A plank produced by cutting through with a saw has the rays and therefore the splits, passing through it. A riven plank has the rays running along inside it and resists splitting through from one side to the other.

Right.

Plank Stability. The wood shrinks as it dries. Most shrinkage is along the annual ring. The longer the annual ring, the more it shrinks. If the rings are different lengths, they distort the wood. In the sawn timber the plank becomes warped. In the riven timber most rings are of similar length and quite short, so shrinkage is more even and the board stays flat.

Surface appearance

The medullary rays are not absolutely straight and flat; they are a bit like wavy sheets of paper running between the vertical cells of the tree. As they pass along inside the riven wood they come up to the surface at an angle. They create blotchy marks on the outside, that are very strongly visible in oak. These marks are called ‘figure’. On the sawn planks the rays are mainly piercing straight through, so the marks are small and almost insignificant. Because of the figure, it is possible for the experienced eye to see at a glance whether any wood has the correct, stable, grain pattern or not. Does it have figure on the sides? If so, it’s good. This has long been understood by craftsmen, and so, traditional oak constructions, such as old table tops or church doors, will be constructed with figured oak. Over time, figure has come to be appreciated as a mark of beauty in oak structures. This picture shows timber samples being tested with surface treatment in the Longshed and the figure is particularly strong on the right-hand sample.

Attaching the garboards

The garboard shapes were determined by making plywood patterns based on the 1:5 model and from the ship itself. They were translated to solid timber and held in place. This is not an easy process at the ends of the ship. The boards are narrow and have to twist quite sharply compared with further amidships. It is sometimes necessary to secure props, above left, to hold the wood in shape until it can be secured with fastenings. In order to cramp one board to another normal ‘G’ cramps have only limited reach across the board. Traditional hand-vices with a much longer reach, above right, made in the workshop, have to be used.

Scarf positions

It is worth recalling that the black frames in the pictures are not be part of the structure of the ship. They are there primarily to ensure that the ship accurately copies the shape of the original craft. Due to this, they had to be positioned so that they would not interfere with the actual ribs of the ship which would be added later. The scarf joint, which attaches boards end-to-end, has to avoid the location of the rib. This scarf is 1:6 and was cut by axes on the two planks separately. It is usual to make scarfs by working both joining ends of the boards involved together. The topic of scarfs is explained in more detail earlier in Ship > Investigation 8 > Planks and Scarfs on this website. In that account the question was raised as to which style of scarf - tapered or half-lap, would be easiest to make using an axe rather than a plane, which is the usual tool used for this process. Practical experiments on the ship's planking showed that the tapered scarf can be made with the axe.