Category (en-gb)
Planks The next log was picked up in Dyvig. It was split outside the shipyard and the two halves were dragged into the hall. It turned out that the 20 cm diameter heartwood did not fit with the other wood. This meant that the side planks had to be laid further down the trunk, which reduced the width. We could not achieve the width indicated in Johannessen's drawings. It was decided to cut the side planks to the width they could get and possibly increase the width of the stringer planks if necessary.It turned out that the root end could provide an acceptable width, while the planks would become increasingly narrow towards the top end. It was decided to leave the wide end of the side planks facing forwards. The trunks were orientated with the split side down, the bark was peeled off the trunk halves and a series of transverse notches, 10-15 cm deep, were cut into the trunk at a distance of approximately 1 m. The wood between the notches was then removed by splitting. The two side planks being worked on. Photo: H.P. Rasmussen. The outside of the side planks were cut to final shape. According to the construction team's templates, the side planks were to be cut slightly curved in a transverse direction over the middle 6 metres.After the outer sides of the side planks were smoothed, they were turned over and the inner side was processed. To determine and mark the exact position of the cleats, the planks were temporarily placed on the boat.The bottom edge of the side planks was chosen to have a wall thickness of 2 cm, while their top edge was given a wall thickness of 1.5 cm.It turned out that the top end of both planks had some cracks that were caused by the felling. The tree had fallen so badly during felling that the top splintered. The finished side planks weighed 56 kg each. The cross curvature of a side plank is being checked. Photo: H.P. Rasmussen. Mounting side planks Before assembly, the bow of the keel line had to be decided, a decision that would determine the shape of the entire boat. At the same time, we had to ensure that the boat was symmetrical about the longitudinal centre plane. A string was suspended over the line that would be the longitudinal centre of the boat. A series of plumb lines (3 pieces) were attached to this.The bottom plank, which was very flexible, was blocked up and clamped to the ground so that the keel arch had an arrow height of 37 cm, calculated over the distance between the two inner locking boards.Using our computer data, the transverse profiles were drawn at a scale of 1:1, this time not for each frame, but at the centre point between the individual frames. These drawings were glued to 10 mm chipboard and profiles were sawn out. These were then mounted on the bottom plank. This allowed us to shape and mount the side planks and the railing planks so that the boat had the right shape. Auxiliary profiles were used to ensure the right shape. The green line at the top of the image is the reference centre line for hanging the three plumb lines. Photo: H.P. Rasmussen. This method was obviously not used by our predecessors 2,350 years ago. They probably used a series of rods that were lashed together to create the internal shape. However, we did not feel experienced enough to use such arrangements. We became more and more convinced that the ancient shipbuilders who had built the original Hjortspring boat must have built it in a series of increasingly refined boats. Every detail of the boat suggests this. The boat was probably also built according to the size of the available logs.In order to achieve a boat shape that best represented the find, we had to deviate from the ancient method in this respect as well. The plank edges are adjusted during sewing. Photo: H.P. Rasmussen. The side planks were sewn to the bottom plank by first attaching them both amidships and then sewing them both to the bottom plank at the same time, moving forwards and backwards, while the edges were trimmed and the woollen cords were continuously laid in the joint. The sewing holes were sealed by pressing ox tallow into them. Reparation As mentioned, the two side planks had some cracks in the aft section. After the installation of the side planks, these cracks were repaired using the same method as in the original boat. These repairs have been interpreted as the boat being old and worn, but they could just as easily be seen as repairs to cracks originally contained in the trunk.The cracks were stopped with woollen string, saturated with ox hide, and then 6 cm wide strips of ash wood were sewn over the cracks. Cracks in the side planking were repaired with sewn-on strips as in the original boat. Photo: H.P. Rasmussen. After installation, the front and rear rows of cleats were cut on site to ensure they were horizontal.A new milestone had been passed. Sources Hvad Haanden former er Aandens Spor. Language The text in this article has been translated from Danish to English using the free DeepL translation programme.
- Details
Keel stretching party We had decided to organise a keel stretching party once both bows had been mounted on the bottom plank. Locking pin for securing the stem and bottom plank. Photo: H.P. Rasmussen. In addition to sewing the lower edge of the bow to the bottom plank, the bow was also attached to the bottom plank with a square pin that ran across the bow and the edge of the bottom plank that guided the bow. Finally, there was a vertical locking plank that started in an elongated recess at the beginning of the bottom plank to the lower horn and ran up through a slot in the bow extension towards the upper horn.This locking plank was secured with transverse square pins that ran through the base of the two horns and through the locking plank. This was also fitted with a boss at the top that stepped down towards the top of the upper horn. Woolen cords saturated with ox tallow were placed under this boss as a gasket to prevent water from penetrating the horn.Both the locking planks and the pins were made of oak. Assembly of stem and keel plank. Photo: H.P. Rasmussen. Once the sewing was complete and the locking planks had been fitted, the keel stretching party was organised with speeches and mead drinking. It took place on 23 March 1996. It was an open house and was attended by both members and local residents. One of the speeches claimed that we were now halfway to the completion of the boat. We needed to get smarter. The keel stretching party with keel plank and bows. Photo: H.P. Rasmussen. Sources Hvad Haanden former er Aandens Spor Language The text in this article has been translated from Danish to English using the free DeepL translation programme.
- Details
Bast, cording and sewing While the woodworking exercises were taking place, one group was studying the making of cords for sewing or lashing together the boat's parts. As previously mentioned, there was some uncertainty as to what material the sewing cords were made of. No traces of cord were found in the stitching, only impressions in the adhered sealing material. However, there was a bundle of cords at one end of the boat that had been interpreted as being made from lime bast. This material was already known to have been used for cordage production in the Stone Age.However, some wooden parts were found in a couple of sewing holes that could be interpreted as cords made from birch roots, for example. However, they could also have come from wooden sticks that were driven into the sewing holes after sewing to ensure that the cords remained tight, as seen on Maori boats in New Zealand.We had our doubts. We finally decided in favour of limewood, thinking that there should have been boats sewn together with cords of this material by now.One member had a son who was a forester on Zealand. He had a plantation of young lime trees that needed thinning. The member travelled to Zealand one day in June, when the trees are easiest to debark due to the sap rising. The trees, which had a diameter of 10-12 cm, were felled and cut into lengths of 1.5 metres. The bark was scratched with longitudinal scratches at a distance of 15 mm and peeled off. The strips of bark were then soaked in water for a month in vats. This allowed the glue (pectin) between the individual bark layers to rot away and the thin strips of bark could be peeled off. These were washed and hung to dry and finally bundled into 10 cm diameter bundles. The bast harvest totalled 13 kg. The bast strips were twisted into cords using a spinning reel that constantly rotated the finished cord piece.As the twisting progressed, new bast strips were added. The number of bast strips simultaneously present somewhere on the cord determines how strong the cord will be.The finished cords were then twisted together to form cord. This was done using a so-called shearing machine, a machine where three hooks rotated at the same speed and direction. The twisting of the cords into string was controlled by a top.The method used is no different from ordinary rope weaving. A method of spinning cords and twisting rope using only hands and fingers was tried. This method, which had probably been used in ancient times, gave the same good results, but it was significantly slower. The finished bast strips ready for cord making. Photo: H.P. Rasmussen. Twisting corded cord. Photo: H.P. Rasmussen. The find indicates that two-strand cord was used to sew the boards together. For the lashings of the ribs or wraps for the boards' cleats, we chose to use three-strand cord, as such cord was found in a bundle in the find. A total of 600 metres of cord was used to sew the boards and for the lashings.By moistening both the bast strips and the cords before tying, we achieved a much more beautiful and uniform result. A series of measurements were taken with two-stranded cord with a density of 6.8 g/m. After normalisation, where the cord was stretched, we found the following values: Shrinkage: from dry to wet: 1%. Coefficient of elasticity: 0.005 %/N. Breaking strength: 250-300 N (25-30 kg).We found that the breaking strength was on the lower side, which is why we increased the number of bast strips in the cords from 2-3 to 3-4. This increased the density of the two-stranded cords to 11 g/m, corresponding to a breaking strength of 40-50 kg, and it was with this cord that the boat was sewn together.Tests were also conducted on the strength of the sewing of the boards. Before we report on these tests, we need to look at the seal between the boards. According to Rosenberg, the find indicates that the seams on both the outside and inside were covered with an organic mass, which Rosenberg interpreted to be resin. Later research into the parts of the find revealed a lump of organic material that was analysed and found to contain animal fat and linseed oil.After discussions with the National Museum in Roskilde, we decided that the seal between the boards should consist of rolls of carded sheep wool dipped in a mixture of beef tallow and linseed oil. Some tests were carried out with different mixing ratios, and we found that an 80/20 mixture of ox tallow/linseed oil gave the best results in terms of sealing. The density of the saturated sheep's wool roll was 20 g/m.Two boards were now manufactured to represent tables. Over a length of 60 cm, these boards were sewn together with the sealing material in the gap between the boards. The final cord is wound (3-stranded). Photo: H.P. Rasmussen. Sewing together test pieces. The use of the ‘S-shaped’ tool is a hypothesis. It works, but is a bit cumbersome. Photo: Ib Stolberg-Rohr Notice the curved wooden tool used to tighten the stitching. It is a copy of a part from the find and is thus interpreted as a sewing tensioner. The stitching of the test piece was subjected to shear forces and we measured coherent values of shear force and mutual movement. The figure shows this correlation. Up to a force of 0.6 N/mm seam length, the pieces did not move relative to each other.0.6 N/mm corresponds to 36 kg force over the entire length or 60 kg force per metre of seam.At higher forces, we got a shear, but it was slow and hesitant, suggesting that the seal acts as a very viscous fluid. This displacement was permanent, i.e. it did not disappear when the force was removed.A stitch was sewn every 7.5 cm. Consistent with the finding, the stitch was double with a course that caused the knot to be self-locking, so that cutting a string would not cause the entire seam to unravel. Forskydningskraft og forskydning. Tegning: K.V. Valbjørn Self-locking double stitch. Photo: H.P. Rasmussen. Sources Hvad Haanden former er Aandens Spor. Language The text in this article has been translated from Danish to English using the free DeepL translation programme.
- Details
Wood A possible stem piece. Photo: H.P. Rasmussen. Adaptation To carve the stems, you needed a short piece of whole log with a length of about 2 metres and a diameter of up to 1 metre. The Polish logs were of insufficient diameter, so we had to get local wood for these stems. One member donated a linden tree that was close to his house. This tree had a sufficient diameter and no large branches in the first 5 metres. This tree was a large-leaved lime (Tilia platyphyllos), while the rest of the boat was built from small-leaved lime (Tilia cordata). Rough adaptation of a possible stem piece. Photo: H.P. Rasmussen. An undecided design of the upper horns' connection to the stems caused a lot of discussion. There were three possibilities. The first possibility was that the upper horn was part of the stem and bent in heat after carving. The second possibility was that the upper horn had been contained in a branch from the trunk. And finally, the third possibility was that the horn was a separate part that had been mounted on the stem, as was the case with the lower horn.The possibility of a branch was first investigated, but neither of the first two bows we roughly carved could be used, as rot appeared in the gap between branch and trunk and in the core wood of the trunk.We decided to try to form the curved horn by steaming the horn that was carved as part of the log. A steam generator in the form of an oil-fired boiler was constructed by removing the overboiling fuse and we let the steam from the boiler flow into a hose from a tractor tyre that enveloped the horn. After a few hours in the ‘smoke and steam’, we began to curve the horn according to a doctrine that included the curved shape. It went well for a while, but then the horn broke at a crack. A dramatic attempt had failed. At this point we learnt that there had been heartwood in the middle of the bow as well as in the upper horn. This meant that either the upper horn was formed by a branch extending from the trunk containing the bow, or the upper horn was a loose part attached to the bow. We chose the latter solution.The actual carving of the bows was a lot of work, not least because of the above-mentioned experiments and because the bows had to be carved quite far before the quality of the wood could be ascertained.Two logs had to be discarded along the way. One finished log weighed 20 kg. An almost finished stem piece. Photo: H.P. Rasmussen. Sources Hvad Haanden former er Aandens Spor. Language The text in this article has been translated from Danish to English using the free DeepL translation programme.
- Details
As mentioned earlier, the four large lime logs arrived at the end of January 1994. The three logs were transported to Dyvig and lowered there using pieces of iron to keep them fresh, while the largest of the logs, which seemed best suited for the production of the bottom plank, was towed into the hall at the Linde shipyard. Splitting The trunk had a length of 18 metres, a diameter at the root end of 90 cm and an age of 170 years. It was orientated so that the largest crack was vertical and the trunk was supported with wooden wedges.The splitting was done by driving steel wedges into this crack from the root end of the trunk. As a crack appeared on the upper side of the trunk where the bark had been removed, this crack was opened by driving wedges vertically into the crack. Initially, steel wedges were used, but as the crack grew larger, beech wedges were used, which did not hurt the soft linden wood as badly as the steel wedges. Linden wood does not split as readily as oak or beech, and there were always ‘runners’ from one half of the trunk to the other. These were cut with hatchets.The splitting took about three hours. The first stem is almost completely split. Photo: H.P. Rasmussen. Shaping We now had to decide how to orientate the bottom board in one half of the trunk.After roughly levelling the split side and removing the bark and sapwood from the entire trunk, we realised that the maximum width of the bottom board over the middle 6 m could not be contained within the trunk. It was decided to cut a hole in the log and then stretch or open it to a wider and flatter shape using boiling water. While the chopping was going on, two 1 metre long, half-round vessels were made from lime wood, open at the ends and with a wall thickness of 3 cm. They were to be used for stretching experiments, as the plan was to boil them in water before stretching. According to the rule of thumb, a plank that needs to bend strongly should be boiled for one hour per inch of thickness. After boiling sample 1 for two hours, it stretched and cracked lengthwise in several places. Test piece 2, which was boiled for a further hour, also cracked. We now realised that heating with boiling water was not a viable method. After an unsuccessful attempt to open the hollow trunk, we had to cut it flat and glue planks to increase the width.Drawing: K.V. Valbjørn However, the traditional method in ancient times was to heat boat shells to be stretched by heating the wood over a fire. It is claimed that the wood needs to reach a temperature of 140 degrees to become suitably plastic. To achieve this temperature, the boat shell was filled with tar, which was brought to a boil by heating it in the fire. We had to abandon this method, partly because we had no experience and partly because we realised that the already fairly hewn keel line would result in an excessively curved keel line when the boat shell was opened amidships, causing the ends to lift. We realised that our predecessors had not used this method for the same reason. They must have had a large enough trunk so that they could cut the bottom plank into shape without stretching.As we had already started cutting the plank, we had to increase the plank width by gluing. We found a glue commonly used in wooden shipbuilding (an epoxy from System West) and a carpenter experienced in using this glue was contacted.Some experiments with gluing planks together and subsequent loading resulted in no difference in strength or elasticity, whether the planks were whole or glued together. The hollowed-out plank was chopped almost flat over the middle 6 metres, creating 10 cm wide glue surfaces. As the glue could hardly withstand more than 0.7 mm thick glue joints, these surfaces had to be planed very flat. 13 boards, 8 cm thick, were carved from the other half of the log. The boards were moulded with flat gluing surfaces. The actual gluing was carried out by ship carpenter Arne Wahl, Fåborg. As the boards were placed, the joints were clamped together with grease-lubricated screws that could be removed after the glue had hardened. The gluing process is described in detail in the member folder in section 4.5.After curing and removing the screws, the plank with the glued boards was ready for further processing. Smaller planks are adapted before gluing. Photo: H.P. Rasmussen. The processing The plank was turned over so that the outside could be worked down to its final shape. After the final operation that left the surface smooth (without using sandpaper), the plank was turned over and the inside was carved and cut out. A particular challenge was the carving of the cleats and the through hole through each of them. This hole was found to be square with an edge length of 10 mm.Some U-shaped chisels were made, the handles of which were shaped so that they could pass over the neighbouring cleat. The square holes in the cleats were carved with a U-shaped chisel. Photo: H.P. Rasmussen. The edge of the bottom plank was left raw so that it could be adapted to the side plank during assembly. The ends of the bottom plank contained the tongue-and-groove joint between the bottom plank and the lower horn, also a wood carving job of dimensions.After completing the bottom plank, we realised that the wood present largely determines the workflow.The weight of the finished bottom plank was 96 kg. The bottom plank is finished. Photo: H.P. Rasmussen. Sources Hvad Haanden Former Language The text in this article has been translated from Danish to English using the free DeepL translation programme.
- Details