For the purpose of this research a crop intended for fibre production has been analyzed and photographed. Such varieties have long stalks, branch very little and yield only small quantities of seed. However, the stalk’s section varies greatly between the different types of produce crop. From selective breeding it is possible to develop desired traits within the plant. For the production of stalks for construction, the breeding program could produce a tall crop with a low leaf stock while the stalk itself would have a wide diameter with relatively thick shiv layer.
The structure of the stalk can carry a great weight of leaves ensuring the plant can grow high. It has a tubular structure which proves very efficient at taking vertical loads. Much like bamboo, the stalk’s hollow core reduces waste growth while maximizing its diameter which is more effective at taking both horizontal and vertical loads.
When densely planted the plants are enclosed by their neighbours, encouraging vertical growth. This produces a more uniform crop with straight stalks, minimizing the likelihood of failure occurring at weak points such as bends in its length.
By taking regular sections through a stalk down its length it is possible to see how its structure varies. The image to the left shows that, from base to top, the stalk’s diameter (blue line) decreases. However, the diameter of the hollow core (red line) increases dramatically over the first 10 centimeters then stays relatively constant for the next 120cm. The distance between the red and blue line shows how the shiv and fibre that thin out along the stalk’s length. After 125 cm both the inner core and stalk diameter reduce at the same rate, retaining the same thickness of shiv and fibre until the end of the sample.
The green line shows the surface area of the sections through the hemp stalk. Even with the varied cross sections, the graph shows the surface area decreasing at a relatively constant rate from bottom to top. The dashed yellow lines show the points at which the stalk is segmented and from where the secondary branches grow. As the data shows, the hollow core reduces drastically at these points with the overall diameter also thinning.
The shiv would take the majority of the load under compression and likewise the fibre in tension while both would act when in bending. It therefore follows that, due to the reduced loads experienced towards the top of the plant while it is growing, the stalk’s structural capacity also reduces towards its top.
The capacity of hemp in compression has not been exploited in the same way it has in tension, reflected by the sheer amount of products made from the fibre and relatively few from the shiv. This would suggest that the material’s most efficient use is in tension yet does not determine that it can never be used in compression. It is in the structure of the stalk that its abilities in compression may lye.
The samples performed much as expected, with the larger segments taking greater loads. Sample A reached a peak loading capacity of 2.33kN at 1.9mm of compression. At this point the sample failed, taking less and less load as it is compressed further. After being compressed over 11mm the test was stopped and reversed while still measuring load. Samples B and C performed in similar ways, however only reaching peak loading capacities of 1.7kN and 0.9kN respectively.
A compression load through a column will be taken differently to one through a smaller section of the material. A hemp stalk that is not supported from all directions along its whole length would experience loading as a column element. A column test allows the possibility of a bending or buckling failure to occur.
The compression results showed a very promising load bearing capacity for a natural stick material, especially in comparison to its sectional surface area. The peak loading capacity of the column samples were only slightly less than the loads taken by the equivalent section of 25mm compression samples. This would suggest that the column samples nearly reach their maximum potential loading before failing. All of the samples showed similar stiffness up until failure(24).
The same correlation between the 25mm compression and column compression results can be found in the sample’s stress calculations. The results of the column testing show an average maximum loading capacity of just over 14N/mm2
The hemp stalk has great potential in tension. Its external fibres are long and run down the length of the stalk. Through processing, the fibres are usually extracted, refined and re-spun into materials that exploit their potential, such as ropes and textiles. Reaching tensile strengths after processing of 514N/mm2, these same fibres must have a certain degree of strength while in their most natural state; in the hemp stalk. This could prove valuable for the stalk’s use within construction as an inherent tensile strength would be necessary to withstand bending forces in a structure or it could also perform as tensile bracing.
Although the shiv core does contribute, it is the outer layer of fibre where most of the tensile strength is found. Sectional analysis of the hemp stalk shows that the thickness of the fibre layer does not vary a great deal along the stalks length, however, the diameter of the layer reduces drastically. It would follow that the smaller the diameter the lower amount of fibre is present and therefore the lower the tensile strengths that could be achieved.
The results show that hemp stalk can take a much larger load in tension compared to compression with even the weakest samples recording higher loads than that of the strongest in compression. Surprisingly, the samples from the central section of stalk took higher loads than those from the base. This may be due to the greater proportion of shiv in the base samples which reduces the effectiveness of the end joints, specifically where the stalk is clamped between hose clips and coach bolt. It is therefore likely that the base segments would take greater load than the others however further testing would need to take place to confirm this.
To calculate strain, an extensometer was used to measure the extension of the hemp stalk when under tensile loading. However the when plotting stress against strain it was found that the smaller sections carried greater load per mm2. This is due to the larger ratio of fibre to shiv, with the smaller samples having much less shiv than the larger ones. Samples with a higher proportion of fibre are much more efficient at taking load.
The modulus of elasticity was calculated from the gradient of the first portion of the Stress-Strain graph. The modulus of elasticity is solely a property of the material unlike stiffness which is dependent on the element’s structure as well as the material. It shows a variation through the stalk, from 8360N/mm2 in the base sample to 9054N/mm2 in the top sample.
If the stalk is to perform in structural applications it will need to withstand loading which places it in bending. This could be anything from a wind load on a vertical element to a dead load on a spanning element. What follows is an account of three-point bending tests on samples of hemp stalk.
Samples taken from lower on the stalk took higher loads, with samples TA1 and TA3 reaching 140N. All the samples deflected over a millimeter before failing with some deflecting over 2mm. Sample TA1 was recorded while being unloaded. Even after failure it still moved back towards its original shape, but did not straighten out completely and could not take load to the same capacity.
The calculated flexural strengths of all the samples, which take into account each ones different sectional structure, were all fairly similar. A maximum flexural strength of 18.4N/mm2 and an average of 16.7N/mm2 was recorded.
Hemp can be compared closely to timber, achieving compressive and tensile strengths and an elastic modulus very similar to that of wood. Its weight is slightly lower than wood’s while its flexural strength is much lower. The compression results show that the hemp stalk is only slightly weaker than timber, which would be expected due to its lower density. Its tensile strength is at least as strong as timber, potentially much greater; however further testing would need to be carried out to confirm this.
The results show that hemp stalk may have potential in compression and tension but less so in bending. The stalk appears to take forces more efficiently when they run in the direction of the long fibres on its exterior. The data can be used to make approximate comparisons with building elements. For example, a single standard 50 by 100mm timber stud has a loading strength of 125kN where as with hemp you would require 67 stalks to take the equivalent load.
As the results show, a few samples achieved much higher strengths than those used to calculate the hemp equivalent to a timber stud. Further research is required to investigate what factors determine the load bearing capacity of any particular stalk. Can climate, soil type or rate of growth affect it and is it possible to selectively breed a stronger crop? It is likely that such research would drastically reduce the number of hemp stalks required to take an equivalent load to timber. This research provides indicative data on the strength of hemp stalks and is limited in its breadth of analysis. A more thorough investigation which analyses a larger sample of stalks is required to give definitive loading strengths, especially for tension, which could potentially be a lot higher than recorded here.