Duralumin kyudo arrows: Suetsugu 2024 study

In kyudo an archer may draw the string too far, or use an arrow that is too short, so that the arrowhead ends up inside the bow stave instead of projecting towards the target: this is called hikikomi. If the string is released at that point, the arrow is trapped between string and bow and takes a heavy load along its axis. The authors measured in the laboratory how much load a duralumin arrow can bear before it buckles, and what happens next.

Two shafts were tested: A, recommended for bows of about 98 N (10 kg) and up, and B, for bows of about 147 N (15 kg) and up. Under slow loading, A buckled at about 50 N and B at about 65 N, roughly half the draw force of the bows they are sold for, or less. The values calculated with the classic buckling formula (46.8 and 62.2 N) and with a simple finite element model (46.70 and 62.01 N) are very close to the measured ones. After buckling, both shafts kept compressing for about 250 mm and then snapped in half, leaving sharp metal edges. Under rapid loading, meant to resemble the release, the behaviour was the same: failure came in about 200 ms, and again the shafts split in two at the centre. The initial load peak corresponded to a stress of about 30 MPa, far below the alloy's yield strength of about 500 MPa: the shaft does not fail from the impact itself but because it buckles.

The shafts are 7000-series duralumin tubes: A has an outside diameter of 7.541 mm and an inside diameter of 6.880 mm, B 7.938 and 7.176 mm, with masses of 18.06 and 22.30 g per 0.86 m of length. They were cut to 850 mm, a typical draw length for an archer about 170 cm tall, with the nock removed and the fletchings left on. The elastic modulus, measured with a strain gauge in a tensile test, came out at 70.3 GPa.

In the static test the ends were capped with cones seated in funnel-shaped cups made on a 3D printer, so that the shaft was free to pivot at both ends, as it is when trapped between string and bow; load was applied by a Shimadzu universal testing machine at 10 mm per minute. In the dynamic test a compressed-air cylinder drove the shaft at high speed, a laser sensor tracked the piston rod, a load cell at the other end measured the force and a high-speed camera filmed at 800 frames per second. The finite element calculation used half a shaft divided into 10 beam elements.

The tests were run on a machine, not with a bow: hikikomi is reproduced as a shaft compressed between two supports, without a real string pushing and oscillating. The paper does not say how many shafts of each type were tested, and the charts appear to show a single test per type and condition. Fragments flying towards the archer's face are a reasoned conclusion, not an observation: there was no archer and no bow in the laboratory, and the path of the pieces was not measured. The rapid load comes from a pneumatic cylinder whose speed and force profile is not compared with that of a bow at release.

Only two duralumin shafts were tested; bamboo, carbon or aluminium arrows of other diameters were not, and the claim that point and nock do not alter the buckling load is the authors' own, reasonable but untested. The paper is in Japanese, with an English abstract.

The study measures a figure that explains why hikikomi is dangerous: a duralumin shaft can bear end-on only about half the draw force of the bow it is recommended for, or less, and when it gives way it snaps in half with a sharp edge. For kyudo practitioners the authors' message is that this mistake must be avoided altogether, which is why the paper was written. In Western archery, as the authors themselves note, the anchor under the chin and the shorter draw make the situation less likely. By analogy, similar mechanics may apply whenever an arrow that is too short ends up trapped between string and bow at release, but no recurve or compound arrow was tested, so the numbers apply only to these two shafts.