Impact Forces From Artificial and Real Birds on a Large Diameter Hopkinson Bar

Forces were measured from normal impacts with two different artificial bird recipes and two species of real birds on a very rigid flat surface. The tests were conducted by launching the soft body projectiles axially into a Hopkinson bar in accordance with the SAE AS6940 test standard at three different nominal impact conditions: (a) a 1 kg projectile at an impact velocity of 49 m/sec; (b) a 1.8 kg projectile at an impact velocity of 110 m/sec; and (c) a 1.8 kg projectile at an impact velocity of 310 m/sec. At each condition two artificial birds and one real bird projectile were tested with at least three test repeats. Simulations were conducted to assess the effects of projectile orientation and velocity on predicted forces.

The Hopkinson bar consisted of an Aluminum 6061 solid cylindrical bar with a diameter of 305 mm and a length of 7.31 m, made up of two 3.66 m long sections that were in axial contact with each other. The bar was instrumented for strain measurement at two locations, 457.2 mm and 609.6 mm (1.5 and 2 diameters) from the impacted face. Forces were calculated from the measured strain. In addition, digital image correlation (DIC) was used to measure the velocity of the free end of the bar.

The real bird projectiles were 1 kg Mallard ducks and 1.8 kg Golden Comet chickens that were prepared per ASTM F330-21. Prior to testing, the real birds were placed in muslin bags to minimize tumbling during flight. Two artificial bird formulations were tested, one produced by the University of Dayton Research Institute (UDRI) and the other by the German Aerospace Center, DLR. The UDRI projectile was a homogeneous, relatively soft solid, formulated with a mixture of water, gelatin and phenolic micro-balloons to achieve a nominal density of 0.95 g/cc. The DLR projectile consisted of a printed plastic shell with internal plastic ribs and filled with a gel material. The projectiles were designed such that the overall average density was the desired value of 0.95 g/cc.

Initial testing showed that contact between the two Hopkinson bar segments was not sufficient to allow full transmission of the waves through the interface. For the 49 m/sec and 110 m/sec tests this resulted in reflections returning to the strain measurement site prior to the end of the impact. For these tests only a portion of the force pulse could be accurately measured. For the highest velocity tests at 310 m/sec the force pulse was complete by the time reflected waves interfered with the response. Despite this limitation, useful force data were obtained for all velocities.

Impact forces were very sensitive to the impact orientation and velocity. The force history generally consisted of an initial region with varying amplitude followed by a relatively constant amplitude response. The steady state region was similar among all projectiles and similar to what would be predicted for an inviscid fluid with a density of 0.95 g/cc. The initial part of the force response exhibited varying degrees of test-to-test repeatability associated with each projectile type. Forces calculated from DIC velocity measurements of the free end of the bar compared favorably those from the strain measurements.

In general, the artificial projectile impact forces were quite repeatable from test to test, with some exceptions which could be explained by issues associated with either the projectile itself or with how the projectile exited the gun barrel. For all impact velocities the initial portion of the impact force from the bird projectiles exhibited large test-to-test variability.

In this paper, impact forces for three types of projectiles at three impact velocities will be presented with an emphasis on the test-to-test repeatability of the results. Based on test and simulation results a proposed method for normalizing the impact force to minimize effects of differences in projectile orientation, impact velocity and density will also be discussed.