br Introduction Understanding characteristics of explosive s

Introduction Understanding characteristics of explosive sympathetic detonation is very important to use explosives safely. The charges with and without shells for sympathetic reaction test are used to study the sympathetic action of explosives. In the sympathetic reaction tests of bared explosives, the acceptor is initiated by shock waves or products from donor detonation. Because of shock wave is attenuated rapidly in the air, the critical distance of sympathetic detonation is not longer than the gap of two bared explosive charges. The charges with shells will produce many fragments besides shock wave and detonated products. These fragments may fly a long distance to impact on the acceptor. The critical distance of sympathetic detonation can be measured in the tests, but detonation details cannot be obtained. Through the numerical simulation of sympathetic reaction tests, the detonation details can be analyzed to help in reducing the number of tests. In 1982, Howe et al. [1] simulated the sympathetic reaction tests with Eulerian code 2DE, calculated the acceptor initiated by shock waves from donor, and analyzed the influence of distance between charges, shell width, board width between donor and acceptor on tests were analyzed. Lu et al. [2,3] calculated the sympathetic reaction tests of PBXN109 explosives without shell in 2006. In 2010, Fisher et al. [4] studied safety of PBXN29 explosive charge with shell in the packaging containers by sympathetic reaction tests and numerical simulations. Shell expanding and impacting acceptor had been considered in calculation, while the distance between donor and acceptor was very close. Because there is a long critical distance for charges with shell in sympathetic reaction test, it hoechst is a challenge how to describe shell deformation, fragments forming and fragments action on the acceptor in calculation.
Experiments The sympathetic reaction test device consists of booster, donor, acceptor and witness plates. Fig. 1 is a photo of the sympathetic reaction test. Donor and acceptor were left standing on the ground with some distance apart. The witness plate was set under acceptor. The donor was initiated by a booster. The shock waves, detonation products and shell fragments of the donor acted on the acceptor. The reaction extent of acceptor was assigned based on the damage to the witness plate and the remnants of any un-reacted material. The sympathetic detonation tests were conducted at different distance between donor and acceptor, and the critical distance for acceptor exploding could be found. Here, the charge distance means the shortest distance between charges\' boundary of donor and acceptor. The critical distance means the charge distance for sympathetic detonation. The donor and acceptor were identical cased cylindrical GHL explosive charges. All charges were 60 mm in diameter and 240 mm in length, filled in the steel cylinder shells, while the shell was 3 mm in thickness, and its top and bottom covers were 3 mm in thickness and 30 mm in height. The boosters were 25 mm in diameter and 25 mm in length. The witness plates were 12 mm in thickness and made of 45# steel.
Simulation of sympathetic reaction When the donor blasted, its shell expanded and ruptured to form the irregularly shaped, high-speed fragments with different sizes, and then these fragments acted on the acceptor. If the charge distance from donor to acceptor was long, the actions of shock wave and detonation products were too weak to act on the acceptor, and the acceptor was mainly initiated by these high-speed fragments. The key to simulate the sympathetic reaction was to describe shell deformation, fragment random production and fragments action on the acceptor. In usual calculation, the shell of the donor was developed in continuous model, and some elements were deleted when elements get to its failure values, as seen in Fig. 2. The early expansion effect of shell could be simulated, but shell deformation, fragments formation and fragments impacting on acceptor could not be calculated. Herein, the elements-apart method and nodes random-failure method were used in the model to describe the progress of shell expansion and fragment random production. In order to describe the progress of shell fragments, the donor shell was modeled as an accumulation of equal volume elements. The nodes at the same location recomposed to a node group with a failure strain value, as seen in Fig. 3(b). If the failure strain value was got, the node group would fail, the adjacent elements would be separated to make the shell rupture and form fragments. The whole process of shell rupture, fragments dispersal and action on acceptor could be simulated. Fig. 3(c) showed the failing progress of the node group failed and the formation of the fragments. In practical tests, the donor would produce fragments with different sizes, so the failure strain values were defined in random normal distribution. As shown in Fig. 3(d), some of node groups got to their failure strain values and the elements separated, while some of them did not. The process of random size fragments forming and fragments action on the acceptor was simulated using the two methods.