Nonsteady-State Detonations

The first observation of build-up of detonation was in the the explosive PBX-9404. Build-up of detonation has also been observed in the explosives Composition B and TNT. 

In 1965, Craig ° first discovered the nature of the nonsteady behavior of explosives. He studied the interaction of the explosive 9404 in one-dimensional plane geometry at four charge lengths with Dural, magnesium, and Plexiglas plates. If the 9404 behavior were steady-state, the experimental data should have scaled as a function of plate vs. explosive thickness. The data did not scale and they clearly indicate that the effective C-J pressure Pecj) for underinitiated detonation 9404 increases, or builds up, as the detonation wave runs. 

The result was not unexpected, because at about the same time, Davis et al. had shown by other experimental studies that the steady-state C-J theory did not accurately describe the behavior of real explosives. The exact nature of the nonsteady-state behavior shown by Craig s data ° (a 25% change in Pecj with less than 1% change in detonation velocity) was a surprise to most detonation scientists however, it permitted one to understand why thin layers of explosives acted so differently from the behavior indicated by the simple calculations calibrated with data from thicker explosive charges. 

There are clues, to be discussed later in this chapter, indicating the processes that result in failure of the steady-state detonation model. They are currently of little value to the numerical engineer who wishes to treat an explosive as realistically as practicable. He would like a description of the explosive behavior that would work in divergent and convergent geometry as well as in plane geometry. The study is limited to one-dimensional flow. 

The build-up model assumes that a real nonsteady-state detonation can be approximated adequately by a series of steady-state detonations with instantaneous reaction whose effective C-J pressures vary with the distance of run. This empirical model depends completely upon experimental data for its calibration. If the magnitude or duration of the initiating pulse is changed, or the the explosive is one for which experimental data are not available, new experimental data must be generated and the model must be calibrated for the new system. 

---

In addition, mixed semiconductors with graded compositions have, as described earlier relevant to gradations due to pressure, graded electronic properties for example, graded band gaps. A graded mixed crystal explosive should have predictable, nonsteady-state, detonation characteristics. The nonsteady time dependence and partial directionality of the detonation may be of interest. 

Because the nonsteady-state detonation process requires new concepts, some appropriate definitions are given. 

Charles L. Mader and B. G. Craig, Nonsteady-State Detonations in One-Dimensional, Plane, Diverging and Converging Geometries , Los Alamos Scientific Laboratory report LA-5865 (1975). 

---