The Cylinder Test

Adjustments of the metal wall strength and the explosive equation of state could give late-time motions within 1% of the observations however, comparison of the calculated initial wall velocity of a copper cylinder driven by 9404 with Campbell and Engelke s precise measurements showed 10% differences. Using one-dimensional cylindrical calculations with the explosive decomposing at constant volume will give late-time wall motions that agree as well with the experimental data as the two-dimensional calculations. 

The numerical model used to interpret cylinder wall expansion experiments must include a realistic description of build-up of detonation, Forest Fire burn and resulting detonation wave curvature. A problem in numerical simulation of long cylinders of explosive confined by thin metal walls is to obtain sufficient numerical resolution to describe the explosive burn properly and also to follow the simulation of long cylinders. The NOBEL code includes the necessary physics and will numerically model cylinder tests as described in Chapter 6. 

The error bars show the range of experimental C-J pressures measured in various geometries. The correlation is good for all the high performance explosives except PETN. PETN is the most oxygen-rich, has the fastest kinetics, smallest reaction zone, fastest Forest Fire rate, and least build-up of any of the explosives. 

The BKW equation of state was shown earlier in this chapter to be adequate to describe the expansion of an explosion under water from several hundred kilobars to a tenth of a bar. The BKW equation of state was also shown to be adequate to describe the plate dent test from over 500 kbars to less than 10 kbars. While it increases the confidence in using the BKW equation of state for describing such integral experiments, there is nothing unique about the BKW equation of state. The experimental data is also described, within experimental error, by explosive equations of state that describe the pressure-volume-temperature-energy relationship for detonation products quite differently. The cylinder test is used to calibrate the JWL equation of state for PBX-9404/9501 assuming C-J pressures varying from 380 to 305 kbars in the Livermore explosive data compendium. They all are forced to fit the cylinder test data  

Modeling of the cylinder test using the NOBEL code is described in Chapter 6. Animations of the cylinder test are on the CD-ROM in the /NOBEL/BUNOBEL directory. The PowerPoint NOBEL.PPT in the /NOBEL/BUNOBEL directory contains a discussion of the cylinder test as does APP.PPT in the /CLASS.PPT/CHAPT5 directory on the CD-ROM. 

---

Kury et al (Ref 7) consider the cylinder test to be the most versatile for determining relative performance. The test was developed by Kury et al in I960 and described in conf paper (Ref 3) which was not used as a source of information. The test was later improved and its modified version was described in unclassified Ref 7... 

Sphere Test. A metal acceleration test for measuring the relative performance of expls (Ref 1). Kury et al found the cylinder test to be more versatile (Ref 2)... 

Small scale, unconfined detonation rate sticks were tested at 3.00 0.03 mm and 6.35 0.03 mm pellet diameters. The ETN pellets were pressed to an initial average density of 1.74 0.01 gem 3 (98% TMD), and pellet expansion was not measured thereafter. These measurements gave a detonation velocity of 7.90 0.13 and 8.03 0.04 mmps 1 at 3.00 and 6.35 mm respectively, which are within uncertainty of the adjusted detonation velocities observed in the cylinder tests, given in Table 1. These velocity measurements demonstrate a critical diameter for ETN of <3 mm, which is consistent with the PETN critical diameter of<1 mm [9]. 

Another test which shows considerable promise for exploring time-dependent adia-bats with small amts of expl is the cylinder test. The std cylinder test geometry consists of a 1 inch diam, 12 inch long expl chge fitted into a Cu tube with a 0.1022 inch thick wall. A plane wave lens and 0.5 inch thick Comp B booster are used to initiate the test expl at one end. The radial motion of the cylinder wall is measured in a plane perpendicular to the cylinder axis 7 inches from the booster end. A streak camera records the motion, using conventional shadowgraph techniques. In addn, the deton vel of the expl is measured by placing pin switches 9 inches apart on the surface of the cylinder (Ref 2)... 

H Initial height of sludge in graduated cylinder in the cylinder test... 

A BKW equation of state in a one-dimensional hydrodynamic simulation of the cylinder test can be used to estimate the performance of explosives. Using this approach, the novel explosive 3,6-diamino-l,2,5,6-tetrazine 1,4-dioxide has been analyzed . 

The Gurney energy is most frequently obtained experimentally from cylinder test data. Additionally, on the basis of more detailed theoretic consideration of the detonation products and the wall material expansion process, the cylinder test makes it possible to calculate the detonation pressure and the heat of the detonation of an explosive. 

Besides, on the basis of the cylinder test data, if the detonation parameters at the CJ point as the initial point of the products expansion are known, it is possible to deduce constants in the Jones-Wilkins-Lee equation (known as JWL equation of state) of the detonation products isentropic expansion. 

Although the cylinder test appears to be a simple method, the treatment of the data provides a lot of important information regarding the tested explosive. It gives us the Gurney energy (from cylinder wall velocity) and the equation of the detonation products isentropic expansion. 

The direct experimental output of the cylinder test is the increase of the cylinder external radius as a function of time, i.e., the cylinder expansion curve (Figure 5.16). 

There are several ways described in the literature for the determination of the parameters in the JW equation of state from the cylinder test data. One of the possible ways is given by the block diagram (Figure 5.18). 

An example of the cylinder test data cylinder expansion curve, wall velocity-time curve, acceleration-time curve, and JWL isentropic expansion curve for Comp B are given in Figures 5.19 and 5.20. 

W. Fickett and L. M. Scherr, Numerical Calculation of the Cylinder Test , Los Alamos Scientific Laboratory report LA-5906 (1975). 

Using the build-up to and of models in NOBEL described in this chapter determined from experimental observations of PBX-9501 and PBX-9502 described previously to model the cylinder test of the explosives are shown in Figure 6.46 for PBX-9501 and in Figure 6.47 for PBX-9502. 

---