You are on page 1of 8



L. G. Green, E. L. Lee June 27, 2006

13th International Detonation Symposium Norfolk, VA, United States July 23, 2006 through July 28, 2006

Disclaimer This document was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor the University of California nor any of their employees. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or the University of California. . express or implied. makes any warranty. or process disclosed. does not necessarily constitute or imply its endorsement. process. or assumes any legal liability or responsibility for the accuracy. or otherwise. product. or service by trade name. trademark. manufacturer. or usefulness of any information. recommendation. and shall not be used for advertising or product endorsement purposes. completeness. or favoring by the United States Government or the University of California. Reference herein to any specific commercial product. apparatus. or represents that its use would not infringe privately owned rights.

This lowered the CJ pressure slightly more than one percent for the highest PETN density experiments. and the detonation velocity at 8. Lawrence Livermore National Laboratory Livermore. the Riemann integral results.the PETN density at 1. We undertook a painstaking review of the detonation pressure measurements reported in an earlier work that was presented at the Fifth Detonation Symposium and found that corrections were required in determining the shock velocity in the PMMA witness material. and EOS parameter values derived from the revised results. The equation of state used for the PMMA was the same as used in the data reduction. The first had to do with reducing the data using the quadratic equation as given in the paper. et al1 as presented at the Fifth Detonation Symposium. The third correction was to determine the CJ pressure from the measured shock velocity in the PMMA reference material by means of the Riemann integral instead of the acoustic approximation. The reaction zone width is less than currently detectable. Introduction This work is a reanalysis of work in “Equation of State of Detonation Products” by Hornig.DETONATION PRESSURE MEASUREMENTS ON PETN LeRoy G. For reasons to be explained. Lee Energetic Materials Center. The second correction was to ignore entirely the 0. The chemical composition is such that little or no carbon is produced in the detonation products. Green and Edward L. PETN is widely recognized as an example of nearly ideal detonation performance. There were three major corrections to the original presentation.5GPa. (<1 ns) Observations on PETN have thus become a baseline for EOS model predictions.4 GPa.765grams per cubic centimeter. This approach resulted in an average calculated CJ pressure that was about 5 percent lower than that reported for the higher density experiments. The quadratic extrapolation to determine entering shock velocity in the PMMA and the Riemann correction were verified by simulating the experiment with a very high resolution computer calculation where the input CJ pressure was set at 31.0 point.0 point in the data reduction. We present analysis of the records. The three corrections applied to the computer-generated results yielded a value of PCJ of 31. . CA 94550 Abstract. Our review indicates that the CJ pressures previously reported for full density PETN require an average lowering of about 6 percent. this point is thought to have a systematic error. supporting hydrodynamic simulations. We also refined the impedance calculation to account for the difference between the usual “acoustic” method and the more accurate Riemann integral. It has therefore become important to characterize the detonation parameters as accurately as possible in order to provide the most exacting comparisons of EOS predictions with experimental results. The lower densities require progressively smaller corrections.274 km/sec. well within probable error. which was originally applied with the quadratic term on the distance instead of the time variable. This resulted in artificially elevated values for the shock velocity in the PMMA and also concealed a systematic offset in the 0.

For the highest density PETN samples.4< ρ0 <1. It is our belief that the offset seen in the 0. There were large deviations in the data. In that paper. From this.4 These fitted results were sed in combination with the revised Chapman Jouguet ressure results to yield the Chapman Jouguet derivative parameters. In addition the results for pressures and derivative values in most cases exceeded realistic limits. When the 0. One can either accept all of the data or come up with a reason why some should be rejected. the x.6)2 for ρ0 > 1.4 mm thick samples. There is also a different anomaly that the reanalysis brought to light.6 D=3. D=3.In order to obtain continuous values for dD/dρ0 we modified the formula for fitting the detonation velocity as a function of density. The computer-generated data was fit perfectly by the quadratic equation with no significant value to the constant. This is a substantial error and had to be resolved.7mm samples are low because of the very steep Taylor wave that follows the extremely thin reaction zone of the detonating PETN. it was discovered that there was a systematic offset in the 0. Figure 1.4)2 for ρ0 < .48 g/cc using optical techniques.0 points and all of the experimental data was reduced omitting the 0.7(ρ0-. the ones that are only 12.27 mm thick).t data points from the experiments on high density PETN were properly fit to quadratic equations. That is. the CJ pressure was then determined by use of the acoustic approximation. An examination of the actual shock electric output from the experiment gave a possible clue to the problem.7(ρ0-.0 point was taken out of the calculation. there is a constant in the fit on the order of 7 nanoseconds.5%.5(ρ0 − . Experimental arrangement for PCJ measurement by shock electric effect.4) for . Then. The PETN detonation pressure was obtained from the shock velocity transmitted to a reference material (PMMA) by a plane detonation wave in the PETN.t data generated was used to back calculate the input pressure to the computer with the quadratic equation being used to calculate the entering shock velocity in the PMMA.5 GPa.7 mm samples should be rejected. The measurements reviewed here were those where the shock velocity in the PMMA was obtained by using the shock electric effect described by Hayes2 to observe the shock arrival time at each interface of a stack of five accurately measured discs (1. Judging that reliable results at these lower densities must await future experiments we have not included these results in this paper.0 points. The method chosen was to simulate the experiment with a very high-resolution computer calculation. If all values were to be averaged. This means that at x=0 mm the time has already progressed 7 nanoseconds. the CJ pressure would be 31. It is our opinion that the results from the 12. Evaluation of the PMMA shock measurements When the x. The signals from the shock transit of the PMMA interfaces were all very large and of one polarity while the signal from the shock entering the PMMA from the explosive was very small and of the opposite polarity.0 point arises from this divergence in signal strength and polarity to the oscilloscope used to record the data in the experiment.7 mm thick yield a CJ pressure statistically lower than the 25.30-3.4) – 3(ρ0 – 1.0 points. it was concluded that there was an error in the values of the experimental 0. Experimental description The work is described in detail in the referenced paper1.27 to . the entering shock velocity in the PMMA was the same as with it in. The Fifth Symposium paper included observations at charge densities .6 D=3.7(ρ0−.30-3.30-3. .4) + 3. The modified formula shown here retain the previous values to within 0. This is because it is quite likely that the CJ pressures measured for the 12.

P = PCJ + (dP/du)(u.(C/ρ)S where C is the sound speed (1) The integral. The Riemann integral is derived from the relation du/dρ = . The deviation in pressure from the acoustic result can be calculated to within 1% as λ(uPPMMA – . in general. The integral was carried out over the interval V(CJ detonation products) to V( witness PMMA state) by transforming (1) from C(ρ) to C(V) . leads to an expression which cannot be evaluated in closed form for all but simplest expressions for the isentropic pressure PS. This is found to apply for the higher densities.5(d2P/du2)(u – uCJ)2 . For PETN (ρ0 = 1..while the PETN CJ pressure from 25. 3 Comparison of Riemann and acoustic approximation.4. . Even for relatively low impedance witness materials such as the PMMA in these experiments it is generally assumed that the rarefaction of the detonation products from the CJ state follows a linear path in the P.4 mm thick samples averages 32. For the highest density PETN charges the results from the Riemann integral analysis shows a measurable correction to the acoustic result. Figure 2. The correction to be applied at the lower densities can be determined by noting the deviation of the Riemann integral values for the pressure from the acoustic values and accounting for the slopes at the intercept as illustrated in figures 2 and 3.04.uPPETNCJ)2/2 where λ = 2. must be evaluated by iterative technique since the equation for C C2 = (dP/dρ)S dP/du =.uP plane which is the negative of the slope of the hugoniot locus of the products. The derivative (dP/du) is obtained from the JWL form of the equation of state for PETN at ρo = 1.0 GPa if an errant data point not included in Table 1 is omitted.uP(CJ) = Int [ (1/Vo)(dP/dV)]1/2dV/V V = v/vo (4) The acoustic estimate is given by Pcj = (PPMMA + ρoDuPMMA)/2 (5) The deviation of the Riemann result from the acoustic result is illustrated in figures 2 and 3.. Fig. Acoustic vs Riemann PCJ calculation The customary method for obtaining the value of the Chapman Jouguet pressure from shock wave impedance measurements is to apply the “acoustic” approximation to the shock wave observations as the means to connect the shock response of the witness material to the shock wave CJ state in the detonation products.762 to calculate C. This is in fact the first term in the expansion of the Riemann result. For the lower density PETN charges due to decrease in both the particle velocity difference and the coefficient lthe deviation becomes negligible. .762) the correction is . Comparison of Riemann and acoustic approximation.Cρ (2) (3) uP(PMMA) .0 kb.uCJ) + . This approximation to the exact Riemann integral result is quite adequate for witness materials whose shock impedance and particle velocity closely matches that for the detonation products. .

Thus the deviations in comparisons for Γ and for γ vary mainly from the pressure deviations. although. in general. The agreement with detonation velocities is very nearly within experimental variation (no shown here). Pressure vs.Adiabatic Γ parameter The Γ values for the adiabatic derivative Γ = (dlnP/dlnV)S are calculated as ΓCJ = PCJ/(ρoD2) . This work was performed under the auspices of the U. Thus.1 Results of the calculations are plotted in Fig. Fig. be used as a test comparison with calculational models. density Figure 5 summarizes and Table 1 lists the results of our recalculations. The values for adiabatic gamma are thus. The values obtained for the thermodynamic derivatives Γ and γ are roughly in agreement with the trend of results from the CHEETAH calculations. Lawrence Livermore National Laboratory under contract W-7405-Eng-48. Summary Our revised results for Chapman Jouguet pressures are in almost all cases lower than previous results. Acknowledgments The authors are indebted to Craig Tarver and David Hare for their patience and perseverance in setting up and running computer simulations of the detonation electric effect experiments on high density PETN. with confidence. Gruneisen γ parameter The well known Jones relation provides a means to calculate the Gruneisen parameter for the CJ states at various densities. 4. Some comparisons with molecular model code calculations show that for this “baseline” explosive the molecular model yields values for pressures lower by approximately 7% than our results as illustrated in figure 5. ρ0/ρCJ = Γ/(Γ+1) we can likewise derive the dependence γ(ρCJ) shown in Figure 4. greater than earlier estimates. 4 PETN CJ derivatives. density with recalculated values and values from Cheetah calculations. Symp. . values of PETN CJ pressure vs.5 Comparison of 5th Det. S. Fig. as compared to the earlier results the current revised results for pressure are closer agreement with model predictions. γρ0) and since. Department of Energy by University of California. We believe that the current results can. the differences are still significant. γ is defined as V(dP/dE)V γ = Γ( Γ−1−2Ζ)/ Γ−Z) Γ is the adiabatic derivative at the CJ state Z is the logarithmic derivative of the detonation velocity = dlnD/dlnρ0 We are thus able to derive the dependence.

830 6.0 33.86 20.980 7.263 8.930 4.4 25.4 x 25.711 1.78 26.63 29.250 5.877 6.92 7.160 6. Comparison of recalculated values of PETN PCJ with values from 5th Detonation Symposium.4 25.4 x 25.980 6.1 14.316 5.360 5.43 31.4 25.010 6.4 x 12.5 34.32 12.006 5. (g/cc) (km/sec) (mm) (km/sec) (GPa) (km/sec) (GPa) (GPa) 1.704 6.95 28.4 x 12.732 6.88 31.506 30.227 0.26 31.76 32.74 6.1 25.762 1.3 6.540 5.43 6.450 5.8 x 25.065 5.600 6.825 6.Table 1.265 5.6 25.927 0.4 x 25.62 .232 1.49 20.4 25.4 x 38.4 x 25.1 30.8 x 25.73 30.534 4.75 6.740 7.4 x 25.532 1.4 25.86 13.866 6.812 6.8 x 12.3 30.222 7.3 33.9 30.710 1.931 0.4 x 25.39 30.58 14.703 1.7 31.763 1.56 29.65 13.4 x 25.21 25.59 32.01 32.535 4.266 8.375 6.762 1.2 14.01 13.448 1.4 x 25.591 1.4 x 25.886 0.5 19. x Length PMMA uS PETN PCJ PMMA uS PCJ acoustic PCJ Rm int.810 6.4 25.61 14.56 26.980 7.8 33.8 30.21 7.758 1.44 26.830 6.082 8.144 6.4 25.63 20.687 4.990 7.261 8.74 19.030 7.877 0.4 x 25.4 x 25.18 31.574 6.331 5.013 6.231 1.56 31.76 28.630 4.840 6.755 7.4 x 12.47 6.32 31.600 1.567 4.010 7.799 6.103 8.439 1.42 6.85 6.4 25.77 30.623 4.8 x 12.106 8.6 13.63 29.263 8.8 7.763 1.610 4.95 7.4 x 25.4 25. PETN Detonation DIMENSIONS 5th DET.05 20.41 20.4 13.364 6.830 6.263 8.063 8.910 4.7 50.7 6.898 6.35 28.8 x 25.4 x 25.4 x 25.8 20.65 21.4 25.13 12.4 x 25.4 x 12.0 7.343 5.4 x 25.45 29. SYMPOSIUM1 RECALCULATED VALUES Density Velocity Diam.7 14.7 7.7 25.371 6.16 12.460 1.1 6.795 6.5 8.673 6.9 13.371 6.448 5.45 25.4 x 25.4 33.4 25.41 32.49 6.83 32.21 7.1 7.604 1.610 6.483 5.250 8.6 30.16 32.228 1.4 25.4 50.160 6.680 33.775 6.070 6.2 8.4 25.86 19.230 1.098 5.520 6.949 0.030 6.3 33.706 7.569 6.470 5.83 12.8 19.4 25.210 5.4 25.7 25.4 25.420 5.764 1.932 8.578 5.7 50.9 22.489 6.366 5.661 6.108 8.74 6.4 25.4 25.686 1.261 8.6 26.770 4.4 x 25.560 4.4 x 38.7 25.97 13.230 1.050 5.71 31.697 1.712 1.4 x 25.990 6.990 0.263 8.883 6.4 25.4 x 25.34 12.7 25.178 7.4 x 25.4 25.763 1.10 28.84 6.75 6.4 x 25.102 6.763 1.4 25.570 6.4 26.29 29.4 25.488 7.535 4.4 50.4 25.378 6.32 26.338 4.53 13.560 5.21 29.8 34.4 25.268 50.700 4.16 31.860 8.4 x 25.

18-21 1970. K. B. L. P. 38. and P. A. Hayes.References 1. . 28. L. Vol. Plus additional references from the original paper. Pasadena. W. ‘Electric Signals Generated by Shock and Detonation Waves. and J. version 4. 2. Aug. Hayes. M. “Equation of State of Detonation Products”. 4. Cheetah Code. M. Vtello. H.” J. Lee. Fried. Kurrle. E. Finger. 3. E. E. 1968. 2005. Oct. R. Fifth Symposium (International) on Detonation. UCRL-CODE-155944.. CA.Appl.503-511 (1970). C. pp. B. 507-511. Feb. Phys.0.” presented at the Western States section of the Combustion Institute. Glaesemann. pp. Hornig. C. Howard. Souers. “The Detonation Electric Effect. 1967.