European Journal of Scientific Research ISSN 1450-216X Vol.46 No.4 (2010), pp.554-562 EuroJournals Publishing, Inc. 2010 http://www.eurojournals.com/ejsr.

htm

Analytical and Numerical Study on Explosion into/on Cohesion less Soils

Hamed Niroumand Department of Geotechnical Engineering, Faculty of Civil Engineering University Technology Malaysia E-mail:niroumandh@gmail.com Khairul Anuar Kassim Deputy Dean, Faculty of Civil Engineering, University Technology Malaysia Abstract The article reviews the analytical and numerical studies of explosion into/on cohesion less soils during the last forty years. Various numerical methods for estimating of behavior of soils during explosion have been developed. This paper discusses different theories and numerical studies on explosion into/on cohesion less soils by previous researchers. Analyses were beginning from E. B. Polyak and E. N. Sher (1978), A. T. Rodionov and A. G. Terent'ev (1985) until the most recent analysis such as H. Niroumand (2009) are reviewed. This analysis was pioneered by Absil et al (1997), Dorn et al. (1999), Williams et al. (2000), Laine, L., et al. (2001), Niekerk (2001), Wang (2001), Cheng et al. (2002), Fairlie et al. (2002), Gupta (2002), Jacko et al. (2002), Persson et al. (2003), Rhijnsburger (2003), Fierov et al (2004), S.O. Olofsson(2007) and H. Niroumand(2009) will also be discussed. The results include most recent theories and numerical studies that accompanied by experimental results.

Keywords: Explosion, Numerical Analysis, JWL, Cohesion less Soil, Analytical, Euler, Lagrange

Introduction
The last years, some researchers have been tries to research explosion and different loads in order to safety the design underground structures and design of different vehicles used in military. Simulation of the performance of soil subjected to close-in explosion of explosives materials a challenge for different reasons. Its important to study the interaction between the explosion and soil response because different structures affected by soil deformation subject to detonations. In recent years it implemented the Eulerian mesh and multi-material option which has extended its modeling capacity and enabled a comprehensive technique to become feasible. The simulation and numerical analysis is very important in this research. The numerical simulations allow the approach to a reasonable configuration; minimize the number of the experimental cases, saving considerable amounts of cost and time. Finite element method (FEM) studies are widely used in defense related engineering analyses, such as high velocity impact and penetration. Protection of army vehicles and personnel against landmine threats is an important issue in the area of defense research. Responses of the vehicle being subjected to various explosions loading can be iteratively modeled using codes and appropriate

Analytical and Numerical Study on Explosion into/on Cohesion less Soils

555

numerical techniques. This approach reduces the cost of expensive experimental tests. Researchers use numerical simulation to research various tasks with different focus such as (a) injury of occupants, (b) design of vehicle components and (c) attenuation material. Numerical analysis and simulations dealing with explosion are summarized in Table1. The table focuses on parameters depend to blast loading. Analytical and Numerical analyses of the effect of sand conditions and the depth of buried explosive undertaken by Laine et al.(2001) presented the following conclusions: A buried explosive gives a much lower maximum pressure acting over a long duration than the flush and surface-laid explosive. In contrast, the incident impulse of a buried explosive was the highest at distances above the ground less than 1m. Blast loading caused by a buried explosive is concentrated in the vertical direction more than a surface explosive that is more spatially spread. Thus, it can be stated that the effect of soil condition has a great in flounce on the pressure and impulse during explosion. Although, these numerical simulations were not validated by experimental data and details on soil were not presented. A methodology for simulating soil explosion is presented by Fairlie et al (2002). Firstly, two approaches in AUTODYN-2D were investigated for modeling high explosive: (1) Multi-material Euler solver employing Jones-Wilkins-Lee, (JWL) equation of state (EOS) for detonation products, (2) Single material Euler solver employing ideal gas equation of state for detonation products. It was observed that prediction of the aim momentum is consistent. Therefore, further modeling and analysis used the ideal gas equation of state approach for detonation products. He used in the experiments conducted by Bergeron et al (2000) was modeled in AUTODYN-3D. The Lagrangian grid was filled using the cohesion less soil model derived by Laine et al (2001).By comparing the numerical results with experimental data, it was concluded that the impulse from a mine buried in dry cohesion less soil was over estimated by 24% respectively by the numerical model. Therefore, varying sand parameters, namely moisture content and initial density amplifies the impulse generated by the mine. Numerical results obtained by Wang (2001) were compared with field results performed by Bergeron et al (2000). Comparison shows that the numerical predictions of positive phase duration and the displacement of ejecta front over estimated experiments by 36% and 10% respectively. The findings are on the same plane in the case of a surface explosion as well. Both field investigation and numerical analyses play significant roles in this research. The analysis helped minimize the number of the experimental tests required, which are usually very costly, and also help to interpret the test results. Once verified by the field tests, it can be used as a design tool for the consequent improvement of the system under research. Modeling of landmine explosion is highly complicated, involving an explosion causing shock wave propagation in soil and air and then interaction with a structure. A simulation modeling tool needs to incorporate adequately these challenging factors. Ls-Dyna3d software appeared to be a suitable code currently available for its research. Since a number of assumptions and numeric approximation techniques are employed in Ls - Dyna, verification is important for each problem. Wang in his report described a benchmark case applying Dyna to simulate an explosion in soil and air. The simulation is compared with results from a well-defined landmine-explosion field. The agreement was reasonably good. This work had provided a base for further modeling and analysis of a system involving a structure, such as an army vehicle, subject to a landmine explosion.

556

Hamed Niroumand and Khairul Anuar Kassim

Figure 1: 100gr Simulated Anti-Personnel Mine Change Configuration of Field Tests by J. Wang (2001)

Figure 2: Finite Element Mesh (DOB = 3cm) by J. Wang (2001)

A part from the peak pressure and positive phase duration, the modeling agrees well with field results. Further analysis was suggested by Wang in area of cell size sensitivity and the feasibility of an empirical JWL EOS. The response of a box-like structure to an explosion was studied by Cheng et al (2002).The cohesion less soil was considered as a rigid surface which allowed reflection but no refraction. Researchers assumed that this simplification would result in a greater loading of the structure over a model that incorporates cohesion less soil in its parameters. In the contrast, Laine et al (2001) Bergeron et al (2000) observed that buried explosion have given higher loading and larger deformation than the surface explosion. Numerical analysis had not been supported by experimental evidence. Unfortunately details about the structure and metallic material are not available.

Analytical and Numerical Study on Explosion into/on Cohesion less Soils

557

Williams et al (2000) investigated the floor deflection of a vehicle using a simple empirical impulse model for cohesion less soil. This was obtained to generate an initial velocity boundary condition, in spite of their using finite element method software, LS-DYNA-3D. The only variable sand parameter of the model was density (2170kg.m).No other soil properties were incorporated in the model to predict pressure loading and momentum transfer. This setting resulted in a plate deformation of 578mm, while experimental results presented a deformation of 287mm. This obvious discrepancy was due to the fact that energy absorption mechanism of sand was not implemented into the empirical model used. Another notable effort, Gupta is an algorithm derived from the empirical-based code CONWEP and implemented into the finite element method software LS-DYNA, in order to generate explosion overpressure loading on a panel. When results of simulations that focused on the response of a plate are not in good agreement with field data ,i.e. the value of deformation, it has been concluded that further study of the plate model needs to be carried out or simple scale-up of the explosive material is done. T. Rodionov and A. G. Terent'ev (1985), One of the simple mathematical models in the theory of the deformation of continuous media in an blast is the solid-liquid model. This does not describe the dynamics of the soil and so enables us to determine only approximate characteristics of the crater. This model has now been used to study a wide range of problems in determining a crater in a continuous medium with various tensile characteristic and various positions of the explosive. Authors considered below within the framework of the solid- liquid model boundary-value problems in determining a crater in the blasts of point explosives and uniformly distributed explosives on the surface and deep within an isotropic cohesion less soils with angular and curvilinear free boundaries. E. B. Polyak and E. N. Sher (1978), The basic idea proposed in i of the solid-liquid model of an explosion in the ground consists in that in regions close to the blast the tensile forces are small in comparison with the pressure and the inertial loading, and the medium here can be assumed to be ideal. High velocities and displacements (liquid zone) were characteristic for that zone. These areas were separated by a transition layer. As a first approximation, it is assumed that it layer is infinitely thin and is a sealed boundary for the moving ideal medium. If the compressibility of the soil is neglected, then for the incompressible ideal medium in the liquid zone of the blast a pulsed formulation is applicable, in which the action of the blast is described in terms of the pressure pulse exerted on the medium by the blast products.
Figure 3: The Craters Profiles by E. B. Polyak and E. N. Sher (1978),

It can be seen from Figure2 that this theory enables the experimental shapes of the craters obtained in soils by the blast of a surface cord charge to be described satisfactorily.

558

Hamed Niroumand and Khairul Anuar Kassim

S.O. Olofsson(2007) modeled impact of explosive load that it considede in the design of civil defense structures. Blast occurring below soil level generates a ground shock wave that effects buried structures in the proximity of the blast point. The program code FLAC suggested an opportunity to model the propagation of soil shock waves as well as interaction soil-structures. These researchers present the procedure setup of a numerical model against field test for a buried explosion. Explosion effected from a conventional bomb were estimated by using CONWEP that it is a field data compiled by the U.S. Army (US Army, 1986).
Figure 4: Location of Experimental Points Relative to the Center of Explosive by S.O. Olofsson(2007)

Figure 5: Model Dimensions and Boundary Conditions by S.O. Olofsson(2007)

Fierov et al (2004), investigated an approach that mirrors this requirement was presented in its research. The blast of an antitank mine is modeled and analyzed by using the non-linear dynamics analysis code program, AUTODYN. The initial simulation setup included of a hemispherical charge laid on a perfectly reflective plane. Two equations of state for explosive materials were investigated,

Analytical and Numerical Study on Explosion into/on Cohesion less Soils

559

with the first one being the commonly obtained to empirical equation of state, Jones-Wilkins-Lee (JWL). The second research applied the ideal gas equation of state, often advanced for simplification in complicated models. The mesh sensitivity research was carried out. Two parameters of explosion waves, namely maximum pressure and specific impulse, are evaluated and compared with accessible experimental data used from CONWEP. Consequently, a blast of a mine laid on cohesion less soil was modeled using JWL EOS and explosion parameters were compared with the last model.
Figure 6: Material Location for Explosion in Various Deployments in Dry Sand by Fierov et al (2004)

H. Niroumand(2009), simulated the dispersion behavior of sand subjected to explosion on the surface of a sand layer. The simulation was conducted using AUTODYN. Explosion effects from an explosive were achieved by using the code program Conventional Weapon Effects Backfill (CONWEB), which is based on field data compiled by the U.S.army (US Army, 1986). Three different governing equations were used for air, sand and explosive. Ideal gas equation was used to equate the movement of air and the dry sand was based on compaction effort. For the explosion, the JWL (JonesWilkins-Lee) equation was used. Author presented the effect of explosion on the crater depth, crater diameter and overpressure exerted on sand and the surrounding air. The results had shown that crater depth and diameter increase with time during explosion. The experimental data on crater depth however were initially lower than the numerical simulation but increased more than the numerical simulation after 30ms. The overpressure showed a reducing trend with time. The numerical simulation based on AUTODYN predicted higher crater depth and overpressure at the initial stage but showed a good agreement with the experimental data with time.

560

Hamed Niroumand and Khairul Anuar Kassim

Figure 7: Material Position During Explosion in 1.36ms by H.Niroumand(2009)

Analytical and Numerical Study on Explosion into/on Cohesion less Soils

Table 1: Analytical and Numerical Studies on Explosion into/on Cohesion less Soils
Mathematical model Mathematical model AUTODYN FLUENT and LS-DYNA LS-DYNA AUTODYN MSC. Dytran LS-DYNA AUTODYN and MSC. Dytran AUTODYN LS-DYNA AUTODYN AUTODYN LS-DYNA AUTODYN FLAC AUTODYN 2 8 10 25 10 Solid Liquid model Solid Liquid model 475gr Composition B 7.5kg C4 10.4 kg Composition B 800gr Pentolite 100gr C4 5 kg TNT 1 kg C4 907.2gr Pentolite 500gr TNT 0.125/0.5/1/4 kg PETN 10kg TNT 100gr TNT 100gr TNT Rigid surface Rigid surface Cohesion less soil Cohesion less soil Cohesion less soil Rigid surface Cohesion less soil Rigid surface Cohesion less soil Cohesion less soil

561

E. B. Polyak and E. N. Sher (1978) A. T. Rodionov and A. G. Terent'ev (1985) Absil et al (1997) Dorn et al. (1999) Williams et al. (2000) Laine, L., et al. (2001) Niekerk (2001) Wang (2001) Cheng et al. (2002) Fairlie et al. (2002) Gupta (2002) Jacko et al. (2002) Persson et al. (2003) Rhijnsburger (2003) Fierov et al (2004) S.O.Olofsson(2007) H. Niroumand(2009)

Conclusion
Some of last researchers were reported dealing with numerical / simulation analysis of the limiting different numerical methods and code programs in explosion into/on cohesion less soils. The investigation shows that sand properties and depth of buried explosive significantly influence explosion output. So, it is important to investigate the sand properties along with the blast parameters. Some of the research work reported did not include the cell size, detail of modeling. There for, it is need to more research for prediction of behavior of cohesion less soils by new methods in numerical analysis during explosion.

Reference
[1] Absil, L.H.J., Verbeek, H.J. and Weerheijm, J. Combined experimental and numerical study of mine detonations in the vicinity of vehicles.In 15th International symposiumon the Military Aspects of Blast and Shock(DRDB, Defence Research Establishment Suffield, Banff, Canada, September1997). Bergeron, D.M., and Tremblay, J.E. Canadian research to characterize mine blast output. In 16th International Symposium on the Military Aspects of Blast and Shock (Oxford, UK, September 2000), pp.501511, Bergeron, D.M., et al. Numerical and experimental study of in-soil explosions from a multi phase perspective. In 8thInternationalSymposiumon Interaction of the Effects of Munitions with Structures (McLean, USA, April1997), pp.1134, Bergeron, D.M., et al. Assessment of foot protection against anti-personnel landmine blast using a frangible surrogate leg. In UXO/Counter mine forum (New Orleans, USA, April2001), Dorn, M.R., Rees, S.J., and Docton, M.K. Improving vehicle resistance to blast. In 8th European Attack and Survivability of AFVs Symposium (Shrivenham, UK, March1999). Fairlie, G., and Bergeron, D. Numerical simulation of mine blast loading on structures. In 17th International symposium on Military Aspects of Blast and Shock (LasVegas, USA, June2002). Fierov,D., etal.. Numerical simulations as a reliable alternative for landmine explosion studies: The AUTODYN approach. Journal of Impact Engineering.

[2]

[3]

[4] [5] [6] [7]

562 [8]

Hamed Niroumand and Khairul Anuar Kassim Gupta, A.D. Modeling and analysis of transient response in a multilayered composite panel due to explosive blast. In 20thInternationalSymposiumonBallistics (Orlando, USA, September2002). Laine, L., and Sandvik, A. Derivation of mechanical properties for sand. In 4thAsia-Pacific Conference on Shock and Impact Loads on Structures (Singapore, November 2001),pp.361 368. Niekerk,B. Land mine protection dealing with the uncertainties. In 10th European AFV Attack And Survivability Symposium (Shrivenham, UK, May 2001). Niroumand, H., Simulation Comparison of Dispersion Behavior of Dry Sand Subjected to Explosion, International Journal of Geomechanic, ASCE, under review Vovk, A.A., and Demeshchuk, L.I. Applied geodynamics of explosions. International Journal of Fluid Mechanics Research (1996), 116 121.ISSN1064-2285. Wang, J. Simulation of landmine explosion using LS-DYNA-3D software: Benchmark work of simulation of explosion in soil and air. Technical report DSTO-TR 1168,DSTO, Aeronautical and Maritime Research Laboratory, Melbourne, Australia, June2001. Williams, K., and Fillion-Gourdeau, F. Numerical simulation of light armoured vehicle occupant vulnerability to anti-vehicle mine blast. In 7th International Ls-Dyna Users Conference (Dearborn, USA, May2002),pp. 67 Williams, K., and Poon, K. A numerical analysis of the effect of surrogate anti-tank mine blasts on the M113. Technical report DREV TM-2000-007, Defence Research Establishment, Valcartier, Canada, March2000.

[9]

[10] [11] [12] [13]

[14]

[15]