Performance of prestressed concrete targets against projectile impact

ARTICLE IN PRESS

JID: IE

[m5G;December 5, 2016;20:17]

International Journal of Impact Engineering 000 (2016) 1
11

Contents lists available at ScienceDirect

International Journal of Impact Engineering journal homepage: www.elsevier.com/locate/ijimpeng

Performance of prestressed concrete targets against projectile impact TagedPD3X XM.A. IqbalDa4X X , D5X XAbhishek RajputDa, 8X X *, D9X XN.K. GuptaDb10X X a

TagedP Department of Civil Engineering, Indian Institute of Technology Roorkee, Roorkee 24677, India b Department of Applied Mechanics, Indian Institute of Technology Delhi, New Delhi 110016, India

TAGEDPA R T I C L E

I N F O

Article History: Received 30 July 2016 Accepted 26 November 2016 Available online xxx TagedPKeywords: Prestressed concrete Computations for prestressing Scabbing Ballistic limit Perforation

TAGEDPA B S T R A C T

Energy dissipation in prestressed concrete targets has been studied against the impact of long rod steel projectiles. Experiments have been carried out wherein prestressed concrete plates of thicknesses 60, 80 and 100 mm were subjected to impact by 1 kg steel projectiles at normal incidence velocities close to ballistic limit. An initial prestress of 10 and 20% of unconfined compressive strength 40 MPa has been induced in the target through pre-tensioning of 4 mm diameter high strength (1646 MPa) steel wires. The reinforcement has also been provided in the prestressed concrete targets to enable a direct comparison of their performance with the equivalent reinforced concrete targets. The prestressing in concrete has been found to be effective in globalizing the induced damage and thus enhancing ballistic resistance. The influence of prestress has become more prominent with increase in target thickness and decrease in projectile velocity. The experimental findings have been reproduced through finite element simulations on a commercial finite element code to understand the individual characteristics of prestressing wire, reinforcement and concrete. The prestressing force has been numerically transferred in concrete by introducing initial stress in the strands and then performing quasi-static simulation. The simulations for projectile perforation have been subsequently carried out employing Holmquist
Johnson
Cook model for concrete and Johnson
Cook elasto-viscoplastic model for both reinforcement and prestressing strands. The finite element simulations predicted the ballistic limit of the targets within 11% accuracy. © 2016 Elsevier Ltd. All rights reserved.

1. Introduction TagedPConcrete is the second largest material used by human beings after water. It is highly durable, fire and corrosion resistant and nonporous. Introduction of steel as reinforcement enabled the concrete to cover large space and sustain intensive loads under tension, flexure, torsion and shear. A consistent improvement in strength and performance over last few decades transformed the concrete into most suitable construction material for nuclear, strategic and protective structures. Prestressed concrete is generally employed to reduce sections and crack width in structural elements by eliminating flexural tension; and therefore it is preferred in pressure vessels and large span structures. Studies on the perforation capacity of 250 mm thick full-scale (2 m £ 2 m) reinforced and prestressed concrete walls against 47 kg hard missiles described that the just perforation velocity for prestressed concrete wall was higher than that of the reinforced concrete wall but the prestressed concrete wall suffered 35% more rear surface damage [1
3] than the reinforced concrete wall. The provision of transverse reinforcement had no influence on the perforation capacity of reinforced concrete wall due to localized damage however it increased the capacity of prestresssed concrete *

Corresponding author. E-mail address: [email protected] (A. Rajput).

TagedPby 10
15%. Holquist and Johnson [4] introduced radial and hydrostatic prestress in the silicon carbide ceramic targets and simulated their ballistic response against cylindrical steel/tungsten projectiles. The ceramic tiles of thickness 12.7 and 20 mm were introduced two different levels of prestress through metal confinement. Prestressing thin targets delayed the tensile failure at the rear ceramic-metal interface by facilitating the interaction of projectile for a longer duration and thereby improved the ballistic performance. For thick targets the radial prestress improved ceramic dwell behavior because of high surface pressure. The study of blast load resisting capacity of the concrete has demonstrated that an initial prestressing resulted in reduced deflections (both maximum and residual) in concrete elements, and has also been found effective in delaying the appearance and growth of flexural cracks [5]. The prestressing in concrete has improved the flexural capacity of the beams by reducing tensile damage. As the strength increased from 42 to 70 MPa, and the initial stress in strands, 315 to 630 MPa, the capacity to perform under blast loads improved. It has been noted however, that increasing the prestress level in beams may increase the diagonal shear cracks toward the support which might prove to be detrimental under blast loading. TagedPThe prestressed concrete has attracted very limited attention in the open literature possibly due to the fact that the introduction of prestress in concrete is a complicated process (through both

http://dx.doi.org/10.1016/j.ijimpeng.2016.11.015 0734-743X/© 2016 Elsevier Ltd. All rights reserved.

Please cite this article as: M.A. Iqbal et al., Performance of prestressed concrete targets against projectile impact, International Journal of Impact Engineering (2016), http://dx.doi.org/10.1016/j.ijimpeng.2016.11.015

ARTICLE IN PRESS

JID: IE 2

[m5G;December 5, 2016;20:17]

M.A. Iqbal et al. / International Journal of Impact Engineering 00 (2016) 1
11

TagedPexperimental and finite element procedures). Authors could not find any study available on the performance of prestressed concrete against projectile impact. TagedPThe present experimental and finite element investigations aim to explore the possible influence of the magnitude of prestress on the energy absorption capacity of prestressed concrete against projectile impact. The prestressed concrete plates of thicknesses 60, 80 and 100 mm were subjected to impact by 1 kg ogival nosed steel projectiles at normal incidence velocities close to ballistic limit. An initial prestress of 10 and 20% of unconfined compressive strength 40 MPa was induced in the target through pre-tensioning of 4 mm diameter high strength (1746 MPa) steel wires. The deformed reinforcing steel bars of f8 mm @ 100 mm c/c have also been provided to enable a direct comparison with the non prestressed concrete targets. The effect of the magnitude of prestress has been studied on the experimental results and the results have been compared with the non prestressed concrete targets. The experimental findings have been reproduced through finite element simulations on ABAQUS/Explicit finite element code to obtain further insight of prestress concrete behavior. The prestressing force has been numerically transferred in concrete by introducing initial stress in the strands and then performing quasi-static analysis. The simulations for projectile perforation have been subsequently carried out by employing Holmquist
Johnson
Cook model [6] for concrete and Johnson
Cook elasto-viscoplastic model [7
8] for reinforcement as well as prestressing strands. 2. Preparation of prestressed concrete specimens TagedPA concrete mix was designed for obtaining 28 days unconfined compressive strength of 48 MPa in accordance with the requirement of Indian Standard; IS10262. Trials were conducted with various configurations of cement, potable water, river sand and coarse aggregate. The final composition of the mix had 440 kg cement, 0.4 water cement ratio, 730 kg dried river sand and 1050 kg basalt coarse aggregate of average size 10 mm in one cubic meter concrete as per the requirement of IS456 (2000). The concrete was casted in the controlled laboratory environment at 20
30 °C temperature. The slump of concrete was measured between 95 and 110 mm. In each casting, five cubes of 150 mm were also prepared along with the plate specimens. The typical uniaxial compression tests performed on cube specimens after 28 days curing in potable water at normal temperature resulted an average compressive strength 46
51 MPa. TagedPThe concrete specimens of span 450 mm £ 450 mm were introduced a unidirectional prestress of 10% and 20% of unconfined compressive strength with the help of f4 mm high strength (1650 MPa) steel wires, stretched across target span at the center of thickness to enable development of uniform compressive stress at the cross-section. The target was also reinforced with f8 mm deformed steel bars of tensile strength 415 MPa @ 80 mm c/c both ways with a clear

cTagedP over of 15 mm. The strength of the prestressing wires was measured by carrying out tension tests on controls universal testing machine. An especially designed prestressing bed of mild steel girders enabled casting of multiple specimens simultaneously. The prestressing strands were inserted through the square shaped steel target molds fixed with the prestressing bed to introduce initial stress in the concrete specimens. A total number of 13 and 22 strands were inserted in the target to induce 10% and 20% prestress respectively, see Table 1. Each strand, anchored at one side with the I-section girder, was stretched from the other side with the help of a hollow hydraulic jack and introduced an initial tension of 10, 14 and 16.5 kN for inducing 10% prestress and 13, 16 and 18 kN for inducing 20% prestress in 60, 80 and 100 mm thickness respectively. Thus, an initial stress of 4.8, 5.0 and 4.76 MPa corresponding to 10% and 10.5, 9.7 and 8.8 MPa corresponding to 20% prestress was induced in 60, 80 and 100 mm thickness respectively at anchorage take up, see Table 1. The strands were held in position with the help of steel wedges. The total losses in pretress due to elastic shortening, friction, creep and shrinkage were assumed to be 15% of the initial stress (at anchorage take up) as per the recommendation of IS 1343 (2012). Authors, also explored literature on this subject and found that the 15% loss of prestress is quite an established assumption and has been justified in the previous studies on prestressed concrete and also recommended in the standards on prestressed concrete, please see [9
11]. The effective prestress in the target after deducing the losses was calculated to be 4.09, 4.24 and 4.05 MPa corresponding to 10% and 8.9, 8.3 and 7.5 MPa corresponding to 20% prestress for 60, 80 and 100 mm thickness respectively, Table 1. TagedPThe concrete was poured in the square steel molds, carefully compacted with the needle vibrator avoiding contact with the strands, and surface D1X X D12X X finished. The curing of concrete was done with the help of wet gunny bags for 28 days. The wedges were subsequently released to enable transfer of stress in the body of concrete. 3. Ballistic experiments TagedPThe experiments were conducted with 1 kg ogival nosed hardened steel projectiles impacted on prestressed concrete targets at incidence velocities in the range 90
225 m/s. A pneumatic gun comprising of a single stage compressor of 60 kg/cm2 working pressure, a pressure reservoir and 19 m long steel barrel, capable of launching 1 kg projectile up to a velocity of 300 m/s was employed to carry out the ballistic experiments, see Fig. 1. In the present study, the air pressure used to launch the projectile was up to 50 kg/cm2. A pneumatic actuator which enabled opening a mechanical ball-valve system was used to release the air pressure and thus launch the projectile on the targets at normal incidence. A robust steel fixture was designed for rigidly holding the targets and maintaining fixity at the (target) edges. The incidence and residual velocities and the perforation phenomenon were recorded with the help of a high speed

Table 1 Calculations for inducing initial pre-stress in concrete targets. Calculation of effective stress for inducing 10% preDstress 1X X in the target Target thickness (mm)

Effective cross-sectional area (mm2)

No. of wires

Force in each wire (kN)

Initial stress in target (MPa)

Losses (15% of initial stress) (MPa)

Effective stress in target (MPa) D initial stress in target
losses

60 80 100

27,000 36,000 45,000

13 13 13

10 14 16.5

4.81 5.0 4.76

0.72 0.76 0.71

4.09 4.24 4.05

10.50 9.78 8.80

1.58 1.46 1.30

8.9 8.3 7.5

Calculation of effective stress for inducing 20% preDstress 2X X in the target 60 80 100

27,000 36,000 45,000

22 22 22

13 16 18

Please cite this article as: M.A. Iqbal et al., Performance of prestressed concrete targets against projectile impact, International Journal of Impact Engineering (2016), http://dx.doi.org/10.1016/j.ijimpeng.2016.11.015

JID: IE

ARTICLE IN PRESS

[m5G;December 5, 2016;20:17]

M.A. Iqbal et al. / International Journal of Impact Engineering 00 (2016) 1
11

3

Fig. 1. Schematic of pneumatic gun.

TagedPvideo camera, Phantom V411, operated at a frame rate, 15,000
20,000 per second, at D13X X a resolution, 640 £ 200 pixels. The projectile after perforation was safely recovered through a bullet catcher placed at 2 m behind the target. The projectile followed its central axis and struck at target center. Non flickering LED lights were used to achieve enough illumination for video recording the perforation process. It was ensured during the specimen preparation that the prestressing strands and the reinforcing bars do not pass through target center and hence not to come in contact with the projectile.

4. Material model TagedPThe Holmquist
Johnson
Cook (HJC) elasto-viscoplastic material model [6] was employed for simulating the perforation of concrete. The HJC strength model predicts the normalized equivalent strength, s *, of concrete as a function of pressure and strain rate through a simplistic uncoupled approach taking into account the damage cracking and compaction;
 s  D Að1¡DÞ C BPN ½1 C Clnɛ  ð1Þ

TagedPmaterial parameters for HJC model were taken from Holmquist et al. [6] for 48 MPa concrete, see Table 2. TagedPThe material behavior of reinforcing steel and prestressing strands was incorporated in the ballistic impact simulations using Johnson Cook (JC) elasto-viscoplastic material model [7
8] capable of predicting the yielding, plastic flow, strain hardening, strain rate hardening, softening due to adiabatic effects and damage. The model, best suited for predicting perforation in metals assumes the equivalent von-Mises stress s in the following form; " !# pl  n i h i   h m ɛ_ pl 1 C Cln ð4Þ s ɛ pl ; ɛ_ ; T^ D A C B ɛ pl 1¡T^ ɛ_ 0 where A, B, n, C and m are material parameters ɛ pl is equivalent plaspl tic strain, ɛ_ is equivalent plastic strain rate, ɛ_ 0 is a reference strain ^ rate and T is non dimensional temperature defined as; T^ D ðT¡T0 Þ=ðTmelt ¡T0 Þ

ð5Þ

where T is the current temperature, Tmelt is the melting temperature and T0 is the room temperature.

where, s  D s =fc; ; s is actual equivalent stress, fc; is uniaxial compressive strength, A is the cohesive strength, D is material damage (0  D  1), B is pressure hardening coefficient, P D P=f 0 c is the normalized pressure, N is the pressure hardening exponent, C is the strain rate coefficient and e* D eeq/e0 is the normalized strain rate, eeq is equivalent strain rate and e0 is the reference strain rate. The HJC model includes a scalar damage formulation wherein the damage is accumulated from equivalent plastic strain increment (Deeq) initialized from plastic shear deformation and equivalent plastic volumetric strain increment (Dmeq) initialized from plastic crushing of the air voids during a cycle of integration. The damage evolution is expressed as; " # X Dep C Dmp DD D ð2Þ efp C mfp where efp C mfp is the plastic strain to fracture and is defined as; efp C mfp D D1ðP  C T  ÞD2

T0  T  Tmelt

ð3Þ

where D1 and D2 are damage parameters, T  D T=f 0 c is the normalized maximum tensile hydrostatic pressure and T is the maximum tensile hydrostatic pressure the material could withstand. The

Table 2 Material parameters for concrete of unconfined compressive strength 48 MPa; Holmquist et al. [6]. Density (kg/m3) Specific heat (J/kg K) A B N C f’c (GPa) SMAX Shear modulus (GPa) D1 D2 EFMIN Pcrush (GPa) mcrush K1 (GPa) K2 (GPa) K3 (GPa) Plock (GPa) mLock T (GPa)

2440 654 0.79 1.60 0.61 0.007 0.048 7.0 14.86 0.04 1.0 0.01 0.016 0.001 85 ¡171 208 0.80 0.10 0.004

Please cite this article as: M.A. Iqbal et al., Performance of prestressed concrete targets against projectile impact, International Journal of Impact Engineering (2016), http://dx.doi.org/10.1016/j.ijimpeng.2016.11.015

ARTICLE IN PRESS

JID: IE 4

[m5G;December 5, 2016;20:17]

M.A. Iqbal et al. / International Journal of Impact Engineering 00 (2016) 1
11

TagedPThe failure is assumed to occur when the damage parameter D exceeds unity;   X Dɛ P P D ɛ P ; ɛ_ ; T; s  D ; T; s  Þ P ɛ f P ðɛ_

ð6Þ

where Dɛ P is increment in accumulated equivalent plastic strain during an integration cycle and ɛ f P is the equivalent plastic failure strain. TagedPThe fracture model proposed by Johnson and Cook [7] takes into account the effect of tri-axial state of stress, strain rate and temperature on the equivalent failure strain. The equivalent failure strain ɛ pl f is expressed as; " !# pl h  i s  h i ɛ_ m _ pl ; T^ D D1 C D2 exp D3 s m 1 C D 1 C D5 T^ ɛ pl ; ɛ ln 4 f s s ɛ_ 0 ð7Þ are material parameters, ssm is the stress triaxiality

where D1 ¡ D5 ratio and s m is the mean stress. All the material parameters employed for JC flow and fracture model taken from Borvik et al. [12] are enlisted in Table 3. TagedPThe concrete damaged plasticity (CDM) model was employed for inducing initial stress in the prestressed concrete. The CDM model is based on the concept of isotropic damaged elasticity in conjunction with isotropic tensile and compressive plasticity to represent the post elastic behavior of concrete and could be used for the analysis of plain as well as reinforced concrete subjected to monotonic, cyclic and dynamic loading. The response of concrete under uniaxial compression was incorporated using the stress-strain relations provided in Iqbal et al. [13] and Sadique et al. [14]. The behavior of concrete in tension was predicted using the Hillerborg et al. [15] fracture energy criterion by employing the energy required to open a unit crack area, Gf, as a material parameter, see Iqbal et al. [13]. The material behavior of reinforcing steel and prestressing strands has been considered to be linearly elastic for simulating D14X X induction of prestress in the concrete.

5. Computational model TagedPThe prestressed and non prestressed concrete targets and the impacting projectile were modeled in ABAQUS/CAE. Initially the plan dimensions of the square targets, 450 mm £ 450 mm, were modeled and extruded to obtain the required thicknesses, 60, 80 and 100 mm of the three dimensional deformable continuum. The reinforcement was modeled as f8 mm three dimensional wire, translated at the required spacing (80 mm c/c) in both orthogonal directions, merged as a single part and placed in the body of concrete after assigning specified cover; 15 mm. The prestressing wires were Table 3 Material parameters for reinforcement and prestressing strands; Borvik et al. [12]. Density (r) (kg/m3) Young's modulus (E) (N/m2) Poisson's ratio A (N/m2) B (N/m2) n m Melting temperature (K) Transition temperature (K) D1 D2 D3 D4 D5 Strain rate (s¡1)

7850 20 £ 1010 0.33 490,000,000 383,000,000 0.45 0.94 1800 293 0.07005 1.732 ¡0.54 ¡0.01 0 0.0005

TagedP lso modeled as f4 mm three dimensional wires (13 for 10% and 22 a for 20% prestress) translated at the required spacing and placed at the designated position in the body of concrete (center of target thickness). The projectile was modeled as analytical rigid body with a reference point at its centroid to assign mass, rotary inertia and initial velocity. The surface to surface contact between the projectile and the contact region (f20 mm) of concrete was defined using kinematic contact algorithm available in the code assuming projectile as master and target as node based slave surface. The coefficient of friction between the concrete and the projectile surface was considered to be 0.05. The contact of projectile with reinforcement and prestressing strands was also modeled by employing the kinematic contact algorithm considering projectile as master and wires (reinforcement and strands) as node based slave surfaces, assuming negligible friction. The reinforcement and prestressing strands were embedded in the body of concrete by assigning tie constraints considering concrete as host and wires as guest elements. TagedPThe core region of the target which constitutes the contact region of diameter 20 mm was meshed with 1 mm £ 1 mm x 1 mm continuum three dimensional eight node reduced integration (C3D8R) brick elements. The outer region of the target was also meshed with C3D8R elements of size 5 mm £ 5 mm £ 3 mm. The region between the core and the outer region was assigned with the six node tetrahedral elements (C3D4) of edge varying from 1 to 3 mm for maintaining compatibility between the core and the outer mesh, see Fig. 2(a). The number of elements at the thickness of core was 60, 80 and 100 mm, and in the outer region, 20, 27, and 33 for 60, 80 and 100 mm thick target respectively resulting in a total number of elements of concrete, 260,781, 391,049, 437,049 respectively. The reinforcement and prestressing wires were meshed with two node three dimensional truss elements (T3D2) of length 5 mm, see Fig 2(b), and their size was considered constant for all thicknesses. Total number of elements of reinforcement and prestressing strands was 640 and 1040 for all target thicknesses. The mesh convergence was studied by varying the size of element in the inner core from 5 mm x 5 mm £ 5 mm to 0.8 mm £ 0.8 mm £ 0.8 mm and in the outer region from 10 mm £ 10 mm £ 10 mm to 3 mm £ 3 mm £ 3 mm. Considering different mesh configurations, the simulations were performed at various incidence velocities on 100 mm thick non prestressed concrete targets and the results converged at the current mesh configuration of element size 1 mm £ 1 mm £ 1 mm in the core and 5 mm £ 5 mm £ 3 mm in the outer region. A typical explicit simulation with 100 mm thickness took about 18 CPU hours. When the size of element in the inner core was reduced to 0.8 mm £ 0.8 mm £ 0.8 mm, the residual velocity remained almost same but the CPU hours increased to 23. 6. Numerical simulation for inducing prestress in the concrete TagedPThe simulation for predicting ballistic performance of prestressed concrete was carried out in two steps. In the first step, the prestressing force was induced in the concrete by performing quasi-static analysis in ABAQUS/Standard, and in the second step, the simulation for ballistic impact was performed in ABAQUS/Explicit. TagedPIn order to induce the prestsress in the concrete, the prestressing strands were initially embedded in concrete and assigned an initial tensile stress obtained after deducting all the prestsress losses. The resultant stress was thus calculated to be 85% of the initial force (Table 1) divided by cross-sectional area of wire. The embedded constraints (Tie constraints) were applied between the prestsressing wires and the concrete by considering the wires as the guest and the concrete as the host body, see also Mercan et al. [9]. The embedded constraints ensure perfect contact between the guest and the host body, and do not allow the slip of wire. The model was then subjected to quasi-static analysis in ABAQUS/Standard to enable transfer of prestress in concrete. Due to the introduced stress in the wire,

Please cite this article as: M.A. Iqbal et al., Performance of prestressed concrete targets against projectile impact, International Journal of Impact Engineering (2016), http://dx.doi.org/10.1016/j.ijimpeng.2016.11.015

JID: IE

ARTICLE IN PRESS

[m5G;December 5, 2016;20:17]

M.A. Iqbal et al. / International Journal of Impact Engineering 00 (2016) 1
11

5

Fig. 2. Typical discretization of prestressed concrete target (a) planar and sectional views of 60 mm thickness (b) detailing of reinforcement, prestressing wire and element size in 100 mm thickness.

TagedPthere will be development of strain (in the wires) due to which the stress will be transferred to the concrete. The damaged plasticity model (CDP) was used for predicting the behavior of concrete and the linear elastic properties were assigned to the prestressing strands and reinforcement. Typical results obtained are shown in Fig. 3. The stresses induced in the concrete have been found to have minor variation across target span. Nominal tensile stresses have also been noticed toward the target support due to boundary condition effects. However, away from the supports the stresses are in compression and their magnitude

TagedPcorresponds to experimentally induced prestress, see Fig. 3(a)
(c) for 60, 80 and 100 mm thickness respectively. The corresponding resultant stresses induced in the prestressing wires are shown in Fig. 3(d)
(f) respectively. The theory of prestsress is based on transferring the stress from the prestressing wire to the body of concrete. The transfer of stress in the concrete is due to the introduced stress in the wire that would not remain constant throughout its embedded length due to Hoyer's effect [16]. The similar variation in the stress profile will be there in the body of concrete with some shear lag [16].

Please cite this article as: M.A. Iqbal et al., Performance of prestressed concrete targets against projectile impact, International Journal of Impact Engineering (2016), http://dx.doi.org/10.1016/j.ijimpeng.2016.11.015

JID: IE 6

ARTICLE IN PRESS

[m5G;December 5, 2016;20:17]

M.A. Iqbal et al. / International Journal of Impact Engineering 00 (2016) 1
11

Fig. 3. Induced initial prestress in (a) 60, (b) 80 and (c) 100 mm thick concrete targets and (d)
(e) resultant stresses induced in the prestressing wires of corresponding target thickness.

Fig. 4. Typical failure modes of 100 mm thick target (a) reinforced concrete, (b) prestressed concrete with 10% induced stress, (c) prestressed concrete with 20% induced stress.

Please cite this article as: M.A. Iqbal et al., Performance of prestressed concrete targets against projectile impact, International Journal of Impact Engineering (2016), http://dx.doi.org/10.1016/j.ijimpeng.2016.11.015

ARTICLE IN PRESS

JID: IE

[m5G;December 5, 2016;20:17]

M.A. Iqbal et al. / International Journal of Impact Engineering 00 (2016) 1
11

7

Fig. 5. Typical failure mode of 80 mm thick prestressed concrete with 20% induced stress at different impact velocities.

7. Results and discussion TagedPIn case of 100 mm thickness, the projectile stuck in the target during perforation at 160 m/s velocity. The projectile has also seen to have deviated from its central axis. In all the other cases however where perforation has occurred, the projectile has followed its central axis. TagedPThe projectile perforated the targets by forming a circular hole equal to its diameter and also removed the material from front and back surfaces. The magnitude of scabbing has been found to be higher than that of the spalling in both prestressed as well as non prestressed concrete targets, see Figs. 4
6. The pattern and quantum of damage induced through experiments and finite element simulations has been found to be in close agreement. The equivalent crater diameter at the front and rear target surfaces obtained as the average of four diameters measured in different orientations has also been found to have close correspondence with its numerical imitation, see Figs. 4
6. TagedPThe computations in some cases could not accurately reproduce the size of crater in both reinforced and prestressed concrete targets (see Fig. 4) and there are visible differences in the shapes of actual and predicted craters. However, the simulated contours of high stresses developed in the impact region showed reasonable correspondence with the actually developed craters. The size of equivalent craters obtained in the target specimen has been provided in Table 4. The specimen designation in Table 4 describes the type of concrete and incidence velocity. For example RPSC2, RPSC and RCC referred to prestressed concrete with 20% induced stress, 10% induced stress and non prestressed concrete respectively. The number indicates the projectile impact velocity.

TagedP or a given incidence velocity, the size of front and rear surface F craters was found to be highest for prestressed concrete target with 20% induced stress followed by target with 10% induced stress and non prestressed target respectively. This behavior indicated that the induced prestress played an effective role in spreading deformation to a larger span in comparison to the non prestressed concrete target. Moreover, increasing the intensity of initial stress from 10% to 20% caused further spread of the deformation away from the impact zone. TagedPThe volume of eroded material from front and rear surfaces of the target measured after the ballistic trials has been found highest for the non prestressed concrete target followed by prestressed concrete target with 10% and 20% induced stress respectively, see Fig. 7 (a)
(i). This interesting observation led to the conclusion that the magnitude of damage is reduced by inducing (10%) initial stress in the concrete. Further increasing the level of prestress (20%) has also established this observation. TagedPThe size of crater and the volume of material eroded from the rear surface has also been found to have increased with increase in the target thickness, however, its variation with respect to the type of concrete could not be established. In general, the ejected volume of material increased with the decrease in projectile velocity. This phenomenon seems to have occurred in all the three types of concretes, was more prominent in reinforced concrete. The volume of material removed from the front surface of target was comparatively low and therefore its influence could not be distinguished with respect to the type of concrete. TagedPThe initial prestressing of concrete had significant effect on the ballistic resistance. The ballistic performance of prestressed concrete was found to be higher than that of the non prestressed concrete of

Please cite this article as: M.A. Iqbal et al., Performance of prestressed concrete targets against projectile impact, International Journal of Impact Engineering (2016), http://dx.doi.org/10.1016/j.ijimpeng.2016.11.015

ARTICLE IN PRESS

JID: IE 8

[m5G;December 5, 2016;20:17]

M.A. Iqbal et al. / International Journal of Impact Engineering 00 (2016) 1
11

Fig. 6. Typical failure mode of 60 mm thick prestressed concrete with 20% induced stress at different impact velocities.

TagedPequivalent thickness. Increasing the magnitude of prestress also increased the resistance of each of the three target thicknesses studied, see Fig. 8(a)
(c). Moreover, the effect of prestress has become more prominent with the increase in target thickness; such that the difference in the ballistic resistance between prestressed and non prestressed concrete has further increased. This behavior has also been noticed when the magnitude of prestress increased from 10 to 20%. The ballistic resistance of a given target (prestressed and non prestressed) has increased with the decrease in the projectile

iTagedP ncidence velocity. This effect is more prominent in prestressed concrete compared to non prestressed concrete, and is seen to have enhanced with increase in the magnitude of prestress. TagedPThe prestressed concrete with 20% initial stress offered highest ballistic limit followed by the prestressed concrete with 10% initial stress and non prestressed concrete respectively, see Fig. 9. The 100 mm thick target with 20% induced stress offered 45% higher ballistic limit than the non prestressed concrete target. The same target with 10% induced stress offered 34% higher ballistic limit than non

Table 4 Equivalent crater diameter in different type of concrete target. Target thickness (mm)

20% prestressed concrete

60 60 60 80 80 80 80 100 100 100 100

RPSC2-130 RPSC2-110 RPSC2-90 RPSC2-180 RPSC2-156 RPSC2-135 RPSC2-120 RPSC2-225 RPSC2-205 RPSC2-192 RPSC2-186

Specimen no.

Front diameter (mm) 73.70 83.93 89.04 96.78 104.8 109.4 119.8 121.2 84.37 139.9 78.97

10% prestressed concrete

Non prestressed concrete

Rear diameter (mm)

Specimen no.

Front diameter (mm)

Rear Diameter (mm)

Specimen no

Front diameter (mm)

Rear diameter (mm)

172.5 179.2 246.4 209.3 216.2 225.4 240.1 214.8 226.0 254.0 347.0

RPSC-115 RPSC-92 RPSC-75 RPSC-140 RPSC-120 RPSC-110 RPSC-100 RPSC-220 RPSC-210 RPSC-190 RPSC-175

79.25 90.25 95.75 105.2 114 119 130.3 134.7 93.75 155.5 87.75

185.5 192.75 265 227.5 235 245 261 238.7 251.2 282.3 330

RCC- 87 RCC- 78 RCC-62 RCC-120 RCC- 90 RCC-78
RCC-200 RCC-170 RCC-130


73 111 86 75 98 94
81.5 102.7 163.5


168 275 324 203.2 189 205
290 306 347.2


Please cite this article as: M.A. Iqbal et al., Performance of prestressed concrete targets against projectile impact, International Journal of Impact Engineering (2016), http://dx.doi.org/10.1016/j.ijimpeng.2016.11.015

JID: IE

ARTICLE IN PRESS

[m5G;December 5, 2016;20:17]

M.A. Iqbal et al. / International Journal of Impact Engineering 00 (2016) 1
11

9

Fig. 7. Volume of scabbing and spalling in concrete (a)
(c) 60, 80 and 100 mm thick targets with 20% induced stress (d)
(f) 60, 80 and 100 mm thick targets with 10% induced stress (g)
(i) 60, 80 and 100 mm thick non prestressed concrete targets respectively.

Fig. 8. Ballistic performance of prestressed and non prestressed concrete targets of (a) 60 mm (b) 80 mm and (c) 100 mm thicknesses.

TagedPprestressed concrete target, see Table 5. The simulations predicted the ballistic limit of three concretes with a maximum deviation of 11%, see Table 5. Increment in ballistic limit with the increase in thickness was more prominent in prestressed concrete. Increase in level of prestress however could not make any distinguished effect on the improvement of ballistic resistance with increasing thickness. The ballistic limit of 100 mm thick target was 104, 133 and 94% higher than 60 mm thick target for the concrete with 20% induced stress, 10% induced stress and non prestressed concrete respectively.

TagedPThe energy absorption capacity of each concrete has been plotted in Fig. 10 at the ballistic limit. The energy absorption capacity of the target is seen to have influenced by thickness and type of concrete. In general, the energy absorption capacity increased due to introduction of prestress and also due to increasing level of prestress. However, increasing the level of prestress form 10 to 20% could not significantly enhance the absorbed energy, see Fig. 10. A maximum increase in the absorbed energy due to introduction of prestress (167% and 124% respectively at 20% and 10% induced stresses) was found in 80 mm thick target.

Please cite this article as: M.A. Iqbal et al., Performance of prestressed concrete targets against projectile impact, International Journal of Impact Engineering (2016), http://dx.doi.org/10.1016/j.ijimpeng.2016.11.015

ARTICLE IN PRESS

JID: IE 10

[m5G;December 5, 2016;20:17]

M.A. Iqbal et al. / International Journal of Impact Engineering 00 (2016) 1
11 Table 5 Actual and predicted ballistic limit. Target thickness (mm)

60 80 100

20% prestressed concrete

10% prestressed concrete

Non prestressed Concrete

Actual (m/s)

Predicted (m/s)

Actual (m/s)

Predicted (m/s)

Actual (m/s)

Predicted (m/s)

92.5 127.5 189

89 117 182.5

75 100 175

79 97.5 163.5

67 78 130

73 87 136.5

Fig. 9. Comparison of ballistic limit.

TagedP erformed subsequently for predicting perforation considering p induced stress as the initial state of stress. TagedPInducing initial stress in concrete played an effective role in stimulating the globalization effect in the target deformation. The scabbing in prestressed concrete was found shallow (in depth) but spread over a larger region in comparison to non prestressed concrete. Increasing the intensity of initial stress from 10 to 20% has further enhanced the globalizing effect. TagedPAt a given incidence velocity, the volume of material eroded from the rear surface of 60 mm thickness was reduced by 13% in the concrete with 10% induced stress, and 28.6%, in the concrete with 20% induced stress in comparison to non prestressed concrete. For 100 mm thickness however, the reduction in the volume was 10% and 26% in D15X X the concrete with 10% and 20% induced stress respectively in comparison to non prestressed concrete. MoreoverD,16X X with D17X X respect to the concrete of D18X X 10% induced stress, the reduction of D19X X D20Xvolume X of eroded material in the concrete of D21X X 20% induced stress was found to be 18 and 20% respectively, in 60 and 100 mm thicknesses. TagedPFor a given concrete, the size of crater and the volume of material eroded from the rear surface has been found to increase with increase in target thickness and decrease in incidence velocity. TagedPThe concrete with 20% induced prestress offered highest ballistic limit followed by the concrete with 10% induced prestress and non prestressed concrete respectively. For 60, 80 and 100 mm thick targets, the ballistic limit of concrete with 20% induced prestress was found to be 38, 60 and 45% higher, and that of the concrete with 10% induced prestress was found to be 12, 28 and 34% higher than non prestressed concrete respectively. On the other hand, the ballistic limit of concrete with 20% induced stress was found to be 23, 27 and 8% higher than that of the prestressed concrete with 10% induced stress. TagedPFinite element simulations accurately reproduced the quantity and quality of damage in the three concretes and predicted their ballistic limit with a maximum deviation of 11% from the actual ballistic limit. Acknowledgement TagedPThe financial support provided by Science and Engineering Research Board, Department of Science and Technology, India, through the research grant no. SB/S3/CEE/0032/2014 for carrying out this study is gratefully acknowledged.

Fig. 10. Energy absorption capacity of different type of concrete target.

8. Conclusions TagedPBallistic experiments have been carried out on prestressed concrete targets of different magnitudes of induced initial stress and the results have been compared with non prestressed concrete targets. Finite element simulations have also been conducted on a commercial finite element code employing HJC model for concrete and JC model for reinforcement. Initial tension has been introduced in the strands and quasi-static analysis has been performed for transferring prestress in the concrete. Explicit dynamic simulations have been

References TagedP 1 Orvovic N, Galan M, Mlahoainu A. Hard missile impact tests in order to assess the effect of pre-stressing on perforation capacity of concrete slabs. In: SMiRT-23; 2015. TagedP 2 Galan M, Orvovic N. Quantification of perforation resistance of prestressed walls with transverse reinforcement and liner under hard missile impact based on test results. In: SMiRT-23; 2015. TagedP 3 Sagal G, Orvovic N, Blahoianu A. Numerical simulation of missile impact on reinforced concrete slabs: effect of concrete pre-stressing. In: SMiRT-23; 2015. TagedP 4 Holmquist TJ, Johnson GR. Modeling prestressed ceramic and its effect on ballistic performance. Int J Impact Eng 2005;31(2):113–27. TagedP 5 Chen W, Hao H, Chen S. Numerical analysis of prestressed reinforced concrete beam subjected to blast loading. Mater Des 2015;65:662–74. TagedP 6 Holmquist TJ, Johnson GR, Cook WH. A Computational constitutive model for concrete subjected to large strains, high strain rates and high pressures. Ballistics international symposium 2 warhead mechanisms terminal ballistics by ADPA Arlington, 2; 1993. p. 591–600.

Please cite this article as: M.A. Iqbal et al., Performance of prestressed concrete targets against projectile impact, International Journal of Impact Engineering (2016), http://dx.doi.org/10.1016/j.ijimpeng.2016.11.015

JID: IE

ARTICLE IN PRESS M.A. Iqbal et al. / International Journal of Impact Engineering 00 (2016) 1
11

TagedP 7 Johnson GR, Cook WH. A constitutive model and data for metals subjected to large strains, high strain rates and high temperatures. In: Proceeding the seventh international symposium on ballistics; 1983. TagedP 8 Johnson GR, Cook WH. Fracture characteristics of three metals subjected to various strains, strain rates, temperatures and pressures. Eng Fract Mech 1985;21 (1):31–48. TagedP 9 Mercan B, Schultz AE, Stolarski HK. Finite element modeling of prestressed concrete spandrel beams. Eng Struct 2010;32:2804–13. TagedP10 Hassan T, Lucier G, Rizkalla S, Zia P. Modeling of L-shaped precast prestressed concrete spandrels. Prestressed Concr Inst J 2007;52(2):78–92. TagedP11 Huang T. Prestress losses in pretensioned concrete structural members. Fritz engineering laboratory report no. 339.9, Lehigh University, Bethleham, Pennsylvania, USA

[m5G;December 5, 2016;20:17] 11

TagedP12 Borvik T, Hopperstad OS, Berstad T, Langseth M. Numerical simulation of plugging failure in ballistic penetration. Int J Solids Struct 2001;38:6241–64. TagedP13 Iqbal MA, Rai S, Sadique MR, Bhargava P. Numerical simulation of aircraft crash on nuclear containment structure. Nucl Eng Des 2012;243:321–35. TagedP14 Sadique MR, Iqbal MA, Bhargava P. Nuclear containment structure subjected to commercial and fighter aircraft crash. Nucl Eng Des 2013;260:30–46. TagedP15 Hillerborg A, Modeer M, Petersson PE. Analysis of crack formation and crack growth in concrete by means of fracture mechanics and finite elements. Cem Concr Res 1976;6(6):773–82. TagedP16 Abdelatif AO, Owen JS, Hussein MFM. Modelling the prestress transfer in pre-tensioned concrete elements. Finite Elem Anal Des 2015;94:47–63.

Please cite this article as: M.A. Iqbal et al., Performance of prestressed concrete targets against projectile impact, International Journal of Impact Engineering (2016), http://dx.doi.org/10.1016/j.ijimpeng.2016.11.015