Evaluation of effectiveness of polymer coatings in reducing blastinduced deformation of steel plates

Damith Mohotti , P.L.N. Fernano , Dakshitha Weerasinghe , Alex Remennikov

a School of Engineering and Information Technology, The University of New South Wales, Canberra, ACT, 2600, Australia

b Department of Civil Engineering, Faculty of Engineering, General Sir John Kotelawala Defence University, Ratmalana, Sri Lanka

c School of Civil Engineering, Faculty of Engineering, The University of Sydney, Darlington, NSW, 2006, Australia

d Centre for Infrastructure Protection and Mining Safety, University of Wollongong, Wollongong, NSW, 2522, Australia

Keywords:Polyurea coating Near-field blast LS-DYNA Plate deformation ALE

ABSTRACT Incorporating elastomers such as polymers in protective structures to withstand high energetic dynamic loads,has gained significant interest.The main objective of this study is to investigate the influence of a Polyurea coating towards the blast-induced response in steel plates.As such,Polyurea coated steel plates were tested under near-field blast loads, produced by the detonation of 1 kg of spherical nitromethane charges, at a standoff distance of 150 mm. Mild steel (XLERPLATE 350) and high-strength steel (BIS80)plates with thicknesses of 10 mm were Polyurea coated with thicknesses of 6 mm and 12 mm on either the front (facing the charge) or the back face. The deformation profiles were measured using 3D scanning. Numerical simulations were performed using the non-linear finite element code LS-DYNA. The strain-dependent behaviour of the steel and Polyurea were represented by Johnson-cook and Money-Rivlin constitutive models, respectively. The numerical models were validated by comparing the plate deflection results obtained from the experiments and were then used in the subsequent parametric study to investigate the optimum thickness of the Polyurea coating. The results indicate that back face coating contributes towards an approximately 20% reduction in the residual deformation as well as the absence of melting of the Polyurea layer, while the front-face coating can be used a means of providing additional standoff distance to the steel plates.

1. Introduction

Mitigating the effects of dynamic loads such as blast and ballistic impact by employing metal-elastomer combinations has been in consideration for the last few decade [1-4]. Polyurea is an elastomer that has caught the attention of researchers due to its unique characteristics such as low density, high tensile strength and high deformability[5,6].In addition,under dynamic loading conditions,it has shown to exhibit strain-rate sensitivity, where yield stress and modulus of elasticity increase, with increasing strain rate [7],which has led to its high shock absorption capabilities,as found in research studies involving Polyurea as a helmet suspension pad material[8,9].Thus,it is evident that Polyurea is a viable material in scenarios where shock waves are expected to propagate [10]. Polyurea has demonstrated workability when used in combination with base materials such as concrete, masonry and metals, and hence, highlights its potential to be used as a potential protective coating [11-13].

The motivation for the present study is derived from the fact that most modern critical structures such as bridges, exposed columns in city-buildings,armour vehicles are vulnerable against both blast and impact loads. Fig. 1 highlights two such applications,where a metal-polymer combination would enhance the potential blast and impact resistance of structures.Moreover,the application of polymer onto a metal can be considered as a fairly straightforward retrofitting technique, thus making it practically feasible to employ even in existing structures.

Fig.1. Potential blast resistant applications of Polyurea coated steel.

Several studies have been carried out with the intention of assessing the performance of combinations of polymer plus other materials under blast loads. Raman et al. [14] conducted a comprehensive study on the applicability of Polyurea as a retrofitting coating to improve the resistance against blast loads.The results showed that the Polyurea coating contributed positively towards the resistance of the concrete slab panel against blast pressure by indicating lower deflections and less concrete damaged zones. Bahei-El-Din [15] has reported an experimental investigation performed with conventional and modified sandwiched plate designs subjected to blast loads. The conventional sandwiched plates consisted of a thin outer layer and face sheet made out of fibrous laminates, separated by a layer of structural foam core,while modified plates had a polymer coating (polyurethane and Polyurea) between the foam core and the outer face sheet. The Results indicated that a greater reduction in the total deflection and energy absorption for the sandwiched system with polymer in it.Tekalur et al. [16] conducted an experimental investigation to assess the blast resistance of a composite consisting of Polyurea and E-glass vinyl ester.The blast pressure was generated using a shock tube test setup.The results showed that the application of Polyurea on the impact face of the composite increased the resistance against blast loads.

As for metal-polymer combinations, Ackland et al. [17] conducted a near-field blast trial on Polyurea coated steel plates. The polymer coated plates yielded a higher plastic deflection than the bare steel plates, and this was largely attributed to the debonding between the Polyurea and the metal. Further investigations by Ackland indicated that steel plates with Polyurea coating at the front face performed better [18]. However, this finding is questionable as the Polyurea,if directly exposed to the blast source,can melt under the influence of the heat generated by the blast loads.In fact,an experimental study by Hou et al.[5]indicated that the backface coating of Polyurea on 304 stainless steel plates had the ability to minimize potential tensile failures such as spalling and debonding. Although the inherent physics behind the explosion event may be different,back-face coating of Polymer has shown to improve the blast resistant of steel members during underwater explosion events [19,20]. Moreover, purely numerical investigations carried out by Chen [21] and Samie et al. [22] have shown that back face coating of Polyurea on steel plates,can reduce the potential deflection of the plates.

Based on the above findings, it is evident that there is a clear research gap to assess the influence of front and back face coating of Polyurea on steel plates,using a comprehensive experimental and numerical program. Most of the previous experimental programs have been conducted using small quantities of explosives(less than 100 g) at stand-off distances of between 100 and 150 mm. Hence,the absence of experimental evidence for the performance of polyurea coated steel plates under higher explosive loads at similar standoff distances must be addressed. Moreover, most of the previous experimental studies have been conducted using low strength grades of steel (with yield stress less than 400 MPa),However, high strength armour grade steel is often used is most modern applications involving blast resistant structures. Hence,this clearly highlights the requirement of experimental evidence for the performance of Polyurea coated high strength steel plates,when subjected to blast loads.

This paper presents a comprehensive study performed on Polyurea-coated metallic plates subjected to a close-range blast load generated by 1 kg spherical nitromethane charge.This mass of explosives was chosen to ensure the plate only deformed and did not indicate any major failure in terms of spalling and scabbing.The deformation profiles for uncoated and coated plates were measured through a 3D scanning process. In addition to its ability to mitigate structural damage, melting of the coating under the influence of the fireball created by the blast source and the delamination of metal and elastomer layers were also investigated.Advanced numerical models were developed and were validated using the experimental findings so that they can be used for the subsequent parametric studies.

2. Experimental setup

The experimental program was designed to compare the blast response of uncoated monolithic steel plates and Polyurea coated steel plates. Two types of steel, namely BlueScope XLERPLATE Grade 350 steel (XLEP350) and High-strength steel BISPLATE 80(BIS80) from Bisalloy Steel Group Limited, were chosen for this study.The former is a low-strength steel and was chosen due to its widespread availability and for comparison with the previous blast test results. On the other hand, the latter was selected due to its potential applications in protective structures and armour industry.As for the polymer, the Polyurea named Eraspray ESU630D?,supplied by Era Polymers Pty Ltd., Australia was used. A comprehensive study on the high strain rate behaviour of Eraspray ESU630D? has been reported in Mohotti et al. [6], where high strain rate tensile tests on the Polyurea sample were conducted in order to obtain the stress-strain properties of the material at different strain rates.The relevant strength parameters for the two grades of steel and the Polyurea used in this study are presented in Section 3.2.

All the plates were prepared as 1000 mm×1000 mm and were fixed onto a test rig using 24 M24 bolts,as illustrated in Fig.2.The exposed area of the plates was 700 mm×700 mm.In all the tests,the thickness of the steel plates was 10 mm.For the coated plates,Polyurea was applied on either the top face (which is the side directly exposed to the blast load)or the bottom face in thickness of 6 mm and 12 mm.Table 1 provides a summary of the uncoated and coated plate configurations tested under this experimental program.

Sensitized nitromethane ignited with the aid of an electric detonator was used to create the near field blast-loading scenario.As illustrated in Fig. 2, the liquid explosive of nitromethane was filled in a plastic spherical casing of 120 mm diameter, manufactured with Nylon using 3-D printing technology. The spherical charges were suspended directly above the centre of the tested plates with the aid of a Nylon string and the distance between the top of the plate and the centre of gravity of the spherical casing was maintained at 150 mm.The TNT equivalency based on the mass of NM can be considered as 1.0 [23] and hence, the scaled standoff distance of the blast event considered in the present study was 0.15 m/kg1/3.

Fig. 2. The experimental test set up.

Table 1 Summary of test configurations.

The main output from the experiments was the deformation profiles of the uncoated and coated plates, especially the residual central deflection of the plates.The former was observed at site and the latter was carried out subsequent to the tests, using 3D scanning. In addition, the nature of the plate surfaces and their behaviour at the edges were observed after the experiments.

3. Numerical modelling

A comprehensive, non-linear three-dimensional numerical analysis using Arbitrary Lagrangian-Eulerian (ALE) method in LSDYNA [24] was used to numerically model the near-field blast experimental setup. The aim of the numerical simulations was to develop validated models based on the experimental findings,which was then used for further parametric studies,as presented in Section 4.4 of this manuscript. This Section discusses key aspects related to the development of the models and the defining of material models, both of which are critical towards creating accurate numerical models.

3.1. Model development

A detailed view of the finite element model is given in Fig.3(a).The model consists of an air domain,plates and the two flanges of the blast rig. The air was modelled as an Eulerian domain using solid elements, whilst the remaining parts were modelled with Lagrangian shell/solid elements. Considering the symmetry of the test setup, a quarter model was used in the analysis (with an air domain of 500 mm × 500 mm × 350 mm, as shown in Fig. 3(b)),which significantly enhanced the computational efficiency of the model.

The main output of interest from the numerical simulations was the out-of-plate deformation of the plates. To this end, having a sufficient number of elements in the thickness direction of the plate is important, in order to accurately capture the bending and membrane effect of the plates.Since the minimum thickness of the tested plates was 10 mm, the height of elements in the plate was assigned as 1.25 mm, thereby ensuring that there would be a minimum of 8 elements in the thickness direction of the plates.The other dimensions of the plate elements were maintained at 5 mm,which corresponded with the element size of the air domain(5 mm × 5 mm × 5 mm). This ensured the nodes in the plate matched perfectly with the corresponding nodes in the air domain,which is critical towards ensuring a strong coupling effect between the Lagrangian and Eulerian parts.The*CONSTRAINED_LAGRANGE_IN _SOLID card was used to produce the interaction between the Eulerian air domain,the plate and the flanges(Lagrangian bodies).In order to model the 24 bolts-employed to connect the plates to the flanges in the blast rig - in a simplified manner, line elements were used with the keyword*CONSTRAINED_SPOT WELD between the top and the bottom flanges to represent the bolts.This keyword models the bolt as a rigid beam that connects the nodal points of the nodal pairs; thus, assigning rotational and translational constraints to the connected parts.This is in good agreement with the experimental evidence, as no fracture, shear or slippage were observed near the bolts.The bottom face of the bottom flange was assigned fixed conditions to replicate the supporting rig underneath. As for the connection between the steel plate and the Polyurea, since failure of the polymer is expected to occur before possible debonding at the interface between polymer layer and the steel plate,the common nodes at the interface of steel and polymer parts in the numerical model were merged to replicate this. The exposivechargewasmodelledusingthekeyword*INTIAL_VOLUME_ FRACTION_GEOMETRY and its detonation was initiated using the keyword *INITIAL_DETONATION.

Fig. 3. Numerical model: (a) full model of the rig and the air domain and (b) quarter model.

As could be seen in Fig. 3, the modelling of air domain was limited to the extents of the plates. Hence, in order to simulate realistic conditions where the air is free to move beyond this region,the keyword *BOUNDARY_NON_REFLECTING was defined for the faces of the air domain, to represent the flow-out conditions.Therefore,due to the flow-out conditions defined by this keyword,as soon as a pressure wave reaches the boundary of the air domain,it will get excluded from the analysis.These measures ensured that the pressure does not reflect from the outer boundaries of the air domain or diffract to the back of the plate.

Fig.4 illustrates a typical sequential capture of the propagation of the part corresponding to explosives, obtained from the numerical models. It can be observed that its interaction with the plate causes the plate to deform. Moreover,it can be seen that the explosives have not propagated below the plate,thus indicating no leakage through the Lagrangian elements, which is an important consideration in assessing the accuracy of ALE models.

Fig. 4. Sequential capture of the propagation of explosives in the ALE model.

Table 2 Mechanical properties of XLER350 and BIS80 steel used in the numerical models.

3.2. Material properties

Modelling the strain rate dependent behavior of the steel and Polyurea is critical towards accurately simulating the deformation of the plates.Especially,since near-field blast loads are expected to generate stresses in excess of the yield strength of materials, the material models must be capable of capturing the non-linear stress-strain behavior beyond the elastic limit.

The Johnson-Cook material model [25] is a commonly used constitutive material model that is available in many advanced finite element codes to represent the non-linear stress-strain behaviour of metallic materials.This is due to its ability to consider strain hardening, strain rate effects, and temperature effects of a material during its deformation process. The constitutive model consists of three different components representing each of the aforementioned effects. This model has been widely used in crash and impact analysis due to its accurate representation of the material behaviour, when compared to other simplified bilinear or perfectly plastic constitutive models [26,27]. Therefore, the Johnson-Cook material model has been used to represent the stress-strain behaviour of the XLERPLATE 350 and BISPLATE 80 steel in all the numerical analyses performed in this study.Table 2 summarizes the typical mechanical properties of the two grades of steel,used in the numerical simulations.

In the derivation of the Johnson-Cook material model parameters A, B and n for the XLERPLATE 350 and BISPLATE 80 steel, the user-defined strain rate was taken as 1 × 10-3s-1. The strain rate parameter (n) was defined by conducting tensile or compression tests at different strain rates. The strain rate parameter (c) and temperature constant (m) were obtained from the published literature. As the test method has been discussed in detail in the authors’previous work and many other publications,for brevity,only the basic material curves and the JC model parameters are presented in the paper [30].

The well-established Mooney-Rivlin material model has shown to be a reasonably good representation of the stress-strain behaviour of polymeric materials under high strain rates.The constitutive material model parameters for Polyurea was obtained from previous studies performed by the authors.However,the Mooney-Rivlin Model in LS-DYNA does not have the capability of considering the strain rate dependency in its constitutive model. Therefore, a predefined true stress-strain curve of Polyurea defined in authors previous work[6]was used in numerical simulations.As such,the curve corresponding to a strain-rate of 1000/s, from the previous work of the authors, was used for the present study [6,31]. Moreover,the material card*MAT_HIGH_EXPLOSIVE_BURN was used to define the initial conditions of Nitromethane and the Equation of State card*EOS_JWL was used to simulate its behaviour during the detonation.The corresponding values are presented in the authors’previous work[32].

4. Results and discussion

As illustrated in Fig. 5, it was observed that the fireball formed due to the explosion interacted with the plate.This is indicative of the fact that the shock front produced by nitromethane was well formed and impinged on the plates, which caused the residual deflection in them.The results presented in this Section focuses on the overall performance of the plates, with a special emphasis on the deflection profiles. The results from the uncoated plates (Section 4.1) were used as the reference to assess the performance of the coated plates and to identify the most effective plate configuration(Sections 4.2 and 4.3).Based on these findings,a parametric study was carried out using the validated numerical models to identify the optimum coating thickness (Section 4.4).

4.1. Uncoated XLEP350 and BIS80 plates subjected to blast load

The deformation profiles obtained for both BIS80 and XLEP350 plates are presented in Fig. 6. These deflections were measured between edges of the exposed area of the plate(700 mm),through the centre(at 350 mm)of the deformed plate.

Fig. 5. Series of high speed video plot showing the fire ball interaction with the blast rig.

Fig. 6. Deformation profiles of the uncoated BIS80 and XLEP350.

Fig. 7. (a) 3-D scanned profiles of the plates after the test and (b) deflection shape from the numerical simulations.

A closer investigation of the deflection profiles in Fig. 6 also shows a smooth variation of deflection in XLEP350 plate, but a change in the curvature of profile, at 100 mm and 600 mm in the BIS80 plate. This can be due to the formation of a plastic hinge at these locations.This led to a sharp pointy deflection at the centre of the BIS80 plate, when compared to the smoother variation in the XLEP350 plate.This is further confirmed by the 3D scanned outputs of the plates,given in Fig.7(a).Despite this difference,as evident by Fig. 6, the overall deflection of BIS80 plate is lower than the XLEP350 plate.Since the plates were subjected to similar loads and the boundary condition and the thickness of the plates were the same, the higher deflection values in the XLEP350 plates can be mainly attributed to the higher ductility of the plate material.However, as evident by the 3D scans of the plates, there was no visible failure or crack was observed in either of the plates.This was further confirmed by the deflected shapes of the plates obtained using 3D scanning,as well as the numerical simulations.These are presented in Fig. 7.

The residual plastic deformation at the centre of the plate obtained through 3-D scanning process and the numerical simulations are summarized in Table 3.It can be seen the deflection values from the numerical simulations and the experiments are in good agreement, albeit a minor underprediction in the former, for both materials. This is in addition to the accurate capturing of the deflection profiles, as evident by the deflection contours from the numerical simulations given in Fig. 7(b). This highlights the accuracy of the material cards as well as the modelling technique used in the present study.Therefore,these numerical models were used to develop the scenarios involving coated plates,as presented in the subsequent Sections.

Table 3 Residual deformation of uncoated plates from the experiments and numerical models.

Fig.8. (a)and(b)Deformation profiles and(c)3D scanned images of the XLEP350 and BIS80 plates and with a 6 mm thick Polyurea coating.

Table 4 Residual deformation of coated plates- 6 mm coating.

These deflection values further confirm the enhanced ductile behaviour of XLEP350.However,it must be mentioned that higher ductility with controlled deflection and without failure is a desired feature in protective structures [32,33]. One way of enhancing the ductility of an existing plate is through the introduction of ductile material such as polymers.Therefore,the performance of Polyureacoated plates will be considered in the next sections.

4.2. Plates coated with 6 mm polyurea

A Polyurea coating of 6 mm was applied on the front face and back face of both the XLEP350 and BIS80 plates and tested for a similar blast load to that of the uncoated plates.The performance of the plates was assessed using three indicators, namely plate deformation, melting of the Polyurea coating and delamination at the polymer-metal interface.

Fig. 8(a) and (b) provide the deflection profiles of the coated plates and Table 4 summarizes the residual plastic deformation at the centre of the plates. Furthermore, Fig. 8(c) presents the 3D scanned images of the tested plates.

Kissing him on the forehead, I said my good-byes and started toward the door. Then I suddenly realized something. If the adults in his life are not aware of the real power of this technique, they might discourage him. Not wanting limiting beliefs to swallow up Timmy s possibilities for a miracle, I spun9 around and returned to his bedside.

Compared to the uncoated plates,the residual plastic deflection at the centre of the plate is less by 4%and 15%in the XLEP350 and BIS80 front face coated plates, respectively. The corresponding reduction for back face coated plates are 7% and 6%. This is an interesting finding because the overall deflection in the coated plates have reduced, despite the higher thickness and the weight,when compared to the uncoated plates. Moreover, as evident in both Fig.8(a)and(b),unlike in the case of the uncoated plates,the variation in the deflection profiles are similar for both XLEP350 and BIS80 coated plates.Thus,it can be concluded that the presence of the Polyurea layer at the front of the plates,directly influenced the behaviour of the coated plates.

Fig.9 presents the state of the front and back faces of the coated plates, after the tests. It was observed that the Polyurea coating near the centre of the plate was melted,possibly due to its contact with the fireball created by the blast.Polyurea has a comparatively low melting temperature (in the range of 120°C) than metallic materials such as steel (Melting temperature in excess of 1000°C)and therefore,can easily reach the melting point during a near-field blast event. However, the area that was affected by high temperature is significantly small and no damage was observed in the other areas of the coating. Furthermore, as identified earlier,even with this melted coating, the coated plates indicated a lower residual plastic deformation, which highlights the effectiveness of the Polyurea coating.

Another important aspect to consider in a metal-polymer coating system is the possible delamination of polymer coating from the base material.However,Mohotti et al.[4]have previously highlighted that a good bond between the Polyurea coating and the steel plate can be achieved by following the appropriate surface preparation techniques and a suitable bonding agent.This creates a strong bond between the polymer and the metals,where the bond strength exceeds the magnitude of the tensile stresses,which was shown to be the main cause of delamination in multi-material systems [34]. A close observation of the coated plates after the tests confirmed that there was no visible delamination between the Polyurea coating and the steel surface.This confirms the feasibility of fabricating the coated plates, with a strong bond between the Polyurea and steel.

4.3. Plates coated with 12 mm polyurea

The XLEP350 and BIS80 steel plates were coated with a 12 mm thick Polyurea layer at both the front and back faces and the performance of these plates were compared with the uncoated and the coated plates with Polyurea layer of 6 mm at the front face.

The deflection profiles of these coated plates steel plates are presented in Fig. 10 and Table 5 provides the residual plastic deformation at the centre of the plates, obtained by the experiments and numerical models.

The 12 mm Polyurea coating at the front face reduced the deflection of the coated plates by 21%and 26%for the XLEP350 and BIS80 respectively.The corresponding reduction when the Polyurea coating was applied at the back face of the steel plate was 16%and 30%for XLEP350 and BIS80 plates respectively.While the reduction in deflection using a Polyurea layer of 12 mm at the front face was greater than the reduction obtained through a Polyurea layer of 6 mm at the front face, the greatest reduction was achieved when the 12 mm coating was applied at the back face.The increase in the aerial density of the 12 mm Polyurea coated plates when compared to the uncoated plates was 7.5%. Therefore, it is evident that the increase in Polyurea thickness significantly enhances the shock absorption capability of the coated plates.Moreover,its application at the back face of the plate was more advantageous.

Fig. 9. XLEP350 plate coated with 6 mm Polyurea coating at the front face (a) before and (b) after the test showing the melted region of the coating.

Fig.10. Deformation profiles of XLEP350 and BIS80 plates with 12 mm thick Polyurea coating at the (a) front and (b) back faces and (c) contour profiles from the numerical model of back face coated BIS80 plate.

The state of the front and back faces of the plates after the tests were similar for both XLEP350 and BIS80 coated plates and Fig.11 illustrates the state of the XLEP350 coated plates.Similar to plates with 6 mm coating,significant melting of Polyurea was observed at the centre of the front face of the 12 mm coated plates(Fig.11(a)).However, no melting of the Polyurea was observed, in the plates where the Polyurea was coated at the back face of the plate. This can be because the Polyurea coating did not face the blast source directly.The duration of the fireball is considerably smaller than the time taken by the materials to transfer heat from the front face to the back face.However,a discolouration was observed in the front face of the steel,which faced the blast directly.This was similar to the discolouration observed in the blast trials carried out for the uncoated plates,which was presented in Section 4.1.Moreover,no debonding was observed between the Polyurea and steel layers(see Fig.11).

The above discussion clearly highlights the superior performance of steel plates with a Polyurea coating at the back face. In fact,these findings are in line with the observations made by Amini et al.[35],albeit using a shock tube setup as opposed to a blast trial.It was where it was observed that steel plates with Polyurea at the back face yielded better performance in terms of prevention of failure and energy absorption.

In addition to the aforementioned advantages of the Polyurea coating at the back of the plate, further benefits can be identified,especially when employed in applications previously identified in this manuscript. One such benefit is that the Polyurea present on the back face of the plate increases the tensile capacity of the plate.During the dynamic deformation of the plate, it creates tensile stress at the back face.However,the Polyurea coating at the back of the plate has the potential to reduce the magnitude of tensile stresses generated at the back face, and hence enhance the tensile capacity of the coated plate.This can be attributed to the potential of stress reduction through an impedance-graded multi-material system,as identified by Fernando et al.[32].Since the impedance of polymer is less than the impedance of steel, the steel-polymer combination is effectively an impedance-graded system.

Another additional advantage of having an elastomer coating at the back face is that it can act as a fragment catcher in the event of a complete fracture of the preceding metal component of the coated plate.Such fracture at the back face of a metal plate can occur due to tensile failures and this can lead to the ejection of metal particles at high velocities. Hence, these particles can act as high-speed projectiles. When considering the performanc of a protective shield against a blast load,these projectiles pose a greater threat than the pressure induced damages. Hence, the application of Polyurea at the back face of metallic plates can enhance the level of protection offered by such a coated plate system.

Therefore, it can be concluded that the findings of the experimental and numerical investigations suggest that the steel plate with a 12 mm thick Polyurea coating at the back face indicated the best overall performance.Furthermore,it was clear that the use of high strength steel(BIS80)as the base metallic plate,generated less deflection in the coated plates.

4.4. Parametric study to investigate the optimum coating thickness of polyurea on the back face

To further explore the effectiveness of the coated plates,several Polyurea coating thicknesses at the back face were considered.This is of particular importance because the areal density of the coated plates increases when the thickness of the Polyurea coating increases. Therefore, beyond a certain thickness of the Polyurea coating, it is possible that the gain in terms of deflection can beoffset due to the additional weight imposed by the Polyurea coating.Therefore,for the purpose of practical application,it would be imperative to identify an optimum thickness or a range of optimum thicknesses of the Polyurea coating, in order to the coated plates to be effective. To this end, the validated numerical models were used to conduct a parametric study to find the optimum thickness of the Polyurea coating.

Table 5 Residual deformation of coated plates- 12 mm coating.

Fig.11. State of the XLEP350 plates after the tests for 12 mm Polyurea coated plates at the (a) front and (b) back faces.

Steel-Polyurea plates with the thickness of the Polyurea coatings at the back face ranging between 6 mm and 20 mm were numerically modelled. The material properties of BIS80 steel were used for the steel parts of the model, because the previous findings of this manuscript indicated its superior performance.A similar blast load to that of the blast trials, presented previously in this manuscript,were considered.The residual plastic deflection at the centre of the coated plates were obtained from the numerical simulations.The uncoated 10 mm thick BIS80 steel plate was used as the reference and the increase/decrease in the deflections of the coated plates were compared against the increase in aerial density.Table 6 and Fig.12 summarizes the findings of this parametric study.

Fig. 12. The variation between aerial density and residual plastic deformation for different thicknesses of back face Polyurea coating of BIS80 plates.

It is clearly visible that increasing the thickness of the back face Polyurea coating contributes towards the reduction of the permanent plastic deformation in the BIS80 plates. In fact, beyond a coating thickness of approximately 7 mm(which corresponds to a 9% increase in the aerial density) the percentage gain in terms of the reduction in the plastic deformation outweighs the percentage increase in aerial density.However,for smaller coating thicknesses,the advantage is not so pronounced because the metallic deformation dominates the behaviour of the plates. Therefore, taking into account the total thickness of a coated plate for practical applications, it is recommended to use a Polyurea coating thickness between 7 and 15 mm to reduce the residual deformation by 8-30%in high strength steel plates, when subjected to near field blast loads.

5. Conclusions

This study evaluated the performance of two grades of steel plates coated with Polyurea at the front and back faces, when subjected to near-field blast loads. The experimental study was complimented with comprehensive numerical simulations.The key conclusions of this study are summarized below.

· The high-strength steel plates (BIS80 plates) showed less deformation when compared to the commonly available XLEP350 plates, for both uncoated and coated plates. This highlights the suitability of employing high-strength steel, for applications related to protective structures.However,the high-strength steel grades are less ductile and hence,are susceptible to sudden fracture if excessive loads are applied.Therefore,it is desirable to combine such steel grades with non-metallic lightweight material such as polymers to design a ductile multi-material system,without compromising on the enhanced strength properties of such metals.

Table 6 Comparison of aerial density and residual deflection for different Polyurea thicknesses in back-face coated BIS80 plates(150 mm stand-off).

· The application of Polyurea coating at the front face reduced the plate deformation when compared to the uncoated plates, by absorbing the initial impulsive load.However,the Polyurea layer closer to the centre of the plate melted due to the sudden rise in temperature caused by the fireball. Therefore, application of Polyurea at the front face of a metallic system can cause serious concerns regarding its structural integrity.

· On the other hand, the application of Polyurea at the back face further reduced the residual plastic deflections of the steel plates.Moreover,the Polyurea layer showed no signs of melting.In addition to the membrane effect from an additional coating at the back face, it also contributed to reduce the possibility of ejection of metallic particles due to fragmentation,by acting as a catcher. Therefore, it was concluded that coating of Polyurea at the back face of the steel was the more effective configuration for the coated plates.

· The parametric study performed to identify the optimum thickness of the Polyurea layer at the back face, indicated that having a 7-15 mm coating in a BIS80 high strength steel plate can reduce the plate deformation by 8-30%when compared to the uncoated plates.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

The authors would like to thank the technical staff at the laboratories of the University of Wollongong.