JOURNAL OF MATERIALS SCIENCE 40(2005)5483-5490 Robust design and manufacturing of ceramic laminates with controlled thermal residual stresses for enhanced toughness NINA ORLOVSKAYA Drexe/ University, 3141 Chestnut St. Philadelphia, PA 19104, USA E-mail: orlovsk @drexel.edu MYKOLA LUGOVY, VLADIMIR SUBBOTIN, OLEKSANDR RADCHENKO Institute for Problems of materials science. 3 Krzhizhanovskii st. 03, Kiev. Ukraine JANE ADAMS Army Research Laboratory, Aberdeen Proving Ground, MD 2 1005, USA MUNJAL CHHEDA. JAMES SHIH Ceradyne Inc, 3169 Redhill Ave, Costa Mesa, CA 92626, USA JAG SANKAR SERGEY YARMOLENKO North Carolina a&T State University, 1601 E Market St, Greensboro, NC 27411, USA Boron carbide-silicon carbide ceramic composites are very promising armor materials because they are intrinsically very hard. However, their fracture toughness is not very high. Their ballistic performance could be significantly increased if the brittleness of these materials could be decreased. Here we report development of boron carbide-silicon carbide layered ceramics with controlled compressive and tensile stresses in separate layers Such B4C-SiC laminates with strong interfaces can provide high apparent fracture toughness and damage tolerance along with high protection capabilities The theory of heterogeneous layered systems was used to develop optimal design parameters allowing the evaluation and maximization of apparent fracture toughness. The layered composites were designed in a way to achieve high compressive residual stresses in thin B4c-sic based layers and low tensile residuals stresses in thick B4C layers. The residual stresses were controlled by the phase composition of layers and the layers hickness. The estimated apparent fracture toughness was calculated for both three layered and nine layered composites. B4 C-30 wt%SiC/B4C laminates were made based on the optimized design for high apparent fracture toughness. Processing of laminates involved preprocessing of powders, forming green tapes and hot pressing Work is in progress to measure fracture toughness of laminates as well as their strength, hardness and the ballistic performance. C 2005 Springer Science Business Media, Inc. 1. Introduction method to control cracks and brittle fracture by defied Ceramics offer a number of attractive properties. These tion, microcracking, or internal stresses [5-7].Lami include high specific stiffness, high specific strengths, nates with strong interfaces, combined with excellent low thermal conductivities, and chemical inertness in fracture toughness and damage tolerance, can poten ny environments. Ceramics and ceramic compos- tially provide the highest ballistic performance. The due to low density, superior hardness, and compres- is to control the level of residual stresses in the indi- sive strength values relative to metals. As a result, ce- vidual layers. It is also a way to increase the failure ramics have been subjected to a multitude of ballistic strength of ceramics by creating a layer with and dynamic behavior investigations [1-4]. However, sive stresses on the surface that will arrest the the widespread usage of ceramics is currently ham- cracks and achieve higher failure stresses [8]. The layer pered by their lack of the requisite toughness. The lat- composition, as well as the systems geometry, allows est developments in ceramic composites show that the the designer to control the magnitude of the residual use of layered materials is perhaps the most promising stresses in such a way that compressive stresses in the 0022-2461 o 2005 Springer Science+Business Media, Inc. DOI:10.1007/s10853-005-1923-x 5483
JOURNAL OF MATERIALS SCIENCE 4 0 (2 005) 5483 – 5490 Robust design and manufacturing of ceramic laminates with controlled thermal residual stresses for enhanced toughness NINA ORLOVSKAYA Drexel University, 3141 Chestnut St., Philadelphia, PA 19104, USA E-mail: orlovsk@drexel.edu MYKOLA LUGOVY, VLADIMIR SUBBOTIN, OLEKSANDR RADCHENKO Institute for Problems of Materials Science, 3 Krzhizhanovskii St., 03142, Kiev, Ukraine JANE ADAMS Army Research Laboratory, Aberdeen Proving Ground, MD 21005, USA MUNJAL CHHEDA, JAMES SHIH Ceradyne Inc., 3169 Redhill Ave., Costa Mesa, CA 92626, USA JAG SANKAR, SERGEY YARMOLENKO North Carolina A&T State University, 1601 E. Market St., Greensboro, NC 27411, USA Boron carbide-silicon carbide ceramic composites are very promising armor materials because they are intrinsically very hard. However, their fracture toughness is not very high. Their ballistic performance could be significantly increased if the brittleness of these materials could be decreased. Here we report development of boron carbide-silicon carbide layered ceramics with controlled compressive and tensile stresses in separate layers. Such B4C-SiC laminates with strong interfaces can provide high apparent fracture toughness and damage tolerance along with high protection capabilities. The theory of heterogeneous layered systems was used to develop optimal design parameters allowing the evaluation and maximization of apparent fracture toughness. The layered composites were designed in a way to achieve high compressive residual stresses in thin B4C-SiC based layers and low tensile residuals stresses in thick B4C layers. The residual stresses were controlled by the phase composition of layers and the layers thickness. The estimated apparent fracture toughness was calculated for both three layered and nine layered composites. B4C-30 wt%SiC/B4C laminates were made based on the optimized design for high apparent fracture toughness. Processing of laminates involved preprocessing of powders, forming green tapes and hot pressing. Work is in progress to measure fracture toughness of laminates, as well as their strength, hardness and the ballistic performance. C 2005 Springer Science + Business Media, Inc. 1. Introduction Ceramics offer a number of attractive properties. These include high specific stiffness, high specific strengths, low thermal conductivities, and chemical inertness in many environments. Ceramics and ceramic composites are attractive materials for use in armor systems due to low density, superior hardness, and compressive strength values relative to metals. As a result, ceramics have been subjected to a multitude of ballistic and dynamic behavior investigations [1–4]. However, the widespread usage of ceramics is currently hampered by their lack of the requisite toughness. The latest developments in ceramic composites show that the use of layered materials is perhaps the most promising method to control cracks and brittle fracture by deflection, microcracking, or internal stresses [5–7]. Laminates with strong interfaces, combined with excellent fracture toughness and damage tolerance, can potentially provide the highest ballistic performance. The way to achieve the highest possible fracture toughness is to control the level of residual stresses in the individual layers. It is also a way to increase the failure strength of ceramics by creating a layer with compressive stresses on the surface that will arrest the surface cracks and achieve higher failure stresses [8]. The layer composition, as well as the system’s geometry, allows the designer to control the magnitude of the residual stresses in such a way that compressive stresses in the 0022–2461 C 2005 Springer Science + Business Media, Inc. DOI: 10.1007/s10853-005-1923-x 5483
outer layers near the surface increase strength, flaw tol- 2. Thermal residual stresses erance, fatigue strength, fracture toughness and stress and its calculation corrosion cracking. In the case of symmetrical lami- In this work the two-component brittle layered com- nates, this can be done by choosing the layer compo- posites with symmetric macrostructure are considered sitions such that the coefficient of thermal expansion The layers consisting of different components alternate ( CTE)in the odd layers is smaller than the CtE of one after another, but the external layers consist of the the even ones. The changes in compressive and tensile same component. Thus, the total number of layers N stresses depend on the mismatch of CTE's, Young's in such a composite sample is odd. The layers of the moduli, and on the thickness ratio of layers(even/odd). first component including two external(top) layers are However, if the compressive stresses exist only at or designated by index 1(= 1), and the layers of the near the surface of ceramics and are not placed inside second component(internal)are designated by index the material, they will not effectively hinder internal 2(j= 2). The number of layers designated by index cracks and flaws [ 9] I is (N+ 1)/2, and the number of layers designated Boron carbide is an important ceramic material with by index 2 is(N-1)/2. The layer of each omponent many useful physical and chemical properties. After cu- has some constant thickness, and the layers of same bic boron nitride, it is the hardest boron containing com- component have identical thickness pound [1o]. Its high melting point, high elastic modulus, There are effective residual stresses in the layers of large neutron capture section, low density, and chem- each component in the layered ceramic composite Dur- ical inertness make boron carbide a strong candidate ing cooling, the difference in deformation, due to the for several high technology applications. Due to its low different thermal expansion factors of the components density and superior hardness, boron carbide is a very is accommodated by creep as long as the temperature romising material for light-weight ballistic protection. is high enough. Below a certain temperature, which Boron carbide exists as a stable single phase in a large is called the"joining "temperature, the different com- homogeneity range from B4C to B1o 4C [11]. The most ponents become bonded together and internal stresses that hardness of stoichiometric B4C is the highest one component material are leen l么、sm的。e table boron carbide structure is rhombohedral with a appear In each layer, the total strain after sintering is stoichiometry of B13C2, B12 C3, and some other phases the sum of an elastic component and of a thermal com- close to B12C3 [12, 13]. The Vickers hardness of B4 C is ponent [22, 23]. The residual stress of in the range of 32-35 GPa [ 14]. There is an indication perfectly rigid bonding betw in comparison with boron rich or carbon rich boron car- bide compositions [15-17]. However, B4C-based com E1E2f2(ar2-ar1)△T osites have a relatively low fracture toughness of 2.8- (1) 3.3 MPa-m/2[18]. While high hardness is one of the EIfi+ E?f2 very important requisite indicators for a materials bal listic potential, toughness might play an equally impor- tant role in realizing that potential. Thus, materials with E2E1far1-aT2)△T both high hardness and high fracture toughness are ex- E1f1+E22 pected to yield the best ballistic performance [1, 19 Therefore, a significant increase in fracture toughness where E=Ej/(I-vi), fi=2u, f2=2D2 of boron carbide based laminates has the potential for Ej and v are the elastic modulus and Poisson's ratio of realization of improved armor material systems j-th component respectively, II and l2 are the thickness be controlled by designing the distribution of resid- aT2 are the thermal expansion coefficients(CTE)of ual stresses, i.e., placing the layers with high compres- the first and second components respectively, AT is sive stresses into the bulk of the material. The sign and the difference in temperature of joining temperature value of the bulk residual stresses have to be firmly and current temperature, and w is the total thickness of established by theoretical prediction [20]. A signifi- the specimen cant increase in ballistic protection of B. C based lam Equations l and 2 give the residual stresses in layers inates may be achieved by designing high compressive which have an infinitive extent. Far away from the free stresses placed into the bulk of the materials. The goal of surface, the residual stress in the layer is uniform and this research was to develop the design and processing biaxial. In the bulk of layers, the stress perpendicular to of boron carbide-silicon carbide ceramic laminates with the layers is zero. At the free surface of the laminates, controlled residual stresses. In this article we demon- the stresses are different from the bulk stresses Near strate a laminate design concept by determining the the edges, the residual stress state is not biaxial because prospective combination of layers, their geometry and the edges themselves must be traction-free. Highly lo- microstructure for the B, C/B4 C-30 wt%SiC system, as calized stress components perpendicular to the layer well as a laminates manufacturing route. The appar- plane exist near the free surface and it decreases rapidly ent Klc of three layered composite was measured to be from the surface becoming negligible at a distance ap 7.42+0.82 MPa. m/, but the detailed report on the me- proximately on the order of the layer thickness. These chanical properties, such as Youngs modulus, fracture stresses have a sign opposite to that of the equibiaxial toughness, hardness, and ballistic performance of the stresses deep within the layer. Therefore, if the bulk developed laminates will be presented elsewhere [21]. stress is compressive within the material, the tensile 5484
outer layers near the surface increase strength, flaw tolerance, fatigue strength, fracture toughness and stress corrosion cracking. In the case of symmetrical laminates, this can be done by choosing the layer compositions such that the coefficient of thermal expansion (CTE) in the odd layers is smaller than the CTE of the even ones. The changes in compressive and tensile stresses depend on the mismatch of CTE’s, Young’s moduli, and on the thickness ratio of layers (even/odd). However, if the compressive stresses exist only at or near the surface of ceramics and are not placed inside the material, they will not effectively hinder internal cracks and flaws [9]. Boron carbide is an important ceramic material with many useful physical and chemical properties. After cubic boron nitride, it is the hardest boron containing compound [10]. Its high melting point, high elastic modulus, large neutron capture section, low density, and chemical inertness make boron carbide a strong candidate for several high technology applications. Due to its low density and superior hardness, boron carbide is a very promising material for light-weight ballistic protection. Boron carbide exists as a stable single phase in a large homogeneity range from B4C to B10.4C [11]. The most stable boron carbide structure is rhombohedral with a stoichiometry of B13C2, B12C3, and some other phases close to B12C3 [12, 13]. The Vickers hardness of B4C is in the range of 32–35 GPa [14]. There is an indication that hardness of stoichiometric B4C is the highest one in comparison with boron rich or carbon rich boron carbide compositions [15–17]. However, B4C-based composites have a relatively low fracture toughness of 2.8– 3.3 MPa·m1/2 [18]. While high hardness is one of the very important requisite indicators for a material’s ballistic potential, toughness might play an equally important role in realizing that potential. Thus, materials with both high hardness and high fracture toughness are expected to yield the best ballistic performance [1, 19]. Therefore, a significant increase in fracture toughness of boron carbide based laminates has the potential for realization of improved armor material systems. Brittleness of boron carbide ceramic laminates can be controlled by designing the distribution of residual stresses, i.e., placing the layers with high compressive stresses into the bulk of the material. The sign and value of the bulk residual stresses have to be firmly established by theoretical prediction [20]. A signifi- cant increase in ballistic protection of B4C based laminates may be achieved by designing high compressive stresses placed into the bulk of the materials. The goal of this research was to develop the design and processing of boron carbide-silicon carbide ceramic laminates with controlled residual stresses. In this article we demonstrate a laminate design concept by determining the prospective combination of layers, their geometry and microstructure for the B4C/B4C-30 wt%SiC system, as well as a laminates’ manufacturing route. The apparent KIc of three layered composite was measured to be 7.42±0.82 MPa·m1/2, but the detailed report on the mechanical properties, such as Young’s modulus, fracture toughness, hardness, and ballistic performance of the developed laminates will be presented elsewhere [21]. 2. Thermal residual stresses and its calculation In this work the two-component brittle layered composites with symmetric macrostructure are considered. The layers consisting of different components alternate one after another, but the external layers consist of the same component. Thus, the total number of layers N in such a composite sample is odd. The layers of the first component including two external (top) layers are designated by index 1 (j = 1), and the layers of the second component (internal) are designated by index 2 (j = 2). The number of layers designated by index 1 is (N + 1)/2, and the number of layers designated by index 2 is (N − 1)/2. The layer of each component has some constant thickness, and the layers of same component have identical thickness. There are effective residual stresses in the layers of each component in the layered ceramic composite. During cooling, the difference in deformation, due to the different thermal expansion factors of the components, is accommodated by creep as long as the temperature is high enough. Below a certain temperature, which is called the “joining” temperature, the different components become bonded together and internal stresses appear. In each layer, the total strain after sintering is the sum of an elastic component and of a thermal component [22, 23]. The residual stresses in the case of a perfectly rigid bonding between the layers of a twocomponent material are [7]: σr1 = E 1E 2 f2(αT 2 − αT 1)T E 1 f1 + E 2 f2 (1) and σr2 = E 2E 1 f1(αT 1 − αT 2)T E 1 f1 + E 2 f2 , (2) where E j = E j /(1 − νj), f1 = (N+1)l1 2w , f2 = (N−1)l2 2w , E j and νj are the elastic modulus and Poisson’s ratio of j-th component respectively, l1 and l2 are the thickness of layers of the first and second component, αT 1 and αT 2 are the thermal expansion coefficients (CTE) of the first and second components respectively, T is the difference in temperature of joining temperature and current temperature, and w is the total thickness of the specimen. Equations 1 and 2 give the residual stresses in layers, which have an infinitive extent. Far away from the free surface, the residual stress in the layer is uniform and biaxial. In the bulk of layers, the stress perpendicular to the layers is zero. At the free surface of the laminates, the stresses are different from the bulk stresses. Near the edges, the residual stress state is not biaxial because the edges themselves must be traction-free. Highly localized stress components perpendicular to the layer plane exist near the free surface and it decreases rapidly from the surface becoming negligible at a distance approximately on the order of the layer thickness. These stresses have a sign opposite to that of the equibiaxial stresses deep within the layer. Therefore, if the bulk stress is compressive within the material, the tensile 5484
Compressive layer Compressive layers B.C+30wt%sic Tensile layer e B C+30wtysiC 3 Figure I Schematic presentation of symmetric 3 layered and multilayered composite. stress components appear at or near the free surface of residual stress, and Kc is the intrinsic fracture toughness of a material in the layer. If a condition of a crack growth onsets fulfilled then Ka= Kc-K, is the apparent frac- ture toughness. If o, is compressive, then K, <0 and 3. Laminate design for enhanced Ka increases. The more rI, the more Ka. The more fracture toughness a, the more Ka. The largest value of a crack length in The schematic presentation of symmetric three-layered compressed layer is I1. The maximum apparent frac- and nine-layered composites are shown in Fig. l. The ture toughness can be obtained for such crackUnfor proposed design targeted a fracture toughness increase tunately, small cracks have Ka close to Ke B4C-SiC composites and was based on the prelimi A schematic presentation of factors that affect an ap- nary results both from our work [24-26]and from the parent fracture toughness is shown in Fig.2.Note, that work of others [9, 27-30] the contribution of a residual stress to the maximum ap- In case of non-homogeneous(particularly, layered) parent fracture toughness is K,=Y(,/w)o, 1/,where materials, So-called apparent fracture toughness should be considered. This is the fracture toughness of some .(i/w)is a geometrical factor. The factor Y(/w)//2 effective homogeneous specimen. If we measure frac increases as lI increases(Fig. 2a). The compressive residual stress decreases as li increases. It can be cal- ture toughness in bending, the effective sample param- culated using Equation 1. In addition, the residual stress eters should satisfy the following conditions:(1)the depends on the number of layers in the sample(Fig 2b) specimen has to have the same dimensions as a real The final dependences of Ka onI, for various numbers depth equal to that of the real layered specimen; (3) non-monotonic curves with a maximum that depends under the same loading conditions the specimen has on a number of layers in the laminate. The labels w/5 to demonstrate the same load to fracture as that of the real layered specimen. Under these considerations the w/4, w/3 and w/2 designate the maximum thickne parent fracture toughness is the fracture toughness of the top layer for symmetrical layered structures with 9, 7, 5 and 3 layers, respectively. It can be seen that the calculated from a testing data of the layered sample highest apparent fracture toughness can be obtained for approach does not meet the fracture mechanicA uch an the three-layer specimen. Thus, the study of the layers onsidering this specimen as"homogene ment of taking into account all features of stress dis- tribution near crack tip in layered media, but it is still a useful characteristic allowing an effective contribu- ion of such factors as residual stresses and a material inhomogeneity to be accounted for. The compressive residual stress or in the top lay- ers of a laminate shields natural and artificial cracks in the layer. Therefore, the effective(apparent) frac ture toughness of such a structure increases. The more oppressive residual stress induces, the more shielding occurs. Another important factor that contributes to the 3 layers apparent fracture toughness increase is a crack length a A longer crack promotes more shielding. A maximum length of a transverse crack in a top compressive layer is limited by the layer thickness I1. These two factors de- termine the apparent fracture toughness of the material. In general, a condition of a crack growth onset is Ka+kr=Kc. where Ka= Ka(da, a) is the applied stress intensity factor that can be measured, oa is the w/5w/4w/3w2 distribution of applied stress resulted from bending, Figure 2 Factors affecting laminate design for maximum apparent frac- Kr= k(or, a) is the stress intensity fac a ture toughness
Figure 1 Schematic presentation of symmetric 3 layered and multilayered composite. stress components appear at or near the free surface of a layer. 3. Laminate design for enhanced fracture toughness The schematic presentation of symmetric three-layered and nine-layered composites are shown in Fig. 1. The proposed design targeted a fracture toughness increase of B4C-SiC composites and was based on the preliminary results both from our work [24–26] and from the work of others [9, 27–30]. In case of non-homogeneous (particularly, layered) materials, so-called apparent fracture toughness should be considered. This is the fracture toughness of some effective homogeneous specimen. If we measure fracture toughness in bending, the effective sample parameters should satisfy the following conditions: (1) the specimen has to have the same dimensions as a real layered specimen; (2) the notched sample has a notch depth equal to that of the real layered specimen; (3) under the same loading conditions the specimen has to demonstrate the same load to fracture as that of the real layered specimen. Under these considerations the apparent fracture toughness is the fracture toughness calculated from a testing data of the layered sample considering this specimen as “homogeneous”. Such an approach does not meet the fracture mechanics requirement of taking into account all features of stress distribution near crack tip in layered media, but it is still a useful characteristic allowing an effective contribution of such factors as residual stresses and a material inhomogeneity to be accounted for. The compressive residual stress σr in the top layers of a laminate shields natural and artificial cracks in the layer. Therefore, the effective (apparent) fracture toughness of such a structure increases. The more compressive residual stress induces, the more shielding occurs. Another important factor that contributes to the apparent fracture toughness increase is a crack length a. A longer crack promotes more shielding. A maximum length of a transverse crack in a top compressive layer is limited by the layer thickness l1. These two factors determine the apparent fracture toughness of the material. In general, a condition of a crack growth onset is Ka + Kr = Kc, where Ka = Ka(σa, a) is the applied stress intensity factor that can be measured, σa is the distribution of applied stress resulted from bending, Kr = Kr(σr, a) is the stress intensity factor due to a residual stress, and Kc is the intrinsic fracture toughness of a material in the layer. If a condition of a crack growth onset is fulfilled then Ka = Kc−Kr is the apparent fracture toughness. If σr is compressive, then Kr < 0 and Ka increases. The more |σr|, the more Ka..The more a, the more Ka..The largest value of a crack length in compressed layer is l1. The maximum apparent fracture toughness can be obtained for such crack. Unfortunately, small cracks have Ka close to Kc. A schematic presentation of factors that affect an apparent fracture toughness is shown in Fig. 2. Note, that the contribution of a residual stress to the maximum apparent fracture toughness is Kr = Y (l1/w)σrl 1/2 1 , where Y (l1/w) is a geometrical factor. The factor Y (l1/w)l 1/2 1 increases as l1 increases (Fig. 2a). The compressive residual stress decreases as l1 increases. It can be calculated using Equation 1. In addition, the residual stress depends on the number of layers in the sample (Fig. 2b). The final dependences of Ka on l1 for various numbers of layers are shown in Fig. 2c. These dependences are non-monotonic curves with a maximum that depends on a number of layers in the laminate. The labels w/5, w/4, w/3 and w/2 designate the maximum thickness of the top layer for symmetrical layered structures with 9, 7, 5 and 3 layers, respectively. It can be seen that the highest apparent fracture toughness can be obtained for the three-layer specimen. Thus, the study of the layers’ Figure 2 Factors affecting laminate design for maximum apparent fracture toughness. 5485
relative thickness and layers numbers reveals that the ple. The maximum possible apparent fracture tough maxImum crad ielding will be achieved for three- ness of the corresponding layered structure is also de layer composites with an edge crack extending to the termined in all iterations as an indicative parameter of first interface. However, the multilayered design is also the design. The determination of the apparent Klc uses very important to meet specific ballistic requirements. the compressive residual stress and the thickness of a It is essential during impact loading to have more bar- top layer as a crack length at any given iteration. These riers to arrest cracks. In our case it is the number of two parameters( the compressive residual stress and the nly one such barrier that is a top compressive layer. site directions. A decrease in the top layer thickness can compressive layers For a three-layer design there is thickness of the top layer)have trends acting in opp The top layer plays a key role for the projectile defeat, increase the residual stress in the layer, but it decreases however multilayered design is of further importance the length of the maximum crack. Therefore, the max- to stop cracks more effectively. Therefore, we designed imum apparent fracture toughness was always used to and manufactured both three-layered and nine-layered analyze the correct thickness ratio composites in this work. The input parameters of laminate design are the co- 4. Processing of laminates efficients of thermal expansion, Youngs moduli, Pois- The material systems selected for the proposed study sons ratios, and densities of the constituents of lay- were B4C and BC-30 wt% SiCbecause of theirpromise ered composite. A very important but experimentally for ballistic applications [31-33]. Table I shows therel unknown input parameter is also AT-a" tem- evant material properties used in the design calculations perature. The output parameters are layers thickness (compiled from literature), and Tables II and Ill show and composition. The step-by-step design technique to the corresponding calculated residual stresses in the obtain the enhanced fracture toughness of a layered B4C/B, C-30 wt%SiC laminates. The maximum possi- omposite is as follows ble apparent fracture toughness for corresponding lay ered structures is also presented in the Tables Il and Ill 1. The compositions of layers are selected depend- The layers under tensile stress have higher CTE, and in ing on a future application of the composite. Then, the this case they are B4C layers. The layers under com- relevant material constants entering the design are de- pressive stress have lower CTE; here they are B. C- termined 30 wt%SiC layers. A temperature T= 2150C 2. The effective coefficients of thermal expansion, used for the majority of the calculations, when resid an effective Young's modulus, an average density and ual stresses appeared in the layers upon cooling from a thickness ratio of layers are determined using the rule the hot pressing temperature. There is no liquid ph of mixture present during the sintering of B4 C/B4 C-Sic ceram- 3. The next step of design is the selection of the ics [34], therefore, the hot pressing temperature wa er's number It can be any appropriate number de- used as a "joining "temperature AT for calculations nding on the required total thickness of the tile. It should be noted that all laminates were designed in To obtain the enhanced fracture resistance of layered such a way that the tensile stresses had been maintained composite, the factors affecting the apparent fracture at low values toughness should be taken into account. Usually, the B4C and a-SiC powders with a grain size of thickness of the thinnest possible layer is limited by 2-5 um were used for laminates manufacturing. BC the manufacturing technology. Note that a compressive 30 wt% Sic mixtures were made by ball milling the layer should be thin enough to reach high level of resid- respective powders in acetone in a polyethylene bot- ual stress tle using b4c milling media 48 h. The laminates were 4. The ratio of tensile and compressed layer thick- produced via rolling of tapes followed by hot pressing ness(thickness ratio)is determined. Any appropriate The formation of a thin ceramic layer is of specific im- thickness ratio can be used as a first approximation portance, as the sizes of residual stress zones(tensile 5. Tensile layer thickness is found and compressive)are directly connected to the thick- 6. The calculation of residual stresses is fulfilled us- ness of layers. The advantage of rolling, as a method of ing(1)and(2). The total thickness of the sample is also green layers production, are that it allows easy thick determined at this step for a given layer's thickness ratio ness control, achieves high green density of the tapes, aking into account the selected number of layers and requires a rather low amount of solvent and organic 7. The thickness ratio is changed after analysis of the additives as compared to other methods like tape cast- residual stress and the total thickness of the specimen. ing [35]. Additional powder refinement, giving a higher Note that increasing ratio of tensile layer thickness to sintering reactivity, might occur due to large forces ap- oppressive layer thickness decreases tensile residual plied in the pressing zone during rolling. The model stress. However, it can result in increasing total thick- ing of rolling was recently performed that potentially ness of sample. TABLE I Properties of ceramics used in the stress calculation changing thickness ratio, the calculation is re- Such iterations are continued to find a unique Composition E(Gpa) Poissons ratio CTE(10-6K-) layer thickness ratio that produces the maxi ossible compressive residual stress, low tensile sic 483 0.17 5.5 411 0.16 residual stress, and required total thickness of the sam- 5486
relative thickness and layers’ numbers reveals that the maximum crack shielding will be achieved for threelayer composites with an edge crack extending to the first interface. However, the multilayered design is also very important to meet specific ballistic requirements. It is essential during impact loading to have more barriers to arrest cracks. In our case it is the number of compressive layers. For a three-layer design there is only one such barrier that is a top compressive layer. The top layer plays a key role for the projectile defeat, however multilayered design is of further importance to stop cracks more effectively. Therefore, we designed and manufactured both three-layered and nine-layered composites in this work. The input parameters of laminate design are the coefficients of thermal expansion, Young’s moduli, Poisson’s ratios, and densities of the constituents of layered composite. A very important but experimentally unknown input parameter is also T – a “joining” temperature. The output parameters are layers thickness and composition. The step-by-step design technique to obtain the enhanced fracture toughness of a layered composite is as follows: 1. The compositions of layers are selected depending on a future application of the composite. Then, the relevant material constants entering the design are determined. 2. The effective coefficients of thermal expansion, an effective Young’s modulus, an average density and a thickness ratio of layers are determined using the rule of mixture. 3. The next step of design is the selection of the layer’s number. It can be any appropriate number depending on the required total thickness of the tile. To obtain the enhanced fracture resistance of layered composite, the factors affecting the apparent fracture toughness should be taken into account. Usually, the thickness of the thinnest possible layer is limited by the manufacturing technology. Note that a compressive layer should be thin enough to reach high level of residual stress. 4. The ratio of tensile and compressed layer thickness (thickness ratio) is determined. Any appropriate thickness ratio can be used as a first approximation. 5. Tensile layer thickness is found. 6. The calculation of residual stresses is fulfilled using (1) and (2). The total thickness of the sample is also determined at this step for a given layer’s thickness ratio taking into account the selected number of layers. 7. The thickness ratio is changed after analysis of the residual stress and the total thickness of the specimen. Note that increasing ratio of tensile layer thickness to compressive layer thickness decreases tensile residual stress. However, it can result in increasing total thickness of sample. After changing thickness ratio, the calculation is repeated. Such iterations are continued to find a unique optimal layer thickness ratio that produces the maximum possible compressive residual stress, low tensile residual stress, and required total thickness of the sample. The maximum possible apparent fracture toughness of the corresponding layered structure is also determined in all iterations as an indicative parameter of the design. The determination of the apparent KIc uses the compressive residual stress and the thickness of a top layer as a crack length at any given iteration. These two parameters (the compressive residual stress and the thickness of the top layer) have trends acting in opposite directions. A decrease in the top layer thickness can increase the residual stress in the layer, but it decreases the length of the maximum crack. Therefore, the maximum apparent fracture toughness was always used to analyze the correct thickness ratio. 4. Processing of laminates The material systems selected for the proposed study were B4Cand B4C-30 wt%SiC because of their promise for ballistic applications [31–33]. Table I shows the relevant material properties used in the design calculations (compiled from literature), and Tables II and III show the corresponding calculated residual stresses in the B4C/B4C-30 wt%SiC laminates. The maximum possible apparent fracture toughness for corresponding layered structures is also presented in the Tables II and III. The layers under tensile stress have higher CTE, and in this case they are B4C layers. The layers under compressive stress have lower CTE; here they are B4C- 30 wt%SiC layers. A temperature T = 2150◦C was used for the majority of the calculations, when residual stresses appeared in the layers upon cooling from the hot pressing temperature. There is no liquid phase present during the sintering of B4C/B4C-SiC ceramics [34], therefore, the hot pressing temperature was used as a “joining” temperature T for calculations. It should be noted that all laminates were designed in such a way that the tensile stresses had been maintained at low values. B4C and α-SiC powders with a grain size of 2–5 µm were used for laminates manufacturing. B4C- 30 wt%SiC mixtures were made by ball milling the respective powders in acetone in a polyethylene bottle using B4C milling media 48 h. The laminates were produced via rolling of tapes followed by hot pressing. The formation of a thin ceramic layer is of specific importance, as the sizes of residual stress zones (tensile and compressive) are directly connected to the thickness of layers. The advantage of rolling, as a method of green layers production, are that it allows easy thickness control, achieves high green density of the tapes, and requires a rather low amount of solvent and organic additives as compared to other methods like tape casting [35]. Additional powder refinement, giving a higher sintering reactivity, might occur due to large forces applied in the pressing zone during rolling. The modeling of rolling was recently performed that potentially TABLE I Properties of ceramics used in the stress calculation Composition E (Gpa) Poisson’s ratio CTE (10−6 K−1) B4C 483 0.17 5.5 SiC 411 0.16 3 5486
TABLE II Three layered composite design. A total thickness of a tile- 10.5 mm Thickness of Layers (um) Apparent Composition B4C-30wt%SiC (MPa) FIens(MPa) le(MPam/2 B4C-30wt %SiC/B4C 32 TABLE III Nine layered composite design. A total thickness of a tile-10.35 mm Thickness of Layers(um) B4C-30wt%SiC a comp(MPa) Ptens(MPa) Kle(MPam/-) B4C-30wt%SiC/B4C 2250 TABLE IV Some properties of the powders and green tapes after rolling Additive density Poured density Poured density Relative Thickness Comp f the powder f the granulas density 0. 0.095 0.45 B4C-30wt %SiC/B4C 1.5 0.186 0.14 allows optimizing the process of roll compaction [36]. Powders are continuously supplied in the bunker and In our case there is a challenging problem to produce furtherinto the deformation zone in between rolls. Pow- thin tapes with a small amount of plasticizer and suffi- ders are supplied to the deformation zone due to both cient strength and elasticity to handle green layers after the gravitational force and friction between rolls and rolling Crude rubber(1-3 wt%)has to be added to the powders. The relative density of the tape (pr)can be mixture of powders as a plasticizer through a 3% solu- calculated from tion in petrol. Then the powders were dried up to the 2 wt% residual amount of petrol in the mixture. After C2r sieving powders with a 500 um sieve, granulated pow ders were dried up to the 0.5 wt% residual petrol. The schematic presentation of rolling is shown in Fig 3. where p, is a relative powder density, i is a drawing coefficient, a is an intake angle, and R is a roll diame- ter. A roll mill with 40 mm rolls was used for rolling The velocity of rolling was in the range of 1-1. 2 m/min Working pressure was varied from O I ton/cm- for rela- tive density of tapes 64%to l ton/cm- for 74% density propertie rolling ed in Table iv Samples of ceramics were prepared by hot press ing of the rolled tapes stacked together. The hot press- ing conditions were as follows:(a) a heating rate was 100C/min;(b)a hot pressing temperature was kept at 2150oC during hot pressing of a majority of the tiles, Bunker. 2. Powders 3 Rolls 4. Transmission. 5. Motor. 6 Bottom support 7. Tape and some hot pressing was done at 2200.C to ensure that fully dense materials were obtained; (c)a pressure B4C·30wt% SiC tapes B was kept at the level of 30 MPa; and(d)a dwell time at hot pressing temperature was 50-60 min Graphite dies were used for the hot pressing of laminates with graphite surfaces coated by B layer in order to prevent a direct contact between graphite and ceramic material. 90 x 90 x 10 mm tiles were produced as a result of hot pressing. Dense(97-100% of density)laminate sam- ples were obtained 100mm 5. Microstructure of laminates During hot pressing of laminates the shri and B,- 30 wt%SiC rolled tapes. The thickness of an individual tape individual layers occurred, and their thickness become after rolling is between 0.4-0.5 mm. 0. 15 mm after hot pressing. The interfaces between
TABLE II Three layered composite design. A total thickness of a tile – 10.5 mm Thickness of Layers (µm) Apparent Composition B4C-30wt%SiC B4C σcomp (MPa) σtens (MPa) KIc (MPam1/2) B4C-30wt%SiC/B4C 900 8700 632 131 44 TABLE III Nine layered composite design. A total thickness of a tile – 10.35 mm Thickness of Layers (µm) Apparent Composition B4C-30wt%SiC B4C σcomp (MPa) σtens (MPa) KIc (MPam1/2) B4C-30wt%SiC/B4C 150 2250 662 99 32 TABLE IV Some properties of the powders and green tapes after rolling Green Tape Additive density Poured density Poured density Relative Thickness Composition d50 (µm) (g/cm3) of the powder of the granulas density (mm) B4C 2.5 2.52 0.111 0.095 0.71 0.45 B4C-30wt%SiC/B4C 1.5 2.69 0.186 0.146 0.74 0.47 allows optimizing the process of roll compaction [36]. In our case there is a challenging problem to produce thin tapes with a small amount of plasticizer and suffi- cient strength and elasticity to handle green layers after rolling. Crude rubber (1–3 wt%) has to be added to the mixture of powders as a plasticizer through a 3% solution in petrol. Then the powders were dried up to the 2 wt% residual amount of petrol in the mixture. After sieving powders with a 500 µm sieve, granulated powders were dried up to the 0.5 wt% residual petrol. The schematic presentation of rolling is shown in Fig. 3. Figure 3 (A) Schematic presentation of rolling. (B) Photograph of B4C and B4C-30 wt%SiC rolled tapes. The thickness of an individual tape after rolling is between 0.4–0.5 mm. Powders are continuously supplied in the bunker and further into the deformation zone in between rolls. Powders are supplied to the deformation zone due to both the gravitational force and friction between rolls and powders. The relative density of the tape (ρr) can be calculated from ρr = ρp λ 1 + α2R hs (3) where ρp is a relative powder density, λ is a drawing coefficient, α is an intake angle, and R is a roll diameter. A roll mill with 40 mm rolls was used for rolling. The velocity of rolling was in the range of 1–1.2 m/min. Working pressure was varied from 0.1 ton/cm2 for relative density of tapes 64% to 1 ton/cm2 for 74% density. The properties of the powders and green tapes after rolling are presented in Table IV. Samples of ceramics were prepared by hot pressing of the rolled tapes stacked together. The hot pressing conditions were as follows: (a) a heating rate was 100◦C/min; (b) a hot pressing temperature was kept at 2150◦C during hot pressing of a majority of the tiles, and some hot pressing was done at 2200◦C to ensure that fully dense materials were obtained; (c) a pressure was kept at the level of 30 MPa; and (d) a dwell time at hot pressing temperature was 50–60 min. Graphite dies were used for the hot pressing of laminates with graphite surfaces coated by BN layer in order to prevent a direct contact between graphite and ceramic material. 90 × 90 × 10 mm tiles were produced as a result of hot pressing. Dense (97–100% of density) laminate samples were obtained. 5. Microstructure of laminates During hot pressing of laminates the shrinkage of the individual layers occurred, and their thickness become 0.15 mm after hot pressing. The interfaces between 5487