Studies on FRP composites with modified Epoxies : Structural and mechanical evaluation of syntactic foam sandwich composites

Abstract

Epoxy syntactic foams made of glass microballoons, by virtue of their superior structural and dielectric characteristics, have proved to be efficient core materials in sandwich composites especially for microwave transparent applications in the aircraft industry. These composites comprise of thin FRP (Fibre reinforced plastic) skins bonded to relatively thick foam core resulting in a light weight structure of greatly enhanced flexural rigidity. While the superior mechanical properties of syntactic foams are distinctly advantageous, those epoxy foams which employ aromatic diamine hardeners for curing may suffer from poor crack resistance under impact load, due to the inherent brittleness of the cured matrix. This leads to a reduced damage tolerance of those sandwich composites which employ them as the core. Any attempt to improve the impact toughness of this special class of composite sandwich panels either through structural or material modifications should ensure no deterioration in the dielectric characteristics too. The present investigation is aimed at improving energy absorbing properties of a typical syntactic foam sandwich composite used in microwave transparent applications,through suitable modifications in the epoxy binder of the foam core. This will be the first of its kind to report on the subject of foam core modification by liquid rubbers. The specific objectives , in particular, are a. To fabricate three-phase syntactic foams of resin volume fractions varying between 0.04 to 0.20 made from glass microballoons and epoxy resin binder with and without liquid rubber modifier and study their mechanical properties, b. To fabricate foam sandwich panels with glass/epoxy skins as facings and syntactic foams (modified and unmodified) of typical density values as the core and evaluate them for tensile, flexural and impact characteristics in order to assess the degree of improvement in the desired properties, c. To check the conformity of experimentally obtained mechanical property values of sandwich panels with those values obtained from theoretical design equations and validate them, and d. To propose suitable model equations for estimating tensile and shear modulus values of three-phase syntactic foams by extending available models based on particulate filled composites and evaluate their validity. Two sets of epoxy syntactic foams one with 5 parts by weight of liquid rubber modifier ATBN (Amine Terminated Poly Butadiene-co-Acrylonitrile) in the binder formulation and the other without the modifier were fabricated using Bisphenol-A based epoxy resin and aromatic diamine hardener (Araldite LY556 and Diaminodiphenyl sulphone). Glass microspheres of bulk density 0.18 gm/cc were used for the purpose. Foam samples of resin volume fraction varying between 0.038 to 0.20 (corresponding to a density variation of 0.23 to 0.43 gm/cc) were fabricated for study. The volume fraction of glass microspheres was found to be nearly constant in the foam samples irrespective of variation in the densities. All the samples were tested for compressive and flexural properties (strength and modulus). Sandwich composite panels with glass/epoxy facings were fabricated from woven E-glass fabric and epoxy matrix of the same base system as the foam core (i.e Araldite LY 556 and Diaminodiphenyl sulphone) but without any modifier. Both modified and unmodified syntactic foam cores of density values 0.25, 0.30 and 0.35 g/cc were employed for the fabrication of sandwich, by wet layup and vacuum bagging process. The thickness of the facing was 0.6 mm and that of foam was approximately 6mm. Scanning Electron Micrographs (SEM) were recorded for the fractured surfaces of the foam core after mechanical testing wherever necessary. The compressive as well as flexural strength and modulus values of the foams were found to increase with increasing density in both the modified and unmodified foams. It was observed that at lower initial densities (values less than 0.3 gm/cc) the rubber modified foams, contrary to expectations, showed improved compressive strength and modulus values in comparison to the unmodified foam samples of corresponding density values. In the density range above 0.3 gm/cc there was a reduction in the values as anticipated. The observation could be explained based on the microstructure of the foam samples. Both the flexural strength and modulus values of the modified foams were lowerthan the corresponding unmodified foams. Experimentally determined tensile strength values of sandwich panels varied marginally, affected neither by density nor by rubber modification of foam core. The modification of the core in sandwich samples had resulted in a significant reduction, to an extent of 10 -12% in the tensile modulus values in comparison to the unmodified ones. A notable observation was that an increase in the foam density did not result in enhancing the modulus values of the panels. Instead, the modulus values showed a marginally decreasing trend. SEM examination revealed that the tensile deformation of the foam core is dominated more by the binder than the microballoons which could explain these observations. The percentage elongation of the modified samples was more in comparison to the unmodified ones. In the three-point flexural tests, modified beams with 0.3 and 0.35 gm/cc foam core, showed increased absorption of total/ maximum energy and also increased values of maximum displacement resulting in an enhancement of fracture toughness. The effects were maximum in the beams with 0.30 gm/cc foam density wherein the total energy absorbed was almost doubled. In these beams there was a reduction in both the failure energy and failure displacement but an enhanced residual energy and post-failure displacement. This clearly demonstrated the improvement in the damage tolerant properties of the sandwich beams. On the other hand in the beams with 0.25 gm/cc foam core, liquid rubber modification increased the failure energy and displacement resulting in a deterioration of post-failure behaviour. The various failure modes of syntactic foam sandwich samples were studied. SEM pictures revealed that the glass microballoons could absorb energy by crushing along with the matrix cracking. In general, there was a reasonably good agreement between the flexural rigidity values determined experimentally and calculated theoretically from the sandwich design equations using the material properties of the constituents. The experimental and theoretical values of flexural rigidity indicated that there was no significant reduction in the flexural rigidity of the beams due to liquid rubber modification of the foam core. This leads to the inference that improved toughness characteristics were achieved without sacrificing flexural rigidity of the beams. The shear rigidity values estimated theoretically from the experimental data showed an increase for the modified panels. Instrumented impact test results of the panel samples indicated that the post peak load and deflections were larger for the modified panels the maximum being for the 0.30 gm/cc foam panel. The damage area of the panels estimated through C-Scan experiments were larger for the modified panels the maximum being for the 0.30 gm/cc foam panel. This indicated that a larger volume of material was participating in energy absorption in the modified panels for similar impact energies. Though the normalised peakload values were lower for the modified panels with decrease being maximum in the 0.30 gm/cc foam core panel, the reduction in peak load energy values were not as significant. The various modes of fracture in the damaged zone were identified and presented. Shear and tensile modulus values of three-phase syntactic foams were estimated through a modeling approach based on particulate composite models of two-phase syntactic foams available in the literature, with suitable modifications and extensions. The estimations were compared with experimental values of the investigation. While there is a good agreement for the tensile modulus values, shear modulus value predictions require further improvement. Corrections based on imperfect adhesion is suggested for the proposed models for closer agreement of tensile modulus values. The suitability of the sandwich panels for microwave transmission was evaluated through attenuation measurements in the 1-12 GHz region, which conclusively proved the effectiveness of the material modification approach adopted in the investigation, in fulfilling the objectives.