Search NASA⌕ Search

Engineering topics

Hahn, H. T.

Publications and source records attributed to Hahn, H. T..

17 records
All sourcesNASA NTRSDOE OSTI
NASA NTRS1994

Structural design criteria for filament-wound composite shells

Advanced composite cylinders, manufactured by filament winding, provide a cost effective solution to many present structural applications; however, the compressive performance of filament-wound cylinders is lower than comparable shells fabricated from unidirectional tape. The objective of this study was to determine the cause of this reduction in thin filament-wound cylinders by relating the manufacturing procedures to the quality of the cylinder and to its compressive performance. The experiments on cylinder buckling were complemented by eigenvalue buckling analysis using a detailed geometric model in a finite element analysis. The applicability of classical buckling analyses was also investigated as a design tool.

Hahn, H. T.↗
NASA NTRS1992

Interfacial stress state present in a 'thin-slice' fibre push-out test

An analysis of the stress distributions along the fiber-matrix interface in a 'thin-slice' fiber push-out test is presented for selected test geometries. For the small specimen thicknesses often required to displace large-diameter fibers with high interfacial shear strengths, finite element analysis indicates that large bending stresses may be present. The magnitude of these stresses and their spatial distribution can be very sensitive to the test configuration. For certain test geometries, the specimen configuration itself may alter the interfacial failure process from one which initiates due to a maximum in shear stress near the top surface adjacent to the indentor, to one which involves mixed mode crack growth up from the bottom surface and/or yielding within the matrix near the interface.

Kallas, M. N.↗
NASA NTRS1990

The effect of weak interface on transverse properties of a ceramic matrix composite

Experimental studies conducted at NASA Lewis on SiC reaction-bonded Si3N4 composite system showed that transverse stiffness and strength were much lower than those predicted from existing analytical models based on good interfacial bonding. It was believed that weakened interfaces were responsible for the decrease in tranverse properties. To support this claim, a two-dimensional FEM analysis was performed for a transverse representative volume element. Specifically, the effect of fiber/matrix displacement compatibility at the interface was studied under both tensile and compressive transverse loadings. Interface debonding was represented using active gap elements connecting the fiber and matrix. The analyses show that the transverse tensile strength and stiffness are best predicted when a debonded interface is assumed for the composite. In fact, the measured properties can be predicted by simply replacing the fibers by voids. Thus, it is found that little or no interfacial bonding exists in the composite, and that an elastic analysis can predict the transverse stiffness and strength.

Shimansky, R. A.↗
NASA NTRS1986

Compression failure mechanisms of composite structures

An experimental and analytical study was conducted to delineate the compression failure mechanisms of composite structures. The present report summarizes further results on kink band formation in unidirectional composites. In order to assess the compressive strengths and failure modes of fibers them selves, a fiber bundle was embedded in epoxy casting and tested in compression. A total of six different fibers were used together with two resins of different stiffnesses. The failure of highly anisotropic fibers such as Kevlar 49 and P-75 graphite was due to kinking of fibrils. However, the remaining fibers--T300 and T700 graphite, E-glass, and alumina--failed by localized microbuckling. Compressive strengths of the latter group of fibers were not fully utilized in their respective composite. In addition, acoustic emission monitoring revealed that fiber-matrix debonding did not occur gradually but suddenly at final failure. The kink band formation in unidirectional composites under compression was studied analytically and through microscopy. The material combinations selected include seven graphite/epoxy composites, two graphite/thermoplastic resin composites, one Kevlar 49/epoxy composite and one S-glass/epoxy composite.

Hahn, H. T.↗
NASA NTRS1986

The effect of resin toughness and modulus on compressive failure modes of quasi-isotropic graphite/epoxy laminates

Compressive failure mechanisms in quasi-isotropic graphite/epoxy laminates were characterized for both unnotched and notched specimens and also following damage by impact. Two types of fibers (Thornel 300 and 700) and four resin systems (Narmco 5208, American Cyanamid BP907, and Union Carbide 4901/MDA and 4901/mPDA) were studied. For all material combinations, failure of unnotched specimens was initiated by kinking of fibers in the 0-degree plies. A major difference was observed, however, in the mode of failure propagation after the 0-degree ply failure. The strength of quasi-isotropic laminates in general increased with increasing resin tensile modulus. The laminates made with Thornel 700 fibers exhibited slightly lower compressive strengths than did the laminates made with Thornel 300 fibers. The notch sensitivity as measured by the hole strength was lowest for the BP907 resin and highest for the 5208 resin. For the materials studied, however, the type of fiber had no effect on the notch sensitivity.

Sohi, M. M.↗
NASA NTRS1986

Compression failure mechanisms in unidirectional composites

Compression failure mechanisms in unidirectional composites were examined. Possible failure modes of constituent materials are summarized and analytical models for fiber microbuckling are reviewed from a unified viewpoint. Due to deficiencies in available models, a failure model based on nonlinear properties and initial fiber curvature is proposed. The effect of constituent properties on composite compression behavior was experimentally investigated using two different graphite fibers and four different epoxy resins. The predominant microscopic scale failure mode was found to be shear crippling. In a soft resin, shear crippling was in the form of buckling of fibers on a microscopic scale. However, stiff resins failure was characterized by the formation of a kink band. For unidirectional laminates, compressive strength, and compressive modulus to a less extent, were found to increase with increasing magnitude of resin modulus. The change in compressive strength with resin modulus was predicted using the proposed nonlinear model.

Hahn, H. T.↗
NASA NTRS1986

Buckling of a fiber bundle embedded in epoxy

Buckling of a fiber bundle embedded in epoxy resin was studied to gain insight into compressive failure mechanisms in unidirectional composites. The fibers used were E-glass, T300 graphite, T700 graphite, and P75 graphite. These fibers were combined with two different resins: Epon 815/V140 and Epon 828/Z. In both resins the failure mode of the bundle was found to be microbuckling of fibers for the first three types of fibers; however, the high-modulus P75 fibers failed in shear without any sign of microbuckling. The strains at which microbuckling occurred were higher than the compressive failure strains of the corresponding unidirectional composites. In the soft resin, Epon 815/V140, fibers buckled at lower strains than in the stiff resin, Epon 828/Z. The buckling strains and the segment lengths followed the trends predicted for a single filament embedded in an infinite matrix.

Hahn, H. T.↗
NASA NTRS1986

Kink band formation in unidirectional composites under compression

The kink band formation in unidirectional composites under compression was studied with the help of optical and scanning electron microscopy. Materials studied included graphite/epoxy, S-2 glass/epoxy, and Kevlar 49/epoxy composites. The kink bands in the graphite/epoxy and glass/epoxy composites were initiated by fiber microbuckling while those in the Kevlar/epoxy composite were triggered by microkinking in the fibers themselves. Fracture surfaces of the graphite/epoxy composites showed fibers broken in bending and fiber segments tilted in the direction of slip. The kink band formation depended not only on the constituent material properties but also on the kinematic constraints imposed. Both primary and secondary kink bands were observed in the graphite/epoxy specimens, but only primary kink bands were present in the other two composites because of higher lateral flexibility of the specimen during testing.

Sohi, M. M.↗
NASA NTRS1984

Effects of Constituent Properties on Compression Failure Mechanisms

Compression failure mechanisms were investigated through the analysis of matrix-embedded fiber bundles. The use of fiber bundle specimens can provide much needed information on compression failure mechanisms because failure of the bundle is well contained and can be monitored during testing. The method can clearly distinguish between buckling-induced failure and shear-induced failure. The results indicate that WY and T300 graphite fibers and E-glass fiber fail in buckling while the high-modulus P75S graphite fiber fails in shear. Buckling-induced failure becomes more evident with low-modulus fiber in softer epoxy.

Hahn, H. T.↗
NASA NTRS1984

Compression failure mechanisms in unidirectional composites

Compression failure mechanisms in unidirectional composites were examined. Possible failure modes of constituent materials are summarized and analytical models for fiber microbuckling are reviewed from a unified viewpoint. Due to deficiencies in available models, a failure model based on nonlinear properties and initial fiber curvature is proposed. The effect of constituent properties on composite compression behavior was experimentally investigated using two different graphite fibers and four different epoxy resins. The predominant microscopic scale failure mode was found to be shear crippling. In a soft resin, shear crippling was in the form of buckling of fibers on a microscopic scale. However, stiff resins failure was characterized by the formation of a kink band. For unidirectional laminates, compressive strength, and compressive modulus to a less extent, were found to increase with increasing magnitude of resin modulus. The change in compressive strength with resin modulus was predicted using the proposed nonlinear model.

Hahn, H. T.↗
NASA NTRS1983

Failure mechanisms

Data, charts, and graphs relating to longitudinal tension, crack growth modes in unidirectional composites, strength predictions, and compression are given.

Hahn, H. T.↗
NASA NTRS1983

Residual stresses and their effects in composite laminates

Residual stresses in composite laminates are caused by the anisotropy in expansional properties of constituent unidirectional plies. The effect of these residual stresses on dimensional stability is studied through the warping of unsymmetric (0 sub 4/90 sub 4)sub T graphite/epoxy laminates while their effect on ply failure is analyzed for (0/90)sub 2s Kevlar 49/epoxy laminate. The classical laminated plate theory is used to predict the warping of small and large panels. The change of warping does not indicate a noticeable stress relaxation at 75 C while it is very sensitive to moisture content and hence to environment. A prolonged gellation at the initial cure temperature reduces residual stresses while postcure does not. The matrix/interface cracking in dry (0/90)sub 2s Kevlar 49/epoxy laminate is shown to be the result of the residual stress exceeding the transverse strength.

Hahn, H. T.↗
NASA NTRS1982

Modelling of dimensional stability of fiber reinforced composite materials

Various methods of predicting the expansion and diffusion properties of composite laminates are reviewed. The prediction equations for continuous fiber composites can be applied to SMC composites as the effective fiber aspect ratio in the latter is large enough. The effect of hygrothermal expansion on the dimensional stability of composite laminates was demonstrated through the warping of unsymmetric graphite/epoxy laminates. The warping is very sensitive to the size of the panel, and to the moisture content which is in turn sensitive to the relative humidity in the environment. Thus, any long term creep test must be carried out in a humidity-controlled environment. Environmental effects in SMC composites and bulk polyester were studied under seven different environments. The SMC composites chosen are SMC-R25, SMC-R40, and SMC-R65.

Hahn, H. T.↗
NASA NTRS1981

Warping of unsymmetric cross-ply graphite/epoxy laminates

Warping in unsymmetric graphite/epoxy laminates was studied with particular attention given to the change of residual stresses resulting from long term environmental exposure. Square, cured prepreg sheets were measured for edge deflection with a cathetometer, then quartered and remeasured. Two postcuring durations were then used, 7.5 and one hr at 177 C; varying cooldown rates after curing were used for other samples, and one set was stored in vacuum at 75 C. Maximum deflections and weight changes were measured periodically at room temperature. Average curvatures, the effect of postcure, and the effect of long-term exposure were determined. Larger panels exhibited cylindrical warping and smaller panels underwent anticlastic warping. The deflections were related to weight changes, i.e. moisture absorption, and the lower the moisture content, the higher the deflection. Relaxation of residual stresses at 75 C was neglibible after 220 days.

Hahn, H. T.↗
NASA NTRS1976

Residual stresses in polymer matrix composite laminates

Residual stresses in composites are induced during fabrication and by environmental exposure. The theory formulated can describe the shrinkage commonly observed after a thermal expansion test. Comparison between the analysis and experimental data for laminates of various material systems indicates that the residual stress-free temperature can be lower than the curing temperature, depending on the curing process. Effects of residual stresses on ply failure including the acoustic emission characteristics are discussed.

Hahn, H. T.↗