ISSN:0272-8397    

DOI:10.1002/pc.750130402

出版商:Society of Plastics Engineers    

年代:1992

数据来源: WILEY

摘要:

AbstractOrthotropic, fiber‐reinforced composite laminae under biaxial plane stress loading conditions are considered. A new strain‐based strength criterion for such laminae is proposed. This failure criterion, which is cast in parametric form, is completely general. A parametric equation that is expressed as a Fourier series is proposed for describing the failure surface for the laminae. This equation is shown to be valid for most fiber‐reinforced composite laminae through comparisons with experimental data taken from the literature. The advantages of the failure theory are discussed in comparison with the quadratic form of the tensor polynomial theory. Also presented is a biaxial apparatus for testing cruciform specimens under biaxial loading conditions. Efforts under way to characterize composite laminae and laminates under complex stress states are disc

ISSN:0272-8397    

DOI:10.1002/pc.750130403

出版商:Society of Plastics Engineers    

年代:1992

数据来源: WILEY

摘要:

AbstractThe Tsai‐Wu Quadratic Failure Criterion provides satisfactory strength predictions for fiber‐reinforced composite materials but requires five experimental tests to determine the strength parameters. This paper presents two modifications of this criterion, which employ micromechanics to determine these parameters. Experiments on uniaxial and multiaxial vinyl ester/fiberglass composite laminates show that the first modified failure criterion, which is based solely on fiber and resin properties, predicts strength within an average absolute error of 25.4% in comparison to the experimentally determined strength. The addition of a single longitudinal tensile test to the modified expression (second modified failure criterion) reduces the average absolute error to 15.9%. This compares well with the Tsai‐Wu Failure Criterion, which gives an average absolute error of 9.4%. The proposed modified criteria are shown to provide satisfactory failure predictions, while greatly reducing the amount of testing req

ISSN:0272-8397    

DOI:10.1002/pc.750130404

出版商:Society of Plastics Engineers    

年代:1992

数据来源: WILEY

摘要:

AbstractThe large deformation elastic response of a plane woven Kevlar fabric is investigated analytically and experimentally. The analysis assumes the undeformed geometry to be a sequence of interlaced arcs of circles that reverse at each yarn midpoint, and each yarn is modeled as an extensible elastica subject to certain compatibility conditions. Deflection‐force relations for the fabric are determined in terms of the initial weave geometry and the elastic properties of the individual yarns. The theoretical results agree well with the results of experiments performed on a fabric woven from 400 denier Kevlar yarns under conditions of uniaxial loading in both warp and fill direction

ISSN:0272-8397    

DOI:10.1002/pc.750130405

出版商:Society of Plastics Engineers    

年代:1992

数据来源: WILEY

摘要:

AbstractVoids or bubbles in polymer composites have a detrimental effect on the properties of the material. Such voids are found in PMR 15/graphite fiber (fabric type) composites. The matrix, PMR 15, is formed in‐situ during the molding process and undergoes two major stages of reactions, viz., imidization and crosslinking. Volatiles are formed as by‐products of these reactions and may remain trapped in the matrix to form voids or bubbles. Photomicrographs of composite cross sections suggested the classification of voids as two different phases in two different directions, each aligned with the fibers, resulting in a five phase model. The composites were tested in three‐point bending tests to obtain the flexural modulus. By combining Weng's inclusion approach (1) and Ishikawa and Chou's crimp method (2), an analytical model has been developed to predict the elastic moduli of such a multiphase composite material. Thus, the voids have been treated as inclusions in an enclosing matrix, which in turn is reinforced by woven fibers in two directions. The stiffness behavior obtained analytically was within 10% of the experimental v

ISSN:0272-8397    

DOI:10.1002/pc.750130406

出版商:Society of Plastics Engineers    

年代:1992

数据来源: WILEY

摘要:

AbstractRandom glass mat thermoplastic composites (GMT), which can be thermostamped to form complex deep‐drawn parts with ribs and boxes, are complex material systems in which the local elastic modulus and local strength vary widely and randomly across the material (the tensile modulus can vary by a factor of two over a 12.7‐mm length scale). And the values of these local properties depend on the length scale of measurement. The random, large‐scale point‐to‐point variations in their properties cannot be described by a single number. The broad distribution of elastic moduli in GMT has been modeled by a four‐parameter probability density function. Moments of this distribution function provide numerical measures that can be used for comparing data sets representing properties of different material samples. This statistical characterization is used to establish the consistency and the random nature of previously obtained elastic moduli data sets. The framework is also used to predict the effect of the gage length used to measure the local elastic modulus on the shape of the modulus probability densi

ISSN:0272-8397    

DOI:10.1002/pc.750130407

出版商:Society of Plastics Engineers    

年代:1992

数据来源: WILEY

摘要:

AbstractRandom glass mat reinforced composites have large variations in mechanical properties, requiring a statistical characterization of their elastic properties. The broad distribution of elastic moduli is modeled by a probability density function that determines the statistical characteristics of the material property data. This characterization is used to develop a framework for predicting the stiffness of parts made of these composites. The effective elastic modulus and the structural stiffness of parts have been defined in a statistical sense, and the main properties of the effective moduli have been discussed. The effective elastic modulus, if it exists, is shown to be a system property that must be considered in conjunction with the geometry of the part and the boundary conditions.

ISSN:0272-8397    

DOI:10.1002/pc.750130408

出版商:Society of Plastics Engineers    

年代:1992

数据来源: WILEY

摘要:

AbstractThis paper sets out the theory and numerical methods used to simulate filling and fiber orientation is simple injection moldings (a film‐gated strip and a center‐gated disk). Our simulation applies to these simple geometry problems for the flow of a generalized Newtonian fluid where the velocities can be solved independently of fiber orientation. This simplification is valid when the orientation is so flat that the fibers do not contribute to the gapwise shear stresses. A finite difference solution calculates the temperature and velocity fields along the flow direction and through the thickness of the part, and fiber orientation is then integrated numerically along pathlines. Fiber orientation is three‐dimensional, using a second‐rank tensor representation of the orientation distribution function. The assumptions used to develop the simulation are not valid near the flow front, where the recirculating fountain flow complicates the problem. We present a numerrical scheme that includes the effect of the fountain flow on temperature and fiber orientation near the flow front. The simulation predicts that the orientation will vary through the thickness of the part, causing the molding to appear layered. The outer “skin” layer is predicted only if the effects of the fountain flow and heat transfer are included in the

ISSN:0272-8397    

DOI:10.1002/pc.750130409

出版商:Society of Plastics Engineers    

年代:1992

数据来源: WILEY

摘要:

AbstractExperimental measurements of fiber orientation are reported for two parts injection molded from nylon 6/6 reinforced with 43 wt% of glass fibers. The parts are a center‐gated disk and a film‐gated strip. Orientation is measured from polished cross sections and reported as a function of position, both across the thickness and in the flow direction. Both parts have a layered structure, with outer shell layers of flow‐aligned fibers surrounding a central core of either random‐in‐plane (strip) or transversely aligned fibers (disk). The disk also has surface skins with less alignment. The experiments are compared with predictions of the simulation presented in Part I. The simulation predicts the presence, nature, and location of the layers very well. However, it overpredicts the small out‐of‐plane fiber orientation and places the core‐shell transition too close to the midplane. A comparison with selected experimental results suggests that the major source of error is the closure approximation used by the fiber orientation equation. The simulation is exercised for a variety of cases to show the importance of material and process parameters. A polymer matrix with a small power‐law index or a large heat of fusion gives a thicker core and is less likely to have a skin. Injection time is an important parameter, but injection temperature and mold temperature have little effect on

ISSN:0272-8397    

DOI:10.1002/pc.750130410

出版商:Society of Plastics Engineers    

年代:1992

数据来源: WILEY

ISSN:0272-8397    

DOI:10.1002/pc.750130401

出版商:Society of Plastics Engineers    

年代:1992

数据来源: WILEY