Analyzing composite properties by rheological testing.
Thermoplastic and thermosetting resins are used in conjunction with a large variety of reinforcements and manufacturing technologies to produce composites offering wide versatility in design, manufacturing, and functionality. Typical reinforcements include graphite (carbon fibers), fiberglass, and aramid fibers. In continuous form, these fibers are available in a variety of formats, ranging from unidirectional tapes to carbon tows and fiberglass rovings, to monofilament bundles, to single or multilayer woven mats (also available in a number of weaves, such as plain, twill, and 8-harness satin).
In a composite, the polymer matrix is responsible for transferring induced mechanical shocks and stresses to the high-strength fibers. The matrix must also strongly bond the fibers and plies together to prevent failure of the system. Epoxies are commonly used for their ability to impregnate fibers and fabrics, and to form a strong interphase. This affords effective distribution of the stresses from the matrix to the high-modulus fibers. Epoxies are also characterized by high heat and chemical resistance; however, they are susceptible to moisture.
Woven fabrics are one of the more commonly used reinforcing forms. Their advantages include better impact resistance, reinforcement in two directions with one fabric layer, and increased formability. Generally supplied as pre-pregs, the fabric can have many types of fibers, thicknesses, and weave patterns. The composition of a weave consists of warp yarns, which lie in the lengthwise or machine direction (0|degrees~), and fill (or weft) yarns, which lie at right angles to the warp yarns. Fabrics having heavy warp yarns with fine fill yarns are considered unidirectional fabrics because of their contribution to mechanical properties in the warp direction.
The total fabric weight, strength, and stability are dependent upon the orientation of the warp and fill lines. A plain weave consists of one fill line interacting with one warp line in an immediate over-under repeating pattern. This produces a very stable weave. An 8-harness (or end) satin weave has one fill line under seven and over one warp yarns. This produces a stronger weave because of the high warp yarn count, while the irregular pattern increases its formability.
Dynamic mechanical rheological testing (DMRT) involves a controlled deformation of the material and measurement of the response to that mechanical deformation. Rheological qualities are: G'=in-phase or storage (elastic) modulus (referred to as modulus in this article); G"=out-of-phase or loss (viscous) modulus; and |delta~=tan delta=the loss tangent=G"/G'.
Rheological testing variables include temperature, time, frequency, and strain amplitude of deformation. By providing insight into how the material responds to these variables, DMRT gives a detailed analysis of its viscoelastic behavior.
By use of torsion rectangular geometry (ASTM 4065, Project X-10-157), important thermomechanical characteristics can be generated: modulus as a function of temperature; indications of maximum use temperature; trends for predicting creep and impact behavior; and any of the material's thermal transitions. Compared with ASTM D 648, Distortion Temperature Under Load (DTUL), which produces a single data point, DMRT produces a blueprint of the failure mechanism--crucial to the engineer when selecting a candidate material.
Four composite weaves were evaluated to determine the effects of reinforcement type and orientation:
* Kevlar/epoxy, unidirectional lay-up (Kevlar aramid fiber is a product of the Du Pont Co.);
* fiberglass/epoxy, unidirectional lay-up;
* fiberglass/epoxy, isotropic lay-up;
* graphite/epoxy, isotropic lay-up.
The samples were made by cutting in three directions, 0|degrees~, 45|degrees~, and 90|degrees~, with respect to the warp fibers. For the isotropic systems, a tracer line was incorporated in the warp direction of the top ply during fabrication. Modulus as a function of temperature was correlated to composite structure following established ASTM dynamic mechanical protocols.
The Rheometrics Mechanical Spectrometer, Model System 4, was run in the dynamic mode using torsion rectangular tooling at a heating rate of 3|degrees~C/min from room temperature to generally 20|degrees~C to 40|degrees~C above the glass-transition temperature (Tg). The rate of oscillation was 6.28 radians/sec at a forced, constant strain of 0.04%. All samples were preloaded in 10% tension to allow for thermal expansion during testing.
The fiberglass/epoxy system produced higher modulus values as a function of temperature in each orientation than the Kevlar/epoxy system. The Tgs were slightly higher in the fiberglass/epoxy system because of the higher heat resistance of the fiberglass. While the 90|degrees~ oriented samples showed higher moduli than the 0|degrees~-oriented samples in the Kevlar system and lower in the fiberglass system, these differences in moduli as well as the Tg and type of failure are insignificant.
In the 45|degrees~ orientation, modulus values for both unidirectional systems were more than twice as large as those in the 0|degrees~ and 90|degrees~ orientations. In the Kevlar system, the modulus was constant up to the glass-transition zone, and then decreased sharply through the zone, after which it remained constant through the completion of the temperature ramp. This can be related to an increase in damping force of the fabric at this orientation, as shown by the tan delta curve in Fig. 2.
The 45|degrees~ modulus values exhibited a gradual decline over the temperature range in the fiberglass system; this decline was only slightly steeper in the glass-transition zone. A sharper decrease in tan delta values after the glass-transition zone resulted in a decreased damping force, and subsequent inability to retain consistent modulus values at elevated temperatures.
Because of the isotropic lay-up, modulus values at the various orientations should fall in close proximity. Values were comparable within both the epoxy/fiberglass and epoxy/graphite systems, but only at the 0|degrees~ and 90|degrees~ orientations.
Both systems exhibited similar modulus curves for the 45|degrees~ orientation, showing neither a significant modulus decrease throughout the temperature range, nor a glass-transition zone. This resistance to failure may result from the applied stresses' being effectively distributed by the particular weave orientation in the matrix. The lower initial modulus values indicate a lower concentration of warp fibers in this direction, suggesting that the systems were not perfectly isotropic.
The isotropic and unidirectional epoxy/fiberglass systems were compared to determine whether either system offered a functional advantage. Modulus values were comparable at the 0|degrees~ and 90|degrees~ orientations, but at 45|degrees~, the modulus values for the unidirectional system were more than twice as high as those for the isotropic lay-up. This can be attributed to the single warp yarn direction in the unidirectional system.
Three additional benefits to materials, processing, and design engineers from DMRT are: the ability to define the ceiling-use temperature for functional performance; the property retention index of the construction; and time-temperature superpositioning for long-term prediction of properties.
The temperature corresponding to an arbitrary modulus has been shown to be an excellent indicator of the material's performance at an elevated temperature |M.E. Takemori, SPE ANTEC Tech. Papers, 24, 216 (1978)~. These data are automatically generated by DMRT, eliminating the need for the use of ASTM D 648 DTUL. Ceiling performance temperatures for selected moduli are given in the Table.
TABLE. Ceiling-Use Temperatures, |degrees~C, at a Modulus of 5x10|sup.9~ dynes/|cm.sup.2~. Tg at Tan delta Sample Temp. |delta~ peak peak Unidirectional epoxy/Kevlar 0|degrees~ 167 189 0.272 90|degrees~ 158 197 0.241 45|degrees~ 275+ 187 0.294 Unidirectional epoxy/fiberglass 0|degrees~ 231 199 0.195 90|degrees~ 225 196 0.200 45|degrees~ 275+(*) 205 0.119 Isotropic epoxy/fiberglass 0|degrees~ 275+(**) 191 0.219 90|degrees~ 275+ 190 0.103 45|degrees~ 275+ 199 0.103 Isotropic epoxy/graphite 0|degrees~ 265 217 0.205 90|degrees~ 270 215 0.202 45|degrees~ 320+ 210 0.197 * 190|degrees~C ceiling-use temperature at 5x10|sup.10~ dynes/cm|sup.2~ modulus for 45|degrees~ unidirectional epoxy/fiberglass sample. ** 32|degrees~C ceiling-use temperature at 5x10|sup.10~ dynes/cm|sup.2~ for 0|degrees~ isotropic epoxy/fiberglass sample.
The property retention index (PRI) is an indicator of the retention of physical properties after controlled aging in user-defined environments (ASTM Project X-10-179). PRI's significance is that a complete rheological fingerprint of the candidate material is generated, rather than only one property at a single temperature, which often does not reflect how the material is used in actual service.
Time-temperature superpositioning by DMRT is straightforward. The material's modulus as a function of temperature is first determined. Based on the identified Tg, a series of discrete thermal steps are selected for isothermal frequency sweeps. The results of these relatively fast analyses are then used in the Williams-Landel-Ferry (WLF) equation to generate a master curve for predicting long-term performance over several decades of time and at any reference temperature.
In agreement with previous studies, higher modulus values over the temperature range and at each orientation were observed for the fiberglass/epoxy unidirectional weave system than for the Kevlar/epoxy system. In both the isotropic 8-harness satin weave systems and the unidirectional weave systems, there were no significant differences between properties in the 0|degrees~ (warp) and 90|degrees~ (fill) directions.
Unique properties were observed at the 45|degrees~ orientation. Twofold increases in modulus were obtained in the unidirectional systems. In the isotropic systems, lower initial readings were recorded; however, a high retention of modulus was observed over temperature range.
The author wishes to thank John "Otis" Veilleux of A&M Engineering for providing the materials and John Mackey of USCI for preparing the test specimens; and special thanks to John Macaluso and Professor Stephen Burke Driscoll of the University of Massachusetts-Lowell for their support and assistance throughout this study.
|Printer friendly Cite/link Email Feedback|
|Date:||Dec 1, 1992|
|Previous Article:||Blow molded vehicle bumper beams.|
|Next Article:||Kunststoffe '92 attract 260,000+ to Dusseldorf.|