In an age in which many composite processes and products benefit from some form of computer assistance, sizing development remains primarily an experimental, laboratory-driven pursuit. New sizings are formulated and tested. Test results are evaluated. Adjustments are made. More tests follow.
"The industry has grown up on empirically knowing what works," explains Mike Gunther, president of Paradox Technical Consultants (Parker, Colo.), noting that it nevertheless has advanced size chemistry to a very sophisticated status. "You have to give the empirical guys a ton of credit."
That said, computer modeling of the complex interface between glass and resin, now more precisely called the "interphase region," is in development — although such work is still almost exclusively an academic venture.
Empirical development persists among fiber manufacturers, in part, because there is little demand for continuous improvement in existing formulations. If a new formulation can improve performance only by a few percent, says fiber/sizing consultant Bob Schweizer (Granville, Ohio), the industry isn't going to buy it. "It's too expensive for our customers to requalify the new formulation, then requalify with their customers." Developmental work in established applications is typically prompted only by innovations in conversion/fabrication technologies that necessitate a different solution. "The industry is going to bulk handling and pneumatic conveyance for chopped strands," Schweizer illustrates, "so size chemists are seeking to produce better strand integrity and dry handling characteristics without negatively affecting composite properties or glass dispersion."
Improvements in an end product's long-term properties also propel some sizing developments, reports Phil Schell, R&D director for Saint-Gobain Vetrotex (Wichita Falls, Texas). "Our technology is close to optimum for the end product's initial mechanical properties," he asserts. "So the next thing to work on, besides improved processing, is optimizing fatigue performance, water resistance and other long-term performance characteristics." Such efforts are especially challenging, he notes, because a detrimental impact on initial mechanical properties might accompany the improved long-term performance.
Gunther believes that these and other future advances will depend on computer modeling. One such area of research, he says, is to "learn more about the actual dynamics at the exact instant when the strand meets the sizing applicator." With a better understanding of the interphase region, he contends, size chemistry will advance to the point that most of the sizing is functional when handling requirements or adhesion requirements predominate, compared to current formulations in which the handling components are deliberately designed to dissolve during wetout (see "Sizing Up Fiber Sizings," p. 18).
Although glass manufacturers have not yet taken up such theoretical research themselves, most now maintain close relationships with universities involved in advance molecular modeling. Johns Manville, for example, has relationships with several universities in Europe and the U.S., D'Silva points out, and the company already takes advantage of their computer modeling expertise. "If we encounter a problem," he says, "we ask them to look at the molecular dynamics and identify which component is causing the problem." And while glass manufacturers focus on innovations expected to enter the commercial market in a three- to five-year window, D'Silva notes, "Our relationships allow us to leverage the university laboratory efforts on more fundamental research issues."