|
Cycle Test Evaluation of Various Polyester Resins and a Mathematical
Model for Projecting Flexural Fatigue Endurance.
Introduction
An analysis of the static physical properties of a general purpose orthophthalic, corrosion resistant isophthalic, or
vinyl ester resin does not immediately reveal why a particular type of resin should be chosen. In fact, economics and laminate durability often tend to be overriding factors in resin selection. We believe our data
comparing unsaturated polyester laminating resins by chemical type, static physical properties and dynamic physical properties is useful information for the marine laminate designer and enables them to make a more
informed decision about which resin type is most suitable.
A boat hull flexes millions of times in its use, and laminates which are more capable of resisting fatigue will prove themselves superior in durability. Further, the superior resin may
allow for reduced thickness and weight in the laminate and composite. Conversely, the marine laminate that is less capable of flexural fatigue will deteriorate faster and must to be heavier to approach the
capabilities of superior resins. Our data demonstrates the superior performance of CoREZYN brand vinyl ester resins for suitability in boat design and building. A mathematical model also is shown fitting the fatigue
data to a curve.
 |
  Orthophthalic laminating resins Fig. 1, isophthalic Fig. 1, dicyclopentadiene and vinyl ester resins Fig. 1 are used in marine laminate construction. Comparisons of standard physical testing results Fig. 2 reveals there are adequate strengths in each system. A graphic presentation of static testing is depicted at Fig. 3a and b  . This pictorial example of "toughness" (integral of area under the stress-strain curve) illustrates that the polymer type is relevant to desired structural properties.
Industry trends toward lighter composites require us to consider cyclic testing for fatigue durability. In conjunction with the
United States Testing Company, Inc., we completed a series of laminate evaluations in "Flexural Fatigue Endurance Limit" tests. The samples were evaluated in accordance
with the procedures outlined in ASTM Test Method D671, "Constant Stress Fatigue Testing." Type A  specimens were used in this evaluation Fig. 4. A constant amplitude of force was
applied to the specimen in a fixed cantilever type testing machine and the number of cycles to failure were recorded. Evans correctly emphasizes the requirement for
choosing physical test procedures relevant to the intended application. We compared fatigue resistance in similar laminates changing over the resin type. We also
recommend reviewing Loveless, et al, Hofer, et al, and Knoebel, as they have presented excellent data and arguments for accelerated fatigue testing to compare to expected performance.
Preperation of Composites
 The laminates were designed as represented in Fig. 5. The orthophthalic laminating
resin, CoREZYN 1097-121S; the isophthalic laminating resin, CoREZYN 9503; and the bisphenol-A epichlorohydrin epoxy-based vinyl ester, CoREZYN VE8300 (Fig. 1), are
all commercially available resins used in marine laminations. We also compared the CoREZYN VE8119 pre-promoted, thixotropic resin to a "standard" CoREZYN VE8300
vinyl ester to ensure that their physical properties have not deteriorated as a result of formula modifications.
 The ASTM D671 construction was accomplished as in Fig. 6 using type-A flexural test coupons described in Fig. 4 (contrasting to flexural test coupons used in ASTM
D790). We constructed the laminate by alternating layers of 0.75-ounce mat with 24-ounce woven roving, beginning and ending with the 0.75-ounce mat. The resin-to-glass ratio was held at 75:25.
TEST Results by ASTM D671
  A graphic representation of the months of continuous testing
conducted at United States Testing, Inc., is portrayed in Figs. 7a, b and c. This summarized data is reported in U.S. Testing
Report 88110-2. To be useful, this data has to be further analyzed to compare strength at one cycle (static testing of this laminate  design) and the results in psi at specific numbers of
cycles. We show 8,500; 10,000; 11,500 and 13,500 psi.
We applied the data in the standard formula for exponential curve fit:
An obvious use of the results would be to graph the projected cycles-to-failure using the applied stress in psi as the other  variable Fig. 8. A graphic
example of this shows the results at 8,500 psi becoming increasingly obvious. As one changes from orthophthalic to isophthalic to vinyl ester, so increases the projected
cycles-to-failure in comparison to applied stress.
A Model for Data Projection
The ASTM D671 flexural fatigue test results provide useful information on the relative
ability of these four resins to resist developing cracks or general mechanical deterioration due to repeated
stress and strain. Rather than fit a curve by eye on semi-logarithmic graph paper, an exponential curve fit can be calculated where:
where:
x = psi of the applied stress level
y = number of cycles to failure
rē = coefficient of determination
(0 is poor and 1.00 is a good fit of the data of the curve).
Then the following would occur:
From further calculation, we can derive the projected failure due to cyclic flexural testing.
Calculated ASTM D-671 Results from Curve Fit Equation:
Other Corollory Considerations
Aoki, et al suggests that the permeation of distilled water into stressed laminates was faster and that the type
of resin, as well as the corrosion resistant capability, determined the percentage of flexural strength retention
over time. Amoco and Eastman would further suggest that there is a correlation between corrosion resistance of a marine laminating resin and the cosmetic/structural expectations of a boat laminate. At Interplastic
Corporation, we know there is a direct correlation between resin type and the percentage of water absorption; and resin type and the onset of gel coat blistering at elevated temperature.
Conclusion
Cyclic flexural testing of specific polyester resin types resulted in predictable data that was oriented by
polymer description, i.e., orthophthalic was exceeded by isophthalic, and both were vastly exceeded by vinyl
ester resins. Little difference was observed between the standard vinyl ester and the pre-promoted, thixotropic vinyl ester.
A mathematical model is useful to compare cycle test results at an exact loading. The regression analysis used
was above the 95% confidence limit in three cases and above 90% confidence limits in the fourth. These results should allow the marine engineer to design higher strength, lower weight laminates with a reduction or
elimination of gel coat blisters in the boat hull.
References
Revised 1/01 from Cycle Test Evaluation of Various Polyester Types and a Mathematical Model for Projecting
Flexural Fatigue Endurance, by P. Burrell, T. McCabe and R. de la Rosa. The original work, including full references, may be obtained from Interplastic Corporation.
|
