【正文】 h the catalyst content of the vacuum degassed batch (batch 6) and its control (batch 5) were essentially the same, the water content of the vacuum degassed material was approximately 25% less. Hexcel39。s results for highpressure liquid chromatography (HPLC) testsconducted on the finished prepreg are shown in Table 4, along with the initial mix positions for the major epoxy (TGMDAtetraglycidyl methylenedianiline), the curing agent (DDS4,439。diaminodiphenyl sulfone) and minor epoxy no. 1 (novolac). Since HPLC gives the free or unreacted quantities for TGMDA, DDS and the novolac, the remaining amounts of these quantities are bined or reacted, during neat resin cooking and subsequent prepregging operations. Therefore, the percent change from the initial position to that after prepregging should indicate the amount of reaction occurring during the mixing and Table High Pressure Liquid Chromatography Results Batch Number 2 (I) 3 4 5 (2) 6 Variation Low Normal High High Normal Vacuum Flow Flow RC Flow Flow Degas BF3 COli/em (WI. %) 0 Major Epoxy ? Initial % ? Prepreg % ? % Decrease Curing Agent ? Initial % ? Prepreg % ? % Decrease Minor Epoxy ? Initial % ? Prepreg % ? % Decrease + + (1) Control for Batches 1,3 and 4 (2) Control for Batch 6 228 prepregging operations. All batches showed decreasing amounts of unreacted TGMDA and DDS for the prepreg samples. The data for the novolac epoxy did not show as clear a trend. However, HPLC is not as sensitive a test for the novolac as it is for the TGMDA and DDS. Doubling the BF3 catalyst content (batch 1) had the greatest effect. The greater catalyst content, along with the longer processing time for this batch, resulted in larger reductions in both the TGMDA major epoxy and DDS curing agent. These data suggest that the BF3 catalyst significantly accelerates the curing reaction, even at the relatively moderate temperatures used during the mixing and prepregging operations. Thermal Properties Both dynamic and isothermal DSC scans were performed on neat resin samples. Dynamic scans were made at three different heating rates (2, 5and 10176。C mirr+). In addition, scans were made at 5176。C minI on resin samples aged for 30 days at room temperature. The dynamic DSC results are summarized in Table 5 and representative scans (at 5 176。C minI) are shown in Fig. 3 for the high, low and normal flow resins. The normal flow 35016 dynamic DSC curve exhibits two distinct exotherm peaks. The first or minor peak is a direct result of the BF3 catalyst. Table Dynamic DSC Test Results Second exotherm (0C) Enthalphy (cal/grn) Second exotherm (0C) Enthalphy (caUgm) Second exotherm (0C) Enthalphy (caUgm) Scan Rate of 5 Crmin Second exotherm (0C) 207 Enthalphy (caUgm) 97 (1) Control for Batches 1 and 4 Prepreg Aged for 30 Days at RT prior To testing 229 The Effect of Chemical Composition and Processing on Carbon/Epoxy Laminate Quality: A binaison of Effects 170 190 210 Ttmperiture C 230 270 290 Fig. 3. Dynamic DSC Scan Comparison The highflow batch with no catalyst exhibited only the single large exothermic peak characteristic of resins lacking a BF 3 catalyst, such as Hexcel39。s 3502 and Cytec39。s batch also had much higher total heat of reaction, /)JJ. Again, this was a function of the absence of the catalyst, since the catalyst reacts rapidly at low temperatures, thereby reducing the total heat of reaction. On the other hand, for the lowflow batch with twice the normal catalyst content, the first exothermic peak (usually the minor) was larger than the second peak. When DSC scans were run on the prepreg material, the second peak again became the predominate peak. Since the BF3 catalyst reacts vigorously at relatively low temperatures, the heat from the prepregging operation probably resulted in enough reaction to suppress the minor exothermic peak. Further evidence of the importance of the BF3 catalyst can be seen by examining the heat of reaction data. The total heat of reaction for the lowflow batch (% BF3) was lower than the normal flow batch ( % BF3)39。 while that of the highflow batch (0% BF3) was much higher. In addition, the batch with no catalyst was scanrate independent, while all batches with BF3 catalyst exhibited a decrease in the total heat of reaction with increasing scan rates. Rheological Properties Dynamic viscosity measurements were performed at 1,2 and 5 176。C minI on each batch of neat resin using RDS 7700. In addition to viscosity vs. time curves, flow numbers: Table Neat Resin Dynamic Viscosity Test Results (1 176。C/min) Balch Number Varialion BF3 Content (WI. %) Minimum Viscosity (poise) Temperature at Minimum Viscosity (OC) Gel Temperature at 1,000 poise eC) Flow Number (min/poise) (1) Control for Batches land 4 were calculated from each curve. Flow is the reciprocal of viscosity when integrated as a function of time between the starting time (to) and the time to gellation (tgel). The test results for the viscosity tests at a heating rate of 1 176。C mirr39。 are summarized in Table 6. The dramatic effect of the BF3 catalyst was again evident. The low flow system (batch 1) exhibited a flow number of approximately onehalf that of the normalflow material (batch 2) and an order of magnitude less than that of the highflow system (batch 4).The rest of the viscosity data further emphasize the effect of the BF3 catalyst. For example, the minimum viscosity for the highflow material (batch 4) was much lower than that of the lowflow material (batch 1). The gel temperature also showed the effect of varying the