Detecting Resin Pre-Gelation in Hydro Generator Stator Bar Insulation
Excerpt from Carolyn Payne's (GE Energy) work published paper from the 28th International Thermal Conductivity Conference
Thermal effusivity is evaluated as a method for detection of resin pre-gelation in generator stator bar insulation. If the resin within the insulation gels before the system is pressurized, the resin does not flow and wet out the mica. This creates a dry porous insulation. Dry, porous insulation not onlya has a different density than compact, well-impregnated insulation, but its thermal conductivity and specific heat should also be different. Therefore theoretically, thermal effusivity should pick up the difference between pre-gelled and properly cured insulation.
The modern stator bar (fibure 1) consists of insulated Roebelled copper strands, over wrapped with an epoxy resin-rich mica tape. Tape application and curing are the main factors affecting the performance of the insulation system. A solid, void free ground wall insulation that meets stringent electrical, mechanical and thermal specficiations is required.
Picture Above: Stator Bar
Currently the only foolproof way to determine whether there is resin pre-gelation is disection. Disection can be done on selective bars; however because of the variance in temperature throughout the autoclave, picking the correct candidates for disection is not guaranteed.
Picture Above: Cross-Secion of Dissected Pre-Gelled Stator Bar
Thermal effusivity is being explored as a non-destructive Quality Control (QC) method to detect resin pre-gelation in stator bars before the bars leave the manufacturing site.
Experiments were carried out in the laboratory with the insulating tape. The tape was exposed to four different vacuum temperatures for five hours. These samples were given the same cure. Several tests, including thermal effusivity, were performed to determine the effects of the vacuum cycle on the properties of the cured laminates and whether the thermal effusivitycould indeed differentiate between pre-gelled and compact insulation.
Results & Discussion
|Resin Flow (%)||11.0||11.7||9.8||0|
|DSC Tg (◦C)||133||141||138||154|
|Young's Modulus (Mpsi)||8.7||7.6||8.4||0.4|
|Three Point Breaking Load (lbf)||345||338||360||53|
|Thermal Effusivity (Ws½/m²K)||911||859||898||711|
There was no resin flow for the samples with the five-hour vacuum cycle at 125◦C, indicating resin pre-gelation had occurred before the pressurized cycle took place. Samples which experienced 95◦C, 105◦C and 115◦C vacuum temperatures, had excellent resin flow. They produced compact laminates with higher three-point breaking loads and higher Young's modulus. Their thermal effusivity values were 17 to 22% higher than the laminates produced with the 125◦C vacuum cycle.
Thermal effusivity has the potential of detecting a pre-gelled stator bar in order to optimize production quality.
To learn more about how C-Therm's TCi Thermal Analyzer is applied for non-destructive testing visit here.