OEMs recommend periodic inspection of pinned finger turbine blade attachments for detection of service-induced damage. Some designs require removal of the pinned finger blades for inspection of the blade fingers and the mating disk finger attachment. This article provides an example where Structural Integrity detected cracking on one of the disk finger attachments and provided an engineering assessment to support continued operation and to identify a re-inspection interval for the LP rotor. Pinned finger attachments are commonly used to secure last-stage (L-0) or next-to-last stage (L-1) blades in low-pressure (LP) turbines. This approach could be applied to other pinned finger blade attachments to determine suitability for service.
Structural Integrity Associates (SI) performed inspections of an LP rotor from a 380MW unit. Inspections included fluorescent magnetic particle inspection of the exposed L-1 and L-0 finger attachments on the turbine-end and generator-end rows, as well as phased array ultrasonic inspections of L-2 and L-3 tangential entry dovetail attachments. Those inspections identified numerous shallow indications on the L-1 attachments, with engineering condition assessment requested for one notable indication on the generator end.
The one indication of note was identified on the L-1 generator end, circumferentially oriented at the base of the step below the outermost pinhole for two adjacent buckets. The indication is 1 inch long and is apparent on both the admission and discharge side of finger #4, therefore the indication is assumed to have propagated through the finger (approximately 0.2 inch thick). A photo of the indication is shown in Figure 1.
Figure 1. Photo of indication on disk finger #4, admission side of generator-end L-1 stage.
A proposed modification to the pin in the vicinity of the crack is a reduction in the pin diameter only at the location of disk finger #4. The nominal pin diameter would be maintained at the adjacent blade-side fingers and all other disk-side fingers. The pins for the affected adjacent buckets, bound the current circumferential extent of the crack, and the objective of the modified or recessed pin geometry is to reduce radial loading adjacent to the crack location. We evaluated effects of the modified pin on applied stresses in the vicinity of the crack.
We created a finite element model and performed stress evaluations to determine operating stresses in the L-1 stage finger pin attachments. Stress evaluations were also performed to determine the impact on operating stresses of cracking detected on disk finger and proposed pin modifications.
Stress analyses were performed to evaluate the effect of the crack in the finger on redistribution of stresses to pin holes at adjacent blades. We considered two crack sizes for analysis. The as-found crack length (1 inch) was explicitly included in an FE model comprising 6 blade widths. A second flaw was modeled having a conservative projected flaw length after 8 years operation. Separate models were created to evaluate the effects on finger ligament stresses for the nominal top pin case and the proposed pin modification, i.e. reduced diameter top pin for the two blades adjacent to the as-found crack and nominal top pins for the other blades. Stresses due to disk RPM and blade loading were evaluated separately, with the respective FE models illustrated in Figure 2 and Figure 3.
Figure 3. Finite element model for cracked L-1 disk finger showing boundary conditions and applied loads due to blade loading.
Stress results for the nominal pin geometry showed excessively high radial stresses on path 1 in the crack tip region for both crack models (not shown here). For the reduced top pin case, radial stresses versus distance along thickness transitions (paths 1-3) and at pin holes (paths 4-6) for the as-found flaw (flaw 1) are shown in Figure 4 for paths 1-3 and Figure 5 for paths 4-6. The corresponding stress distributions for the 8-year projected flaw (flaw 2) are shown in Figure 6 and Figure 7.
Figure 4. Radial stress vs. distance along thickness transitions, reduced top pin, as-found crack (flaw 1).
Figure 5. Radial stress vs. distance at pin holes, reduced top pin, as-found crack (flaw 1).
Figure 6. Radial stress vs. distance along thickness transitions, reduced top pin, 8-year projected crack (flaw 2).
Figure 7. Radial stress vs. distance at pin holes, reduced top pin, 8-year projected crack (flaw 2).
Finite element stress analyses show that stresses in the disk remain within acceptable levels with the use of the recessed pin. Analysis of the detected crack on the generator-end L-1 disk finger at its current size (1 inch long) and conservative 8-year projected crack demonstrate that stresses remain within acceptable limits. Critical crack lengths were also calculated with consideration for overspeed operation. The results of these analyses show that the use of a modified pin geometry could be used to alleviate stresses in the vicinity of the crack tips without raising stresses to unacceptable levels elsewhere in the disk, demonstrating that the detected crack produces little risk of attachment failure for an 8-year operating period from the unit’s return to service following the inspections.