Compressor Vane Clocking Effects on Embedded Rotor Performance

Abstract

VANE clocking is the circumferential indexing of adjacent vane rows with similar vane counts. The clocking configuration can be arranged so that either all downstream vanes are positioned in the wakes shed from the upstream vanes, or the upstream vane wakes pass through every passage of the downstream vane row without any interaction with the vane surface. Researchers have shown that vane clocking can impact discrete frequency noise, the unsteady forces acting on the embedded rotor, and stage efficiency. Stage efficiency, in general, can be affected by changes to stator loss, rotor loss, or work done on the flow by the rotor. Researchers tend to focus on how the downstream stator loss is affected by vane clocking, but the role, if any, the rotor flowfield plays in vane clocking effects on stage efficiency has not been fully investigated. Many clocking studies have been performed in turbine environments, but fewer compressor clocking studies have provided conclusive results because the more common low-speed compressor research facilities generate a small pressure rise per stage, leaving the even smaller changes associated with vane clocking indiscernible. One of the few vane clocking studies that report a measurable and repeatable vane clocking effect on stage efficiency has been performed in the intermediate-speed Purdue University three-stage axial compressor research facility [1,2]. With similar vane counts for the IGV, Stator 1, and Stator 2, the effects of vane clocking were isolated to the embedded stage by clocking Stator 1 with respect to Stator 2. The objective of this Note is to investigate how vane clocking affects rotor loss and work done on the flow. This is accomplished with both steady total temperature measurements and unsteady tangential flow angle measurements. Two loading conditions were investigated: design point loading and a high loading condition. Vane clocking effects were more than three times stronger at high loading than at design loading. In addition, the larger potential field associated with the high loading condition would make any vane clocking effects on rotor performance, if present, more obvious. Three spanwise locations of the flowfield were interrogated: 30, 50, and 70%span. Thesewere chosen because at design loading, vane clocking effects were strongest at 70% span. At high loading, the hub variations were out-of-phase with the tip variations because of the significant skew in the Stator 1 wakes by the time they convected to the Stator 2 leading edge plane. Therefore, the clocking trends at 30% span were opposite of the trends at 70% span, and thus, interrogating these spanwise locations would allow for this type of trend to be identified in the rotor flowfield data, if present. The experiments were performed in a compressor featuring geometry representative of the rear stages of a high-pressure compressor, with engine representativeMach numbers and Reynolds numbers [3]. The two clocking configurations in this Note, CL3 and CL6, represent offsets of 32 and 83%vane passage (vp), respectively. These include the configuration leading to wake impingement on the downstream vane row, and the other, half a passage out of phase with the first, allows the wakes to pass through downstream passages. For these studies, the relative position of the IGV and Stator 1 was held constant, as was the relative position of Stator 2 and Stator 3. All results are acquired at 100% corrected design speed, 5000 rpm. To determine the effects of vane clocking on stage 2 total temperature ratio, measurements were acquired at the exit of Stator 1 and Stator 2 with seven-element Kiel head total temperature rakes. The thermocouple sensor and thermocouple-grade extension wire was designated as “special limits of error,” and for type T thermocouples, the manufacturer specified uncertainty was 0.5 °C. In an effort to reduce this value, a temperature bath was used to verify that the uncertainty for this group of wires was actually only 0.1 °C for the temperature range of interest. A miniature cross-film sensor was used to acquire instantaneous flow angle and velocity information in the axial-tangential plane. The uncertainty in the tangential flow angle measurements acquired with the cross-film sensors are 1 deg . All time-accurate data were acquired at a sampling frequency of 120 kHz (Rotor 2 blade pass frequency is 2.75 kHz at design speed), with 200 revolutions of data used to characterize an ensemble-averaged (EA) revolution.