The printable version is no longer supported and may have rendering errors. Please update your browser bookmarks and please use the default browser print function instead.

Vortex ropes in draft tube of a laboratory Kaplan hydro turbine at low load

Application Area 6: Turbomachinery Internal Flow

Application Challenge AC6-15

Test Data

Description of Experiment

The experiments were performed in an open aerodynamic test rig (Fig. 1a). Air supply into the draft tube model was provided by a vortex air blower MT 08-M1S-7.5 with a maximum flow rate ${{\mathcal {Q}}_{max}}$ = 550 m³/h and pressure excess of ΔP = 0.4 bar. The air flowrate was controlled by an ultrasonic flowmeter "IRVIS-Ultra" and frequency converter of the blower with feedback. For capturing the LDA signal the air flow was seeded with tracers produced by a standard-type generator of the paraffin oil aerosol with a mean particle diameter of 1-3 μm.

Figure 1 provides a photo and a sketch of the experimental turbine model. The airflow with tracers was supplied to an axisymmetric chamber through six radial inlet channels. To suppress the influence of turning the radial inflow into axial flow and to align the airflow entering the guide vanes, two honeycombs and a profiled nozzle were applied. Air flows through the cascade of blades and enters the draft tube cone. The draft-tube model and a pair of blade rows "guide vanes - runner" were manufactured using a 3D printing technology. Servo drive SPSh10-3410 ensured precise setting of the rotation frequency of the runner in the range from 0 to 3000 min-1. The control of the experiment was ensured by a computer. Using the original software it was possible to maintain the present flow regime with an accuracy of 1.5% and 0.5% for the flowrate and the runner rotation frequency respectively.

The experiments included the measurements of the velocity fields in the conical part of the draft tube using a two-component LDA system LAD-06i, whereas pressure pulsations were recorded by acoustic sensors mounted on the cone walls (Fig. 2). The pressure fluctuations were recorded using a Bruel&Kjaer analyzer and a sound-level meter Type 2250 complemented by a microphone Type 4189 with high output characteristics, a frequency range of 6.3 Hz - 20 kHz, a dynamic range 14.6 - 146 dB and a sensitivity of 50 mV/Pa. The microphone head was set half-way along the draft tube cone mounted flush with the wall (45 mm below the swirler), Fig. 2.


Figure 2: Locations of the LDA measuring cross section and of the microphone mounted in the casing. Dimensions are given in mm.

The output signal from the microphone unit was digitized by an analog-to-digital converter L-card E-440 (Litvinov, Shtork, Kuibin, Alekseenko, & Hanjalic, 2013). The experiments were conducted at a fixed runner rotation frequency f=2432 min-1 and a flow rate changeable in the range from Q=65 m3/h to 216 m3/h with a step of 0.4 m3/h. At each operation regime, the signal detected by the acoustic sensor was digitized for 5 seconds with a sampling rate of 2 kHz, which ensured sufficient resolution of the typical 0.2 Hz frequency for calculating the pressure spectra (note that the typical PVC (precessing vortec core) frequency was above 15 Hz).

Profiles of the averaged velocity and the rms (root mean square) velocity fluctuations of the flow were measured using the two-component laser Doppler anemometer "LAD-06i" operated in the back scattering mode. The LDA measurements of the axial and tangential fluid velocity components were performed at one cross-section in the draft-tube cone located at 3 mm below the cowl of the swirler, Fig. 2. The LDA optical head with focal distance 0.5 m formed a measuring volume with size 0.05x0.05x1 mm. To eliminate the directional ambiguity, the frequency shift based on a Bragg cell was applied. The LDA system was mounted on an automated x-y-z traversing system for the adjustment of the measurement position with a precision of 0.1 mm for all axes. A PC based software controlling both the LDA unit and the traversing system was employed to run the measurement process in fully automated mode. The number of velocity samples for each measuring point was not less than 5 thousand bursts ensuring the uncertainty in the LDA velocity measurements of the mean velocity and the rms velocity fluctuations to be not larger than 5% and 2%, respectively.