AirSpeed Calibration Facility by using LDV and A Wind Tunnel at CMS Cheng-Tsair Yang Center for Measurement Standards Industrial Technology Research Institute/Taiwan

Content Selection of Wind Tunnel Calibration of A LDV Measuring Probe Implementation and Characteristics of A Wind Tunnel for Air-Speed Standard Estimation of Uncertainty for Anemometry Calibration

Type of Wind Tunnel Closed Circuit Wind Tunnel (CCWT) directly re-circulate (low Power). Noise is significantly lower. Open Circuit Wind Tunnels (OCWT) Relatively small footprint Low construction cost

Concerns of a wind tunnel Diffuser: to lower the air flow speed, consequently reduce the pressure loss due to friction Fan drive/Blower: provide a pressure increase of flow, to overcome the pressure loss in the tunnel circuit. Steady driving force is mostly required. Test section: provide desirable flow condition and space for calibration anemometer Contraction Nozzle: to accelerate the flow speed to reach the desirable level in the test section, reduce the turbulence intensity.

Required flow fields in Wind tunnel test sections A uniform-flow, steady and low turbulence Restriction due to blockage effects Size effect that induce flow disturbance. Hard to estimate the uncertainty (a) ideal uniform flow (b) symmetric flow resulting fromwall friction

Timeline of air speed standard at CMS ’98-’99: Small wind tunnel + a TSI LDV 2000: LDV Probe adjustment & calibration* Traceability of air-speed measurement Anemometry calibration Service 2003: Design a new wind tunnel with expanded LDV probe 2004: Implementation of new system *Optics and Lasers in Engineering, 38, 2002

Measurement Facilities-1 Wind tunnel : Open-loop with expanded test section Outlet of contraction nozzle: 200 mm Motor controlled by a frequency converter Designed air speed: ~30 m/s Flow patterns simulated at test section

Characteristics of axial fan volume flow rate and pressure rise Fan selection is based on the volume flow rate and pressure rise. the size (diameter=0.56 m). power (energy loss of flow in the tunnel) is proportional to the cube of the air speed.

CFD simulation vs.LDV measurement Outlet of contraction nozzle Suction section Air LDV measurement velocity at 10 mm downstream LDV measurement velocity at 100 mm

Measurement Facilities-2 LDV : TSI Probe 9832 with beam expander 2. Focal length: 450 mm 3. Nominal fringe spacing: 1.921 m for beam of 514.5 nm, 1.822 m for 488nm

Measurement Traceability of Anemometry Calibration Length Standard Frequency Standard (Time) Vernier Caliper Universal Counter Spinning Disc Laser Doppler Anemometry + Wind tunnel Check Anemometry Anemometry under Test

Spinning disc traceability Pedestal -Disc eccentricity adjustable -Horizontal disc -rotary encoder for rotational speed - Local fringe spacing detectable

Obtain of Fringe Spacing Basics of calibration β Vldv= df  fD Vdisc= r   cos Obtain local fringe spacing df = r ×  / fD× cos  <1 degree 13

Obtain of Fringe Spacing Comparison between (a) favorable and (b) improper Doppler bursts. Both the sequences show the change of burst signals when slightly moves the traverse stage. 14

Calibration of LDV Probe-1 Why is adjustment and calibration of LDV probe necessary? Measurement Volume df * *Miles, P. C., ”Geometry of the Fringe Field Formed in the Intersection of Two Gaussian Beams,” Applied Optics1996;35:5887–5895.

Calibration of LDV Probe-2 Step 1. Steer beams to optimize beam crossing Z=1850 m Z=1000 m Z=0 Minimal variation of fringe spacing in measuring volume Change state of beam crossing 16

Calibration of LDV Probe-3 Step 2. Determination of coefficient of air speed (coefficient of speed: mean fringe spacing) 1.Repeat measurement of fringe spacing at different rotation rates 2.Determined mean fringe spacing: 1.9225 m with U= 0.1%. (Nominal value: 1.921)

Calibration of LDV Probe-4 Step 3. Evaluation of uncertainty of air-speed coefficient df = r ×  / fD× cos + ε

Available region for installation of anemometry-probe Potential-core Region is considered for locating anemometry probe and was examined of its flow characterestics. Air flow d D X X = 100 mm D = 200 mm d = 80 mm (region I) 140 mm (region II) : region of interest : anemometry probe

Characteristics of Wind Tunnel Considered uncertainty sources of air speed measured by LDV in wind tunnel: 1. Characteristics of flow* A. Particle Lag B. Turbulence C. Velocity Bias (Sampling bias) D. Fringe Bias 2. Characteristics of wind tunnel A. steadiness & stability B. uniformity of velocity profiles C. variations of axial velocity *Ref. Fry, DTNSRDC, 1985

Uncertainty analysis of air speed in interested region Vtunnel : Air speed in wind tunnel Vldv : Measured air speed by LDV δ : correction factor due to flow characteristics ε : correction factor due to tunnel characteristics

Uncertainty due to Flow characteristics in interested region 1. Turbulence intensity Tu = 1.73 % - 0.81 % for V=0.5 m/s – 25 m/s Measured at 100 mm downstream centerline of nozzle 2. Velocity bias (sampling effect) by comparing weighted (residence-time) with un-weighted velocities (Fry, 1985): , estimated to be < 0.01%

Effects due to velocity distribution in interested region 1. Flow profiles along vertical radius (Z-axis flow uniformity) r (mm) V1 (m/s) V2 V3 V4 Mean Std. Dev. (m/s) Rel. Std. Dev. % For region I： r = (-40 ~ 40) mm, X = (0~100)mm Standard Uncertainty = 0.14 % For region II： r = (-70 ~ 70) mm, X = (0~100)mm Standard Uncertainty = 0.22 %

Effects due to velocity distribution in the interested region 2. Flow profiles along horizontal radius (Y-axis flow uniformity, the same process as 1.) 3. Variation of velocities along flow direction (X-axis flow uniformity) Std. dev. = 0.129% (obtained from repeat measurement at different air speed)

Uncertainty analysis item Sources u(x )/x % ν LDV facility 0.050 9.5 Flow Property 0.0557 1066.7 1 Particle lag ∞ 2 Velocity bias 0.01 3 Turbulence int. 0.0548 999 4 Fringe bias Case A : 0.2304 23.9 Wind-tunnel Property Case B : 0.3474 26.5 Case A : 0.1412 8 Velocity Profile (vertical axis) Case B : 0.2245 14 Case A : 0.1290 (horizonal) Case B : 0.2319 7 V along x-axis 0.1284 9 Case A : 0.2422 23 Air speed in interested region Case B : 0.3554 25.5 Case A : 2.074 k Case B : 2.059 Case A : 0.50 Case B : 0.73

Concluding Remarks Beams in LDV probe can be steered to reach optimal crossing Fringe spacing in measuring volume could be calibrated. Expanded Uncertainty of df = 0.1% Characteristics of wind tunnel dominate the measurement uncertainty. ie. Good wind tunnel is the key to calibration and measurement capability. For the present system, U = 0.5% for air speed in region I, and U = 0.73% in region II.

Comparison with other Lab.

Operation principle of RPTM module The spinning disc outputs a reference signal for tuning time delay. By superimposing the RPTM signal on the Doppler signals, a favorable burst can be frozen. Then, tuning the time delay and gate width to enclose interested region.