Vibrating-Wire, Supercooled Liquid Water Content Sensor Calibration and Characterization Progress

www.nasa.gov 1

Vibrating-Wire, Supercooled Liquid Water Content

Sensor Calibration and Characterization Progress

Prepared by: Michael King

National Aeronautics and Space Administration, Glenn Research Center, Cleveland, OH

John Bognar, Daniel Guest, & Fred Bunt

Anasphere, Inc., Bozeman, MT

8th AIAA Atmospheric and Space Environments Conference Washington, D.C.

www.nasa.gov

Presentation Overview

•

Background

•

SLWC Sensor Description

•

Calibration

•

Ice Accretion Analysis

•

Sensor Data Drift

–  Data Analysis

–  Laboratory Testing and Analysis

–  Correcting Data

•

Summary and Conclusions

www.nasa.gov 3

•  Inflight icing of aircraft remains a hazard to the aviation community.

•  Providing operators with real-time information of atmospheric icing conditions is a means

to reduce hazard by operationally minimizing exposure.

•  NASA has been developing icing remote sensing technology to provide the necessary

information, but the technology requires validation through in-situ measurements.

•  NASA funded Anasphere, Inc., through the SBIR program, to develop and calibrate

weather balloon-borne, supercooled liquid water content (SLWC) sensors.

•  NASA conducted a field campaign at the NASA Glenn Research Center (GRC) in

Cleveland, OH during winter 2015 to develop a validation database using the Anasphere SLWC sensors.

•  Significant effort has been made to calibrate and characterize these sensors including

facility development and experimental testing.

•  Progress has also been made to understand an unexpected behavior observed in the

www.nasa.gov

SLWC Sensor Description

•  The weather balloon instrument package used during the

field campaign included the iMet-1-RSBN Radiosonde and the vibrating-wire, SLWC Sensor.

•  The measurement principle behind the SLWC sensor is

that ice accretion along the wire leads to a decrease in the natural frequency of the wire.

•  Vertical profiles of SLWC can be derived from the time

history of the measured wire vibration frequency and the ascent speed.

•  The 0.6 mm steel wire is periodically “plucked” using a

servomotor with a magnet attached to a short support.

•  A thin film piezoelectric sensor, sandwiched between a

silicon rubber structure, measures the wire vibration frequency.

•  An Anasphere, Inc. proprietary signal processing method

is used to determine the natural frequency every 3 seconds.

4

Instrument Package – SLWC Sensor highlighted in images

www.nasa.gov

Facility Development

•  Anasphere, Inc. developed a low speed icing

tunnel for calibration testing of the SLWC sensors. •  The facility is a closed circuit, fan driven, icing

tunnel with a 30 cm by 30 cm test section.

•  Steady airspeeds around 5 m/s are achieved in the

test section

–  The nominal ascent rate of the winter 2015 weather

balloons is 5 m/s.

•  The tunnel is insulated and the heat exchanger is

capable of cooling to subfreezing temperatures between -15 °C and -10 °C.

•  A single Spraying Systems Co. 1/4J-SS

air-atomizing nozzle is used to create the supercooled cloud.

–  Air and water pressure are controlled independently.

–  Water supply is chilled to ensure supercooling

5

www.nasa.gov

Calibration:

Tunnel Testing

•  Testing was carried out in two separate

phases

–  1st Phase: SLWC Sensor and PDI

–  2nd Phase: Icing Blade

•  The SLWC Sensor frequency data and

drop size and speed distributions were

acquired during the 1st _Phase.

–  Vibration plane of SLWC Sensor was

orthogonal to the flow direction

–  PDI was non-intrusive and situated

upstream of the SLWC Sensor housing

•  Independent measurements of SLWC

were acquired using an icing blade during the 2nd _Phase.

–  Blade used due to concerns about hotwire

instrument operability, data fidelity and heat load on the facility

–  0.64 mm thick blade

–  Spray times were varied to obtain

approximately 0.6 mm thick ice accretion

6

1st _{Phase Testing Setup}

Icing Blade with ice accretion from 2nd _Phase

Flow Direction

PDI – Transmitter Side

SLWC Sensor

6.35 mm 9.53 mm

www.nasa.gov

Data Reduction – SLWC Sensor

•  SLWC is calculated using the frequency

profile

–  The time derivative of the frequency, df/dt, is the driving term

–  The terms f, f₀, ε, D and u are the frequency,

clean-wire natural frequency, collection efficiency, wire diameter and ascent speed, respectively

–  The term b is a coefficient related to the

physical characteristics of the wire

•  The trends for all case are linear and all have

R2 _{values greater than 0.996.}

–  Trends were used instead of typical

generalized central differencing method that have been used for atmospheric soundings.

1 _{Hill, G. E., “Analysis of Supercooled Liquid Water Measurements Using Microwave Radiometer and Vibrating Wire Devices,” Journal of}

Atmospheric and Oceanic Technology, Vol. 11, 1994, pp. 1242-1252. 7

Select SLWC Sensor Frequency Data from 1st _{Phase Testing}

SLWC =

2bf

ε

Duf

df

dt

Natural Frequency, Hz _{Case 1} Case 2 Case 5 Case 7

www.nasa.gov −200 −15 −10 −5 0 5 10 15 20 25 30 soft rime hard rime glaze Temperature, °C Wind Velocity, m/s

Calibration:

Data Reduction – Icing Blade

•  SLWC is calculated using the blade

equation (Ref. 2)

–  The terms ρ, S, ε, u and t are the ice

density, the ice accretion thickness, collection efficiency, flow speed and time, respectively

•  ρ is calculated using the Macklin

empirical equation (Ref. 3)

–  The terms d_v0.5 and T_s are the MVD and

surface temperature, respectively –  Soft rime was observed during testing

–  Soft rime ice densities range from 200 to

600 kg/m3

2 _{Ide, R. F. and Sheldon, D. W., “2006 Icing Cloud Calibration of the NASA Glenn Icing Research Tunnel,” NASA/TM-2008-215177, 2008.}

3 _{Macklin, W. C., “The Density and Structure of Ice Formed by Accretion,” Quarterly Journal of the Royal Meteorological Society, Vol. 88, 1962, pp. 30-50.} 8

SLWC =

ρ

S

ε

ut

ρ

= 110 −5 ×10

d

u

T

#

$

%

&

'

(

Ice accretion formation regimes with the testing and the typical operating envelopes

highlighted

Testing Envelope

www.nasa.gov

Results (1 of 2)

•  The calculated SLWC profiles rise to varying

degrees, between 3% and 17%.

•  This effect may be due to general form of the

SLWC equation.

–  f 3 in the denominator, where f is a decreasing

term with increasing ice accretion.

•  Other effects such as cloud recirculation may

also be contributing factors, but were not investigated in this work.

•  The linearity of the frequency profiles suggest

test section SLWC remained constant.

9

Gradual increase in calculated SLWC

0 10 20 30 40 50 60 0 0.25 0.5 0.75 1 1.25 1.5 1.75 2 2.25 2.5 Time, Sec Sensor SLWC, g/ m 3 Case 1 Case 2 Case 3 Case 4 Case 5 Case 6 Case 7 Case 8 Case 9 Case 10

www.nasa.gov

Calibration:

Results (2 of 2)

•  The correlation between the two measurements is promising, agreeing to well within

±25% in several cases.

10

Case Sensor Blade P_H2O Marker

--- g/m3 _g/m3 _{psig ---} 1 1.15 0.71 10 2 0.74 0.92 20 3 0.64 0.46 4 0.35 0.34 5 0.19 0.15 6 1.39 1.24 30 7 0.79 0.94 8 0.49 0.89 9 0.37 0.75 10 0.29 0.59 0 0.5 1 1.5 2 2.5 3 0 0.5 1 1.5 2 2.5 3 +25% −25% 1:1 Blade SLWC, g/m3 Sensor SLWC, g/ m 3 P H 2O = 10 psig P H 2O = 20 psig P H 2O = 30 psig

Sensor and Blade SLWC Comparison Plot Sensor and Blade SLWC Comparison Table

www.nasa.gov

L0

L1

ice

airflow

•  Images of the ice accretion on the

SLWC Sensor were taken during testing.

•  A ‘shadow zone’ exists which has not

be previously documented.

•  The length of the wire with no ice

a c c r e t i o n , L₁, i n c r e a s e d b y approximately 11%.

•  Effect on the calculated SLWC was

found to be negligible.

–  Increased SLWC in all cases by 0.75%

–  Parametrically, SLWC increases less

than a few percent for L₁ values approaching 50% of L₀ 11 L0 L1 ice airflow 30 mm

Ice Accretion along SLWC Sensor wire from 1st

Phase Testing

Previously assumed idealized ice accretion

Current assumed idealized ice accretion highlighting shadow zone

www.nasa.gov

Sensor Data Drift:

Data Analysis (1 of 2) 12 400 500 600 700 800 900 1000 1050 Pressure, mbar 42 43 44 45 46 7.2 5.6 4.2 3.0 1.9 1.0 0.1 Frequency, Hz Pressure Altitude, Km

•  The gradual increase, or drift, in the frequency profiles is observable.

•  The drift is linear when plotted on a logarithmic pressure scale, and does not appear to be

affected by the presence of icing conditions.

400 500 600 700 800 900 1000 1050 Pressure, mbar 42 43 44 45 46 7.2 5.6 4.2 3.0 1.9 1.0 0.1 Frequency, Hz Pressure Altitude, Km 03/20/2015 – 01 400 500 600 700 800 900 1000 1050 Pressure, mbar 42 43 44 45 46 7.2 5.6 4.2 3.0 1.9 1.0 0.1 Frequency, Hz Pressure Altitude, Km 03/11/2015 – 01

www.nasa.gov

Data Analysis (2 of 2)

•

The primary suspect sources include:

–  Increased modulus of elasticity of the silicone rubber clamping structure, reducing the effective

length of the wire by decreasing the clamp’s structural pliability.

–  Increasing modulus of elasticity of the steel wire, thus stiffening the wire.

13 NEED TITLE!!!! 0 0.25 0.5 0.75 1.0 44 44.2 44.4 44.6 44.8 45 Decreasing Effective Length, mm Natural Frequency, Hz 0 1.0 2.0 3.0 Increasing Modulus of Elasticity, GPa

1% Length Decrease 1% Stiffness Increase

www.nasa.gov

Sensor Data Drift:

Laboratory Testing and Analysis

•  A SLWC sensor was tested in a

controlled laboratory environment to better understand this behavior.

–  A thin thermocouple used to measure the temperature of the silicone rubber square.

–  The sensor was placed in a freezer with a controllable temperature, and the thermocouple temperature and vibration frequency were recorded over a period of 30 minutes.

•  The data clustering near -60 °C for

some cases is due to the weather balloons reaching the isothermal layer near the tropopause.

•  The trends for the balloon soundings

and the laboratory test are similar indicating the effect is thermal.

14 43 43.5 44 44.5 45 45.5 46 −70 −60 −50 −40 −30 −20 −10 0 10 20 Raw Frequency, Hz Temperature, °C 03/11/15 01 03/17/15 01 03/25/15 02 03/25/15 03 Freezer Test isothermal layer Ice accretion on probe wire

www.nasa.gov

Correcting Data

•  The data was corrected by fitting a trend and removing the difference between the trend

and clean wire natural frequency

•  The maximum SLWC only increases marginally, approximately 3%, but the integrated

liquid water, ILW, increases by 20% for this case.

15 −20 −10 0 10 20 400 500 600 700 800 900 1000 1050 Temperature, °C Pressure, mbar Isotherms (°C) 0 −10 −20 −30 −40 Dry Adiabats (°C) 40 30 20 10 0 − 10 Saturation Adiabats (°C) 16 12 8 4 0 −₄ −₈ −₁₂ −₁₆ Saturation Mixing Ratio (g/Kg)

10 5 2 1 42 44 46 Frequency, Hz Corrected Uncorrected Drift Trendline 0 0.2 0.4 0.6 0.8 1 7.2 5.6 4.2 3.0 1.9 1.0 0.1 SLWC, g/m3 Pressure Altitude, Km Corrected Uncorrected temperature dewpoint isotherms cloud

www.nasa.gov

Summary and Conclusions

•  Significant progress has been made calibrating and characterizing vibrating-wire sensors

—the goal of this work.

•  Preliminary results are promising, based on the agreement between the sensor and

blade.

•  An expanded calibration database is needed.

•  The gradual increase in SLWC without corresponding decrease in frequency suggests

that the equations may need to be modified to handle larger mass formations.

•  The extent of ‘shadow zone’ was identified and shown to have negligible effect on results.

•  Possible causes for the frequency drift in the winter 2015 campaign data were

investigated.

•  A method to correct winter 2015 campaign data was developed and presented.

www.nasa.gov

Questions?