Preliminary results from CVX-2

 

Robert F. Berg

National Institute of Standards and Technology

Gaithersburg, MD 20899-8364

 

22 July 2003

 

The CVX-2 experiment was a clear success. The apparatus worked well, and a preliminary analysis indicates that the data will answer the scientific question: Can a fluid as simple as xenon exhibit shear thinning? Our pleasure is tempered by the loss of Space Shuttle Columbia and its crew, but we are grateful for the scientific opportunity given to us by STS-107.

 

 

Operations

CVX-2 operated for 370 hours, probably the longest of all of Space Shuttle Columbia’s experiments. Due to common communication difficulties, only 85 % of our data was downlinked during the mission. However, most of the missing 15 % was dispersed in small gaps consisting of only one or two points (1-2 minutes), so the coverage of the data set was effectively complete.

 

Before the flight, we had concerns for three possible problems that are unique to an orbiting experiment. None occurred.

·        No problem due to solar exposure

Strong variations of the sunlight hitting the experiment would have compromised the temperature control of the xenon sample.

·        No problem due to vibration

Although the experiment’s sensitivity to vibration was known, the vibration level due to astronaut exercises, motors, fans, etc. was hard to predict.

·        No problem due to particle radiation

In 1997, the CVX experiment on STS-85 was affected by charged particles trapped in the Earth’s magnetic field and by high-energy galactic cosmic rays. Neither caused a significant problem for CVX-2, probably because the orbit of STS-107 was at a lower altitude and farther from the North and South Poles.

 

 

Scientific results

The analysis, which is incomplete at this date, will compare the data to the results of numerical hydrodynamic simulations. First results obtained without the simulations include the following.

·        The force/displacement ratio is about 1% smaller near the critical temperature (Tc) at all frequencies examined so far. This is consistent with the prediction that shear-thinning can occur near the critical point of xenon.

·        Our largest shear rate seems to lower the value of Tc by roughly 0.1 mK. This is qualitatively consistent with theoretical predictions.

·        To within a temperature scale factor, the temperature dependence of the apparent viscosity just below Tc is the same as that above Tc. This suggests that the liquid-vapor “emulsion” created shortly after passing through Tc behaved as a single fluid.

 

Completing the analysis will require improving our understanding in two ways.

·        We now understand the nonlinearity caused by the hydrodynamics when the sample temperature is far above Tc. How does the viscoelasticity near Tc combine with this nonlinearity? Can the combination somehow mimic shear-thinning?

·        What is the correct model for the fluid used in the simulations? Does viscoelasticity suppress shear-thinning or vice-versa?

 

 

Captions for new figures

FREESTAR installed in the payload bay

CVX-2 was one of six experiments on the FREESTAR payload. Note the three cannisters with blue lids mounted on the payload truss. CVX-2 was in the two that are connected by a cable.

 

CVX-2 in Columbia’s payload bay

The two round cannisters on the left hold CVX-2. This rear-facing video image has the Earth in the background.

 

Influence of the lid temperature

The upper plot shows that solar exposure during each orbit made the temperature of the cannister lid go up and down. The lower plot shows the corresponding wiggles in the temperature of the sample cell. The last part of day 5 and the beginning of day 6 were a “critical period” for CVX-2, so Columbia moved to an orientation that minimized the solar exposure. The resulting wiggles in the cell temperature were then only about 10 microdegrees.

 

Temperature of the CVX-2 sample

CVX-2 passed through the critical temperature five times. (The horizontal dotted line indicates Tc.) The passes on days 6, 9, 11, and 13 were very slow, about 1 degree per year, because shear-thinning was expected only within 0.001 degree of Tc. The different colors indicate different frequency combinations.

 

Finding the critical temperature

Raw data from the first pass through Tc show that Tc was near 16.6 degrees C.

 

Consistency with CVX in 1997

The first pass through Tc showed that the data taken in 2003 were consistent with those taken in 1997. This superposition required a horizontal shift of the old data by only 0.006 degree to account for thermometer drift. No vertical shift was required, indicating that no oscillator damage or sample loss had occurred during the intervening years.

 

CVX oscillating screen (picture)

The oscillator used to measure viscosity was a delicate screen made of nickel. This picture shows the (prototype) screen surrounded by four electrode plates. Oscillating voltages on the electrodes caused the screen to oscillate. Viscosity caused the xenon to resist the screen's motion, so viscosity could be derived from the ratio between the screen's motion and the force applied to the screen. CVX's viscometer was designed specifically for operation in the presence of normal Shuttle vibrations caused by motors, thruster firings, and astronaut movements. These vibrations, though necessary for the Shuttle's operation, cause random movements of the oscillating screen which add noise to the viscosity measurements. The low mass of the screen made it comparatively insensitive to the Shuttle's movements while the large area of the screen led to the large viscous resistance required to measure the viscosity with precision.

 

CVX oscillating screen (schematic diagram)

This diagram shows how the viscometer screen oscillated. It exaggerates the thickness of the individual screen wires as well as the amplitude of the oscillation.

 

Maximum oscillations of CVX oscillator at 3 Hz

At maximum, the amplitude at the tip of the oscillator (x0) was much larger than the effective diameter (D) of a single screen wire. It was also much larger than the viscous penetration length (d), which is where most of the fluid motion occurred.

 

Using data below Tc

When the xenon passed below its critical temperature (Tc), it separated into liquid and vapor regions. Due to microgravity and the very slow rate of temperature change, the regions may remained intertwined and acted as a single fluid for several minutes. This hypothesis is supported by the four-step analysis sequence shown in the plots.

1)      Data from the third and fourth passes through Tc nearly superpose even though the fourth pass was twice as fast as the third.

2)      A very small temperature shift makes the superposition nearly exact.

3)      The data below and above Tc are placed on a common positive temperature scale.

4)      Stretching the temperature scale matches the data from below Tc to those above Tc.

 

Deviations of B(xlarge)/B(xsmall) from the calibration curve

These plots show preliminary evidence in support of shear-thinning in xenon. For both plots, the horizontal axis is x0/d, the oscillation amplitude divided by the viscous penetration length. For the first plot, the vertical axis is “magB”, the magnitude of the oscillator’s reduced force measured at a frequency of 5 Hz. The blue circles are data taken far from Tc, where no shear-thinning was expected. Hydrodynamic nonlinearity caused the force to increase as x0 increased. The red dots are data close to Tc. The very closest data, near x0/d = 2.1, show a decrease relative to the blue noncritical data of approximately 1 %. Similarly, the red data on the second plot show a phase shift of approximately 0.01 radian relative to the blue data.

 

On both plots, the red data deviate from the blue data in directions that are consistent with shear thinning. However, further work is needed to rule out the possibility that the deviations were caused by the interaction of viscoelasticity with the hydrodynamic nonlinearity.

 

CVX-2 thermostat found in Columbia debris

Although NASA recovered less than 40 % of the orbiter’s dry mass, this included key pieces of the CVX-2 apparatus.

Left                 The CVX-2 apparatus before it was loaded into the two cannisters.

Lower right     The lower shelf of the experiment cannister.

Upper right     The middle shell of the thermostat, which still holds the inner shell and the xenon sample cell. The surrounding outer shell was not found.

Return to [Fluid Metrology Group | Process Measurements Division | NIST Home]

Last modified: 17 Nov 2005.
Send comments to the webmaster at fluid hyphen science one at nist dot gov.