Skip navigation

Gravity Probe B

Testing Einstein's Universe

WEEKLY UPDATE FOR 10 JUNE 2005:

GRAVITY PROBE B MISSION STATUS AT A GLANCE

Item Current Status
Mission Elapsed Time 416 days (59 weeks/13.64 months)
Science Data Collection 287 days (41 weeks/9.41 months)
Current Orbit # 6,139 as of 4:30 PM PST
Spacecraft General Health Good
Roll Rate Normal at 0.7742 rpm (77.5 seconds per revolution)
Gyro Suspension System (GSS) All 4 gyros digitally suspended in science mode
Dewar Temperature 1.82 kelvin, holding steady
Global Positioning System (GPS) lock Greater than 98.6%
Attitude & Translation Control (ATC)

X-axis attitude error: 283.8 marcs rms
Y-axis attitude error: 253.3 marcs rms

Command & Data Handling (CDH) B-side (backup) computer in control
Multi-bit errors (MBE): 1 (in SRE computer on 6/6)
Single-bit errors (SBE): 8 (daily average)
Telescope Readout (TRE) Nominal
SQUID Readouts (SRE) Nominal
Gyro #1 rotor potential +0.5 mV
Gyro #2 rotor potential -0.2 mV
Gyro #4 rotor potential -0.0 mV
Gyro #3 Drag-free Status Backup Drag-free mode (normal)

MISSION DIRECTOR'S SUMMARY

As of Mission Day 416, the Gravity Probe B vehicle and payload are in good health. All four gyros are digitally suspended in science mode. The spacecraft is flying drag-free around Gyro #3.

Over the past two weeks, we have performed several routine adjustments to various systems on the spacecraft and continued our preliminary calibration tests. These included:

  1. Updating the feed-forward term of the backup drag free controller. The spacecraft is flying drag-free around Gyro #3, which means that this gyro is treated as its center of mass. Since the other three gyros are located in slightly different positions on the quartz block, we must apply small voltages to their suspension systems to fly them drag-free around Gyro #3's position. The feed-forward term is an adjustment factor for anticipating the position of the other gyros in order to prevent error build-up. Our initial calculation of the feed-forward term prior to launch was fairly accurate, and we've been using it thus far in the mission. However, based on 13 months of flight data, we are able to update this term to be even more accurate, which we did this past week. We will now use the updated feed-forward term for the duration of the science (data collection) phase.
  2. Increasing the preload voltage on science gyro #3 to 1.5V for a brief time. The Gyro Suspension System (GSS) for each gyro has a “preload voltage” and a “control voltage.” You can think of the preload voltage as coarse positioning and the control voltage as fine positioning. Generally, we like to keep these voltages about equal. However, right now as part of our calibration tests, we are purposely changing the preload voltage for a brief time to see the effect of this asymmetry. Note that we can perform this calibration test while remaining in drag-free flight.
  3. Starting a new readout calibration cycling plan where we operate the calibration signal on either SQUID's #1 and #2, or SQUIDs #3 and #4, with the other pair off. (We toggle the paris every 24 hours.) This is another of the preliminary calibration tests that we are currently conducting.

On Monday 6 June 2005, a multi-bit error (MBE) occurred in the SQUID Readout Electronics (SRE) computer. A preliminary analysis of this error suggests that the memory location in which the error occurred is not being used, and most likely, this MBE is benign. However, if after further investigation, this is not the case, we will reboot the SRE computer to clear the error. 

Finally, another Heat Pulse meter test to re-check the level of the helium remaining in the Dewar is scheduled for Monday, 13 June 2005. If the results of this test correlate well with previous results, this will be the last heat pulse test of the mission.

MISSION NEWS—THE GP-B PROBE'S REALLY COOL WINDOWS: Part II

In last week's update/highlights, we described the GP-B probe's structure, including the difference between the aluminum shell in the lower part of the probe and the composite shell in the probe neck, the four thermal platforms in the probe neck, and the three 6” round windows that are thermally attached to the first three of the four thermal platforms.

The fourth (topmost) thermal platform has no window connected to it. Rather, the topmost window (Window #4), is embedded inside the top end of a large metal collar, called the "Cross Flange." Window #4, which is about 8” in diameter (2" larger in diameter than the three internal windows), is sometimes called the "warm window," as opposed to the three "cold windows" inside the probe neck. The Cross Flange in which Window #4 is embedded forms the top end of the probe and is bolted with a vacuum seal to the upper end of the Top Hat. It contains two large exhaust valves that were used to evacuate all gases from the probe shortly after launch. The tall, conical Sunshade is bolted to the outside top end of the Cross Flange.

Window #4, along with its unique hermetic vacuum seal, was an engineering tour de force--a synergy of design constraints, ingenuity, and engineering prowess. Because this window forms the vacuum seal at end of the probe, it had to be made of a material that was strong enough to withstand the pressure differential between the high vacuum inside the probe and normal atmospheric pressure and temperature outside the probe without lensing or bulging. In addition, this window had to be a very good thermal conductor, so that it would conduct heat uniformly in all directions from the center outward, with very little temperature rise in the center. Of course, this window also had to be optically transparent, allowing the light from the guide star, IM Pegasi, to enter the probe undistorted, but at the same time, it also had to block electromagnetic radiation. Finally, this window had to perform these functions both in orbit and in extensive testing here on Earth, where the pressure and temperature differential inside and outside the probe is much greater than in space.

What kind of window material meets all of these requirements? The answer is sapphire. Sapphire is extremely strong, it has very high thermal conductivity, it is an electrical insulator, and it is optically transparent. Sapphire has a crystalline structure, that could cause a phenomenon called birefringence or double refractions of a single light beam. However, due to the sapphire's strength, Window #4 could be made relatively thin, and birefringence was not a problem. In addition to an anti-reflective coating, which was also applied to the three fused quartz windows, Window #4 was treated with an electromagnetic interference (EMI) coating to attenuate the ambient low frequency and microwave electromagnetic radiation leaking into the probe through the wide open optical aperture, both in space and on the ground. Because EMI coatings use metal, which is opaque, they reduce the optical transmission of the window. Thus, there was a tradeoff between the effectiveness of the EMI shield and optical transmission requirement. Finally, Window #4 also included a thermal emissivity or “low e” coating that reduced the amount of heat it transmits into the probe.

Perhaps the greatest challenge with Window #4 was its hermetic vacuum seal, which took many design iterations to arrive at a working solution. At the time Window #4 was constructed, rubber O-rings were typically used to create vacuum seals. O-rings would probably have worked in space, but they did not provide a good enough seal here on the ground, where the pressure was much greater. Another option was to braze (very high temperature soldering) a metallic seal onto the outer edge of the sapphire window. However, brazing was deemed too risky, given thermal expansion issues and the very high cost of sapphire. Ultimately, the Stanford-Lockheed Martin engineering collaboration came up with a unique seal that had never before been tried: Window #4 was surrounded by an Indium-coated "C" seal, lined with a thin layer of gold against the sapphire, and held in place by a ring of springy nickel-chromium-iron alloy called Inconel® that is more typically used in formula race car exhaust systems. As of this update, the probe has maintained it ultra low vacuum throughout months of ground testing, plus 13 months in orbit, so this hermetic seal must be working.

After Window #4 was sealed in place, a vacuum was established inside the probe. This vacuum was more than 100 times greater than the vacuum of space in GP-B's orbital region, 641 km (400 miles) above the Earth. Light and some heat enter the probe through Window #4, and some heat from the electronics boxes and heaters located on the spacecraft's frame also enter the probe through the top hat. The visible light passes down the probe, through all four windows, and into the telescope. Much of the heat is initially dissipated to the Dewar's outermost heat exchanger through the topmost thermal platform #4 (which has no window attached to it). Any remaining heat then continues down the probe, eventually striking Window #3, where a large percentage of it is dissipated out through Thermal Platform #3 to the Dewar's second heat exchanger. Likewise, whatever heat still remains travels further down the probe, striking Window #2, where much of it is dissipated out through Thermal Platform #2 and the next heat exchanger. Finally, whatever tiny amount of heat still remains, travels down the probe to Window #1, where most of it is dissipated out through Thermal Platform #1 to the Dewar's innermost heat exchanger.

Thus, moving down through the probe towards the telescope, the average temperature drops significantly from window to window. The actual temperature ranges recorded at each window thus far in the mission are as follows:

  • Window #4: 210 K to 255 K
  • Window #3: 125 K to 155 K
  • Window #2: 72 K to 92 K
  • Window #1: 30 K to 38 K
  • Window #0: 3.3 K to 3.5 K (This is not actually a stand-alone window, but rather the forward plate of the telescope, where the Image Divider Assembly sits.)

The low numbers in each range are from the end of September, when the least Sunlight strikes the Dewar. The high numbers in each range are the current values, when the spacecraft is in full-Sun throughout each orbit. Note that for the inner three windows, the temperatures in the table shown above are measured at the copper mounts, where the window frames connect to the thermal platforms. The actual temperatures of the windows themselves cannot be measured, but they differ slightly from the values shown above. The copper mounts have weak thermal links to both the probe and to the window glass. This way the glass in the three inner windows never experiences any thermal shock from the probe. However, the copper mounts themselves are very good thermal conductors, ensuring that the window glass remains in a very symmetric thermal environment along the axis of the probe. Any axial asymmetry in the windows' thermal environment would cause a distortion in star image.

To minimize ghost images in the field of view of the telescope, formed by light reflecting between windows, each of the four windows was canted (tilted) 2 degrees and clocked perpendicular to the next window. In other words, if Window #4 was canted 2 degrees towards 12 o'clock, Window #3 was canted 2 degrees towards 3 o'clock, Window #2 was canted 2 degrees towards 6 o'clock and Window #1 was canted 2 degrees towards 9 o'clock. The window canting directs reflections between the windows away from the telescope's field of view. Without the window canting, the telescope would see clusters of dimmer stars surrounding a brighter star, instead of just a single star. 

Even with all the special coatings on the windows, the amount of starlight that reaches the telescope's photon detectors is somewhat diminished by each optical surface that it passes through. All told, the starlight encounters some 24 optical surfaces, including the front and back of each probe window, the main telescope plates, the telescope mirrors, the beam splitter, prisms, mirrors, and lenses in the image divider, and beam splitters, mirrors, and lenses in the photon detector packages. Thus, by the time the starlight from IM Pegasi reaches the telescope's photon detectors, only about 8-12% of the light makes it to all 8 photo detectors combined! That is, 2-3% of the initial starlight beam actually falls on each of the four detector pairs (one primary detector pair, plus one redundant backup detector pair in both the X- and Y-axes.) Considering that IM Pegasi is only a 5.6 magnitude star at it brightest, only a very small amount of light actually hits these detectors, but it has been adequate to keep the telescope centered on the guide star throughout the mission.

The table below provides an example of the decrease in light moving down through the probe, surface-by-surface, from Window #4 to the telescope's photon detectors. Note: This table is a simplified model of the probe's optical performance. Actual values differ somewhat and are published in various scientific journals.

Surface

Typical Efficiency

Remaining Light

Number of Paths

Light in Each Path

Notes

raw starlight

n/a

100.0%

1

100.0%

5.8 average magnitude star - pretty dim to start with

window 4 top

88%

88.0%

1

88.0%

Window 4 also acts as an EMI shield and vacuum seal

window 4 bottom

88%

77.4%

1

77.4%

 

window 3 top

96%

74.3%

1

74.3%

Each window reduces the thermal load into the probe.  Four was chosen as the right balance between heat rejected versus light lost.

window 3 bottom

96%

71.4%

1

71.4%

 

window 2 top

96%

68.5%

1

68.5%

 

window 2 bottom

96%

65.8%

1

65.8%

 

window 1 top

96%

63.1%

1

63.1%

 

window 1 bottom

96%

60.6%

1

60.6%

 

telescope forward plate top

96%

58.2%

1

58.2%

The telescope forward plate supports the IDA, DPA's, and secondary mirror

telescope forward plate bottom

96%

55.9%

1

55.9%

 

telescope primary mirror

90%

50.3%

1

50.3%

The telescope mirrors focus the starlight from a 6" wide beam to a 0.001" point for the IDA to divide it

telescope secondary mirror

90%

45.3%

1

45.3%

 

telescope tertiary mirror

90%

40.7%

1

40.7%

 

IDA baseplate bottom

96%

39.1%

1

39.1%

IDA = Image Divider Assembly

IDA baseplate top

96%

37.5%

1

37.5%

 

IDA beamsplitter

68%

25.5%

2

12.8%

The IDA beamsplitter is a "half silvered" plate that makes copies of the beam (1 reflected, 1 transmitted) to be split in the X and Y axes

IDA roof prism

90%

23.0%

4

5.7%

The roof prisms slice the beam into left and right halves; when they're equal, you're pointed at the center of the guide star

IDA fold mirror

90%

20.7%

4

5.2%

Fold mirrors steer the beams to the detectors

IDA lens entrance

96%

19.8%

4

5.0%

after division at the roof prism, the beams are expanding (nearly BB sized at this point); these lenses control the expansion

IDA lens exit

96%

19.1%

4

4.8%

 

DPA fold mirror

90%

17.1%

4

4.3%

DPA = Detector Package Assembly

DPA lens entrance

96%

16.5%

4

4.1%

These lenses refocus the now slightly expanding beams into tight spots on the tiny photodiodes

DPA lens exit

96%

15.8%

4

4.0%

 

DPA beamsplitter

68%

10.7%

8

1.3%

These duplicate the beams onto redundant sets of photodiodes

One final mission note about light and heat from Earth entering the probe: The GP-B spacecraft design includes a shutter at the base of the Sunshade that can be closed to prevent Sunlight from entering the telescope and overpowering its sensitive photon detectors and the telescope readout electronics (TRE) in the event that the spacecraft loses attitude control. The original spacecraft design called for the shutter to remain open throughout each orbit. However, it occurred to members of the design team that the temperature in the cryogenic part of the probe would remain more constant if the shutter were closed during "Guide Star Invalid" periods--anytime the spacecraft moves behind the Earth--thus preventing Earth light (and thus heat) from entering the probe. Earth light includes bright Sunlight reflected off the Earth at mid-day, glancing Sunlight when the spacecraft is pointing at the edge of the Earth, and light originating from Earth, such as city lights and forest fires. The intensity of Earth light is always much greater than light from the guide star, and thus it tends to saturate the TRE detector amplifiers and also results in a very slight warming of the lower, cryogenic section of the probe. Thus, at launch last year, the mission plan called for closing the shutter during the Guide Star Invalid period of each orbit.

The opening and closing movement of the shutter was tested on the ground before launch and found to have no significant effect on the Attitude and Translation Control (ATC) system. However, during the Initialization and Orbit Checkout (IOC) following launch, the GP-B team discovered that shutter movement was significantly detectable by the science gyros, and that its movements could adversely affect their performance. (This issue would have been very difficult, time-consuming, and expensive to test for prior to launch.) Thus, during the IOC phase, we decided to revert to the original design plan of leaving the shutter open during the Guide Star Invalid portion of each orbit and allowing the light and heat radiated from Earth to enter the probe and very slightly increase the temperature in the telescope detector platforms and the quartz block housing the SQUIDs and gyros. This small amount of heat is not a problem. Most of it is drawn off and conducted into the Dewar by the Cryopump, located between Window #1 and the front plate of the telescope. (The Cryopump, itself, is a very interesting technology and will be the subject of a future update.) The temperature of the remaining cryogenic components quickly normalizes when the spacecraft emerges from behind the Earth and re-enters the Guide Star Valid state.

The telescope's photon detectors continue to generate data during each Guide Star Invalid period, but generally, we do not record this data in our ground telemetry passes. However, for a few months of the year, when the Sun is at it closest approach to IM Pegasi and when the alignment of Sun, Earth, Moon, and spacecraft is such that the spacecraft is only receiving light generated by the Earth itself or when the portion of Earth under the spacecraft is bathed in Moonlight, we have collected this data in order to calibrate the sensitivity of the photon detectors. Interestingly, Moonlight is just bright enough to enable the photon detectors to create a nocturnal brightness map of the Earth's surface features--clouds, oceans, forests, deserts, and so on--as seen through the telescope's very narrow field of view. When the Moon and Sun are both on the opposite side of the Earth from the spacecraft, the detectors pick up city lights, forest fires, and any other light-dark patterns generated here on Earth.


UPDATED NASA/GP-B FACT SHEET AVAILABLE FOR DOWNLOADING

We recently updated our NASA Factsheet on the GP-B mission and experiment. You'll now find this 6-page document (Adobe Acrobat PDF format) listed as the last navigation link under "What is GP-B" in the upper left corner of this Web page. You can also click here to download a copy.


Drawings & Photos: Tthe diagram of the GP-B experiment, the drawing of starlight entering the Sunshade & Probe, and the composite photo of the spacecraft flying over North America at night were created by GP-B Public Affairs Coordinator, Bob Kahn using Adobe Photoshop and Adobe Illustrator. The photos of the gyro housing assembly, the SQUID Readout Electronics (SRE) box, and the probe Top Hat and Cross Flange, as well as the drawings of the Dewar and probe are from the GP-B Image Archive here at Stanford. Finally, the photos of the probe windows were taken by Bob Kahn and Ken Bower. Click on the thumbnails to view these images at full size.

MORE LINKS ON RECENT TOPICS

Previous Highlight
Index of Highlights