...
...
...
...
...
- If connected to the web servers, the label ID is used to query for the curated length.
If not connected to the web servers, the length in the label bar code is used. For the WRMSL, knowing the length is not critical, as the track will actually measure the length as part of its section handling process.
If the top of the section is missing because a whole-round sample was removed, enter the length of the missing interval so the sample position can be corrected. DO NOT place any type of spacer in front of the section to make up for the missing interval.
Whole-rounds taken from the bottom of the section will not affect the measurement process.
- When the Section Information window (above) opens, the cursor should be in the SCAN input box. If not, click in the box.
- Hold the scanner over the label and center the dot-code with the red crosshairs while pulling the trigger.
- At the scanner beep, the Sample_ID and the LIMS_ID should correctly display on the screen. If not, rescan.
Upon clicking the MEASURE button, the pusher arm moves forward and pushes the section until the endcap breaks the light beam on top of the section sensor. The pusher then moves the section into the first measurement position and triggers the instruments to measure.
The move and measure process repeats until the pusher arm reaches its motion limits. The pusher retracts to the home position, ready to receive the next section, and the Section Information window opens to repeat the process.
Measurement Sequence
...
Intentions and Assumptions
This section covers the steps needed to take meaningful measurements. Perform these steps prior to starting measurements on a new site. This does not cover physical setup of the hardware or initial deployment of the software.
A Quick Introduction to the IMS Program Structure
IMS is a modular program. Individual modules are as follows:
- INST plug-ins: codes for each of the instruments
- MOTION plug-in: codes for the motion control system
- DAQ Engine: code that organizes INST and MOTION plug-ins into a track system
...
...
...
...
Instrument-Specific Information (digiBase Detector)
...
Standard Information
Select the standard from the list box to use when the calibrate function is executed. Selecting a standard updates the values displayed below. To edit the values, use the GRA Edit Standards function in the Instrument menu.
Calibration DAQ
Parameters used when measuring the GRA standard:
CAL Live Preset: Live time of the measurement.
Core Diameter: Internal diameter of the core liner.
Aluminum Density: Density of the aluminum alloy used to fabricate the standard.
MS: Setup
General Information
Instrument X Offset: Offset along the X-axis (push direction) from the home switch to the center of the MS loop.
Contact Width: Width along the core axis that influences the measured value; used when calculating edge clearance (if enabled).
Units: Set to SI; must match the physical MS meter setting.
Range: Set to 1.0; must match the physical MS meter setting.
Analysis Name: Defined JR_LIMS analysis.
Instrument Group: Defined JR_LIMS instrument group component.
Model: Instrument model number/name of the sensing component.
S/N: Instrument serial number of the sensing component.
Manufacturer's Name: Name of the manufacturer of the sensing component of the instrument.
Menu Name: Name used in the IMS menus.
Full Name: Name used in IMS Reports and displays.
Standard Information
Correction: The vendor provides this value.
- For standard-frequency loops (80 and 90 mm, 565 Hz), this value is 1.000.
- For 10% high loops (80 mm only, 621 Hz), this value is 0.908.
- For 10% low loops (80 mm only, 513 Hz), this value is 1.099.
- For 20% low loops (90 mm only, 452 Hz), this value is 1.174.
PWL: Setup
General Information
Instrument X Offset: Offset along the X-axis (push direction) from the home switch to the center of the PWL transducers.
Contact Width: Width along the core axis that influences the measured value used when calculating edge clearance (if enabled).
Analysis Name: Defined LIMS analysis value.
Instrument Group: Defined LIMS instrument group component.
Model: Instrument model number/name of the sensing component.
S/N: Instrument serial number of the sensing component.
Manufacturer's Name: Name of the manufacturer of the sensing component of the instrument
Menu Name: Name used in the IMS menus.
Full Name: Name used in IMS Reports and displays.
Liner Correction
Liner Thickness: Thickness of the core liner.
Liner Velocity: Acoustic velocity of the core liner (butyrate).
Liner Delay: Calculated value equivalent to the liner thickness divided by the liner velocity.
Velocity Filter
The filter purpose is to remove extremely low or high velocities.
Velocity Filter: Enable or disable filtering.
Min Velocity Filter: Velocities below this value are not reported.
Max Velocity Filter: Velocities above this value are not reported.
Velocity Filter
Waveform Stack: The number of waveforms stacked and averaged. Minimum 50.
...
To access an instrument's calibration procedure, click on the Instrument menu and select the instrument's calibration function. Good practice is to calibrate the instrument prior to receiving cores at each new site.
GRA: Calibration
The GRA standard is an aluminum bar inserted into a core liner, which is sealed and filled with DI water. The aluminum bar is machined into discrete steps that provide composite density values for creating a calibration curve of counts vs. density. Before calibration, use the Standard Editor to enter offset and thickness of each step's center. Under the Instrument menu, select GRA: Edit Standards. This only has to be done once, as the values are retained in the configuration file.
Define two standard sets in the Standard Editor: Density and Water.
...
GRA Calibration Process
- Place the standard on the track with the first step defined in the standard toward the GRA sensor.
- In the Instrument menu, select GRA: Calibration.
- The pusher engages the standard and pushes it until it trips the top-of-section sensor, then moves the standard to all of the defined offsets, taking a measurement at each position.
- After the second position is measured, the calibration data display in the graph along with the calibration line.
- After measuring the last step, the pusher moves the standard to the offset for the water measurement.
- Using the new calibration data, a density value for water is calculated and compared to the known value of 1 gm/cm3. The curve fit provides the coefficients used in the density calculations, ln I = B(¿d) + C; where ¿= Compton attenuation coefficient, d = sample diameter, I0 = gamma ray source intensity, and I = measured intensity of gamma rays passing through the sample. Allowing bulk density to calculated by using, ¿¿ = 1/(¿d) × ln (I0/I).
- Choose to apply the calibration to all future measurements by clicking OK-APPLY or click CANCEL to restore the previous values.
...
- Travel time across the transducer caps
- Electronic delays in the pulser and the digitizer.
...
PWL Calibration process
In the Instrument menu, select PWL: Calibration and the PWL Calibration window will open.
The window opens as seen above. Begin by following the steps on the right sight of the window.
Setting Up a Measurement
Use the controls shown to set the measurement parameters.
Requested Stack: Sets the number of waveforms to average to increase the signal-to-noise ratio.
Threshold: A millivolt value used to find the first arrival peak. For more information on the first arrival determination, see Appendix 1.
Measuring Standard
- Place the aluminum standard between the transducers. Close the transducers. Aluminum standard width is 76.2mm. Select Laser Offset Correction. Confirm that the Distance matches the standard width.
- Enter the Aluminum Standard velocity. 6295.00 m/s. Select Determine System Delay. Verify that the Velocity matches the Aluminum Standard.
...
- Open transducers. Remove Aluminum Standard. Position the Water Standard in place. Close transducers. Turn on Liner Delay. Liner Thickness is 0.275 cm. Determine Liner Velocity. Verify that the water velocity matches the calculated H2O velocity +/- 1% (liner velocity is calculated based on this).
- Open transducers. Accept Changes
...
Analytical Batch
The analytical batch is defined by the number of samples run between each sensor calibration. Each sample in the batch run with the current calibration is associated with that calibration data in the LIMS.
Calibrations and calibration timestamps are accessible through each sensor Instrument Interface screen.
...
GRA
Calibration assumes a two-phase system model for sediments and rocks: minerals and interstitial water. Aluminum has an attenuation coefficient similar to common minerals and is used as the mineral phase standard. Pure water is used as the interstitial water phase standard. A piece of core liner containing a telescoping aluminum rod (5 varying thicknesses) and filled with distilled water is measured as a calibration standard. The largest diameter aluminum rod has a porosity of 0% and a bulk density of 2.7 g/cm3. Water has a porosity of 100% and density of 1.00 g/cm3. Intermediate elements verify the linearity of the log density relationship and the alignment of core and sensor. A linear least-squares fit through 3–5 calibration points yields the calibration coefficient. Total measured counts are divided by the counting time to normalize coefficients to counts per second.
MSL
Absolute susceptibility: Sample cubes are measured using the Kappabridge and results compared with corresponding readings from the Bartington instrument. Empirical correction factors have been calculated.
PWL
Pulse detection settings are checked by IODP technicians on a regular basis and do not require adjustments by the user.
Pulse time is a constant which is included in the total time measured as a result of the threshold peak detection procedure used. This value changes depending on the wiring of the system. The user does not need to make adjustments to this factor.
Transducer displacement and traveltime delay calibrations are performed simultaneously and should be executed once per expedition on a routine basis. Displacement measured in volts is calibrated to millimeters by measuring 3 standard acrylic cylinders. A linear least-squares fit to the points defined by the voltage readings (x-axis) and the known standard thickness in millimeters (y-axis) yields the linear coefficients used in the calculation.
Measured travel distance and time must be corrected for twice the liner thickness. This calibration is performed by IODP technicians. Vendor specifications for the wall thickness of the liner are used.
Accuracy
GRA
GRA accuracy is limited by the assumption that the measured material has the same attenuation coefficient as the calibration standards used. For general sediment-water mixtures, this should be the case and error should be <5%.
MSL
Accuracy of the susceptibility meter and sensor loop is 5% (according to Bartington).
PWL
PWL accuracy can be evaluated by measuring pure water at varying and exactly known temperatures. Past experience shows that for a properly calibrated system and good acoustic coupling, the disagreement with published sonic velocity values is less than ±20 m/s.
Precision
GRA
GRA precision is proportional to the square root of the counts measured, as gamma ray emission is subject to Poisson statistics. Measurements with the system have typical count rates of 10,000 (dense rock) to 20,000 (soft mud). If measured for 4 s, statistical error is <40,000 ± 200 cps, or 0.5%.
MSL
MSL precision is 2 x 10–6 SI. Susceptibility values in natural marine sediment samples over an interval of only a few meters (Milankovitch or millennial scale cyclicity) can range from a few tens to several thousands of 10–6 SI units. Typically, variations are 2–3 orders of magnitude greater than the precision, making magnetic susceptibility one of the most precise proxies for stratigraphic changes.
PWL
...
MDrive Set up
When you select Setup from the Motion menu bar, the window to the right appears.
Track Options
...
...
...
...
...
...
Limit and Home Switches
Click Limit and Home switches to open the window below.
Select Axis: In the case of the WRMSL and STMSL it is always X.
The MDrive can be used with either a dedicated Home switch or a limit switch as a home switch. These tracks use the dedicated Home switch. Select the appropriate setup for the track in use. Use the Utilities to execute the Home command and verify the correct setting. If the home switch position of detection edge changes, verify the instrument offsets and relocate the TOS switch are still correct. Setting the edge from CW to CCW will change the offset by a least 1 cm.
Click Open Utilities to test these settings.
Click Done to save the values or Cancel to return to the previous values.
Motion Profiles
Click Limit and Home switches to open the window below. The profiles are used to set the speed and acceleration profiles used by the track.
Setting the correct values for the motion profile takes a little experimentation to make the track run efficiently and safely.
DAQ Move: This profile controls moves between measurement positions. Set this to a reasonable speed with gradual acceleration so the pusher does not bump the sections.
Limit Seek: This profile finds the limit switch locations. Do not exceed 3 cm/sec and use a large deceleration value or the core could overrun the limit switch and hit the mechanical stop.
Home Final: This profile finds the final location of the home switch. Do not exceed 1 cm/sec and use a large deceleration value.
Load/Unload: This profile moves the pusher back to the load position. Set this to a reasonable but high speed with gradual acceleration and deceleration values. Setting this too slow will waste time, but keep safety in mind.
Push-Slow: This profile allows the pusher to move the new section into contact with the previous section and to locate the top of section. Use a speed a little less than the DAQ Move speed with slightly lower acceleration and deceleration values.
Push-Fast: This profile allows the pusher to move quickly to the TOS switch. Typically, it is set the same as the Load/Unload values.
User Define: This profile is used for testing only in the Motion Utilities program.
Click Open Utilities to test these settings.
Click Done to save the values or Cancel to return to the previous values.
...
To up load the images into the database MegaUploadaTron must be running in background. If not already started do the following:
- On the desktop click the MUT icon on the bottom task bar.
- The following will appear:
...
- Check Automatic Upload in the lower right hand corner and then click the minimize window and MUT will run in background.
Data Management
...
...
Appendix 1: P-wave: Waveform Acquisition and First Arrival Detection
Signal Processing
Figure 2 on the following page summarizes the following signal processing discussion.
Acquisition and Ensemble Averaging
Figure 1The received waveform is 60 µsec long and digitized at 10 nsec intervals. The first 10 µsec of the waveform is replaced with zeros to remove a large spike that occurs at the beginning of the outgoing pulse. We believe that this noise is generated by the electrical "cross-talk" within the pulser-receiver.
A Butterworth Bandpass filter is applied to the waveform to reject frequencies less than 0.4 MHz and above 1.0 MHZ. The results of this filtering is stacked, summed, and averaged (ensemble averaging). The size of the stack is pre-set by the user to a value between 10 and 100.
Even with this initial filtering and ensemble averaging, the resulting signal can still contain significant noise generated by the shipboard environment. Figure 1 is typical of the signal quality.
Picking the 1st Arrival
Attempts to use a cross-correlation method (popular in seismic analysis) have not been successfully applied to IODP data. Coefficients developed for one type of lithology fail when the lithology changes. Therefore, we depend on a threshold-crossing process to determine first arrival time. The weakness of this method is that any high-amplitude noise can cause a false, early pick.
To increase the likelihood of a correct pick, the P-wave Logger software on the Whole Round Multi-Sensor Track (WRMSL) uses a series of mathematical and data manipulation techniques to suppress the noise relative to the peak of the first arrival. Another goal is to eliminate the constant adjustment of the threshold value as the amplitude of the waveform changes due to variation in signal attenuation along the core. As the P-wave Logger is an automated system, constant supervision and adjustment is not practical.
Signal-to-Noise Ratio (SNR) Enhancement
Methods employed are partly based on Dr. Tom O'Haver's 2006 publication: A Pragmatic Introduction to Signal Processing (http://terpconnect.umd.edu/~toh/spectrum/IntroToSignalProcessing.pdf).
Figure 2The ensemble averaging done at the acquisition stage is effective at eliminating nearly all of the random noise. Unfortunately, our signals contain a great deal of systematic environmental noise that shielding signal cables cannot fully eliminate. To suppress this noise, we need to exploit small differences between the noise and the acoustic pulse. The parameters used in the smoothing operators were developed empirically by measuring a variety of materials and by degrading the signal by attenuation at the pulser-receiver. Time was limited on Expedition 361 to select the best smoothing operators and their parameters. Further experimentation is needed to improve the process and validate that it will work for all possible lithologies recovered by IODP.
The SNR enhancement process consists of 3 steps: (1) smoothing, (2) 1st derivative, and (3) smoothing.
- The first smoothing step passes the waveform through a Savitzky-Golay filter that applies a polynomial least-square fit step-wise through the data (similar to a running average). Experimentation shows that using a 6th-order polynomial across a 126 data point window will reduce high-amplitude, high-frequency noise while retaining the shape of the acoustic pulse.
- After the first smoothing operation, we take the 1st derivative (central-difference) which suppresses the low-frequency noise. The downside to the differentiation is that the process adds back in high-frequency noise.
- A second smoothing step using FIR operator is then used to reject this noise. The FIR filter coefficients were calculated using LabVIEW tools.
Figure 3. Waveform as absolute valuesFigure 3 shows the waveform (as absolute values) before and after the 3-step process and illustrates the significant noise suppression achieved. Attempting to find the first arrival on the unprocessed data would have failed because the amplitude of the noise is greater than the first arrival peak.
Normalization
At this point in the process, a copy of the waveform is normalized to scale between the values of –1 and 1. A copy of the processed waveform is converted to its absolute value and then normalized with the waveform scaled between values of 0 and 1. This normalization is critical to making the threshold picking process indifferent to either amplification or attenuation of the signal. Experience has shown that the threshold value needs little adjustment once chosen unless there is a significant lithology change.
1st Arrival Time
The threshold crossing is performed on the absolute copy of the waveform. The crossing time is then used as the starting point to find the first zero-crossing in the derivative waveform. From this time value, we subtract 1.5 µsec to account for ¼ wavelength phase shift caused by the 1st derivative operation and ½ wavelength correction back to the first arrival. See Figure 4 below.
Figure 4: The left-hand graph shows the threshold (Horizontal cursor) crossing the first peak of the acoustic pulse and the 1st zero-crossing (vertical cursor). After correcting the time, the right-hand graph shows the pick on the original stacked waveform.
Waveform Data
...