PP-LWT MEETING MINUTES 3 JUNE 2001 Prepared by Blum, reviewed by Mills, Ledwon, Freeman AGENDA 1. Reports * Legs 193-195 Cruise Evaluations * Legs 193-195 Tech Reports 2. Urgent issues * NGR data transfer * PWS3 hardware * New PWS calibration procedure 3. Action Overview Appendix I: PP application task list, 3 June 2001 1. REPORTS 1.1. LEGS 193-195 CRUISE EVALUATIONS (FROM SCIENTISTS) As on previous legs since the latest dry dock, the space around PP instruments was considered inadequate and a potential safety hazard on all three legs of this period. The problem having the ovens under the auxiliary sampling table was noted again. The ovens should be moved to a different location to avoid heating of the tight work space and drying of cores. The long-standing problem of slow data transfer from NGR and GRA (see detailed description in task list) was noted again by the scientists. MST efficiency (and thus core flow) could be doubled with the appropriate interface upgrade (purchased last year) and software changes (attempted but failed last year and this year). Like on previous legs, pycnometer measurements were hampered by repeated instrument failures on Leg 194, despite having three pycnometers on board and two of them installed to run simultaneously. The slowness of pycnometry was also noted again [and this will not change until ODP has custom instruments built specifically for our purpose]. Pycnometers also tend to become inaccurate with time and need to be re- calibrated [a result of heavy use]. The control measurement taken regularly with calibration spheres are still not being loaded and retrievable from the data base, which limits their usefulness (see also task list). Electrical cross talk flawed NGR measurements early in Leg 194. The problem was fixed but the scientists were frustrated because many hours were spent on figuring out the problem, no explanation was furnished as to the source and nature of the problem, and many cores had to be re-run. As has been reported for a number of periods now (see task list), the PWS3 setup is prone to measurement errors because the sensor separation measurement reference is not fixed. [The replacement part was ordered but ODP apparently received the wrong size.] On Leg 194, PWS signal detection configurations/setting were discovered to be inadequate and was fixed during the leg. On Leg 195, a timing correction was reported to be applied twice, once by the acquisition program and a second time by the database. A software bug was reported with the AVS during Leg 195 (core locator got "lost"). The MAD balance algorithm was criticized for not being as sensitive as the chemistry balance program. The methods used for the two balances are different, but the actual effect on data quality is not apparent to ODP personnel and needs to be further investigated. In the meantime, the chemistry balance program was added to the PP balance program and users can use either one, even though using the chemistry program requires some manual data entry. A rare problem with the overheating and destruction of a TC needle probe was reported on Leg 194. It was again proposed to make an oedometer (consolidometer) available on the ship, a suggestion that reappears about every year. The main problems are resources to buy a complete system and lab space. 1.2. LEGS 190-192 TECH REPORTS No significant issues were reported from Leg 193, which recovered very little core. Few TC, MAD, NGR, and MS measurement were taken PWS data could not be uploaded via "Generic Uploader" and had to be transferred and formatted "manually" via backup protocol. [Old problem, see task list.] A second pycnometer was installed and used to measure dry, porous limestone cubes. [A calculation protocol for this alternate way of getting MAD data from this type of rock was defined and described in the Explanatory Notes - new task is defined to implement the appropriate data report.] Installation of a software patch to accelerate significantly the transfer of NGR (and GRA) data from the MCB to the computer was unsuccessful. [Old and urgent problem, see task list]. No explanation could be provided four the erratic NGR data recorded early in the leg. According to the tech report, the "mysterious" problem ended when the system was rebooted... During Leg 195, significant troubleshooting was required despite relatively low core recovery. ODP technician noticed that switching the MS drift-correction off on the MST resulted in unrealistic MS values. This discovery of a potential bug is new and needs to be investigated. PWL calibrations did apparently not work. The calibration routine and sensor hardware were replaced on Leg 191. When a run was aborted, the PWL transducers didn't stay open, jammed the core, and sometimes snapped the belt. To avoid the jamming, the software had to be reinitialized using a "debug" routine. NGR data can still not be edited in the database. Edits need to be requested from the MCSs. [Old problem.] The latest changes to PWS file structures and calibration protocol (Legs 192-193) was marred by lack of communication, resulting in inconsistent data normalization and uploading and editing problems. The pycnometer ran pretty well. Medium-size cells were mostly used to accommodate rock pieces. The pycnometer needs to be run manually for cell sizes other than small. 2. URGENT ISSUES 2.1. NGR DATA TRANSFER On Leg 194, Mills installed for the second time the new PCI boards and software patches that should dramatically accelerate NGR data transfer and eliminate the significant lost time during MST logging. This has been a high-priority upgrade for a long time. Unfortunately, an unknown problem persisted and the task is not completed. 2.2. PWS The PWS3 device is still measuring the absolute distance between the upper transducer and the fixed frame, which is prone to errors because the lower transducer is not perfectly fixed. The LVDT device needs to be mounted such that it measures the relative displacement between the transducers, analogous to the PWL system. This was planned for Leg 194 but has not happened yet 2.3. PWS DATA HANDLING The new data formats and calibration protocol for the PWS system still cause data consistency, upload, and database editing problems. We need the latest description of the MST output files, including units, precision, etc., so IS can adjust the uploader, database, and data report formats. 3. ACTION OVERVIEW 3.1. TASKS COMPLETED € Cause of chronic MST belt failures was fixed. € Some progress was made with the task: WCMST CONTROL DATA 3.2. OLD TASKS NOT COMPLETED Application development tasks € NGR (and GRA) data transfer [Mills]. € MAD control measurements [ISD/Blum]. € TC user interface [Becker/Blum]. € TC data model [Mithal/Blum]. € Implement bar codes [ISD/Mills]. € WCMST threshold warnings [ISD/Mills]. € MSL drift correction [Mills/Blum]. € WCMST control data [ISD/Mills/Blum]. € Implement PWS4 system [Mills/Ledwon/ISD]. € Simultaneous MSP (MSCL or AMST) [Mills]. € TC database editor [ISD]. € NGR database editor [ISD]. € P-wave fixes [Mills, ISD]. € System-DB links [Mills/ISD]. € MSL background correction [Mills/Blum/ISD]. Other tasks € Fix MAD report [ISD/Blum]. 3.3. NEW TASKS (see descriptions in Appendix I) Application development tasks € Implement MAD Method D (mainly data report) [Blum/ISD]. Other tasks € Look into apparent MST-MS bug [Mills on Leg 197]. € Look into PWL transducer opening control during aborts [Mills on Leg 197]. € Look into PWL calibration problem [Mills on Leg 197]. APPENDIX I: PP APPLICATION TASK LIST, 3 JUNE, 2001 REFERENCE: NGR DATA TRANSFER CONTACT: PP-LWT DATE: 9/21/99; modified 11/20/00 DATA SYSTEM: WCMST-NGR OBJECTIVE: Increase speed (and therefore quality) of natural gamma ray data acquisition. RATIONALE: NGR data acquisition slows down core logging substantially (~200% overhead time) because writing the data to the local disk takes more time than measuring it and writing it to the multi-channel buffer. TASK(S): At least double present rate (or sampling resolution) of NGR (and overall WCMST) data acquisition by upgrading NGR data transfer to present technology. This can be accomplished with the recent change of the WCMST control software to the PC platform (drydock), and the new PCI board installed on Leg 191. A persisting problem should be resolved on Leg 194. RESOURCES: Estimated 8 hr of software configuration. REFERENCE: FIX MAD REPORT CONTACT: PP-LWT DATE: 4/18/00 DATA SYSTEM: MAD OBJECTIVE: Provide accurate and user-friendly data report. RATIONALE: A recent review and discussion with the database group revealed that several format/labeling problems exist in the present MAD report. While accomplishing the minor changes to fix this, some additional modification could be done to enhance the logic and user's grasp of the different methods A, B, and C. TASK(S): Prepare a report according to the example (Excel table) provided by Blum. Re- arrange/eliminate some columns. Use one decimal precision for all % values (water content, porosity) and three decimals for all mass (g), density (g/cm^3), and void ratio values. Fix column headers according to example. RESOURCES: Estimated 10 hr of query writing. REFERENCE: MAD CONTROL MEASUREMENTS CONTACT: PP-LWT DATE: 9/21/99; modified 12/3/99 DATA SYSTEM: MAD OBJECTIVE: Save and make available MAD control (precision) data. RATIONALE: Volume control measurements (using calibration sphere) are routinely acquired with the pycnometer to check on the instrument's performance. These measurements have been used for a quick, real-time verification that the value did not deviate more than a certain empirical amount from the true value. A few years ago, the major pieces (new control program, data model) were put in place to keep a record of these control measurements that would not only allow a quantitative evaluation of the analytical error, but also correction or elimination of data affected by a malfunctioning cell. Today, these data are still not loaded and no web report exists. TASK(S): Load MAD control measurements into database and make them accessible through a web report. RESOURCES: Estimated 24 hr of programming. REFERENCE: WCMST THRESHOLD WARNINGS CONTACT: PP-LWT DATE: 9/21/99; modified 12/3/99 DATA SYSTEM: WCMST-ALL SENSORS OBJECTIVE: Avoid collection of bad data, or loss of data, through real-time warnings. RATIONALE: Instrument problems (electronic failures, calibration problems, etc.) occur on almost every leg. Because WCMST data acquisition is automated and rapid, associated bad data collection is often discovered only days later and entire holes' worth of data may be lost. The original specifications for the new MST software in 1996 included a warning system based on control measurements taken with every section (partly implemented) but the project was never completed. In addition, threshold warnings should also be possible based on the core measurement values. TASK(S): Implement threshold warnings for the WCMST. Add an interface where the operator can enter high and low threshold values for each sensor, for the routine control measurement (default values) as well as for the core measurements (core-specific values), and make alarms go off if the acceptable values are not met. This would drastically reduce the risk of collection bad data by allowing the operators to discover instrument problems immediately. RESOURCES: Estimated 80 hr of Labview programming REFERENCE: MSL DRIFT CORRECTION CONTACT: PP-LWT DATE: 9/21/99; modified 12/3/99 DATA SYSTEM: WCMST-MSL OBJECTIVE: Improve accuracy of magnetic susceptibility drift correction. RATIONALE: The present implementation of the magnetic susceptibility drift correction is rather ineffective because of slack in timing and positioning of the reference measurement. TASK(S): The reference measurement for drift calculation should be taken as soon as possible after the core measurements, i.e., at the end of the last MS measurement rather than at the end of the section run (last NGR measurement) as presently implemented. Furthermore, the reference measurement should be taken at the same position as the zeroing took place. The MST run should therefore be interrupted at the end of the MS run, the boat moved into the zeroing position, and the reference measurement taken, before the section run is resumed. Ideally, zeroing and reference measurements are taken at the center of a 50-cm long "Control 2" standard, a liner filled with pure water and mounted at the top of the core boat. RESOURCES: Estimated 40 hr of Labview programming; fix to hardware configuration REFERENCE: WCMST CONTROL DATA CONTACT: PP-LWT DATE: 9/21/99; modified 12/3/99 DATA SYSTEM: WCMST-ALL SENSORS OBJECTIVE: Make MST control (precision) data consistently available. RATIONALE: Control measurements are used to evaluate and monitor instrument performance, calibrate or correct core measurements, or take test measurement independent of the core sample identification requirement. Control-1 measurements are performed on standards that are run like a core, in a separate run; they may be part of the ODP set of standards or any type of material or standard provided by the user. Control-2 measurements are routinely performed at the beginning of each section run, on a water standard that is mounted at the top of the core boat. Control-3 measurements are routinely performed "through air", at the end of a section run. Implementation of control measurements is at various stages, most problems appear to be related to the web reports and should be easy to fix. TASK(S): Complete implementation of the following control measurements, uploads, and web reports: GRA: Ctrl-1, Ctrl-2, Ctrl-3 MSL: Ctrl-1, (hold), Ctrl-3 NGR: Ctrl-1, (hold), Ctrl-3 PWS: Ctrl-1, Ctrl-2, N/A Specific tasks: 1. Check all Control web report header names and units, there are mistakes". 2. Change names of web reports from "Control" to "Control-1. 3. Create headers on the web report selection page ("Control-1", "Control-2", "Control-3") that are linked to a brief explanatory text about control measurements. 4. When NGR background measurements are taken on the long water core, write the data to the control-1 tables as well so they are available to the user. 5. The Control-2 water standard is too short for the NGR and MSL sensors; don't acquire these data until modifications to the WCMST allow implementation of a longer (50 cm) water standard on the core boat. 6. Control-3 values for GRA appear to be hard-coded or copied for each section rather than measured; the web report does not return any values. 7. For PWL, the Control-3 web report lists values although control-3 measurements are actually not possible with this sensor. RESOURCES: Estimated 200 hr of programming. REFERENCE: IMPLEMENT PWS4 CONTACT: PP-LWT DATE: 9/21/99; modified 12/3/99 DATA SYSTEM: "PWS" OBJECTIVE: Separate instrument for P-wave velocity measurements on cylinders/cubes for better data. RATIONALE: At present, the P-wave velocity sensor 3 (PWS3) is used for measurement on both half-cores and discrete rock cubes or cylinders. The requirement of sensor spring loading for the former has resulted in a lot of bad data for the latter type of measurement because measurement conditions were different from calibration conditions. TASK(S): For this and other operational reasons, implement a separate PWS4 system. It would mostly be a stand-alone copy of the present PWS3 system. Once the hardware and controlling software (Labview) are completed, database model, upload, and data report can be completed almost identical to the PWS3 system. RESOURCES: Estimated 120 hr of Labview programming and testing. REFERENCE: TC USER INTERFACE CONTACT: PP-LWT DATE: 9/21/99; modified 12/3/99 DATA SYSTEM: TC OBJECTIVE: Ensure consistent thermal conductivity data collection. RATIONALE: The present TC program (instrument control) is the generic version written by the manufacturer (Teka, Berlin). It works very well, but it does not include an ODP sample identification interface or data upload protocol. The instrument output files are not consistently returned to ODP from the ship. Instead, spreadsheets prepared by scientists ad-hoc are the product of this system. This is one of the last standard data acquisition systems not integrated at all into the ODP routine and database. TASK(S): Write an ODP-specific user interface with consistent sample identification protocol (use of bar codes is preferred). Add data upload and backup on "Save", and database-editing functions (delete bad run, rename section and interval). Detailed specifications are available from P. Blum. RESOURCES: Estimated 24 hr of ODP programming; $4k for software; $7k for programming. REFERENCE: TC DATA MODEL CONTACT: PP-LWT DATE: 9/21/99; modified 12/3/99 DATA SYSTEM: TC OBJECTIVE: Ensure consistent thermal conductivity data archiving and access. RATIONALE: The present database model only accommodates the final, interpreted TC values. The values are uploaded on shore from inconsistent spreadsheet files prepared by scientists ad-hoc. This implementation is a quick fix and not compatible with the more rigorous data archiving and quality standards for other physical properties measurement systems. TASK(S): Develop new database model that accommodates all measurement parameters and raw data (time-temperature series) that are output by the Teka instrument. Draft specifications are available from P. Blum. RESOURCES: Estimated 40 hr of programming REFERENCE: SIMULTANEOUS MSP CONTACT: PP-LWT DATE: 9/21/99; modified 12/3/99 DATA SYSTEM: AMST-MSP OBJECTIVE: Allow simultaneous acquisition of color reflectance and point-sensor magnetic susceptibility. RATIONALE: At present, the magnetic susceptibility point sensor (MSP) measurements are not feasible on a high recovery leg. The first problem is related to the instruments integration time and is out of our control. The second problem is that MSP measurements require a dedicated run on the AMST. TASK(S): Implement MSP measurements such that they can be taken simultaneously with the color reflectance measurements, slowing the latter down slightly but allowing to take both measurements on moderate-high recovery legs (hardware component!) Integrate data upload and edit functions similarly as for the CR data and as part of the AMST upload and edit component. RESOURCES: Estimated 200 hr of Labview programming and testing. REFERENCE: MSP DATA MODEL CONTACT: PP-LWT DATE: 9/21/99; modified 12/3/99 DATA SYSTEM: AMST-MSP OBJECTIVE: Ensure consistent archiving/access of magnetic susceptibility point-sensor data. RATIONALE: The magnetic susceptibility point sensor (MSP) system is in development. Data acquisition must yet be completed. TASK(S): At this time, create the data model for MSP data so they can be loaded. RESOURCES: Estimated 80 hr of programming. REFERENCE: IMPLEMENT BARCODES CONTACT: PP-LWT DATE: 9/21/99; modified 12/3/99 DATA SYSTEM: ALL PP SYSTEMS IN A FIRST PHASE OBJECTIVE: Avoid operator errors and save time by eliminating typing of section/sample id's. RATIONALE: Shipboard personnel are eager to use barcodes for core section and sample identification. The foundation has been laid a long time ago, and most pieces seem to be in place. However, implementation got stuck somewhere since barcodes cannot be used for any PP station at this time. TASK(S): Complete implementation of the use of barcodes for all PP stations, starting with the WCMST. RESOURCES: Estimated 120 hr of installation and programming. REFERENCE: SYSTEM-DB LINKS CONTACT: PP-LWT DATE: 11/21/00 DATA SYSTEM: ALL PP SYSTEMS, STARTING WITH WC-MST OBJECTIVE: Allow direct retrieval of calibration, standard, and background data from the database as part of a calibration/measurement run. This is a prerequisite for a future 'Direct Upload" project. RATIONALE: The PP measurement systems and database models were designed in such a way that calibration, background, and standard data could be re-used and did not have to be recreated each time. This requires consistent uplaod and time stamping of the relevant values, and their direct retrieval from the database when needed. This link has not been established to date, which results in the need to re-create calibration and background files and causes time stamp mis-matches. TASK(S): Establish direct system-database link. RESOURCES: Estimated 200 hr of programming. REFERENCE: MSL BACKGROUND CORRECTION CONTACT: PP-LWT DATE: 11/21/00 DATA SYSTEM: WCMST-MSL OBJECTIVE: Measure and subtract background (coreboat, dirt) from magnetic susceptibility measurements. RATIONALE: The coreboat carries a magnetic susceptibility signal of a few instrument units. Mud accumulating on the boat can add another few units. This becomes significant in core materials with very low susceptibilities (almost pure carbonate/siliceous oozes). The magnitude of the effect is in the same range as the instrument drift on our MSL system. The background could be removed relatively easily by periodically measuring an empty boat run and then subtracting the values from the core measurements. TASK(S): 1. Implement a background measurement routine in the MST program. 2. Create a table in the data model to hold the data, with a time stamp, and the necessary upload component. 3. Upgrade the data report to subtract the interpolated or selected background values from the core data and display the background values in the first (uncorrected) data column. (The second, corrected data column is to hold the drift- corrected data). Add a column in the report that gives the magnitude of the background correction (placed in front of the column that shows the drift correction). RESOURCES: Estimated 100 hr of programming. REFERENCE: NGR DATABASE EDITOR CONTACT: PP-LWT DATE: 11/20/00 DATA SYSTEM: WC-MST NGR OBJECTIVE: Fix database editor. RATIONALE: Editing capabilities for NGR data seem to be broken as a result of the latest upgrades in NGR data handling. This needs to be fixed otherwise the shore personnel has to be tasked with data clean up. TASK(S): Fix the database editor for NGR data. RESOURCES: Estimated 20 hr of programming. REFERENCE: TC DATABASE EDITOR CONTACT: PP-LWT DATE: 11/20/00 DATA SYSTEM: TC OBJECTIVE: Provide database editing capability for thermal conductivity data. RATIONALE: A data model exists to hold thermal conductivity data. Note that this table only holds the computed final values, and no parameters or raw data (see 'TC Data Model' task for a description of the ultimate data model). ODP personnel has to format the data coming from the Teka04 instrument manually in a spreadsheet and then invoke the upload. This is cumbersome but it works. However, errors in the uploaded data cannot easily be edited because no editor exists. TASK(S): Create a database editor for TC data similar to the MST and PWS database editor. RESOURCES: Estimated 20 hr of programming. REFERENCE: P-WAVE FIXES CONTACT: PP-LWT DATE: 11/20/00 DATA SYSTEM: WC-MST PWL and PWS OBJECTIVE: Adjust data files and data handling to recently installed hardware upgrades; make all P-wave data formats consistent. RATIONALE: The Leg 191 upgrades to transducer displacement measurement systems of the PWL and PWS3 systems require a follow up to make data structure, data formats, database tables, and data upload routines compatible. TASK(S): 1. Create MST data file fields to accommodate new LVDT calibration, eliminate a few fields that were never used (PWL, PWS3); 2. Make all data formats (time and distance) uniform and compatible among all P-wave systems (PWL, PWS); 3. Adjust database tables accordingly; 4. Adjust data uploaders and data reports accordingly. RESOURCES: Estimated 100 hr of programming. REFERENCE: MAD METHOD D CONTACT: PP-LWT DATE: 5/1/00 DATA SYSTEM: MAD OBJECTIVE: Implement Method D of measuring and calculating MAD properties for those materials where total wet mass cannot be measured (saturation cannot be achieved), such as high-porosity, coarse-grained limestones. RATIONALE: The present methods A, B and C are designed for materials that allow the measurement of wet bulk and dry bulk sediment mass, from which the moisture content is calculated. Some of the limestones recovered on Leg 194 did not allow the measurement of wet bulk mass because the combination of high porosity and cementation made some of the pore water drain immediately, i.e., the specimen could not be saturated with water for the purpose of mass determination. The solution is to determine the moisture content from wet bulk and dry bulk volume measurements, whereby the wet bulk volume is estimated from the cub geometry of the rock sample regardless of the fact that it is not saturated. TASK(S): Implement a fourth set of calculations in the MAD web reports, according to the Explanatory Notes of Leg 194 and compatible with the existing calculations for methods A, B, and C. The data model already accommodates all fields that are needed. RESOURCES: Estimated 20 hr of programming and testing.