SEARCH

SEARCH BY CITATION

Keywords:

  • magnetometer;
  • automation;
  • paleomagnetism;
  • instrumentation

Abstract

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Sample Holder and “Flip” Motion
  5. 3. Sample Translation and Rotation
  6. 4. Accurate Sample Positioning
  7. 5. AF Demagnetizer
  8. 6. Programming
  9. 7. Measurement Procedures
  10. 8. Performance and Concluding Remarks
  11. Acknowledgments
  12. References

[1] We have automated a 2G Enterprises superconducting magnetometer to measure and demagnetize standard paleomagnetic samples. After loading a sample and setting the desired demagnetization steps, the operation is performed without further operator attention. Each of a sample's three axes is measured in both directions multiple times. A single solenoid performs three-axis static demagnetization by rotating the specimen 120° about an oblique axis to each orthogonal position. This eliminates potential errors resulting from differences between the fields generated when using two orthogonal coils with different geometry. Each sample is handled only once, minimizing angular alignment errors. For initial sample moments 10–100 times the holder moment (i.e., natural remanent magnetizations greater than 10−3 A/m for standard size samples), very high quality AF demagnetization results can be obtained. The procedure for a complete, 10- to 20-step alternating field (AF) demagnetization takes between 20 and 70 min, depending on the sample's moment. This automated system is complemented by a custom program that controls all system elements, including the magnetometer, AF demagnetizer, and sample handler that shuttles the sample between the two and also rotates the sample to the measurement positions. In addition, the controlling software includes tools for (1) sample parameter input and instant results recalculation upon parameter adjustment, (2) real-time results visualization, (3) integrated Sun compass correction software, and (4) several demagnetization routines optimized for different magnitudes of magnetization. The software uses a very general and flexible, XML-based file structure capable of storing an entire field study in one hierarchical file format, with levels for locality, site, sample, and demagnetization step. It serves as an electronic field notebook for recording many more parameters and comments than those strictly needed to measure the direction of the core.

1. Introduction

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Sample Holder and “Flip” Motion
  5. 3. Sample Translation and Rotation
  6. 4. Accurate Sample Positioning
  7. 5. AF Demagnetizer
  8. 6. Programming
  9. 7. Measurement Procedures
  10. 8. Performance and Concluding Remarks
  11. Acknowledgments
  12. References

[2] The advantages of automated paleomagnetic sample measurement and treatment have been thoroughly discussed in a recent paper by Kirschvink et al. [2008]. In this technical brief we describe our design of an automatic sample handling system that we have built to use with our stand-alone alternating field (AF) demagnetizer [Collinson, 1983] and cryogenic SQUID magnetometer [Goree and Fuller, 1976]. Our goal was a relatively limited, practical one: to be able to perform completely automated measurement of the three components of the magnetic moment of a paleomagnetic sample, each in both the positive and negative direction, for each step of an alternating field (AF) demagnetization sequence in order to reduce the operator time consumed in studying large numbers of samples. Containing costs was also an important consideration. Implementing this multiposition technique required a mechanically more complicated sample holder than is needed for simpler techniques, and also required some sacrifice in speed of measurement, but we prefer it for paleomagnetic studies of lava flows, which typically involve precisely oriented samples that are much more strongly magnetized than our holder. Moreover, we find our system works well with relatively strongly magnetized sediments. In our studies, AF demagnetization is often the paleomagnetic cleaning method of choice, ideal for removal of partial remagnetization by lightning [Cox, 1961], not subject to chemical alteration that often occurs during thermal demagnetization, and often as effective at removing viscous remanent magnetization overprints carried by magnetite as thermal demagnetization [e.g., Coe et al., 2004; Jarboe et al., 2008].

[3] There are many challenges in developing automated systems to measure sample remanence and to carry out progressive AF demagnetization, as indicated by the diverse solutions that researchers have devised. De Sa and Molyneux [1967] built an astatic magnetometer that could measure samples automatically but not demagnetize them. Only a little over a decade later, A. Cox and his group were carrying out static AF demagnetization automatically in their SCT superconducting rock magnetometer (SRM). However, because their AF coils were inside the magnetometer, demagnetization on one of the axes was limited to 30 mT. They were followed closely in the early 1980s by J. Kirschvink at Caltech, who implemented AF capability to higher fields on a vertical prototype of the 2G SRM that was in development by W. Goree and W. Goodman. Around the same time M. Kono built a spinner magnetometer that measured all three components of magnetization of individual samples automatically, and later added fully automated tumbling AF demagnetization [Kono et al., 1984; M. Kono et al., Use of automatic spinner magnetometer-AF demagnetizer system for magnetostratigraphy and paleosecular variation studies, paper presented at 8th Scientific Assembly, International Association of Geomagnetism and Aeronomy, Uppsala, Sweden, 1997]. These pioneering efforts were soon followed by the commercially available 2G automated AF system for semicontinuous measurement and demagnetization of long cores and a variant for U channels that is also suitable for discrete samples. Various versions of these 2G SRM systems are found today in many paleomagnetic laboratories around the world. More than 10 years ago, Giddings et al. [1997] described a fully automated system that is most similar to ours for static AF demagnetization and measurement of all three components in both directions. Finally, in the last decade custom adaptations for the 2G SRM at several laboratories have permitted automatic processing of ever larger numbers of discrete samples. One such implementation uses a simple quartz-glass holder with extremely low moment, enabling measurement of very weakly magnetized samples in vertically oriented models [Kirschvink et al., 2008]. Other implementations developed at paleomagnetic laboratories in Bremen and Utrecht operate with horizontally oriented models (T. Mullender, personal communication, 2008).

[4] To take full advantage of the sensitivity of our 2G Enterprises model 760R-3.0 DC SQUID Superconducting Rock Magnetometer (approximately 10−12 Am2 root-mean-square noise), the materials selected for construction of the portion of the sample handler that is inserted into the magnetometer needed to be as nonmagnetic as possible. Additional considerations involved tradeoffs between accuracy, precision, and measurement speed. Moreover, we wanted the system to be able to run unattended. Our solution to the problem of single-sample, multiposition automated measurement, described below, has proved to be robust and relatively affordable.

2. Sample Holder and “Flip” Motion

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Sample Holder and “Flip” Motion
  5. 3. Sample Translation and Rotation
  6. 4. Accurate Sample Positioning
  7. 5. AF Demagnetizer
  8. 6. Programming
  9. 7. Measurement Procedures
  10. 8. Performance and Concluding Remarks
  11. Acknowledgments
  12. References

[5] One of the triggers that started this project was our purchase of an AGICO JR-5A automated spinner magnetometer. The automated sample holder for this instrument (1) has fairly low magnetic moment (during ordinary use ∼10−9 Am2; ∼10−10 Am2 can be achieved immediately after cleaning), (2) provides multiple degrees of freedom for orienting samples, (3) is very precisely engineered for reproducibility in positioning, and (4) is quite affordable. In a very small space (40-mm diameter), it can rotate a standard paleomagnetic cylindrical sample around a body diagonal (see Figure 1a). In this way, by successive 120° rotations, it can place each of the sample's three orthogonal axes along each of the sample holder's axes. For clarity, we call this motion “flip” to distinguish it from rotation about the holder's long axis (here termed “rotate”). Flip permits use of a single solenoidal coil to AF demagnetize all three sample components. Flip and rotate permit alignment of all six sample directions (±X, ±Y, and ±Z) along the magnetometer's transverse measuring axes (Mx and My), but only three of the six (e.g., +X, −Y and −Z) can be aligned along the axial measuring axis (Mz). As a result, our apparatus reads just the Mx and My axes in the fully automated measuring routines, so that all components of the moment are measured in both positive and negative directions. This enables cancellation of induced components of magnetization and more symmetrical averaging of sample inhomogeneity in the final determination of the magnetic moment.

image

Figure 1. The (a) original and (b) modified AGICO sample holder shows the addition of a drive and two idler pulleys to produce the “flip” motion. It also shows the removal of the aluminum bayonet-mounting collar, which is replaced by a new Delrin® and Pyrex® mounting. Not shown is the removal of the plastic detent parts and elastic spring hidden inside the mounting shaft. A stepping motor, outside the measuring region, pulls on the Kevlar® aramid strings wrapped around the drive pulley. This rotates the sample about its body diagonal, shown by white arrows, producing the “flip” motion that alternately puts each of the sample's three principle directions along the sample holder's axis. Another white arrow shows the axis that produces the movement called “rotate” in the text. A centimeter scale is shown in the bottom.

Download figure to PowerPoint

[6] A 0.35-mm Kevlar® thread wrapped around a 10-mm-diameter Delrin® pulley is used to remotely flip the sample about its body diagonal (Figure 1b). The stepper motor and associated sensors for actuating the flip motion are located outside the AF and sample measuring regions (Figure 2). All of the pulleys and supports in the measuring region are composed of Delrin®, which has previously been tested and selected for low magnetic moment. The parts were machined using tungsten carbide tools to minimize the introduction of magnetic impurities. Before assembly, the parts were washed and then etched for 30 min in 0.1M hydrochloric acid, reducing the moment of the Delrin® parts into the low 10−10 Am2 range. The AGICO holder's ball-and-spring detent and the aluminum bayonet mount were removed to minimize their magnetic contribution. These modifications, combined with AF demagnetization, achieved a sample holder moment of 10−9 Am2 or less.

image

Figure 2. The stepping motor drive pulley and optical interrupter used outside the magnetometer to remotely flip the sample. A flag on the aluminum drive pulley determines the reference starting position for the “flip” motion.

Download figure to PowerPoint

[7] The sample holder is supported by a 25-mm-diameter by 500-mm-long Pyrex® tube attached to a 30-mm-diameter by 800-mm-long, graphite-fiber-reinforced, epoxy tube mounted on the aluminum flip stepping motor support (Figure 3). The Pyrex® and epoxy tubes were selected and machined to be collinear to better than a millimeter.

image

Figure 3. Diagram of the layout of the magnetometer, demagnetizer, and sample handler on a heavily constructed table, which prevents misalignments. The sample handler rides along its track into the demagnetizer and the magnetometer.

Download figure to PowerPoint

3. Sample Translation and Rotation

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Sample Holder and “Flip” Motion
  5. 3. Sample Translation and Rotation
  6. 4. Accurate Sample Positioning
  7. 5. AF Demagnetizer
  8. 6. Programming
  9. 7. Measurement Procedures
  10. 8. Performance and Concluding Remarks
  11. Acknowledgments
  12. References

[8] A 2G Enterprises 2G802 sample translator is used to move the sample horizontally from an accessible loading position to the AF demagnetization position inside the coil and to the measurement position inside the horizontally oriented magnetometer (Figure 3). Samples are also rotated about an axis parallel to the track along which the translation motion occurs. This rotation moves sample components lying in the magnetometer X-Y plane in order to measure each component along all four of ±X and ±Y. The sample translation and rotation are actuated by Vexta PK569-NAA stepping motors and microstepping drivers, which give a resolution of 0.02 mm and 0.04° per microstep, respectively. We also replaced the stock 2G Enterprises 2:1 chain drive on the rotation motion with a 1:1 drive to improve the angular resolution to 0.02°. The flip motion (Figure 1a) is produced by a stepping motor and an Intelligent Motion Systems INT-481 stepping motor driver, which give a resolution of 0.05° per microstep. The microstepping motor drivers were used to divide each motor step into 16 microsteps and produce a smoother motion to help avoid flux jumps in stronger samples.

[9] A Galil DMC-1832 motor controller simultaneously translates, rotates, and flips the sample. This combined move sequence reduced the sample manipulation time by about 40% relative to serially actuated motion. The Galil motor controller allows us to program the speed, distance, acceleration, and deceleration of the sample's motion. Each move sequence starts from a home position, determined by optical interrupters, and all moves are in the same direction to minimize backlash when changing direction.

[10] The flip motion requires some special consideration because any small difference in the diameters of the two pulleys will not exactly flip the sample from one orthogonal position to another when the motor rotates 120°. This is handled by an automated routine that measures the angle that the sample flips per motor step and simultaneously determines the flip motor position that aligns one of the sample directions with the magnetometer's measuring coil axes, as described below.

4. Accurate Sample Positioning

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Sample Holder and “Flip” Motion
  5. 3. Sample Translation and Rotation
  6. 4. Accurate Sample Positioning
  7. 5. AF Demagnetizer
  8. 6. Programming
  9. 7. Measurement Procedures
  10. 8. Performance and Concluding Remarks
  11. Acknowledgments
  12. References

[11] The correct translation measuring position is determined by inserting a standard sample positioned such that a component of magnetization is oriented along each of the magnetometer pickup coils. The correct translation measuring position is determined by the position of the peak of the sum of the normalized Mx and My channel signals. Because the Mx and My response are constant within 1% over an axial distance of ±12 mm in our large-bore magnetometer, the exact translation position is not critical.

[12] The correct rotation and flip measuring positions are determined by measuring an axially magnetized (parallel to sample Z axis) standard sample as it is stepped through the ranges of each motion individually. The rotation motion parameters (degrees per motor step and start position) are determined by the pitch diameter of its gear and the position of its home sensor, though these parameters are regularly verified (see below). The flip position is intrinsically more subject to drift because, over time, the Kevlar® string can stretch or slip and the Delrin® pulley and AGICO sample holder can wear. This drift is readily detected by measuring the standard sample in a full series of positions, and when it becomes appreciable, a recalibration of the motor positions is done. Because of wear, over the life of the instrument we have replaced the string a few times and the pulley and sample holder once.

[13] Determination of the flip positions for the measurement sequence is done by stepping the axial moment standard through approximately one complete flip motion and plotting the arctan(My/Mx) versus the motor step angle. If the system is operating correctly, this will be a straight line with a slope of one and an intercept equal to the angle between the home position and a zero moment in the Y measurement channel, i.e., when the sample's moment is perpendicular to the magnetometer's My axis. If the slope is not equal to one, a correction factor is calculated for the number of flip degrees in a motor step to be used in the next iteration of the flip positioning.

[14] The rotation positioning is very similar to the flip, but the axial standard's moment is first flipped perpendicular to the rotation axis. Determination of the correct flip and rotation positions are not entirely independent because an error in one can be partially compensated by a change in the other, so it is necessary to calibrate them in pairs and to iterate a few times until there is little change between the solutions.

[15] We conduct a full twelve-position measurement on a standard sample on a daily basis and before starting a series of measurements. The mean intensity and direction and the consistency among the 24 component measurements indicate whether the instrument is working properly. In practice the system is quite stable, requiring a minor positioning readjustment only once every month or two.

5. AF Demagnetizer

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Sample Holder and “Flip” Motion
  5. 3. Sample Translation and Rotation
  6. 4. Accurate Sample Positioning
  7. 5. AF Demagnetizer
  8. 6. Programming
  9. 7. Measurement Procedures
  10. 8. Performance and Concluding Remarks
  11. Acknowledgments
  12. References

[16] The AF demagnetization is carried out with a Sapphire Instruments SI-4 model. This instrument uses a simple solenoid to produce an axial alternating magnetic field that can be programmed from 0 to 0.2 T in steps of 0.1 mT, and, if desired, an anhysteretic remanent magnetization (ARM) DC field of up to 0.1 mT. Because the sample can flip to put any one of the three orthogonal sample directions along the solenoid's axis, there is no need to produce a transverse field for three-axis demagnetization: AF demagneization can be achieved easily and well in the uniform demagnetizing field produced by the SI-4's solenoid. The demagnetizing coil is housed inside three nested Magnetic Shield Corp CONETIC AA shields, which reduce the 300-nT residual magnetic field inside our shielded room to insignificance. Since the sample position and all the AF parameters are programmable, any number of demagnetization sequences and steps may be easily accommodated. Currently, we use a sequence that demagnetizes along each of the sample's three orthogonal axes in a different order for each successive step, which reduces the effect of gyroremanent magnetization, but we are in the process of automating the more elaborate procedure of Stephenson [1993] to use when needed to achieve a more complete elimination of GRM. In addition, since we can place many sample directions along the solenoids axis, we also plan to implement an automated sequence to measure anisotropy of anhysteretic remnant magnetization, as P. Roperch (personal communication, 2008) has already done with a JR5 holder.

6. Programming

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Sample Holder and “Flip” Motion
  5. 3. Sample Translation and Rotation
  6. 4. Accurate Sample Positioning
  7. 5. AF Demagnetizer
  8. 6. Programming
  9. 7. Measurement Procedures
  10. 8. Performance and Concluding Remarks
  11. Acknowledgments
  12. References

[17] The control program for the magnetometer/AF demagnetizer system was written using a combination of Java with the Swing API for the user interface and C for the instrument control routines. The Galil motor controller PC card, which controls the sample handler motions, accepts ASCII commands. Computer control interfaces with each of the three magnetometer channels via ASCII commands over an USB-to-quad RS232 serial adapter. The AF demagnetizer is commanded with a sequence of integers sent via RS232.

[18] We chose to store sample data in a new XML-based file format. This file format eliminates the problems that can occur because of fixed space or delimited data fields. A practical advantage is that data writing and retrieval is not sensitive to the number of tabs, spaces, decimal places, etc. used in the various data fields, reducing the potential for text-parsing errors that might otherwise occur if it were to become necessary to manually edit a sample file. It is designed to store data from an entire study area in one file, so a hierarchy of levels is contained within each file. These levels, from the top down, are locality, site, sample, and demagnetization step. The locality is self-explanatory and may be composed of each stratigraphic section within the study area. Within each locality there may be many sites, for example separate lava flows or sedimentary layers. Within each site there may be many samples, generally standard paleomagnetic cores (cylinders) or cubes. When fully analyzed, each sample will contain a series of demagnetization steps. This file format also allows the user to electronically record many more parameters than those strictly needed to measure the direction of the magnetic moment of the core and analyze the results. It is meant to serve as an electronic field notebook, containing space for comments about the locality, site, sample, and demagnetization step. It also permits new parameters to be added to the file format while still enabling older files to be used with the modified program. This allows more features, including analysis functions, to be incorporated into the program. For instance, the program includes a Sun compass routine to correct sample orientations parameters for local magnetic anomalies. Because of the diversity of paleomagnetic data analysis programs and file formats, several file export functions have been written and others can be added.

[19] One of the goals of the system control software was to have an application with a standard Windows graphical user interface, so users could more quickly become familiar with the program. Additional major software features include real-time visualization in Zijderveld, Equal Area, and J/J0 resizable and printable plots; the display may show data in core, geographic, and stratigraphic coordinates. Various measurement routines (CRYO1, 4, 6, or 12 measuring 1, 4, 6, or 12 positions respectively), as well as basic instrument interface routines, are selectable via pull-down menus. These routines permit low-level operations, such as independently moving the sample handler, operating the AF demagnetizer, or taking a single reading on the magnetometer.

[20] All system positioning information resides in a separate text file, allowing easy access to any sample handling parameters for inspection or manual editing. System control software also incorporates many features to maximize measurement accuracy and precision, such as automatic subtraction of a drifting magnetometer baseline (assumed linear), and a user option to subtract the signal of the sample holder at each of the measurement positions in the measuring routine. Recently, we implemented a suggestion by our colleagues at UC San Diego to read all three of the measurement axes of the magnetometer nearly simultaneously instead of in serially. Each of the three channels of the magnetometer is now connected to the computer on separate, individually addressable RS232 cables, instead of being daisy chained as was done previously. This allows read commands to be sent to each channel and results to be returned more quickly in sequence.

[21] The program was designed to be adaptable to work with other magnetometer systems. We are distributing the program and source code under version 3 of the GNU GPL license. For more information, please see http://es.ucsc.edu/∼emorris/cryoslug.

7. Measurement Procedures

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Sample Holder and “Flip” Motion
  5. 3. Sample Translation and Rotation
  6. 4. Accurate Sample Positioning
  7. 5. AF Demagnetizer
  8. 6. Programming
  9. 7. Measurement Procedures
  10. 8. Performance and Concluding Remarks
  11. Acknowledgments
  12. References

[22] We find it is often important to measure complementary pairs of orientations to cancel out induced magnetic components and minimize the effects of inhomogeneities in the magnetization of the sample. The redundancy also helps average out some kinds of positioning errors. By flipping the sample 0°, 120°, and 240° and rotating 0°, 90°, 180° and 270°, we are able to place each of the sample's X, Y and Z moment components along the magnetometer's ±Mx and ±My directions. However, no combination of rotation and flips can place any of the sample's axes in both the +Mz and −Mz axial measurement direction. Because of this limitation, we only use the magnetometer's Mx and My channels.

[23] With this holder, we typically use a 12-position measurement program (CRYO12), which includes four rotation positions 90° apart for each of the three flip positions, and yields 24 independent measurements of the sample components. The instrument baseline is measured both at the beginning and end of the 12-position sequence and a linear drift correction is applied. Consistency among the readings is evaluated in various ways, including an overall angular statistic of quality, Q95, defined by Briden and Arthur [1981] and named α95 by them. We have also implemented a six-position measurement that we can use for well-behaved, strongly magnetized samples. We use the system described here to carry out fully automatic measurements and AF demagnetization of more than half of our standard 10.5 cm3 paleomagnetic samples, those with magnetizations ranging down to a few times 10−4 A/m. By using a sample holder subtraction routine we can measure with good accuracy down to the lower end of the range. In theory, samples stronger than 102 A/m can be measured by slowing down the sample movements, so the magnetometer's electronics do not lose flux counts, but in practice, we measure samples stronger than 30 A/m on a spinner magnetometer.

[24] For weaker samples we measure and AF demagnetize manually using a standard plastic holder with moment 10−11 Am2, and for exceptionally weak samples, a glass holder with an even smaller moment. A four-position measurement sequence, using all three magnetometer channels and involving one manual manipulation of the sample, works well for both standard cylindrical and 2.54-cm cubic samples. Smaller sample sizes are easily accommodated with inserts machined from low–magnetic moment materials.

8. Performance and Concluding Remarks

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Sample Holder and “Flip” Motion
  5. 3. Sample Translation and Rotation
  6. 4. Accurate Sample Positioning
  7. 5. AF Demagnetizer
  8. 6. Programming
  9. 7. Measurement Procedures
  10. 8. Performance and Concluding Remarks
  11. Acknowledgments
  12. References

[25] The automated measuring system produces accurate and highly repeatable determinations of sample magnetization. A test of the automated sample handler comprising 137 twelve-position repeat measurements of a cylindrical, axially magnetized calibration standard yielded an average direction 0.5° from the sample axis: I = 89.5, D = 41.7 in core coordinates, with precision parameter over 500,000 and angular standard deviation of 0.11°. The standard deviation of the absolute moment was 0.05%. The average Q95 for each measurement was 0.8° (which indicates that this quality statistic is a conservative estimate of the measurement precision). Effects of positioning errors due to string stretch, backlash, stepper motor resolution, and random effects are clearly minimal.

[26] For this paper we ran an extraordinarily detailed automated AF demagnetization sequence on a Mono Lake sediment sample that was deposited soon after the geomagnetic excursion recorded there [e.g., Liddicoat and Coe, 1979]. Figures 4a4c show the results of the demagnetization from 0 to 160 mT in steps of 1 mT. Least square fits to the characteristic component from 40 to 160 mT, both unanchored and anchored to the origin, yield tightly constrained directions within 0.5° of each other and fit parameters (MAD [Kirschvink, 1980]) of 1.4 and 0.8°, respectively. The angle δ between the direction of successive steps is small during most of the demagnetization, averaging only 1° up to 138 mT (90% demagnetized). The last stage of demagnetization, from 138 to 160 mT (from 90 to 95% demagnetized), is significantly noisier, with an average δ of 8°, suggesting a modest amount of gyroremanent magnetization imparted by the alternating field [Stephenson, 1980, 1993].

image

Figure 4. (a) Zijderveld diagram in core coordinates showing detailed alternating field demagnetization of Mono Lake sediment sample 1404C from 0 to 160 mT in 1 mT increments. The least squares fitting line is from 40 to 160 mT. Fits that are unanchored and anchored to the origin yield directions within 0.5° of each other with MAD values of 1.4 and 0.8°, respectively. (b) Equal area projection in core coordinates of the demagnetization described in Figure 4a. (c) Normalized magnetization versus alternating field during the demagnetization described in Figure 4a.

Download figure to PowerPoint

[27] Our automated multiposition sample handler for a 2G superconducting rock magnetometer with single-solenoid alternating field demagnetization has seen heavy use for the past 6 years, measuring many thousands of samples. Multiple repeat measurements of standard sample and very detailed AF demagnetization sequences, such as those described above, demonstrate the very high quality data that it can produce. Many more examples of the data that we have obtained with it can be found in the studies of Pluhar et al. [2005, 2006a, 2006b], Busby et al. [2008] and Jarboe et al. [2008] on structure, tectonics and behavior of the ancient magnetic field.

Acknowledgments

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Sample Holder and “Flip” Motion
  5. 3. Sample Translation and Rotation
  6. 4. Accurate Sample Positioning
  7. 5. AF Demagnetizer
  8. 6. Programming
  9. 7. Measurement Procedures
  10. 8. Performance and Concluding Remarks
  11. Acknowledgments
  12. References

[28] The authors would like to acknowledge the enormous contributions of William S. Goree (1935–2007) in developing and improving the 2G SRM and to thank him for his generous help, encouragement, and support for more than 2 decades. We would also like to thank all the researchers that have contributed ideas, encouragement, inspiration, and criticism during this project. We would especially like to thank Darryl Smith, Dave Thayer, and Frits van Dyke of the UCSC machine shop for their expert assistance; Joe Liddicoat for his gifts in support of the UCSC paleomagnetic laboratory; and Shawn Wheelock for his role in the initial programming. NSF grants EAR-0310316 and EAR-0711418 have also provided partial support for this project.

References

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Sample Holder and “Flip” Motion
  5. 3. Sample Translation and Rotation
  6. 4. Accurate Sample Positioning
  7. 5. AF Demagnetizer
  8. 6. Programming
  9. 7. Measurement Procedures
  10. 8. Performance and Concluding Remarks
  11. Acknowledgments
  12. References