Summary

The port call, transit, and tie-up on 396T allowed for many projects, upgrades, and improvements in the Paleomag lab. The main activities were IMS 12 was both tested and implemented, and a new water cooled Haskris was installed for the Cryomech Compressor. In addition, some basic diagnostic experiments were conducted throughout the lab (SRM orientations, fluxgate, and null field protocol).

Measurements

Pmag Play Cubes (P-cubes)

A set of control samples were made on 390C that could be run on both the SRM and the JR-6 for cross calibration, but that also could be standards subjected to magnetization and demagnetization and thus returned to "known" magnetic-states. Samples were collected into J-cubes from two different sources of play core mud (Mud A: MA01-04, Mud B: MB-01-02, and a mixture of the two: MC01-02). These samples were then dried out in the Thermal Demagnetizer at 120°C and the voids were filled in with epoxy. The resulting suite of standards (referred to as Pmag Play Cubes or P-Cubes) can be subjected to various magnetic experiments and afterwards demagnetized and then re-magnetized to the original magnetic state. 

The P-cubes were used to test orientations of working verses archive section halves and verify if the IMS transformations are correct. The samples were given 30 mT IRM and then oriented along (+/-) x, y, and z SRM axes for the experiments (at least sample in each polarity for every axis). The samples were measured as an archive SHLF, then flipped 180° around the z axis, the transformation required to take the upper working half of a whole round after splitting and placing it into to the SRM sample tray. This new W-SHLF was measured as an archive half and then again as a working half. The orientations came out as expect: Z always remained the same, but X and Y were inverted when the W-SHLF was measured "incorrectly" in the archive frame of reference (FOR), but were restored to the correct orientations when measured in the working FOR (Figures 1a-c). 

It appears that IMS is doing the transformations correctly, however if there is an error in the transformation of the drift correction and with samples this magnetically strong, it likely would not be noticeable . This experiment was repeated after the IMS12 update and gave consistent results.

Open	Open	Open
Figure 1a: P-cubes measured in the original orientation as an archive sample showing the "absolute" orientations.	Figure 1b: P-cubes reoriented to be the corresponding working half, but still measured as an archive half sample. Notice that for the relative orientations, the X and Y are negated while Z remains unchanged. This is expected for a 180° rotation about the Z-axis.	Figure 1c: P-cubes still oriented to be the corresponding working half, but now measured as a working half sample. Here X and Y are restored to their correct absolute orientations because IMS applied the correct transformation matrix to the samples [ -1, 0, 0; 0, -1, 0; 0, 0, 1].

Comments and Issues

SRM

During installation of the new water cooled Haskris for the Cryomech Compressor, the system unavoidable warmed to room temperature and requiring re-trapping of the null field. Since there was not any science expedition sailing on 396T and thus immediacy for put the SRM back into service, it was an opportunity opportunity for cross training and experimentation on the field trapping the protocol.

Internal Null Field

The SRM was trapped several times during 396T and the final field trapped in Cape Town after using the modified protocol outlined below was less than +/- 0.5 nT (as measured during trapping and around +/- 1 nT when measured afterwards- Figure 2). This protocol has not been updated to confluence and requires further testing and verification. Also while it seems very promising, is also reaching the limits of the Fluxgate's sensitivity. Furthermore, it should be noted that optimizing for one verify specific ship orientation and zero motion in port does not necessarily mean the system is optimized for the orientations and conditions that will be experienced at sea. An analogy would be tuning a piano: you can either tune (optimize) it exactly according to physics and it will sound fairly off in every key; you can tune it perfect for one key and it will sound perfect in that key but horrible in every other key; or temper the tuning (optimize over an average) and it will never sound perfect but sounds really good regardless of the key it is played in.

Open

Figure 2: The SRM profile SQUID region final trap in Cape Town with IMS 12 and the DAFI output file update. Not the low field values but high gradients.

The cable for the nulling coils were showing wear and tear from being set up and taken down for every field trap. It was also noticed that the typically routing drapes the cable directly over the SQUID region and the cable had minimal shielding. Heavier shielding was added to the cable and routed underneath the SMR. Since the cable now remains in place and the control box unit is turned off and on for SRM vital checks, it is best to not leave the cable connected to the coils. A "dummy" connector box was added to plug the cable's leads into when not in use. The input jacks for the nulling coils were loose again. Washers were added to the inside panel nuts (and lock-tite) so that they could be properly tightened and not come loose from plugging and unplugging the leads. 

Gradient coil

The field gradient across the SQUID region varies substantially between expeditions but was successfully minimize using the gradient coil on 395P (see 395P Pmag tech report). However is a difficult process since the Fluxgate probe needs to be moved to one side of the SQUIDs to measure the field, then moved to the other side to measure again, the gradient approximated, the potentiometer adjust, the gradient remeasured again, and this process iterated until the gradient is minimized. Furthermore this is done while the shield above the superconducting temperature but cooling towards the 7K threshold, and the axes coils still need their own adjustment as well. David Schuler from Applied Physics acknowledged both the importance of adjusting the gradient as well as the difficulty of doing so, and continued to state that most clients had difficulty in accomplishing the task and tending to make the gradient worse. 

To make the process easier, the IMS 10 DAFI code was adapted to allow the measurement region to be rapidly repeated by only clicking start for as many times as desired and all of the data saved into individual DAFI files. (Note: this change to the DAFI code was only made to IMS 10 and not IMS 12 and done with the ability to roll back in case IMS 12 did not function properly.) All 4 voltages supplies to the coils were measured to sweep from -15V to +15V in a total of ~15 turns of the potentiometer (2V/turn). The gradient supply was set to -15V, a DAFI measurement was started and iterated each time increasing the voltage by 1V (1/2 turn). No changes to the fields were observed over the entire voltage range and the gradient was very large. The gradient gradient could be so strong that the adjustment range available was insufficient to affect the actual gradient or something could be wrong with the gradient coil (sense we know that the voltage supply is fine). All 4 coils' resistance were measured and the vertical and horizontal were 10 Ω and the axial was 5 Ω, but the gradient coil was infinite and neither end connected to ground. This implies that it is likely an open circuit and not functioning. A follow up email was sent to Schuler and after leaving the ship, the reply was that there should be resistance across the gradient jacks, and likely a broken connection near the jacks. This is plausible considering the issues with the jacks being loose.

Axes Labeling

The labeling of the axes for the nulling setup are overly complex and very confusing: X, Y, and Z cables are connected to the nulling coil jacks (according to the labels), but then X, Y, and Z on the control unit don't adjust X, Y, and Z on the Fluxgate unit, neither of these correspond to the SRM's axes, and a series of yellow sticker labels translate between each system. In reality only the SRM and the Fluxgate axes matter for adjusting the field and differ by the X and Z axes being swapped (ignoring polarity) and the intermediate connections should just be connected for consistency. To do this the Fluxgate probe was hooked up and centered in the SQUID region. One set of leads from the control unit was plugged into each null coil input and the corresponding pot adjust to observe which Fluxgate axis it adjusted. From this the coil inputs were labeled so that control unit X, Y, and Z adjusts Fluxgate X, Y, and Z. And now since only X and Z swap between Fluxgate and the SRM we can either swap the X and Z BNC cables connecting to DAFI or swap the data in the software or post. Since the point is to avoid confusion, it is better to plug the BNC cables as normally (X, Y, and Z, to X, Y, and Z). Since the DAFI output file is only for the labs own internal use and never uploaded to LIMS or used in any calculations, 3 addition columns of data were added to be in the SRM frame of reference. These columns are the exact same data as the original Fluxgate data column, but just the order switch and the labels changed. The end user can now choose to view the SRM profile in either the Fluxgate or the SRM frame of reference without having to logic through the axes change. This change to DAFI was tested and implemented to the DAFI portion of IMS 12.

An additional benefit to this change is that the USB6008 utility can be simply used to assist in field trapping, with the X, Y, and Z remain consistent from the pot adjusted to the values displayed by the Utility, but the data saved has it in both the Fluxgate and SRM frames of reference. And since the USB6008 allows for changing the averaging time, trapping while experiencing motion at sea becomes much easier (slow adjustments with averaging times longer that the period of the motion) and the concept mentioned earlier of optimizing for range of conditions (orientations and motion) becomes a possibility. (NOTE: there is a slight offset in the LCD display values of the Fluxgate unit and the USB6008, likely due to ADC discrepancies. This can be accounted for by using the USB6008 values when zeroing the problem, which again should be easier with the ability to time average).

Other than the additional columns in the DAFI output file, none of these changes have been made on confluence and the labeling is temporary. This new protocol can be tried out, feedback given, and modified on the next few expeditions before any permanent changes are made, but already is much less confusing than before.

Fluxgate Induced Magnetization

There have been several instances of the Fluxgate probe and cable having remanent magnetization, and the effect this would have on the null field explored in pain full detail on 395P (see 395P Pmag tech report). At the beginning of 396T, Probe A and cable were measured (with cable disconnected and flaked on the tray) and showed the typically slight remanent magnetization of ~10e-1 A/m and then 10e-2 A/m after demagnetization (Figures 3a-b).

Open	Open
Figures 3a: The cable (coupling nuts) and probe showing a moderate remanent magnetization.	Figures 3b: Again after 80 mT demagnetization with the remanent magnetization an order of magnitude weaker.

Again this is should not really be a problem and was shown mathematically to only be an issue if in the samples had a large and variable susceptibility as the induced component (M_i) would be significant and changing. On 396T the probe was mistakenly measured while plugged in and shown to have a large induced magnetization (Figures 4a). The probe was then measured connected with power off, and again with the cable unplugged from the Fluxgate unit, but still plugged into the probe and outside of the SRM (Figures 4b-c). Will the source of this isn't understood, it is a significant magnetization that needs to be considered. 

Open	Open	Open
Figure 4a: The probe measured connected with the power on.	Figure 4b: Measured connected and the power off.	Figure 4c: And again with the cable connected to the probe and dangling outside the SRM.

From the 395P tech report:

B_srm, defined as the internal field of the SRM, can be expressed as

where B_ex is the net field (effective field) of all external fields and B_sh is the field induced and trapped in the super conducting shield.

When we trap the “null-field”, we try to force B_srm => 0, achieved by adjusting B_sh such that  B_sh = - B_ex , so that B_srm => 0.

But since it has been shown that the Fluxgate probe can have a magnetization, then

Where B_fg is the field from the magnetization of the Fluxgate probe itself (probe and cable connectors).

Ideally B_fg = 0,  or is negligible, and for all practical purposed equation 1 holds. However we are now seeing that the fluxgate probe can be actively magnetized.

Thus when trapping the null-field, B_srm = > 0, which now means 

B_srm = B_ex + B_sh + B_fg = 0  (Equation 3)

What was not considered on 395P is that when zeroing the probe, and remanent or active field is also present and thus the probe is zeroed against probe's field so that and that so what is consider "0" is actually B_fg  or 0 = B_fg and equation 3 becomes 

And now when the fluxgate probe is remove, so is its field (subtracted from both sides)

So B_fg from both the remanent and induced magnetization do not matter as long as the probe is accurately zeroed and SRM's field can theoretically go to zero....

VISA Aliases and COMs

VISA Alias(es) are how IMS identifies each device and the COMs and ASRL#s are assigned by the computer. After certain events (i.e. OS updates), the computer may reassign the COMs and ASRL; this does not matter as long as the aliases are correct. This was clarified in the SRM user guide as it has caused confusion in the past. The remaining issue is that Cryowatch is currently not using the "CRYOMECH" alias and has "COM 5" entered somewhere in its vi files. This needs to be changed and the Cryowatch recompiled. Because of this, Cryowatch must be started after IMS when logging in or restarting then computer to avoid COMs issues.

SRM Track Motion

It has been noted that the track jerks abruptly when finding home prior to operations, with hard rock this can cause many issues of sample movement and potential physical damage to the drive train when under a heavy load. A second limit switch was added to the aft end of the SRM track and homing is now done in IMS via the CW limit switch and the home switch and the jerking motion is basically negligible. The changes were added to the user guides.

SRM basic checks

The IMS-10 was used to impart a 30 mT IRM on the P-mag P-cubes to test the Z, Y, and Z of the SRM as well as track position before and after the IMS 12 update in the same manner described above for the orientation experiments. In both cases, all three SRM axes, moments, and track positioning were correct. 

AF Demagnetization Coil

While the three AF demagnetization coils were not explicitly measure with the Gauss Meter (hall probes), they functions correctly and consistently before and after the 

IMS

IMS 12 was extensively tested and then implemented. Various new issues came up and were fixed, and many old issues remain (e.g. "Offline Treatments" not working, etc). The motion control parameters are slightly different and appear correct(reasons related to the new X-Scan) and the parameter values are updated on confluence. Each SQUID (x, y, and z) and the Degausser System Controller are now separate units with their serial numbers logged in IMS.

The orientation experiments on the P-Cubes, as outline above, was repeated after the IMS 12 install.  This verified SRM was recording the correct magnitudes, orientations of the SQUIDs and samples, as well as verifying that the correct transformation operations between the working and archive frame of references.

The DAFI output file was slight modified (as mentioned above) to include 3 additional columns of data, which is the same as the Fluxgate nT columns but relabeled in the SRM frame of reference as discussed as above. No addition calculations were made and the original Fluxgate data columns remain completely untouched. It should also be emphasized that the DAFI output file are for our own internal diagnostic use only and are never uploaded to LIMS.

Cryomech and Chill Water System

The new water-cooled Haskris was installed and the ship chill-water is no longer used to directly chill the Cryomech but instead the Haskris. The air-cooled Haskris remains as the backup system and both systems were plumbed using stainless steel hard lines with 3 way selector valves. A purge line was add to assist in maintenance on the lines instead of only cleaning the reservoirs. The Cryomech and hard lines were extensively cleaned and flushed until all of the build up was removed.

Since both chill systems are closed loops, the Cryomech's low flow warning no longer goes off when the ship's chill-water is lost. Currently only the yellow warning screen happens on Crywatch when something is wrong. Some sort of audible alarm will be necessary in the near future otherwise it's highly likely that most cooling issues will not be caught in time to keep the system super conducting.

All of the changes and new protocols were updated on confluence.

Lab General

Pmag library was organized. All manuals put into individual binders with redundant and legacy manuals retained in one of the cubby spaces. All other literature was put into the respective binder for which it pertains to.

A keyboard and mouse were installed to the sample loading monitor.

Lab Computers

All lab computers were successfully updated. Prior issues with the SRM COMs are better understood and should no longer pose a problem.

ACS Impulse Magnetizer

The IM-10 and IM-10-30 functioned properly giving the P-Cubes IRMs.

Kappabridge KLY-4

This unit was not exercised during 396T but did  successfully respond to its software after the computer updates.

Kappabridge MFK2

This unit was not exercised nor tested during 396T.

JR-6 Spinner Magnetometer

The JR-6 was not exercised during 396T.

D-2000 AF Demagnetizer

The D-2000 was not exercised during 396T.

Thermal Demagnetizer

The thermal demagnetizer not exercised during 396T.