Presenting magnetic susceptibility in SI units

Owned by jr_fackler
Last updated: Jun 22, 2018
2 min read

>> Blum to Barnes, email Exp 341  cataloged on Google Site Oct 8, 2013 <<
(1) My recommendation: use instrument unit (IU) (as reported from LIMS) and forget about conversion because…. 

Even if the “units” option on the meters is set to SI units, magnetic susceptibility from all three shipboard instruments is stored and reported in instrument units. To obtain results in dimensionless SI units, the instrument units need to be multiplied by a correction factor that is a function of the probe type, frequency setting, core diameter, and loop size. Because we are not measuring the core diameter d, and the function to SI is highly dependent on d (power of 3), application of a correction factor only means that we now need to add error bars for the uncertainty in core diameter. If normalized magnetic susceptibility values are needed, the best method is to measure discrete samples from characteristic materials with good accuracy and precision using a Kappabridge and apply a regression. 

(2) If shipboard scientists insist on having the SI label on the plots… 

The formula is on page 57 of 71 on an old operating manual found on the web amongst several Bartington documents and sites (attached): MS(SI) = MS(IU)/KREL*10-5, and KREL= 3.45*(d/D)^3, where 3.45 is a geometric factor for the MS2 loop system, D is the aperture diameter + 8 mm, and d the core diameter. For the 80 mm loop (D=88 mm), and core diameter between 70 and 60 mm for the range of stuffed APC cores to dry XCB cores and firm RCB cores, KREL varies from 0.57 to 0.91, respectively, the difference being 45% of the average. 

So unless you adjust the factor based on some kind of core diameter estimate, at least using a different diameter for each coring system, you have to multiply the IU with 7*10-6 and add an error envelope of +/-22%. 

For the 90 mm loop the factor changes accordingly, and if the frequency was modified from that used in the 3.45 factor, you would have to adjust for that too (don't know how, probably not very significant). 

So, it's much simpler to just live with the IU units until someone does a more rigorous analysis using a good set of Kappabridge measurements on cubes of known volume (or known mass and density). 

However, this is totally up to whoever has the strongest desire out there.