Magnetic particle AC versus DC magnetization
NDT literature including training materials and example exam questions frequently refers to a superiority of AC magnetization over DC magnetization when inspecting for surface breaking (in service) defects frequently with no offer of supporting scientific or empirical evidence.
Supporting evidence often includes a statement that the skin effect "concentrates" the magnetic field at the surface, implying that the total field has moved to the surface. The reality is that the skin effect is a loss to the system caused by the induced eddy currents.
An often quoted reference is the Carl Betz classic "Principles of Magnetic Particle Testing". A demonstration comparing wave forms for magnetizing a hollow pin sporting induced quench cracks of random direction and magnitude is magnetized using a central conductor. The applied current is increased by root mean square (RMS) ratio (1.41:1) and photographs of the indications are tabulated. A declaration the AC case is RMS value, the magnetizing value not having been measured and used for comparison with the DC cases. This places all the AC cases in the wrong lines in the table. If a correction is made to compare magnetizing currents in the process of magnetizing the item the DC indications are remarkably stronger than with the AC indications. This result would reflect the skin effect attenuation (concentration?) occurring from the inner surface close to the magnetizing force through the part to the observed surface.
Can anyone advise other explanations which may support claims of waveform superiority and in some cases requirement in standards or procedures to use AC wave form to inspect for surface breaking defects?
Birring NDE Center, Inc., USA, Joined Aug 2011, 832
Re: Magnetic particle AC versus DC magnetization In Reply to Keith Brewer at 23:33 Sep-06-2019 (Opening).
Re: Magnetic particle AC versus DC magnetization In Reply to Anmol Birring at 20:58 Sep-08-2019 .
Thank you for your response Anmol.
I checked the reference, seems mainly to deal with current flow, but magnetic field is referred to: "Since the skin effect causes a current at high frequencies to flow mainly at the surface of a conductor, it can be seen that this will reduce the magnetic field inside the wire, that is, beneath the depth at which the bulk of the current flows."
This is claiming a reduction with depth rather than an increase at surface and does not really justify the claims for enhancement with AC for surface defects.
My understanding has always been that the field at the surface is dependent on the magnitude of the current flowing within a body and independent of the distribution of that current? See various diagrams for field distribution for instance for hollow and solid conductors both magnetic and non magnetic. So it is the same for AC as for DC?
I do notice that a diagram for distribution in hollow magnetic cylinders magnetized by AC current with central conductor has not been featured in the literature? This analysis would help people to analyse the Betz experiment. I also notice that he refers to the comparison of the results rather that a direct claim that AC is superior?
R & D,
John Deere - Moline Technology Innovation Center, USA, Joined Jul 2011, 191
Re: Magnetic particle AC versus DC magnetization In Reply to Keith Brewer at 01:30 Sep-09-2019 .
NDT practitioners are rarely electrical engineers or physicists who might determine the "scientific" theory behind results (for example, deriving from Maxwell’s equations), and there are several examples of conflicting opinions even when these scientific theories are offered. We can turn to empirical results, but unfortunately these often vary hugely based on specifics of the application, nature of the flaw under study, human factors, and other aspects. An example of this detection variation that comes to mind is a comparison of a 1998 American Petroleum Institute recommended practice that stated an AC electromagnetic yoke could reliably detect surface-breaking defects 0.236" long, while a 2000 preliminary report-out of an Association of American Railroads study suggested that a flaw needed to be longer than 3" to be reliably detected with MT.
From what I've read, the most convincing argument seems to be that AC current’s superiority for surface-breaking flaws is due to its enhanced particle mobility – caused by the strong pulsation in the waveform. Some have concluded that skin effect is a poor argument for AC's surface-flaw superiority, and surface tangential magnetic field strength was independent of waveform. One researcher published a paper showing how the tangential magnetic leakage field at a discontinuity was weaker than that caused by other waveforms (thus supporting the pulsing / particle mobility argument).
Ammeters on wet benches come in a few flavors (peak, RMS, average), so direct comparison of waveforms isn’t trivial. From an empirical standpoint, several authors have compared detectability from AC and various forms of “DC”. One example was done on pipe to seek surface-breaking stress corrosion cracking. That team used a variety of surface preparation techniques, color contrast and fluorescent particles, and two magnetizing waveforms with portable yokes. In their study they found that in general AC could detect flaws of around half the length possible with "pulsed DC". A couple of others discussed how AC required about half the current used with HWDC (FWDC required higher current setting than HWDC).
Taking the pulsation idea a bit further, 1-phase rectified current has more than 3-phased rectified current. In 2006 work published by one equipment manufacturer they showed a strong difference in indication formation between 1-phase and 3-phase rectified current when the wet bench's ammeter showed 500 amps.
New Zealand, Joined Sep 2019, 8
Re: Magnetic particle AC versus DC magnetization In Reply to Rick Lopez at 22:41 Sep-10-2019 .
Thanks for your response Rick
The particle mobility argument is negated by the comparative performance using half wave rectified current and is not relevant at all for wet particles, the carrier fluid providing adequate mobility and dry particles having no mobility benefit if used and are not intended to be used in the wet process. It is hard to find a portable or mobile unit designed for weld testing which offers AC only.
The literature quoting observations of superiority for AC magnetization seem to compare RMS current levels (magnetizing amps????) for AC with DC current levels when the true magnetizing level is peak current flow. If peak values for each case are compared it shows no advantage for AC, and in the case of the quoted Betz demonstration it shows, as expected inferior results for AC (exterior surface examined with central conductor), with lesser indications from the attenuated field which has produced the indications. It is interesting that RMS values were chosen for the demonstration which misplaces each AC results to the next lower DC value, i.e. 420 amperes peak AC is compared to 300 amperes DC. This set up is clearly explained by Betz. If the correct comparison is made by comparing 420 amperes peak AC with 420 amperes DC the difference is quite remarkable and is explained by the skin effect attenuation of the magnetic field with depth through to the outer surface.
The end result is that AC appears to be superior because we are applying the wrong values for the comparison.
The most relied upon publication regarding the practicle application of magnetic particle test is still arguably the 1967 edition of PRINCIPLES OF MAGNETIC PARTICLE TESTING. It is commonly used as source and reference for technician qualification examination questions, and quoted in training documents.
The arguments presented in Chapter 12.5 are worthy of careful consideration.
The originating publication, PRINCIPLES OF MAGNAFLUX by F B Doane quotes “For surface defects, in general, the two methods are equally effective and sensitive.”
A later edition co-authored by F B Doane and Carl Betz introduces experimental results for varying current values inspecting a hollow piston pin using a central conductor. The argument uses meter values for current through the conductor at a series of steps increasing by RMS ratio. The values for AC are stated to be RMS meter reading, the DC options peak reading. The arguments include “The peak of the AC wave is 1.41 times the DC value, and this peak, although extending over a short interval of time, should and does produce a higher magnetization at the surface where there are no complications due to eddy currents.”
This argument fails on the basis that the set up used has the unaffected surface on the bore surface rather than the surface presented in the experiment which is indeed affected by the eddy currents.
The comparison on the basis of RMS for AC vs peak for DC meter readings can hardly be called “normal”, and certainly is not “logical”. The heating from eddy current losses with the accompanying increase in impedance in the AC case is demonstrated simply by the feel of the conductor when AC is used. Obviously a higher voltage is required to achieve the same peak output from AC rather than DC.
“It will be seen that the AC indications are much superior to the corresponding DC indications”. If the demonstrated results are adjusted to reflect peak value comparison, conveniently in this case involving a simple RMS ratio step shift the reverse case of DC being superior at each level, as expected with the skin effect attenuation through the thickness of the subject pin to the viewed surface is demonstrated. The experiment thus serves to show the reason for the restriction in ASTM E 1444 limiting the use of AC current for central conductor inspections to the bore of the item.
The former arguments of the demonstration supporting the conclusion are dropped for PRINCIPLES OF MAGNETIC PARTICLE TESTING. The readers are left to apply their own logic to a series of statements regarding the experiment but notwithstanding the facts available the popular false interpretation is that AC magnetization is superior to DC magnetization for surface breaking defects.
The same publication in Chapter 20 Section 5 (Page 370) includes an example of “high sensitivity techniques – ie., strong direct currents with circular magnetization, and wet magnetic particles.”
NDT literature, including training materials and example exam questions frequently refers to a superiority of AC magnetization over DC magnetization when inspecting for surface breaking (in service) defects often with no offer of supporting scientific or empirical evidence.
Supporting evidence that is offered often includes a statement that the skin effect "concentrates" the magnetic field at the surface, implying that the total field has moved to the surface. The reality is that the skin effect is a loss to the system caused by the induced eddy currents.
New Zealand, Joined Sep 2019, 8
Re: Magnetic particle AC versus DC magnetization In Reply to Keith Brewer at 00:08 Oct-05-2019 .
I have discovered a technical post at "https://physics.stackexchange.com/questions/401056/the-magnetic-field-inside-a-ac-current-carrying-wire" which quotes "Moreover, given the same wire and the same total current, the magnetic field inside the wire carrying an AC current of sufficiently high frequency would be lower than the magnetic field inside the wire carrying a DC current, everywhere in the wire except on its surface, where the magnetic fields of AC and DC currents would equalize, since at that radius all the current in the wire will be encircled in both cases". Perhaps I have missed something in the argument?