Weld metal ferrite secondary standards for calibrating magnetic instruments

TWI Bulletin, July - August 1996

Brian Ginn joined TWI (BWRA) in 1960 after completing a 5 year apprenticeship as a foundry engineer. He is currently employed in the Materials Department as a project leader in the Corrosion and Analytical Section.

Control of the ferrite content in stainless steel weld metals is essential. Brian Ginn describes TWI's role in an international project to produce weld metal ferrite secondary standards used to ensure conformity of measurement.

Control of the ferrite content in welds is required with both conventional austenitic stainless steels and also duplex ferritic-austenitic grades. ^[1] In both cases, the mechanical properties and corrosion resistance of the joint are highly dependent on the amount of ferrite present, while the susceptibility to 'hot cracking' of nominally austenitic steels is largely controlled by the presence or absence of ferrite. There is therefore a need to measure the proportion of ferrite in weld metals, to ensure that specification requirements are met, and optimum properties are obtained.

Determination of the ferrite content of nominally austenitic or of duplex ferritic-austenitic stainless steel alloys and weld metals is commonly undertaken using magnetic instruments. There are other techniques available, some of which are destructive, but they are, in general, indirect or less convenient to use. However, it is accepted that these cannot reliably determine the absolute volume percentage of ferrite present in a stainless steel weld deposit, and, that results obtained are subject to some degree of variability. ^[2]

Magnetic instruments give an indication of the amount of ferrite relative to austenite in a weld deposit by using the differences in electromagnetic properties of the two phases, austenite being non-magnetic. Magnetic properties which may be exploited include ferromagnetism and permeability, with ferrite determined by field strength or induction measurement, for example. The magnetic characteristics of the ferrite are influenced by its chemical composition, which will vary according to the alloying elements present ( ie on the grade of steel), and by the thermal history. This complicates assessment of ferrite level by a magnetic method, but, because a physical property is measured directly, the approach is of considerable potential advantage in avoiding the uncertainties associated with alternative techniques. This has lead to the widely accepted concept of describing the ferrite content of a stainless steel deposit in terms of the force needed to detach a permanent magnet of known strength from the weld metal concerned. The pull-off force is described in terms of a 'Ferrite Number' (FN) which is an empirical scale constructed originally for nominally austenitic materials so that the numerical value of FN was similar to the % volume of ferrite present.

It is necessary to calibrate these pull-off magnet instruments, and for this purpose primary standards having an assigned FN value are available from the American National Institute of Standards and Technology. They are marketed as coating thickness standards, and consist of a non-magnetic coating of specified thickness on a magnetic base. However, the primary standards are not considered sufficiently durable for production shop or on-site use. More importantly, their design and geometry render them unsuitable for calibrating other measuring instruments which use an alternative magnetic characteristic of ferrite and entail a different principle of operation.

A need therefore exists for secondary standards, which can be used to calibrate a wide range of types of ferrite measuring instruments. Secondary standards have been produced previously under the auspices of the International Institute of Welding (IIW), and have been available for a number of years. ^[3] In the past, these were prepared from submerged-arc strip cladding weld deposits, to cover the range 0-30FN, appropriate for nominally 'austenitic' weld deposits. A second series, applicable to duplex ferritic-austenitic weld deposits has now been manufactured by a casting process which gives a microstructure similar to that of a weld metal. This enables measurements of up to about 100FN.

Approach

The rapid growth in industrial application of duplex stainless steels in the early 1980s led to an interest in secondary standards which could calibrate instruments measuring ferrite levels over the range 30-70FN. In 1986, the IIW was instrumental in seeking a further source of suitable samples and in conducting evaluation trials via round robin measurements. It was acknowledged in producing samples that, as the phase balance in duplex stainless steel deposits was strongly influenced by the thermal cycle experienced, single bead specimens were necessary to obtain reasonably homogeneous material.

Samples assessed included those produced from deposits made using the flux-cored wire and the plasma arc processes. Unfortunately, neither type was suitable for calibrating magnetic saturation-type ferrite measuring instruments - a technique commonly used in Eastern Europe. Participation in the work of the IIW by the Russian delegation led to inclusion in the trials of samples produced by a centrifugal casting technique. After a period of evaluation and further series of round robin measurements, the suitability of these samples for use as weld metal ferrite secondary standards was confirmed: they displayed good uniformity of ferrite throughout the material, and gave acceptable reproducibility of ferrite measurement by a number of laboratories in different countries. The IIW therefore requested that secondary standards should be produced from suitable centrifugally cast alloys.

TWI was again asked to undertake all ferrite determinations and to market the boxed sets of standards. Samples were required to cover the ranges 0-30FN for conventional austenitic steels and 30-100FN for duplex grades. The former were necessary to fulfil requests received since stocks of the originally produced standards were exhausted.

Production of standard materials

Samples to which FN values were to be assigned were delivered to TWI, machined to final dimensions of 20 x 12 x 10mm, by Mladis Co (Russia). The samples had been extracted from centrifugally chill cast rings produced from alloys of varying chromium and nickel contents to control the ferrite/austenite balance, chromium ranging from 17.4 to 28.6%, and nickel from 10.3 to 5.4%. The as-cast rings had an outer diameter of 500mm and a wall thickness of 20mm. When supplied to TWI, the samples were grouped according to the manufacturer's approximate determination of ferrite content, marked with a unique reference number, and with one of the 20 x 12mm faces identified as the reference face.

Ferrite measurement on secondary standard samples

In assigning FN values to individual samples, each was subjected to measurement of ferrite content by calibrated Magne Gage instruments (Fig.1). As indicated above, this instrument measures the pull-off force necessary to detach a standardised permanent magnet from a test surface (Fig.2). The magnet is suspended from a balanced beam and the force needed to detach the magnet from the test surface is applied by hand via a coil spring. The suspension beam has provision for attachment of additional counterbalance weights which extend the instrument's working range. Without any additional counterbalance weights, the Magne Gage has a range of 0-30FN. A counterbalance weight of, for example, approximately 8g moves this range to 30-60FN, whilst 16g will, depending on the particular instrument, extend it to 60-90FN.

During the course of assigning FN values to the cast samples, some of which had values in excess of 100, counterbalance weights of up to 23g were used. Knowledge of the exact value of the weight is not important, provided that thesame one is used both to calibrate the instrument with the primary standards and to measure the sample. The value of the counterbalance weight used was, however, recorded on the individual sample data sheets for reference purposes.Calibration of a Magne Gage is by using primary coating thickness standards selected according to the desired instrument working range.

Ferrite measurements were made on individual cast samples at one central location on the reference face. In each case, Magne Gage readings were made in groups of 5, a total of 20 for each sample. These were entered on to a recordsheet, which, once completed, contained details of all measurements made on any one sample (Fig.3). All samples were measured by two operators, each using two separate calibrated Magne Gages. Calculated FN values were enteredon to the sample record sheet, and an average value determined. This value was assigned to the sample, and recorded on the record card accompanying each boxed set of standards assembled (Fig.4).

Conclusions

Acknowledgements

References

Boxed sets of these weld metal ferrite secondary standards are available from TWI. Contact Brian Ginn for more details.