Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
BY A. W. BREWER AND R. L. M O M E N T
ABSTRACT. A technique for con- strength at low t e m p e r a t u r e s , a methods, but is subject to error from
structing standards containing fer- reduction of corrosion resistance in (1) poor etching behavior of some
ritic stainless steel in an austenite certain environments, and enhance alloys. (2) the counting technique e m -
matrix is reported. Such a standard formation of sigma phase at high t e m - ployed, and (3) the aspherical shape
more closely resembles the metal- peratures (Ref. 1). of ferrite particles, particularly in
lurgical structure of a weld than either Thus it becomes extremely impor- welds. While variations on the order of
iron powder in a nonmagnetic matrix tant to be able to measure the ferrite 1 % ferrite for levels up to 4%, and
or various nonmagnetic coatings over content in austenite at levels between 2% for levels near 10% have been
a ferrite base. 1 and 12%. Because no dependable achieved, metallographic methods
Also, a different type of measuring technique for ferrite measurement are time consuming and not applica-
instrument is evaluated which utilizes has been f o u n d , the Welding Re- ble for a nondestructive test.
the interaction of magnetic particles search Council (WRC) has adopted M e a s u r e m e n t of t h e m a g n e t i c
with an alternating magnetic field. an arbitrary procedure which e m - response of a weld has proved to be
Data taken with this device are c o m - ploys the Magne Gage and stan- the simplest nondestructive method
pared with those obtained using the dards made with various thicknesses and is in widespread use. Two instru-
more c o m m o n magnetic type instru- of a nonmagnetic coating over a mag- ments are used in the majority of
ments. The new instrument has three netic substrate (Ref. 2). Ferrite levels tests. The Severn Gage employs a
advantages: (1) faster collection of are reported as ferrite numbers (FN) balance beam and compares the at-
data, (2) greater sensitivity to small (Ref. 3). This standardization of traction of a permanent magnet on
point-to-point fluctuations, and (3) measurement technique allows for one end to the weld and to differ-
ability to be used on large and/or better comparison of results from ent standards. This results in a range
rough weld surfaces. various laboratories. The FN is essen- for the ferrite content bracketed by
tially equivalent to the ferrite content that of two standards. A more precise
Introduction up to about 8%, but differs by in- estimate is obtained by the Magne
creasing amounts above this level. Gage. The force of attraction between
The presence of delta ferrite in the There are numerous methods for a permanent magnet and the weld is
range of 4-8% has proven effective in measuring the ferrite content of a measured against that applied to a
controlling microcracking of welds in weld, all of which, under certain c o n - lever arm by a spring. Values are o b -
austenitic stainless steels during cool- ditions, may lead to incorrect results. tained relative to the standards used
ing (Ref. 1). Partially ferritic cast They can roughly be classified into to calibrate the spring force.
stainless steels have i m p r o v e d empirical, magnetic, and non- The weak link in these ferrite
welding characteristics and in- magnetic categories. Empirical measuring techniques is calibration.
creased strength over purely austen- studies of the ferrite contents of welds The orientation of lenticular ferrite
itic alloys. On the other hand, more as a function of nickel and chromium particles, which affects the magnetic
than 10% ferrite can contribute to a
equivalent alloying constituents have response, will vary within a weld, and
r e d u c t i o n in ductility and i m p a c t
established diagrams showing ap- thus welds made using identical
proximate ferrite content as a function materials will not always give the
of composition. The first such dia- same magnetic response. A homo-
gram was prepared by Schaeffler geneous array of spherical ferrite in a
A. W. BREWER is Senior Development (Ref. 4), and has recently been im- homogeneous matrix will not neces-
Specialist and R. L. MOMENT is Senior proved by DeLong (Ref. 5) and Hull sarily give the same response as a
Research Specialist, Rockwell Interna- (Ref. 6) to include the effect of nitro- weld having the same total ferrite c o n -
tional, Rocky Flats Plant, P.O. Box 464,
gen and other elements. tent. However, this latter type of stan-
Golden, Colorado 80401.
This work was performed under con- Nonmagnetic techniques include x- dard should be more easily repro-
tract AT (29-1 )-1106 for the Albuquerque ray diffraction, Mossbauer-effect and duced than that made from weld
Operations Office, U.S. Energy Research metallography. The latter one has beads.
and Development Administration. been used extensively to check other As noted above, the WRC has
160-s I J U N E 1 9 7 6
800 C. Longer pressing times or <a)
Table 1 — Data on Powder Pressed Ferrite Standards
higher temperatures resulted in the
precipitation of a third phase which
Den- Point
contained manganese and silicon, as
sity, count, Severn Magne Ferrite
identified by electron m i c r o p r o b e
Composition Series % % Gage Gage Meter
analysis. Densities of the pressed
samples ranged from 78.1 to 91.1 1.5% 430 in 316 #2 81.7 1.49 0.5- 1.0 0.35 0.35
percent of the theoretical value. 2.0% 430 in 316 #1 85.0 — 0.5- 1.0 0.3 0.5
Figures 2 and 3 show scanning 3.5% 430 in 316 #1 83.0 — 1.5- 2.5 1.2 1.2
3.5% 430 in 316 #2 79.1 3.2 1.5- 2.5 1.1 1.0
electron m i c r o s c o p e p h o t o m i c r o -
5.0% 430 in 316 #1 79.0 5.0 2.5- 4.0 2.2 1.8
graphs of the metal powders before 2.1
5.0% 430 in 316 #2 78.6 4.46 2.5- 4.0 1.5
pressing. They are essentially iden- 5.0% F e i n 316 91.1 4.4 7.5-10.0 6.1 4.3
tical in shape and size. The uni- 7.5% 430 in 316 #1 84.0 7.5 7.5-10.0 2.9 3.3
formity of distribution of the 430 type 7.5% 430 in 316 #2 78.1 7.5 7.5-10.0 3.2 2.8
stainless steel in the 316 type matrix 7.5% F e i n 316 90.5 7.3 12.0-15.0 9.6 6.5
can be seen in Fig. 4. A series of 10.0% 430 in 316 #1 82.0 — 7.5-10.0 5.0 4.4
photomicrographs were taken of each 10.0% 430 in 316 #2 78.6 10.2 7.5-10.0 5.1 4.2
standard to verify homogeneous dis- 10.0% F e i n 316 88.5 9.9 12.0-15.0 12.0 8.2
persion of ferrite and also to use for
point counting. (a) Two series of standards were prepared using type 430 stainless steel powder mixed with type 316 stainless steel
powder. Samples from each series are noted as #1 and #2. Three additional standards were prepared using powdered Fe
instead of the type 430 stainless steel.
Ferrite M e a s u r e m e n t s
The ferrite content of each stan-
10 I I I
dard was determined by point count-
ing, the Severn Gage, the Magne I SEVERN GAGE
Gage, and the Ferrite Meter. Point • MAGNA GAGE
counting was done on a minimum of I FERRITE METER
two photomicrographs of each stan-
dard using a grid of 1032 points each
time.
Results for both sets of standards
were quite similar and are given in
Table I. A plot of the data for one set is
;
/U
shown in Fig. 5. The Severn Gage / Y y^
brackets the ferrite content while the
other two instruments read a value.
All show a slightly nonlinear response
(with the exception of the Severn 0
A^y2 4 6 8
. IC
w*r%
•%#? FERRITE COMPOSITION FROM POINT COUNT
Gage on the 7.5 percent standard),
though it is apparent that their cal- . Fig. 5 — Measurements of ferrite in stan-
ibrations are not equivalent. The error dards: values obtained using various in-
in the Magne Gage readings is ± 0 . 5
percent ferrite, as determined from
%A > struments are plotted versus the value ob-
tained from point counting
data on the instrument and the graphs Fig. 4 — Micrograph of nominal 10% stan-
used to correct the dial readings to dard showing uniform ferrite (light par-
the ferrite values, while that for the ticles) dispersion; X250, reduced 74% tual concentrations. This leaves the
ferrite meter is ±0.15 percent below 3 Magne Gage and Ferrite Meter as the
percent and ± 0 . 6 percent up to 12 nondestructive measuring instru-
percent. ments to be evaluated. The Severn
less steel powder in order to study the
Eddy c u r r e n t s are sensitive to Gage is much less precise than these
effect of alloying on the magnetic re-
cracks or voids in a metal so there and thus is eliminated from further
sponse of ferrite. The data on these
was c o n c e r n that porosity in the discussion.
samples are listed in Table I and
pressed-powder samples might affect The Magne Gage and Ferrite Meter
plotted in Fig. 7. The Ferrite Meter
the readings. Accordingly, a series of have similar reading errors. On stan-
readings are about twice as high for
specimens was prepared from the dards composed of type 430 ferritic
the iron as for the ferritic stainless,
same 5 percent mixture of type 430 in stainless steel particles in a matrix of
while the Magne Gage readings are
type 316 stainless steel. Varying type 316 austenitic stainless, both in-
slightly less than three times higher. It
pressures were used to obtain spec- struments gave readings within 20
is important to note that the effect on
imens ranging between 55 and 94 percent of each other. These values
the two instruments is not the same
percent of theoretical density. Read- were about half of that obtained by
magnitude. However, one can see
ings from the ferrite meter on these metallographic point counting, and
that the chemical composition of the
coupons are plotted versus density in this discrepancy can be attributed to
ferrite in a standard will affect the cali-
Fig. 6, and while there is some slope the calibration procedures specified
bration of ferrite measuring instru-
to the curve, the variance between 75 for e a c h i n s t r u m e n t . A d d i t i o n a l
ments and thus their ability to accu-
and 95 percent density is less than 0.5 measurements were made on welds
rately measure ferrite in welds.
percent ferrite. Thus, it was con- in type 304 and Nitronic 40 stainless
cluded that the error associated with steel (21-6-9).* The data, plotted in
specimen density variations could be Discussion Fig. 8, show a good linear relation-
neglected. ship with deviations only for data on
Because the ferrite particles in the
Three additional powder metal- p o w d e r metallurgy s t a n d a r d s are
lurgy coupons were prepared using equiaxed, point counting is reason- 'Designation by Armco Steel Corpora-
iron instead of type 430 ferritic stain- ably reliable for determining the ac- tion.
AWS D10.10-75
Local Heat Treatment of Welds in Piping
and Tubing
In the manufacture of welded articles or structures in the shop or in
the field, it may be desirable, for a variety of reasons, to heat the weld
regions before welding (preheating), between passes (interpass heating),
or after welding (postheating). This document presents in detail the
various means commercially available for heating pipe welds locally,
either before or after welding, or between passes. The relative advantages
and disadvantages of each method are also discussed. Although the
document is oriented principally toward the heating of welds in piping
and tubing, the discussion of the various heating methods is applicable
to any type of welded fabrication.
Topics covered include the following:
• Measurement of Temperature
• Induction Heating
• Electric Resistance Heating
• Flame Heating
• Exothermic Heating
• Gas-Flame Generated Infrared Heating
• Radiant Heating by Quartz Lamps.
The price of AWS D10.10-75, Local Heat Treatment of Welds in
Piping and Tubing, is $3.50. Discounts: 25% to A and B members; 20% to
bookstores, public libraries and schools; 15% to C and D members. Send
your orders to the American Welding Society, 2501 N.W. 7th Street,
Miami, FL 33125. Florida residents add 4% sales tax.
164-S I J U N E 1976