Strength and Bauschinger Effect in TMCP Line Pipe Steels

I.Yu. Pyshmintsev1, D.A. Pumpyanskyi2, L.G. Marchenko2, V.I. Stolyarov2

(1. Russian Research Institute for the Tube and Pipe Industries, Russia; 2. TMK, Russia)

Abstract: Loss of strength in TMCP microalloyed low carbon linepipe steels with ferrite-pearlite microstructure due to Bauschinger
effect (up to 70-80MPa of yield stress reduction) has been studied. Mechanical behavior of the steels has been investigated using
tests of full-section flattened specimens taken from both 15,7 or 18,7mm plates and 1420mm spiral welded pipe with specified
minimum yield stress of 461 MPa. The dependence of the microstructure and yielding behavior on position in wall section has been
studied using sub-size 3 mm thickness specimens machined from sub-surface and central parts. Change of tensile curves of surface
and internal layers during forming, flattening and protective polymer coating have been studied using sub-size samples subjected to
three or four-point bending, flattening and annealing. The studies have revealed strong dependence of properties on test methods and
the difference in yielding can be decreased significantly by low temperature annealing.
Key words: TMCP high strength steel, Bauschinger effect, forming, testing, annealing, simulations

1 Introduction during pipe forming. Flattening the specimen for a

Increase in yield stress enables increase in operating tensile strength test introduces additional tension-
pressure and, thus in gas pipeline capacity without compression cycle into external and internal fibers of
weight growth which initiates a developing high metal. As a result the material layer-by-layer acquires
strength steels. Past decades of the studies pointed out different properties. It is fundamental that the bending
significant differences in results of yield stress provides a heterogeneous deformation in section that
measurements for these steels using transverse round determines the presence of high residual stresses.
bar and flattened full-section specimens. The reasons Historically loss of linepipe strength had become a
of the difference are usually associated with work problem during application of control rolled high
hardening, Bauschinger effect and residual stresses. strength steels, in which about a half of the yield stress
While an opinion adopted in the 70th concerning increment due to grain refinement, substructure and
deficiency of softening during pipe forming using new precipitations was lost during pipe making. Nowadays
high strength (API Х70 and higher) steels with it is well known[6] that the type and place of sampling
considerable amount of bainite or martensite in the have a governing influence on results of yield stress
structure now needs to be corrected[1-5]. Nowadays measurement. The wall thickness to diameter ratio t/D
discussion of issues of yield stress measurement for and strain hardening rate during cold expansion
pipe steels is of special interest. The purpose of this influence on the effect value and it’s sign. The general
article is to generalize the known data and tendency is the “softening” growth in new higher
experimental results of the effect during forming and strength steels that is stipulated by higher value of
tests of high strength spiral welded pipes. residual stresses. Nevertheless the value of the effect
at equal strength is determined as well by the type of
2 “Loss” of Strength due to Forming microstructure (specific value of each of the
There are three known methods to determine strength strengthening mechanisms resulted in strain hardening
properties in hoop direction by tension of specimens: behaviour). The published data prove[7] that the
flattened full-section, round bar and ring. There were thermal cycle of polymeric coating of pipes may lead
found significant differences in the results of to significant or even full relief of “softening” in high
measuring the yield stress by these methods for high strength materials due to strain age-hardening and
strength steels[1,2,5]. other effects.
Stress and strain gradients form in the wall section The problem of “strengthening” has a great practical
importance both relatively widely used low pearlitic where f is residual deflection; t is specimen thickness;
steels of X60-X70 type and advanced higher strength l is distance between supports.
materials. The research was aimed at peculiarities of At four-point bend the radius R was calculated using
the effect relating to technology of manufacturing deflection and the strain was determined:
1420 mm spiral welded pipes.
t
δ = 100%
2R .
3 Experimental
Industrial low-carbon control rolled plate steels and
spiral welded 1420 mm diameter pipes of 15.7 and 4 Results and Discussion
18.7 mm wall thickness were used in this study. The The research showed the occurrence of three
range of the chemical composition of used heats for microstructure zones in wall thickness. The 100 mkm
each wall thickness is specified in Table 1. thick surface layer has a very fine grain structure (3-5
mkm) of ferrite. Deeper on the depth up to 2-2.5 mm
Table 1 Chemical composition of steels microstructure with various grain ferrite 3 to 20 mkm
t, Content of elements. wt %
mm
and some amount of pearlite was observed. The
C Mn Si- Nb V Ti Mo N biggest grains are elongated in the rolling direction.
х100 х100 х100 х1000 х1000 х1000 х1000 х1000
15,7 5-6 150-157 26-29 49-60 71-72 3-9 186-226 7-8 Acicular and low-temperature products were not
18,7 6-9 155-156 19-23 50-52 69-73 8-16 220-220 6-9 observed in the structure. The medium size grain is
S-0,003…0,005%; P-0,007…0,016%; Al – 0,028…0,041%
higher in the biggest central area of the plate and
grains are sized 8-30 mkm. Separate bands and
elongated islands of pearlite are observed here in
Flattened full-section specimens for tension tests were
which elongated pearlite particles are separated by
machined from plates in transverse and longitudinal
needle like ferrite which is diagonally oriented
directions. The specimens of the same type were cut
relatively rolling direction. Thus, plates have a natural
out of pipes along and across the axle that
gradient of microstructure and properties that provides
corresponded the diagonal direction in the plates. The
as well for the difference in the stress-strain properties
simulation of pipe forming and testing on strength
and hardening response.
properties were carried out by bending and flattening
Figure 1 gives the results of yield stress measures as
full-section specimens up to deformation
well as the tensile strength in the plates 18.7 mm thick
corresponding to the pipe forming process.
and pipes using full-section transverse specimens. The
The influence of the microstructure type and the level
results of it were compared with the plate mill
of strength properties on the effect for different layers
certificate data. The difference of results of all the
of the plates were studied using flattened micro-
three data groups is obvious. The dependence of
specimens with 3x10 mm cross section cut out of
properties on the direction is more clearly observed in
pipes and original plates at different depth down from
pipes. The yield stress and tensile strength measured
the surface. The orientation of micro-specimens was
at the longitudinal specimens from pipes are
similar to full-section specimens. The simulation of
noticeably higher than at the transverse ones that is
the “forming - testing” cycle was carried out using a
probably caused by the differences in the treatment.
three-point or four-point bend of the specimens.
Tensile curve analysis showed (Fig.2), that due to
The strain of external fibers of micro-specimens
anisotropy the yield stress measured on the transverse
during three-point bend varied in the value of the
specimens from plate are, as a rule, higher by 25-30
residual deflection from 0.5 to 2.2% and was
MPa than in longitudinal direction. The yield stress
calculated by the formula:
for diagonal direction has an intermediate value what
6 ft corresponds to the known data[8]. Yielding plateau on
δ = 100%
l2 , the curves for specimens from pipes was not observed
and yield stress measured for residual strain of 0.2% is
reduced by 40-70 MPa relatively to plates. The
“softening” effect is lower for the diagonal pipe 680

660

Yield stress (after bending), MPa

specimens. Т
640 L
D
620

600
640
580

T.S.
Stress ( for plate specimens) МPа

560

540
600
560 600 640 680
Yield stress (in plate), MPa

а
560
/ transverce 600
/ longitudial

Yield stress (after bending), MPa

plate mill/ pipe mill
Y.S. 560
Т
520 L
D

520

480 480

480 520 560 600 640 440

Stress (for transverce pipe specimen), МPа 440 480 520 560 600
Yield stress (in plate), MPa

Fig.1. Stresses measured using 18.7 mm full-section b

specimens Fig.3 Yield stress in the “plate” and “pipe” at simulation of
forming and testing of full-section specimens from sheets 1
(a) и 2(b) 15.7 mm thick.
550
1
2
Engineering stress, МPа

3 intensely, resulting in substantial strength reduction by

500
4 60-90 MPa.
Figure 4 shows typical tensile curves at small strains
450
for micro-specimens cut off the pipes at different
depth from surface. Difference in deformational
400
behavior of these layers is evident. The surface layer
350
has the maximum yield strength at the presence of the
0,2% plateau at a stress of 600 MPa. The intermediate layer
300
has showed lower yield stress. Central area has the
lowest yield stress, and the yield plateau is minimal
Engineering strain
though transition to plastic yield is clear. Micro-
Fig.2 Typical tensile curves for full section flat specimens specimens cut from the “pipes” (without bending)
of 18.7 mm thick plate (1,2) and pipe (3,4). 1-transverse, 2- indicates the behavior different from those observed
longitudinal, 3–along the direction of sheet rolling (diagonal on transverse full-size specimens of pipes.
in pipe); 4-transverse (in pipe). Tensile curves of micro-specimens cut from pipes and
sheets differ slightly. One can suppose that absence of
The study of the yield stress change during treatment apparent “softening” during measuring on flat micro-
simulation on the full-section flat specimens indicated specimens can be caused by the absence of unbending
different behavior of 15.7 mm steel in sheets 1 and 2 prior to testing, lower fiber strains and change of
(fig.3). Despite the far lower yield stress owning to residual stress during machining both in relation to
features of microstructure the sheet 2 “softens” more full section specimen and initial pipe.
 700
allows to quite accurately simulate the cycle
Surface “forming-testing”, as a result the effect of 20-25 MPa
is observed, that is close to practice.
1,5 mm from surface
600 It is to be noted that “softening” effect was observed
Engineering stress, MPa

middle in pipes before coating. The simulation of thermal

cycle of polymer coating by annealing showed that as
500
a result coating the restoration of yield stress with
formation of the plateau can be found for all
microstructures. The increment of yield stress after
400
annealing can reach 40-60 MPa.
Tensile curves for various layers after “bending-
300 unbending” cycle with strain of 1,5-2%, is similar to
0 1 2 3 4
full size specimens. After the cycle the yield peak
Engineering strain, %
disappears and the yield stress decreases. From the
Fig.4 Tensile curves parts for micro-specimens, cut from data in Table 2, one can be seen that the “softening”
pipes in diagonal direction. depends on the yield stress and maximum was found
in the surface layers with highest yield stress.
Figure 5 shows the correlation of the strength The annealing corresponding to the thermal cycle of
properties, measured with use of full-size specimens coating considerably changes strain hardening
and micro-specimens cut from various pipe layers and behaviour at small strains. As result the upper yield
18.7 mm plate. The simulation of the cycle “forming- stress and yield plateau can be found on the curves.
testing” prior to testing was carried out using three-
point bending up to plastic yield of outer fibers 1-2%. Table 2 The effect of deformation and annealing on
It is apparent that strength, measured with use of microspecimens yield stress, MPa
Plate Upon bending –unbending
micro-specimens cut from central and pre-surface
Specimen

before annealing After annealing

layers, is in good correspondence with properties 200 оС, 10 min
Y.S. Y.S.(0,2) Y.S.−Y.S.(0,2) Y.S.′ Y.S.−Y.S.′
measured at full section. The use of micro-specimens 1 615 565 50 600 15
Plate

2 550 531 19 555 -5

3 520 525 -5 550 0
1 600 521 79 590 10
Pipe

680 2 555 495 60 554 1

3 525 500 25 525 0
660

640 T.S. 5 Conclusions

Stress for sub-size specimen, MPa

620
One can consider that the traditional evaluation of
1
reliability of high grade steel pipes by yield stress
600
2 value, measured with the use of flat full section
580 3 transverse specimens is conditional. It is proved by
Y.S.
560 yield stress measurement with the use of micro-
1
plate before bending
specimens cut off various microstructural zones in the
540
plate after bending pipe wall section which demonstrates no considerable
pipe before bending
520 2 metal “softening”. It is experimentally showed that the
500
3 observed “softening” upon deformation cycle
500 520 540 560 580 600 620 640
“bending-unbending” is practically fully relieved after
Stress for full size specimen, MPa
even short annealing at 200оС, corresponding to the
Fig.5. Correlation of tension in the initial state and upon cycle of polymer coating.
forming simulation with the use of three point-bending of The strength measured with the use of relatively small
micro-specimens, 1 – surface, 2- pre-surface, 3 - central flat or round bar specimens to the greater extent
reflects the real pipe strength rater than flattened full- [4] Ratnapuli R.C. et al. A method of calculating
size specimens. The use of flattened full-size Bauschinger effect in API linepipe steels. Mech. and
specimens for high strength steels can lead to reduced steel processing. 28th MWSP. Conf. ISS, AIME, 1987.
measurement result, and as a consequence, to increase [5] Saikaly W.E. et al. Comparison of ring expansion
of pipelines weight and cost. To a certain extent the vs flat tensile testing for determination linepipe yield
observed yield stress “reduction”, resulting in strength/ Int. Pipeline Conf. Calgary. 1996.
decrease of its ration to tensile strength, can be vol.1.:209-213.
considered positive in terms of criteria based on [6] Liessem A. et al. Influence of heat treatment on
pipeline strain limit. However this effect shall be mechanical properties of UOE linepipe. M.K.Graef,
considered mainly as a measurement result rather than G.Knauf, U.Marewski// 4th Int. Conf. on Pipeline
physical phenomenon. Techn., Oostende. 2004. V.3. pp.1262-1281.
[7] Fluess P., Schwinn V., Bush K. Production and
References: development of pipes for conductors and risers with
[1] Glover A., et al. Yield strength and plasticity of strength level X80 and X100 without pipe expansion
high strength pipelines. 4th Int. Conf. on Pipeline /4th International Conference on Pipeline Techn.,
Techn., Oostende. 2004. V.1.: 65-79. Oostende. 2004. V.2. :809-822.
[2] Millwood N.A., et al. The influence of tensile [8] Baczynski G.J. et al. The influence of rolling
testing method on the measured properties of high practice on notch toughness and texture development
strength steel linepipe. 4th Int. Conf. on Pipeline in high-strength linepipe. Met&Mat Trans., 1999.
Techn., Oostende. Belgium, 2004. V.4.:1857-1879. 30A.: 3045-3054.
[3] Shoemaker A.K. The effect of plate stress-strain [9] Streisellberger A. Correlation of pipe to plate
behaviour and pipemaking variables on the yield properties – Model Calculations and Application in
strength of large diameter DSAW linepipe/ design of X80 linepipe steels. Int. Conf. of pipeline
Eng.Mater.Techn., 1984.106(20):119-126. reliability, Calgary. vol.1., 1992.