Coupling between Welding Conditions and Thermal Cycling for Identification of the Mechanical Heterogeneity of a Weld Joint

The analyses device safety subject to pressure is based on the prediction at break junctions used for the design of this type of devices. The harmfulness analysis of existing defects on these devices makes indispensable the study of the rupture in these components. Various characterization tests (tensile tests, fatigue tests and tensile strength tests) were carried out at room and low temperatures on plates welded end to end and for the different directions of sampling.An estimate of the toughness in the three areas of a weld joint was made by passing from resilience to toughness in the ductile-brittle transition zone of materials. The temperature range of the tests was to provide measurements of the toughness the lower bearing to the beginning of the transition curve. The purpose of this work is to study the state both mechanical and microstructural aspects of the welded junction. The diagnoses used made it possible to deduce that the small thickness of the HAZ, makes the machining of the specimen difficult. Thereby, a mechanical simulation of the HAZ by registration of the thermal cycle that this area undergoes was necessary to be able to reproduce it and compare it with the actual HAZ.

So a microstructural study of the different zones of a weld joint is necessary, and to complete it by a mechanical behavioral analysis is indispensable. As a result, a reproduction of the HAZ by mechanical simulation is an adequate solution to describe the behavior of different parts of this area.

Experimental study 2.1 Material used
The material used in our study is an A48AP steel used in the construction of the pipeline [2]. The process is arc welding carried out according to the API1104 standard, This welding mode has been qualified according to a PQR (Record Qualification Procedure) which includes all tests such as (visual inspection, radiography, hardness, microscopy, traction, folding). The chemical composition and the mechanical characteristics of the base metal are given respectively in Tables 1 and 2.

Mechanical characterization 2.2.1 Tensile tests
For the mechanical characterization of the weld, the standard tensile test specimens were taken in the longitudinal direction according to the standard (NF EN 100002-1), we used three (03) samples in each condition (base metal, weld metal and ZAT) were tested to determine tensile properties. The mechanical characteristics obtained are given in Table 3. The conventional yield strength in the HAZ (480 MPa) is significantly higher than the yield strength in the other areas of the weld joint (MF: 465 MPa and MB: 297 MPa). As for the tensile strength of the base metal is lower (499 MPa) than in the other two zones (MF: 560 MPa and HAZ: 631 MPa). We note that all the values of the apparent yield stress and the tensile strength are higher than the allowable minimum values (manufacturer data). A conventional and rational yield strength limit is generally higher in the HAZ than in the MF feed metal, which is higher than that of the base metal.

Cracking tests and measurements 2.2.2.1 Test Specimens used
The cracking tests were conducted to ambiant air on CT50 test specimens of 08 mm in thickness in accordance with ASTM-E-647 (Fig. 1).
For the filler metal (MF) and the thermally affected area, specimen collection is shown in (Fig. 2).
The stress intensity factor K in the case of a CT geometry is given by the following relation (ASTM Standard E 399):  (1) (2)

The crack growth rate
For the processing of experimental data, a method based on an incremental polynomial method was used. This method uses the smoothing of a series of successive points by a polynomial whose growth is monotonous in this interval of seven points. The equation of the smoothed curve has the following form:

Results and discussions:
The results obtained from the cracking speed in the various test specimens are represented in Fig. 3 and 4. The curves of the cracking speed present an almost rectilinear look on a large part of the explored domain which can be represented PARIS law' of the form [15][16]: The results obtained in the three zones are shown shown in Table 4. For low levels of ∆K a slight difference is generally observed between the HAZ and the MB base metal. However, the gap between these speeds increases when ∆K increases to four times the value. This phenomenon is encountered in the case of carbon steels including the API X60 [17][18][19][20].
For both types of steels it is noticed that the crack did not deviate from its initial plane (direction of propagation), which leads us to conclude that the weld was good and the choice of filler metal which has a yield strength greater than that of the base metal was adequate.

Estimate of toughness 2.2.3.1 Preparation of the specimens
The estimate of the toughness was made from the correlations from resilience tests. These correlations are valid only for the resilience tests carried out on standard 10 x 10 dimensions specimens cut in V the direction of sampling and the dimensions are given in Fig. 5.
To properly locate the notches of the specimens on the zones studied, the welded parts have been mechanically polished to the paper (1000), then a chemical attack with iron perchloride was performed to show the different areas of the weld (Fig. 6).

Test conditions
The temperature range tests were taken to make possible measurements of the toughness from the bottom to the beginning of the transition curve. For each temperature and each zone, three test pieces were used. The liquid refrigerant used to lower the temperature of the specimens is nitrogen associated with alcohol mixed in a calorimeter (Fig. 7). For heating of the specimens we used an oven, and the measurement of the temperature was made using a digital thermometer.

Correlation resilience tenacity
The correlation used allows to determine the tenacity at the desired temperature and select the level of probability at break. This correlation takes into account the thickness of the product and on the transition temperature T27J. The expression xx is applicable to a wide range of steels and validated for different thicknesses.
Kmat: estimated tenacity (N/mm 3/2 ). B: Thickness of the material for which an estimate of Kmat is required (mm). KV: Resilience to the temperature for which it is determined the tenacity (J).

Results and discussion
The results obtained are presented in (Fig. 8) and the following conclusions have been drawn.
Overall, the weakest resilience corresponds to the specimens for temperature ranges below -20 °C.the resilience appears to increase to room temperature (20 °C) and remains virtually unchanged, or a slight rises for temperatures up to 70 °C. The range between (-20 °C and 20 °C) characterizes the transition temperatures, which are as follows: For MB the ductility plateau stabilizes before 0 °C and therefore we speak of stable ductility. On the other hand for the HAZ this transition passes with a hook of Pre-stabiliion HAZ which gives this zone a less stable behavior than the base metal. This phenomenon indicates that there is a change of structure. The molten metal MF lies between the two configurations, this transition is very remarkable in the graphs of micro-hardness filiations.
The results of the toughness presented by (Fig. 9) are obtained by correlation and by introducing an uncertainty factor (distributed according to a Weibull distribution law, (law of the minimum) in the Wallin's correlation. It is noted that for low temperatures a stable average tenacity for the three zones. In the transition phase. We notice a logical evolution but inversely proportional to the laws of behavior of the three zones which characterizes steels working under pressure. From 0 °C the tenacity values stabilize, with higher values in the base metal (BM) than for the molten metal (FM), with less values for the HAZ. Therefore the parameters used by Wallin for Weibull distribution law appear reasonable [21].

Metallographic examination
Weld it's establishing a connection or a metallic continuity, however the welded space does not present a homogeneous structure. Indeed localized fusion heating and cooling who (7)   followed have trained various structures. To put in evidence a structure or process to metallographic examinations. After successive polishing the sample is attacked with a reagent at basis of ferric chloride and water (for macroscopic examination) and at basis of Nutric Acid and Alcohol for microscopic examination.
The microscopic examination is performed at ugly of a scope to observe in a way very thin, the structures in an area very localized.
The transformations suffered by the HAZ are not simulable to the heat treatments applied to steels. In fact after a welding operation, there is appearance of the bainite and the ferrite intergranular in the zone of junction and the zones of transformation. The Fig. 10 shown the difference of structure in all three areas.
-Zone BM: The observation shows that there is presence of ferrite and ferrite-pearlite (Fig. 11).
The modification of the microstructure in the welding area is caused by the superposition of the thermal cycle resulting in each welding pass. This thermal cycle is the cause of microstructural transformations base metal (BM), who is transforming thus in thermally affected zone HAZ or in connection zone. The Figures X and Y, represent the maximum temperatures recorded by the thermocouples at each welding cycle. Thus the equivalent cycles are determined to be able to simulate this HAZ mechanically. -MF area: In the last pass of the MF, the provisions ranges ferritic and constituents carburized comparable to ferrite presented a disposition marked related to solidification. The Observation shows the presence of a dendritic structure with islands (ferrite-perlite) Fig. 12. -Zone HAZ: This zone has a heterogeneous structure variation. The Fig. 13, shows that the structure is ferrito-pearlitic (globular). In the area near the melting line we find an overheated structure with an aspect and a rather special disposition of the ferrite also present islets of bainite in separate slats by ferrite (Fig. 14).

Thermal cycle of the HAZ
The caracterisation of the different areas of the weld joint is a very difficult operation because of geometric discontinuity. Indeed mechanical heterogeneity linked to the presence of three zones presents different mechanical behaviors. Because of the narrowness of the HAZ we have chosen for a thermal simulation of the specimens intended for characterization. The goal of this manipulation is to raise the thermal cycle in the different areas to reproduce later the HAZ on the base material without refilling the plates. The modification of the microstructure in the welding area is caused by the superposition of   For validation of the results by mechanical tests, of the test pieces (pull-ups and micrographs) were taken directly from base metal. The characterization tests and micrographic observations conducted on the simulated ZAT, show that there is a similarity between the two HAZ (real and simulated). We observe an acceptable agreement between the mechanical characteristics of the two HAZ (Table 5), and the Fig. 17, present almost identical appearance of the two microstructures. Indeed, we observe the same provision of the ferrite than observed on the actual HAZ.

General conclusion
This work allowed determining experimentally the mechanical characteristics as well as resistance to ductile tear at room temperature of base metal (BM), Molten Metal (MF) and Heat Affected Area (HAZ) taken from a welded joint made under conditions representative of industrial manufacturing. The profile of the thermal cycle undergoing in the HAZ was established by a mechanical simulation which made it possible to reproduce the characteristics of this zone on the base metal.
In terms of mechanical characterization, the weakest toughness corresponds to specimens for temperature ranges below -20 °C. The resilience appears to increase to room temperature (20 °C) and remains practically unchanged with a slight rise for temperatures up to 70 °C. For the HAZ (real or mechanically simulated) the transition brittle-ductile goes through a pre-stabilization hook which gives this zone a less stable behavior than the base metal. This phenomenon is justified by a change of structure.
The thermally affected zone has a heterogeneous structure variation, this structure is ferrito-pearlitic (globular). In the area near the melting line there is an overheated structure with islands of lapped bainite separated by ferrite.