Soutenance mémoire
PRODUCTION OF FORGING BILLETS (FEEDSTOCK)
Characterization of feedstock
Microstructural analysis of feedstock materials
As-Cast material: In general, as-cast feedstock has an equiaxed grain structure, which is a typical macrostructure, produced in the vertical DC cast billets (Figure 4.1). However, the surface and the core areas of cast billets have distinctive microstructural features due to an uneven solidification process. A modified Poulton etching on samples revealed a uniform cast structure in the inner central zone of the AA6082 cast billet. Surface microstructure Bulk microstructure. Direct chill casting techniques usually induce non-equilibrium freezing within the ingot. Non-uniform solidification introduces a segregation layer in the casting surface. Figure 4.2a shows an inverse segregated layer at the surface of the as-cast feedstock. The presence of many intermetallic compounds was detected as seen in Figure 4.2b. The enriched layer thickness was measured with an average of 129 u.m. a) Presence of inverse segregated layer on the ascast material (Tucker etchant at 100X magnification) b) Presence of intermetallics compounds in the enriched surface layer (no etching, 500X magnification)
As-extruded material: The extruded material has a fibrous and non-recrystallized grain structure except at the surface. Figure 4.3a displays a highly fibrous structure in the inner central location of the extruded bar. oj Metallographic longitudinal profile with elongated and deformed grains (50X magnification) b) Metallographic transversal profile with two different microstructures: bulk and surface zones (SOX magnification) Two different microstructural zones have been detected in the extruded material (Figure 4.3b). They are referred to as specific deformation zones during the extrusion process. The significant thickness of the surface layer and the related through-thickness inhomogeneity in grain structure and crystallographic texture seem to constitute the particular features of the extruded profiles. The extruded material displays a fiber-like and non-recrystallised grain structure. The thin layer at the surface is often called the «Peripheral coarse grain layer (PGL)» (Figure 4.4a). The peripheral grain layer at the surface is produced by high mechanical deformation ratios and significant temperatures that will produce enough energy to redefine the shell zone by a recrystallization process (Figure 4.4b). Quantitative thickness measurements on transversal profiles give a thickness mean equal to 108.15 u.m related to the peripheral coarse grain layer (Figure 4.5).
Fine, round and black Mg2Si constituents are found everywhere in the extruded microstructure but a number of bigger Mg2Si constituents were found around the thinnest zone at the surface. Micrographs revealed that the Peripheral coarse grain layer (PGL) contains larger Mg2Si precipitates (Figure 4.6). Coarse Mg2Si particles in the PGL zone can cause low hardness, increase pick-up on the extruded surface, and influence the corrosion phenomenon . Electrochemical corrosion tests The next section details corrosion results of four metallurgical conditions in a corrosion cell using electrochemical methods. Cast-forged and extruded-forged materials with heat treatment effects were tested to determine their corrosion susceptibility.
Cast-forged surface without T6
The open circuit potential started between -0.76 and -0.85 V /SCE for the cast-forged material without T6. The acquisition rate for all OCP curves was fixed at 1 data per 2 seconds. Starting thirty minutes before the testing began, a system for injecting nitrogen bubbles deaerated the solution of its oxygen content with a 10 SCFH flowrate. The potential evolution decrease was involved by the corrosion phenomenon initiation where samples were submerged in a 3.5 w/v% NaCI solution at 21°C. The higher negative voltage obtained for these fives curves was located between -0.85 V and -1.1 V /SCE. After thirty minutes, all OCP curves tended to increase to obtain a stabilized zone at -0.76 V /SCE. This phenomenon is due to the passivation mechanism of aluminum alloys to form a protective aluminum oxide layer, especially in a corrosive solution. In Figure 4.23, each curve had its own characteristic slope of between 30 to 1000 minutes to access the passive state. A reciprocal phenomenon between corrosion and passivation can sometimes be produced by a breakdown of the passive layer, as can be seen in one of these five curves at 750 minutes. Time (min) with a scan rate of 2 mV/s until the maximum current was reached. The apparatus was able to deliver voltage increments until the maximum current output of 1 ampere. This data is equivalent to a current density of 100 mA/cm2 on the specimen tested in a flat corrosion cell. The presence of regular instabilities in the early stage of polarization is caused by the initial potential that was always manually set below the final value of the OCP test. For example, last OCP measurements of cast-forged samples were identified at -0.76 V/ SCE (see Figure 4.23) and the polarization test (see were produced by a cathodic polarization noise. Polarization voltage began to increase at the last OCP result acquired (near -0.76 V /SCE for the samples). The pitting corrosion potential (EPjt) was estimated from 5 polarization curves of each metallurgical condition. The threshold current density value was established arbitrary as j = 0,1 mA/cm2 , where the current quickly increase. It was recognized that for a given potential, pitting corrosion process proceed at given electrode, if determined value was greater than or equal to threshold current density value. For the cast-forged without T6 material condition, at j=0,l mA/cm2 , the Epit average was determined to -0.63 V /SCE Some instability on polarization curves were visible starting at -0.65 to -0.5 V /SCE. Theses disturbances were produced during the pitting mechanism, referred to as the breakdown potential, to produce continuous pitting corrosion on samples. The passive alumina film contains or is susceptible to the formation of distribution flaws that are potential sites for pitting. For example, the Mg2Si and AlFeSi phases that were identified at the grain boundaries of cast-forged microstructures can cause corrosion pitting and instability in these results (see section 2.2.3, page 33 and section 4.1.2., page 68).
Three out of five curves produced this phenomenon of constant pitting lower than the other two. A continuous pitting phenomenon can be identified after the most active pitting zone Epit. These current variations are associated with initiation and repassivation of pits sites prior to the propagation of stable pits. The pitting phenomenon causes instabilities on curves at Epit Optical and SEM imaging on polarization test specimens Optical imaging revealed many pits produced at the surface during polarization tests especially for castforged without the T6 condition (Figures 4.25 to 4.27). The scanning electron microscope allowed the form evaluation, the dimensions and the depths of the pits as well. Dimensions of the pits were approximately 500
Cast-forged surface with T6
The OCP voltage initiated from -0.85 V /SCE and decreased until a minimum of approximately -1.075 to -1.15 V /SCE was attained for every test. After a specific time for each curve, the voltage was increased to reach a maximum of -0.76 V /SCE, similar to the last condition without T6. After 900 minutes, all tests were stabilized within 0.76 V /SCE ± 0.01 V (Figure 4.28). The cast-forged with the T6 material seems to have the same curve pattern than the cast-forged without the T6. Three similar points were found in theses curves: 1) The voltage drop from the initiating point corresponded to the corrosion initiation on samples during immersion. The voltage became negative and the anodic phenomenon was more active. 2) The curve orientation and the voltage increase are related to the passivation mechanism of aluminum with an aluminum oxide layer at the surface. The corrosion phenomenon was sufficient to produce a protective layer of alumina on the aluminum surface and thus make the potential nobler (more positive). 3) The stabilization of the OCP potential was due to the protective oxide layer formation and this potential was associated with the corrosion potential called « Ecorr ». Time (min)
An anodic polarization action with a scan rate of 2 mV/s was applied after 24 h. of OCP acquisition. The polarization voltage increased at the last Ecorr result acquired (close to -0.76 V /SCE during the OCP test) until a maximum of 0.1 V was reached. Little instability produced by continuous pitting was visible on one out of the five curves and Epit average was determined at -0.54 V /SCE for this metallurgical condition (Figure 4.29). The other four curves presented no important continuous pitting phenomenon. Optical and SEM imagina on polarization test specimen Anoptical macrograph permitted evaluating the corroded surface of samples and detect the presence of many small and some bigger pits. The SEM revealed pit diameters around 10 to 20 u.m. The inside zones of the pits were not observable by SEM making it possible to believe that they are nevertheless deep, i.e., a minimum of a few microns. The optical microscopy analysis was not precise enough to quantify the depth of the pits .
Extruded-forged without T6
Values of open-circuit-potential (OCP) for a forged-extruded condition without T6 were started at -0.76 V /SCE ± 0.01 V and held until the end of 24 hours of acquisition. No important corrosion or passivation mechanisms were apparent during the entire test. Maximal variations during acquisition of all the results were inside an interval of 0.4 V (Figure 4.33). This could be explained by the presence of a protective layer at the surface of this material type. For instance, the peripheral grain layer structure, or the presence of extremely fine dissolved Mg2Si in the matrix, could play a role on the immunity phenomenon. Time [min] At the beginning of each polarization curves, non stable traces are produced by a cathodic polarization noise where the potential was started under the last OCP measurement (see page 75). After this transient zone, Epit was determined for extruded-forged without T6 material at -0.57 V /SCE, where the current sharply increased (at j > O.lmA/cm2 ). No continuous pitting was detected on the polarization graph for this metallurgical condition .
Optical and SEM imaging on polarization test specimens Microscopical observations of corroded samples by electrochemical polarization have permitted to identify a localized corrosion aspect with a lengthened form (Figure 4.35). This asymmetrical form of corrosion seems to follow deformation pattern lines or surface defects produced by the extrusion process. This corrosion presence is not uniform on the surface of samples (Figures 4.36a and 4.36b). Indeed, it is possible to distinguish corrosion spots having significant dimensions (= 500 urn) with others that are smaller. With SEM, it was observed that localized pits have low depths (Figure 4.37).
Extruded-forged with T6
The last condition tested to analyze the evolution of the corrosion phenomenon during 24 hours was the extruded-forged material with T6 heat treatment. Acquisition of five curves made it possible to detect a different tendency than those previously evaluated. The potential start was close to -1.07 V /SCE ± 0.23 and decreased until a maximum of -1.28 V /SCE was reached during the first 30 minutes (Figure 4.38). After this reduction of electrochemical potential, which represents an important phenomenon of corrosion during the specimen immersion time, an increase of potential was carried out to reach equilibrium. In fact, this surface layer stability is explained by the passivation of aluminum where the oxidation phenomenon was too important and created a protective oxide layer. Four out of five curves stabilized Ecorrat-0.8 V/SCE + 0.06. The last curve at Figure 4.38 did not reach the Ecorr potential and finished at -1.06 V /SCE. This more negative result is explained by the fact that the test did not have enough time to reach the electrochemical equilibrium after 24 hours of data acquisition. This significant duration obtained (greater than 24 hours) to stabilize Ecorr of the extruded, forged and heat treated material may be caused by a change in the peripheral coarse grain layer (PGL). Grain size, precipitates or defects in this zone can be modified during the heat treatment step.
|
Table des matières
RÉSUMÉ
SUMMARY
1. INTRODUCTION
1.1. INVESTIGATION ON SURFACE ANOMALIES OF DIRECTLY CAST-FORGED PRODUCTS
1.2. OBJECTIVES OF THESIS
2. LITERATURE REVIEW
2.1. PRODUCTION OF FORGING BILLETS (FEEDSTOCK)
2.1.1. Step 1: Semi continuous – DC casting processes
2.1.2. Step 2: Extrusion or Rotating scalping process
2.1.3. Step 3: Forging process
2.1.4. Step 4: Heat treatment processes
2.2. CORROSION BEHAVIOUR OF ALUMINUM AND ITS ALLOYS
2.2.1. Electrochemical theory and corrosion potentials
2.2.2. Passivity phenomenon
2.2.3. Corrosion forms
2.3. CORROSION TESTING FOR ALUMINUM ALLOYS
2.3.1. Corrosion susceptibility evaluation by cabinet exposure tests
2.3.2. Corrosion susceptibility evaluation by electrochemical tests
2.3.3. Corrosion-fatigue tests
3. EXPERIMENTAL
3.1. MATERIALS PREPARATION
3.1.1. AA6082 aluminum cast billets
3.1.2. Homogenizing and extrusion of cast aluminum billets
3.1.3. Forging of feedstock materials (cast and extruded)
3.1.4. T6 Heat treatment on forged products
3.2. CORROSION TESTING ON FORGED MATERIALS
3.2.1. Electrochemical corrosion tests: OCP and anodic polarization curves
3.2.2. Salt spray corrosion exposure tests
3.2.3. Corrosion-fatigue tests
3.2.4. Design of corrosion-fatigue experiments (DOE)
3.3. MATERIALS CHARACTERIZATION
3.3.1. Microstructural evolution: Optical and SEM microscopies
4. RESULTS AND DISCUSSION
4.1. CHARACTERIZATION OF FEEDSTOCK
4.1.1. Microstructural analysis of feedstock materials
4.1.2. Chemical EDX analysis of secondary phases in feedstock materials
4.2. CHARACTERIZATION OF FORGED PRODUCTS
4.2.1. Microstructural analysis of forged materials
4.3. ELECTROCHEMICAL CORROSION TESTS
4.3.1. Cast-forged surface without T6
4.3.2. Cast-forged surface with T6
4.3.3. Extruded-forged without T6
4.3.4. Extruded-forged with T6
4.3.5. Comparative evaluation of electrochemical corrosion test results of forged products
4.4. SALT FOG CORROSION EXPOSURE TEST.
4.4.1. Cast-forged without
4.4.2. Cast-forged with T6
4.4.3. Extruded-forged without T6
4.4.4. Extruded-forged with T6
4.4.5. Comparative evaluation of salt spray corrosion test results of forged products
4.5. REVERSE BENDING CORROSION-FATIGUE TEST
4.5.1. Fatigue trials in air environment
4.5.2. Fatigue trials in a 3.5 w/v% NaCIsolution
4.5.3. Fatigue fracture surfaces and macroscopic features
4.5.4. Fatigue mechanisms and microscopic features
4.5.5. Comparative evaluation of reverse bending fatigue test results on forged-T6 products
5. CONCLUSIONS AND RECOMMANDATIONS
5.1. CONCLUSIONS
5.1.1. Characterization ofAA6082 aluminum forging feedstock and forged materials
5.1.2. Corrosion tests .-
5.2. RECOMMENDATIONS
REFERENCES
Télécharger le rapport complet
