EP0666333B1 - Fatigue-resistant aluminium alloy thick plate and process for making - Google Patents

Description

The invention relates to heavy plates of Al alloys resistant to fatigue and a process for obtaining them.

We know from the publication of Magnusen et al. in "Journal of testing and evaluation", vol. 18, Philadelphia, US, pages 439-445, that microporosities and intermetallic phases formed at casting are preferential initiation sites for cracks in tired; however the presence of these is inevitable in particular in large section metallurgical products obtained by vertical casting conventional.

So far, and in order to minimize both density and / or size maximum of microporosities and intermetallic phases, we sought to pour very thick plates, in order to fill most of the microporosities and to fragment most of the intermetallic phases, generally concentrated in the vicinity of the large median plane of the plate, during hot rolling using high C = Ei / Ef, Ei being the initial thickness and Ef the final thickness.

We know that the usual process for obtaining thick sheets (thickness > 10 mm) consists of casting plates, scalping, homogenization, their hot rolling. The sheets thus obtained are dissolved and soaked, possibly undergo a traction controlled to reduce the level of internal stresses and then are muries (states T 351 or T4) or returned to states T 651 or T 7x51, depending the nomenclature of the Aluminum Association.

Thus, for example, for sheets having final thicknesses E _f of between 100 mm and 150 mm, one generally starts from plates having a thickness after scalping> 3.5 E _f , or corrugations C> 3.5.

The patent FR 2529578 (= US 4511409) of the applicant and the patent US 5277719 (ALCOA) describe a process for increasing the fatigue resistance of thick aluminum alloy sheets by reducing microporosities, process comprising an initial forging step .
The second of these patents describes an example in which the rate of wrought (forging and rolling) is 2.6. But the introduction of a preforging step requires a double heating of the plates and their transport to two different production sites, which leads to a significant increase in the processing cost.

The present invention which is defined by claim 1, relates to sheets having final thicknesses> 110 mm. It relates to sheets of Al alloy with structural hardening which in the treated state (hardened matured, or hardened and tempered, or hardened hardened and tempered) have good internal health, as well as a fatigue life greater than 100,000 cycles in wavy traction (R = 0.1) under a maximum stress equal to 50% (and even 60%) of the elastic limit Rp0,2.
The present invention also relates to a method for obtaining sheets of thickness between 110 and 250 mm in aluminum alloy with structural hardening as defined by claim 6.
Contrary to the teaching of the prior art, the Applicant has noted, with surprise, that significantly improved fatigue properties could be obtained by limiting the wrinkling to values C <2.4 for sheets having final thicknesses between 110 mm and 150 mm (and therefore with thicknesses of scalped plates <360 mm), and even at values C <2 for sheets having final thicknesses between 150 mm and 250 mm (and therefore with thicknesses of scalped plates <500 mm). This hot working is preferably obtained only by rolling, without prior forging.

It has also been found that the fatigue resistance properties are further improved if the liquid metal at the time of casting has been suitably degassed by methods known to those skilled in the art (degassing with chlorine or with products containing chlorine, argon degassing pockets) so that the hydrogen content in the liquid metal are less than 0.15 cm ³ / 100g. It has also been found that the fatigue properties are further improved if during the hot rolling the last two rolling passes (inlet thickness-outlet thickness) are greater than 25 mm each.

The products according to the invention resistant to fatigue are characterized by good internal health, determined by a US control in the vicinity of the median plane of the thick sheets.
The control device used is a PR 02 transceiver from Synergetics or 5052 from Panametrics with the following settings: Focused US: focal spot (at -6 dB or mid-amplitude) in the shape of an elongated ellipsoid 100 µm x 600 µm
frequency: 50 MHz
Broadband probe (Δf / fo = 70%)

It is associated with digital acquisition and signal processing means, with controlled movement of the probe in xy in steps of 0.02 mm.
The device is calibrated using a standard block of the same type and in the same state of heat treatment as the test product, comprising flat bottom holes 0 50 and 100 μm, to which the maximum amplitude is conventionally assigned. 80% on the control screen, which fixes the overall gain Gb.
The volume examined, on a slice of maximum thickness of 20 mm, is 20 x 20 x 0.6 mm ³ .

The measurements are located on several locations to obtain reliable statistical values.
The porosities are defined by the peaks exceeding an adjustable threshold, and fixed slightly above the background noise.

• le nombre de porosités (de taille équivalente > 20 µm, seuil de détection) par unité de volume
• la moyenne arithmétique des amplitudes maximales Amax, Amax étant la valeur maximum de l'écho sur une porosité,
• la moyenne arithmétique des amplitudes moyennes A, A étant égale à
  
  x et y étant les distances sur l'image de la porosité considérée (voir fig. 7).

We determine:

• the number of porosities (equivalent size> 20 µm, detection threshold) per unit of volume
• the arithmetic mean of the maximum amplitudes Amax, Amax being the maximum value of the echo over a porosity,
• the arithmetic mean of the mean amplitudes AT , AT being equal to
  
  x and y being the distances on the image of the porosity considered (see fig. 7).

Under the conditions defined above, the products according to the invention respect the following limits:
Number of porosities ≤ 800 / cm ³
Average of Average Amplitudes AT ≤ 22%
Average of the Maximum Amplitudes Amax ≤ 50%

L =

sens long

TL =

sens travers long

TC =

sens travers court

R =

contrainte maximale/contrainte minimale d'un cycle lors des essais de fatigue

Kt =

coefficient d'entaille

ΔK =

variation du facteur d'intensité de contrainte lors des essais de fatigue

da/dn =

vitesse de propagation des fissures de fatigue

In the rest of the text we will use the following abbreviations or notations:

L =

long sense

TL =

long cross direction

TC =

short cross direction

R =

maximum stress / minimum stress of a cycle during fatigue tests

Kt =

notch coefficient

ΔK =

variation of the stress intensity factor during fatigue tests

da / dn =

speed of propagation of fatigue cracks

The invention will be better understood using the following examples, illustrated by Figures 1 7.

Figure 1 shows a macrographic section along the L / TC plane (long / short cross) of a 118 mm thick sheet from size A from example 1.

FIG. 2 represents a macrographic section along the L / TC plane a 125 mm sheet from format B of Example 1.

Figures 3 and 4 show the density and distribution of the phases intermetallic sheets from A and B format respectively Example 1, L / TC plan.

FIGS. 5 and 6 represent in an amplitude - number of porosities / cm ³ diagram, the results of the acoustic micrographs carried out on the product of Example 4, the standard being a hole with a flat bottom 100 μm.

FIG. 7 represents the principle for determining the quantities Amax and AT .

The alloys will be designated according to the nomenclature of Aluminum Association.

EXAMPLE 1

7010 alloy plates were developed on the one hand in 420 mm thick formats after scalping (outside the invention: format A) and on the other hand in 260 mm thick formats after scalping (according to the invention: format B ). The thickness of metal removed with scalping is identical in both cases. The hydrogen content in the liquid metal is in both cases <0.15 cm ³ / 100g.

These scalped plates were homogenized under identical conditions (30 hours at 470 ° C), then hot rolled at 390 ° C under identical conditions, with the following sequence of rolling passes: Format A 20-30-40-40-34-34 mm Format B 22-23-26-28-32 mm

The final thickness of the sheets is 118 mm for format A, which leads to a correction factor C of 3.56 for format A and 125 mm for format B, which leads to a correction factor C of 2.08 for format B. We have a plotter of the correction factor revealed by the form factor of the metallurgical grain in an L / TC section (Figures 1 and 2). The form factor of the pouring grain is of the order of 4 to 6 for format A and of the order of 2 to 3 for format B.

The two sheets were then subjected to a solution treatment of 8 hours at 480 ° C followed by water quenching, controlled traction of 2.1% and a 10 hour tempering treatment at 120 ° C followed by 8 hours at 170 ° C, so as to give them the characteristics of state T 7651.

The two sheets were the subject of a complete fatigue characterization with "stair-case" tests at 10 ⁵ cycles in the TL direction at mid-thickness with R = 0.05. The staircase case fatigue test tubes were used to measure the size and number of microposits by image analysis on a micrographic section. The measurements by image analysis were carried out on a polished section in a mid-thickness L - TC plane, under magnification × 100, with examination of 150 fields of 1 mm × 1 mm, ie a total surface examined of 150 mm 2 by test tube. The measured parameter is Dmax, the largest dimension in the observation plane, for all Dmax> 20 microns. The results of the staircase case fatigue characterizations and of the surface density measurements of the microporosities are collated in Table I. The results of crack propagation velocities are collated in Table II. Format Final sheet thickness (mm) Stress at 10 ⁵ cycles (MPa) Number of porosities / cm2 with 20 <Dmax <100 µm Number of porosities / cm2 with Dmax> 100 µm Rp0.2 TL (MPa) A-casting 92990 118 212 ± 15 107 3 455 A-casting 92292 118 210 ± 5 110 10 455 B-casting 93193 125 246 ± 5 67 0 453 Format Final sheet thickness (mm) da / dN at 10MPa √m (mm / cycle) da / dN at 20MPa √m (mm / cycle) A-casting 92990 118 1.8 10 ^-4 10 ^-3 B-casting 93139 125 10 ^-4 10 ^-3

It appears clearly on examining these results that the use of the casting format according to the invention (Format B) makes it possible to very significantly improve the level of stress admissible at 10 ⁵ cycles and the speed of propagation of cracks da / dn at ΔK = 10 MPa √m. This improvement is correlated with a reduction of around 40% of the porosities having Dmax between 20 µm (detection limit) and 100 µm, and with the disappearance of the porosities having a Dmax> 100 µm.

Figures 3 and 4 show the size and distribution of intermetallic in the case of the two formats. There is a size and reduced intermetallic density in the case of format B.

EXAMPLE 2

7050 alloy plates were developed on the one hand in 420 mm thick formats after scalping (outside the invention: format A) and on the other hand in 260 mm thick formats after scalping (according to the invention: format B ). The thickness of metal removed with scalping is identical in both cases. The hydrogen content in the metal is in each case <0.15 cm ³ / 100g.

These plates were homogenized under identical conditions (16 hours at 475 ° C), then hot rolled with the following sequence of the last 2 rolling passes: Format A 28 mm / 32 mm final thickness 204 mm Format B 34 mm / 34 mm final thickness 204 mm

This leads to a correction factor C of 2.05 for format A and of 1.27 for format B.

The two sheets were then subjected to a solution dissolving treatment of 20 hours at 478 ° C followed by water quenching, controlled traction of 1.6% and a 6 hour tempering treatment at 120 ° C followed by 21 hours at 165 ° C so as to give them the characteristics of state T 7451.

The two sheets were the subject of a fatigue life characterization under 242 MPa; R = 0.1; Kt = 1. The logarithmic means of life on 8 test pieces taken in the TL direction at half-width and half-thickness of the sheets are grouped in Table III, as well as the maximum number of porosities per cm2. In this state, the elastic limit is 411 MPa in the TL direction for the sheet from format B (according to the invention) and 420 MPa in the TL direction for the sheet from format A (outside the invention). Format Final sheet thickness (mm) Logarithmic mean of life (in cycles) Maximum number of porosities / (cm2) Rp0.2 TL (MPa) A-casting 96343 204 71000 7 420 B-casting 96381 204 118000 4 411

It is clear from an examination of these results that the use of the casting format according to the invention (format B) makes it possible to improve very significantly the level of fatigue life and reduce the density of microporosities.

EXAMPLE 3

Four 7010 alloy plates were produced in 260 mm thick format after scalping (according to the invention: format B). The hydrogen contents in the cast metal and the values of the last rolling passes are given in Table IV. The elastic limits and the fatigue limit at 10 ⁶ cycles were measured in the T7651 state and appear in Table IV (sense TL). Format H2 cm ^3/100 g Last 2 Passes (mm) Porosities / cm2 (> 20 µm) Breaking stress at 10 ⁶ cycles (MPa) Rp0.2 TL (MPa) B / 1 0.3 7/5 109 115 ± 5.9 446 B / 2 0.3 17/16 90 112 ± 3.8 443 B / 3 <0.15 6/5 16 200 ± 26.1 451 B / 4 <0.15 19/18 7 215 ± 11.3 455

The effect of the hydrogen content is very significant, the effect of reworking passes at rolling is less marked and sensitive only on low hydrogen levels.

EXAMPLE 4

Four 7050 alloy plates were produced in 260 mm thick format (according to the invention: format B).
Two of these plates were produced with a fusion bed containing no recycled chips, the other two plates were produced with a fusion bed comprising 10% recycled chips.
These plates were transformed according to the range described in Example 2, and led to final thicknesses of 200 mm and 150 mm. The logarithmic means of fatigue life under 242 MPa, R = 0.1, Kt = 1 in the T 7451 state were measured in each case. The results are collated in Table V. Landmark Melt bed chip content (%) Final thickness (mm) Log average life (cycles) Rp0.2 TL (MPa) 201 0% 200 118000 411 202 0% 150 145000 436 203 10% 200 93000 413 204 10% 150 112000 447

In this case, we used the characterization technique of ultrasonic porosity at 50 MHz as described above.

The results obtained are grouped in Figure 5. We check well in comparing characteristic areas 201 and 203 a marked effect of the composition of the melting bed for 200 mm thick sheets, the product 201 without chips with the lowest number of porosities correlated with the highest level of fatigue. This difference is attenuated for the least number of porosities on 150 mm thick sheets (comparison of characteristic areas 202 and 204) due to higher work hardening. However, the fatigue level of product 202 remains significantly higher than that of product 204, confirming the importance of the quality of the fusion bed.

Claims (9)

1. Fatigue resistant thick plates of heat-treatable aluminium alloys of a thickness > 110 mm, characterised in that the density of the porosities of an equivalent size between 20 and 100 µm, in the vicinity of the medial rolling plane, is lower than 800 per cm³.
2. Plates according to claim 1, characterised in that the mean of the maximum amplitudes of the defects, measured by the maximum value of the ultrasonic echo on the porosities, is less than 50%, when the maximum amplitude of 80% is conventionally assigned to a flat bottom hole of 100 µm.
3. Plates according to claim 1, characterised in that the mean of the medium amplitudes of the defects, measured by the values

   

   x and y being the distances of the porosities on the image, is less than 22%.
4. Plates according to one of claims 1 to 3, characterised in that in the treated temper (quenched and naturally aged, or quenched and artificially aged, or quenched and naturally and articially aged) they have a fatigue service life under cyclic axial load (R = 0,1) > 100000 cycles under a maximum stress equal to 50% of the yield strength R_p0,2.
5. Plates according to claim 4, characterised in that the fatigue service life exceeds 100000 cycles under a maximum stress equal to 60% of the yield strength R_p0,2.
6. A method for producing fatigue-resistant thick plates of heat-treatable aluminium alloys, having a thickness between 110 and 250 mm, and exhibiting a density of the porosities of an equivalent size between 20 and 100 µm, in the vicinity of the medial rolling plane, lower than 800 per cm³, comprising casting of a plate, scalping of this plate, homogenising, hot working, solution heat treating, quenching, optionally controlled stretching, naturally ageing to obtain the T351 ou T4 tempers, or artificially ageing to obtain the T651 ou T7x51 tempers, characterised in that the hot working factor C is lower than 2,4 for thicknesses of between 110 and 150 mm, and lower than 2 for thicknesses of between 150 and 250 mm.
7. A method according to claim 6, characterised in that the hot working operation is solely a rolling operation.
8. A method according to either claim 6 or claim 7, characterised in that the hydrogen content of the starting liquid metal is less than 0,15 cm³/100 g.
9. A method according to one of claims 6 to 8, characterised in that the last two hot rolling passes are each greater than 25 mm.

Images