WO2007094686A1

WO2007094686A1 - Aluminium alloy with improved crush properties

Info

Publication number: WO2007094686A1
Application number: PCT/NO2007/000057
Authority: WO
Inventors: Trond Furu; Ulf Tundal; Jostein RØYSET; Oddvin Reiso
Original assignee: Norsk Hydro Asa
Priority date: 2006-02-17
Filing date: 2007-02-16
Publication date: 2007-08-23
Also published as: JP2009526913A; EP1987170A1; WO2007094686A8; CN101384741A; US20090116999A1

Abstract

An Al-Mg-Si alloy with improved ductility and crush properties, in particular useful for structural components in crash exposed areas in vehicles. The alloy contains in wt %: Mg 0,25 - 1,2; Si 0,3 - 1,4; Ti 0,03 - 0,4, where Ti is present in solid solution and where the alloy contains in addition one or more of the following alloy components: Mn max 0.6; Cr max 0.3; Zr max 0,25, and incidental impurities, including Fe and Zn up to 0,5 with balance Al.

Description

Aluminium alloy with improved crush properties

The present invention relates to an Al-Mg-Si aluminium alloy having improved ductility and crush properties with good energy absorption and temperature stability, and which is particularly useful for structural components in crash exposed areas in vehicles.

US 4525 326 discloses an Al-Mg-Si alloy where vanadium, V, is added for improving the alloy's ductility. In this patent it is claimed that -additions of V in the range 0.05-0:20 wt. % in combination with a Mn content defined to be' ^'Within A /4 to 2/3 of the Fe content,.'! significantly improves the ductility of a wide range of Al-Mg-Si alloys. Titanium, Ti, is not' mentioned in US 4 525 326, and neither in subsequent process-specific patents by the same inventor (US 5 766 546, EP 1 104 815) where the principle of ductility improvement by adding V in combination with Mn and other elements is applied.

In EP 0 936 278 improved ductility is claimed for Al-Mg-Si alloys with additions of V in the range 0.05-0.20 wt.% in combination with addition of Mn > 0.15-0.4 wt.%. According to this EP application, the preferred Mn/Fe ratio is 0.45 - 1.0, and more preferably 0.67 - 1.0. The role of Ti in EP 0 936 278 is explicitly stated to be as a grain refiner during casting or welding. However, the preferred range for Ti is not more than 0.1 wt.%.

Adding Ti to aluminium casting alloys is commonly known from the production of Al-Si based casting alloys. Thus, it is formerly demonstrated that by adding Ti in excess of the Ti known from the use of Ti in grain refiner, an improved grain refinement is achieved for these alloys as described in CJ. Simensen: Proc. Light Metals 1999, ed. CE. Eckert, TMS, USA 1999 pp 679-684. The benefit of an addition of excess Ti is related to grain refinement and to the properties that are related to grain size in the cast metal. It is not demonstrated, however, that addition of excess Ti is beneficial for properties other than those that are related to grain size.

The present invention is related to aluminium alloys containing Mg and Si as the primary alloying elements. The alloy according to the invention contains Ti in excess of Ti commonly added as grain refiner. The alloy of the invention contains Ti in excess of the Ti-containing particles that are introduced by the grain refiner. The excess Ti contributes to improved ductility of the alloy.

The alloy may contain Cu for additional strength. Further, the alloy may contain Fe and jZη as incidental impurities. Still further, the alloy may contaiη;additional alloying elements including,, but not limited;to, Mn_;,_;Qr, Zr and V for further improving the ductility of the alloy.

The alloy is developed for extruded products where good crush behaviour is requested. The alloy is optimised for productivity and for obtaining good ductility without requiring rapid quenching of the extruded profiles at the extrusion press. However, the alloy may also be used for other products such as rolled sheet or forgings when improved ductility is requested.

The invention is characterized by the features as defined in the attached independent claim 1 and dependent claims 2 -8.

The invention will be further described in the following by way of examples and with reference to the figures, where: Fig. 1 is a diagram showing the total elongation for extruded profiles of aluminium alloys with various Mg and Si contents after age hardening to maximum hardness,

Fig. 2 shows the cross-section geometries of the extruded hollow profiles used in crush-testing in the examples in the specification

Fig. 3 shows the grade in crush-test and yield strength of the specimens of

Example 1 in the specification,

Fig. 4 shows decrease in yield strength (YS) and ultimate tensile strength (UTS) due to thermal exposure for 100Oh at 15O⁰C of age hardened extruded profiles of the alloys O and P of Example 2 in the specification,

Fig. 5 shows Vickers hardness of the alloys of Table 4 in the specification as a function of ageing time at 175⁰C- and at 20O⁰C

Fig. 6 shows the tensile yield strength in T5 condition, and after high temperature exposure of a recrystallised 6005A₁ alloy and a non-recrystallised 6082 alloy,

Fig. 7 shows grade in crush-test and yield strength of the specimens of Example

3 in the specification.

Fig. 8 shows the yield strength and Charpy energy of the water-quenched and age hardened samples according to Example 4 in the specification,

Fig. 9 shows grade in crush-test of the specimens of Example 4 in the specification, were the nominal contents of Mn, V, Cr, Cu and Ti of the different alloys are shown on the depicted chart,

Fig. 10 shows two pictures where the left picture shows the grain structure in alloy variant B1 with 0.15%Mn and 0.1% V, whereas the right picture shows the grain structure in alloy variant B3 with 0.06% Cr in addition to the elements found in alloy B1 of Example 4.

Fig. 11 shows the grade in crush-test of the specimens of Example 6 that were extruded to geometry P3 at an exit speed of 15m/min and where the nominal contents of Mn, V, Cr, Cu and Ti of the different alloys are shown on the depicted chart,

Fig. 12 shows the same as in Fig. 11 , but where the exit speed is 30 m/min and where the yield strength of the specimens is included,

Fig. 13 shows grade in crush-test and yield strength of the specimens of Example

7 in the specification that were extruded to geometry P1 at an exit speed of 15 m/min, and then water quenched prior to age hardening, where the nominal contents of Mn, V, Cr, Cu and Ti of the different alloys are shown on the in the depicted chart,

Fig. 14 shows grade in crush-test of the specimens of Example 7 in the specification that were extruded to geometry P1 at an exit speed of 15 m/min, and then air cooled prior to age hardening, where the nominal contents of Mn, V, Cr, Cu and Ti of the different alloys are shown on the in the depicted chart,

Fig. 15 shows grade in crush-test of the specimens of Example 7 in the specification that were extruded to geometry P3 at an exit speed of 15 m/min, and then air cooled prior to age hardening, where the nominal contents of Mn, V, Cr, Cu and Ti of the different alloys are shown on the in the depicted chart,

Fig. 16 shows grade in crush-test and yield strength of the specimens of Example

7 in the specification that were extruded to geometry P3 at an exit speed of 30 m/min, and then air cooled prior to age hardening, where the nominal contents of Mn, V, Cr, Cu and Ti of the different alloys are shown on the in the depicted chart, Fig. 17 shows Charpy energy vs grade in crush-test for the specimens of Example

6 that were extruded to geometry P3 at an exit speed of 30 m/min and air cooled prior to age hardening,

Fig. 18 shows differences in yield strength and Charpy energy between alloys of

Table 9 and corresponding alloys of Table 8, where all specimens were extruded to geometry P3 at an exit speed of 30 m/min and air cooled prior to age hardening,

Fig. 19 shows yield strength and Charpy energy of the alloys of Table 10 of the specification when the profiles were water-quenched prior to age hardening,

Fig. 20 shows yield strength and Charpy energy of the alloys of Table 10 when the profiles were air cooled prior to age hardening.

Fig. 21 shows a Mg-Si diagram where the Si/Mg ratio of 1.4 is drawn, as well as alloy compositions of particular interest for embodiments of the invention as defined in the claims.

Alloys of the Al-Mg-Si type gain their strength from the precipitation of particles of nanometer size. It is commonly known that the hardening particles have a molar Si/Mg ratio of approximately 1 (G.A. Edwards et al., Mater.Sci. Forum vol 217-222 (1996) pp713-718), and some investigations indicate that this ratio is exactly 1.2 (SJ. Andersen et al., Acta Mater, vol. 49 (1998) pp65-75, C. Ravi and C. Wolverton, Acta Mater, vol. 52 (2004) pp4213-4227). A molar Si/Mg ratio of 1.2 equals a Si/Mg weight ratio of approx. 1.4, and therefore, in the following text, all the given Si/Mg ratios refer to weight ratios. For optimising the strength of the alloys, the Mg and Si content should be chosen so as to ensure that as much Mg and Si as possible is used for making hardening precipitates, or in other words so as to ensure that there is as little surplus Mg or Si as possible after precipitation hardening. By surplus Mg or Si it should be understood the Mg or Si that does not form precipitates. The surplus Mg or Si contributes little to the strength, but has a distinct negative effect on the productivity in extrusion. This approach to selection of Mg and Si content of extrusion alloys has been employed earlier (U. Tundal and O. Reiso, US Patent 6.602.364, M.J. Couper et al. in Proc. Eighth International Aluminum Extrusion Technology Seminar, Orlando Florida, USA, May 18-21 2004 Vol. Il pp51-56) and is considered as well known to persons skilled in the art.

In order to find the optimal Si/Mg ratio of an alloy, one has to consider that some of the Si will be tied up in the Fe-bearing primary particles and other non-hardening particles that form during casting and homogenisation of the alloy. This Si may be considered as "lost" or without effect with respect to age hardening. One may introduce a term "effective Si content", Si_eff, defined by

Sieff = Sitot — Siphp

where Si_tot is the total Si content of the alloy and Si_nhP is the amount of Si tied up-in non- hardening particles. It is not straightforward to calculate Si_nhP, since this is related, both to the composition of the alloy and to the temperature history during homogenisation. However, ϊtrs normally found that the ratio \

Si_nhp [wt.%] / (Fe+Mn+Cr) [wt.%]

lies within the range 0.15 - 0.35.

In conclusion, the Si_eff/Mg ratio should be 1.4 in order to optimise the strength of the alloy.

Optimal composition of Mg and Si for crush ductility and temperature stability.

Extruded profiles of Al-Mg-Si alloys are used as structural components in crash-exposed areas of automobiles. Such components are required to absorb high amounts of energy in the event of a crash, and in order to do so they must deform without fracturing. One of the means of controlling that the extruded profile has the required properties is to test it by axial crushing. In this test a specimen of thin-walled extruded profile, usually a hollow profile with one or more chambers, of pre-defined length is subjected to deformation in the longitudinal direction at a controlled speed, which reduces the specimen length to typically 30-80% of the original length. Good deformation behaviour is characterised by regular folding of the specimen walls, little or no cracking of the specimens and a smooth surface of the deformed areas. Poor deformation behaviour is characterised by limited folding of the specimen walls, extensive cracking or fracturing of the specimens and a rough and uneven surface of the deformed areas.

The deformation behaviour in axial crushing tests depends strongly on the geometry of the tested profile and to some extent also on the testing conditions. Considering the geometry, the cross-section of the profile is particularly important. One general relationship is that good deformation behaviour gets increasingly difficult when the wall thickness increases and when the chamber size decreases. Besides the geometry of the tested profile and the testing conditions there are several factors that have an impact on the deformation behaviour in a crush test including, but not limited to, grain structure, strength and ductility of the extruded profile. The strength follows primarily from the chosen Mg and Si (and Cu) content in combination with the conditions of precipitation hardening. In general a higher Mg and Si (and Cu) content and thus a higher strength leads to lower ductility, for instance as found by the total elongation values in tensile testing as shown in Fig, 1 , and thus also to a poorer behaviour in crush testing. However, the Si/Mg ratio has also some importance for the crush behaviour of the extruded profiles. This is substantiated by the examples following below. In all the examples the alloys have been cast to billets by the DC casting process used in production facilities of the applicant. The billets have been homogenised at temperatures between 570-580⁰C and subsequently cooled at a rate of 300-400⁰C per hour down to room temperature. The preheating of the billets has been performed with an induction furnace to temperatures in the range of 490-500⁰C prior to extrusion.

Example 1.

Tests were performed with alloys as specified in Table 1 below, which are essentially equal except for the Si/Mg ratio. Hollow profiles were extruded from the alloys, with cross-sections as indicated in Figure 2 a). The reduction ratio during extrusion was 24, and the profile exit speed was 15m/min. The profiles were water-quenched at the extrusion press, aged to maximum hardness and cut to specimens of 100mm length. The specimens were reduced to 40mm length in controlled axial crushing, and the deformation behaviour was characterised and given a grade on a scale from 1 to 10. The definitions of the different grades are given in Table 2. For the evaluation of the grades, three samples were crushed for each alloy, and the grade of the alloy is the arithmetic mean of the three samples. Fig. 3 shows the grades given to the individual alloys as a function of Si/Mg ratio. Also shown in Fig. 3 is the yield strength of the different alloys.

Table 1 Composition of the alloys of Example 1

Table 2 Explanation of the grades given to the crush-tested specimens

Even though all alloys show a good behaviour in the crush test, alloy A1 with the highest Si/Mg ratio performs slightly poorer than the other ones and alloy B1 , with a Si/Mg ratio very close to the ideal value of 1.4, performs slightly better than the other ones. The alloys C1 , D1 , E1 and F1 all gained the grade 9 in this test. Considering the strength of the alloys, one should expect the alloys with lower strength to perform better in such a test than alloys with a higher strength. Thus, taking the strength into account one may state that within the alloys C1 , D1 , E1 and F1 the performance in the crush test is improving when the Si/Mg ratio increases from 0.7 to 1.2. In conclusion, there is a benefit of choosing alloys with a Si/Mg ratio close to 1.4 also with regards to the crush test performance.

Some structural components in crash-exposed areas may also be exposed to elevated temperatures. Such exposure may have an influence on the mechanical properties of the alloy. For such applications it is important to select alloys that are less influenced by the thermal exposure. The term "thermal stability" refers to the ability of an alloy to retain mechanical properties after exposure to high temperatures. For Al-Mg-Si alloys it is found that within a given strength class the thermal stability is highest for an alloy with a Si/Mg ratio of approx. 1.4. This is substantiated by the following examples:

Example 2.1

Tests were performed with Al-Mg-Sl alloys as specified in Table 3 below, which are essentially equal except from the Si/Mg ratio. Extruded profiles of the alloys were age hardened by using three slightly different ageing treatments, nominated I, Il and III. The age-hardened specimens were then subjected to a thermal exposure of 1000 h at 150⁰C. The yield strength (YS) and ultimate tensile strength (UTS) achieved by the ageing treatments I, Il and III are given in Table 3, whereas the changes in strength taking place as a result of the thermal exposures are illustrated in Figure 4.

Table 3 Chemical composition of the alloys of Example 2.1 , and yield strength (YS) and ultimate tensile strength (UTS) resulting from the different ageing treatments I, Il and III.

It is evident from the results of the tests as shown in Fig. 4 that alloy P with a Si/Mg ratio of 1.4 has a lower loss in strength, and particular in ultimate tensile strength, than alloy O with a Si/Mg ratio of 0.9. One may therefore state that alloy P has the highest thermal stability of the two alloys.

Example 2.2

Further tests were performed with alloys as specified in table 4 below. The alloys of table 4 are essentially equal, except for the Si/Mg ratio. Samples of extruded profiles of the alloys were artificially aged at 175°C and 200⁰C for times ranging from 0.5h to 20Oh. The Vickers Hardness of the specimens was measured, and the results are illustrated in Fig. 5. For clarity, only the results close to the maximum hardness are shown. From the curves of Fig. 5 it is evident that among the alloys of Table 4, alloy W2 has hardness close to maximum for the widest span of time. It is also consistently found that the decrease in hardness after peak hardness is reached, referred to as overageing, is more delayed in alloy W2 than for the other alloys. Fig. 5 also shows that when the Si/Mg ratio is increasingly deviating from the ideal value of 1.4, such as by going from alloy W3 to W4 and then to W5, the overageing is taking place with an increasingly shorter delay. One may state that alloys which can be subjected to longer ageing times before over aging takes place have a higher thermal stability, and consequently that alloys with a Si/Mg ratio close to 1.4 have the highest thermal stability.

Table 4: Composition of the alloys of Example 2.2

Example 2.3

Still further tests were performed to observe the influence of grain structure on thermal stability for recrystallised vs. non-recrystallised alloys. The compositions of the 6005A and the 6082 alloys used in example 2.3 are given in table 5 below. In both cases the profiles were extruded at 15 m/min and water quenched after extrusion. The Mn content of alloy 6005A is too low to prevent recrystallisation of the material during extrusion, and thus had a recrystallised grain structure. The 6082 alloy on the other hand contain a high number of dispersoid particles due to the high Mn and Cr content and thus had a non-recrystallised grain structure.

Table 5 Composition [wt%] of the alloys of example 2.3

Material from both alloys were aged to a T5 condition giving close to maximum potential strength, see Fig. 6. Upon subsequent high temperature exposure at 180°C the 6005A alloy material appears to be much more stable than the 6082 alloy material. The main reason for this difference is most likely linked to the differences in the grain structure of the extruded material. A non-recrystallised grain structure will have a lot more dislocations and subgrain boundaries, which act as fast diffusion paths for the alloying elements. Thus, the coarsening of the hardening Mg-Si precipitates will occur much faster in the non-recrystallised 6082 alloy than in the recrystallised 6005A alloy.

The influence of cooling rate after extrusion on the behaviour in crush testing. For Al-Mg-Si alloys it is generally found that the ductility after age hardening depends on the cooling rate after the preceding solution heat treatment. For extruded Al-Mg-Si alloys it is not common to apply a separate solution heat treatment, and thus the ductility after age hardening is dependent on the cooling rate of the profile at the extrusion press. A high cooling rate, such as obtained with water quenching, favours a good ductility, whereas a slow cooling rate, such as obtained with air cooling, may lead to reduced ductility.

Ideally the profiles should always be water-quenched after extrusion. However, water quenching leads to distortions of the geometry of the extruded section, and the risk of distortion increases with increasing complexity of the. profile. The general practice is to, cool as fast as possible without introducing geometrical distortions to the profiles. Thus, the majority of the extruded sections are cooled either with forced air or with controlled water-spray.

The reduced ductility that follows from a reduced cooling rate after extrusion will also have a strong impact on the behaviour of an alloy in a crush-test. This is shown in Example 3 below.

Example 3

Consider the alloys of Table 1. The same test as that of Example 1 was performed, but this time with profiles that were cooled in forced air after extrusion. The cooling time between 500°C and 250° was measured to be approximately 2 minutes. Axial crush testing was performed, and grades were given according to Table 2 above. Fig. 7 shows the grades and the yield strength of the individual alloys. For the water-quenched samples of the same alloys, the grades were in the range 8.5 - 9.5 (Fig. 3), whereas in this example the grades are in the range 5-6. This clearly illustrates the impact of cooling rate after extrusion on the crush behaviour of the alloys.

Again, the grades of alloy D1, E1 and F1 should be seen in the light of the lower strength of these alloys. In an overall evaluation, one considers the alloy C1 to have the best crush performance in this example.

The crush behaviour of Example 3 is significantly poorer than in Example 1 , but applying water-quenching as in Example 1 will impose limitations on deliverable profile geometries and geometric tolerances. Therefore the inventors were determined to find alloys that have a crush-behaviour approaching that of Example 1 , but that can be air- cooled after extrusion. It is well known among experts that small additions of the elements Mn, Cr and V improve the ductility of air-cooled extrusions of Al-Mg-Si alloy.

Alloying elements for improving ductility in extruded Al-Mg-Si alloys. Manganese (Mn). and Chromium (Cr) have several known purposes in the -Al-Mg-Si ^: extrusion alloys. Both elements form small particles, referred to as dispersoids, during the homogenisation of the cast material. When present in a sufficient number density, these dispersoids may prevent recrystallisation of the extruded material, resulting in a fibrous microstructure. For lower number densities of dispersoids the extruded material will recrystallise, but the presence of dispersoids has a positive influence on the ductility of the age-hardened material. It has been found that this influence is in part related to the texture of the material. For alloys that recrystallise after extrusion one normally finds a high degree of cube texture in the extruded profiles. The presence of dispersoids leads to a higher degree of cube texture in recrystallised extruded profiles.

Example 4

Tests were performed with alloys as specified in Table 6 below, which are essentially equal except for the Mn and Cr contents. Table 6: Composition of the alloys of Example 4

The alloys were extruded to flat bar profiles. This generates a high degree of cube texture in the alloy material. Some material was subjected to additional thermo mechanical treatment to promote an as texture-free material as possible. The texture intensities obtained are indicated in Table 7.

Table 7: Texture intensities measured in the samples.

The samples were solution heat treated, and either water quenched or air cooled to room temperature. Subsequently the samples were subjected to 5 different ageing procedures, labelled from a1 to a5 in this example. Tensile testing and Charpy testing was performed on the age hardened material, and the results from the water-quenched samples are shown in Fig. 8. By comparing Figures 8 a), c) and e) which are all from texture-free material one finds that there is a positive effect of Mn and Cr on the ductility. However, when comparing texture-free vs. cube-textured material (Fig. 8 a) vs. 8 b), 8 c) vs. 8 d) and 8 e) vs. 8 f)) it is evident that the cube texture has a profound positive effect on the ductility of the alloys as measured in the Charpy test. Similar findings were also found for the samples that were air-cooled prior to age hardening. The amount of dispersoids formed per weight percent element added is significantly higher for Cr than for Mn (O. Lohne and AL Dons: Scand. J. Metall. vol. 12, (1983) pp34-36 ), meaning that a lower Cr addition than Mn addition is required to achieve a specific number density of dispersoids. The dispersoids have three adverse effects on the extrusion process. The first adverse effect is that the hot deformation resistance of the material increases, leading to a decrease in the productivity potential. The second adverse effect is that increasing dispersoid density imposes increasing demands on the cooling rate after extrusion to avoid loss of hardening potential of the alloy. The reason for this is that the dispersoids act as nucleation sites for non-hardening Mg-Si precipitates. The third adverse effect is linked to the grain size of the extruded profile. If the number density of dispersoids is too low to prevent recrystallisation it can still prevent some nuclei from growing to recrystallised grains. With only a few grains able to grow, the result may be a very coarse grain structure in the extruded profile. Upon subsequent forming of the extruded profile the result can be extensive orange peeling. Therefore one seeks to avoid Mn and Cr additions in excess of what is necessary for benefiting from improved ductility. The optimal content of Mn and Cr depends strongly on processing conditions and profile geometry. For many conditions the typical additions for improving ductility are in the range 0.03 - 0.25 wt.% Mn and 0.01 - 0.15 wt.% Cr. If the two elements are combined the amount of each element may have to be reduced in order to keep the total number of dispersoids at an acceptable level.

Zirconium (Zr) is also an alloying element that forms dispersoids during homogenisation of cast material. Zr may form several types of dispersoids. The dispersoid type that forms in highest number density, and thus normally is the preferred type, has a composition of AI₃Zr and an atomic arrangement referred to as LI₂ structure. In alloys of the Al-Mg-Si system the formation of LI₂ AI₃Zr is not always feasible, and other types of dispersoids will form. The other types of dispersoids may contain Si in addition to Zr and Al. The effect of Zr dispersoids on the microstructure of extruded Al alloys is primarily related to the number density and to a smaller degree related to dispersoid type. When the number density of dispersoids is high the extruded material will have a fibrous microstructure, whereas material with a low number density of dispersoids will recrystallise. The presence of Zr-based dispersoids has a similar effect on the texture of the recrystallised material as Mn- and Cr-based dispersoids as explained above, and thus also leads to a higher ductility of the age hardened material. Vanadium (V) has a documented effect in increasing the ductility of Al-Mg-Si alloys. V may form dispersoids in Al-Mg-Si alloy, but for additions of up to 0.1 wt.% and possibly higher one finds that no appreciable amount of dispersoids is formed.

Titanium (Ti) is normally added to Al alloys together with boron (B) or carbon (C) for the purpose of refining the grain size of the alloy during casting. Ti and B or Ti and C are not added individually to the melt, but as pre-prepared Al-Ti-B or Al-Ti-C alloys. The pre- prepared Al-Ti-B or Al-Ti-C alloys are commonly referred to as a "grain refiner". The Al- Ti-B grain refiner often contains two classes of particles, one class consisting essentially of Ti and B and these particles are in the following denoted as (Ti₁B) particles, and another class consisting essentially of Ti and Al and these particles are in the following denoted as (AI₁Ti) particles. The Al-Ti-B grain refiners are often characterized by the weight ratio between the Ti and B content, and the Ti/B ratio is normally in the range 2- 10. When the Al-Ti-B grain refiner is added to a melt, the (Ti,B) and (AI₁Ti) particles are dispersed in the melt and upon casting they act as nucleation points for aluminium grains during the solidification. The Al-Ti-C. grain refiners work essentially in the same manner, except that they contain (Ti,C) particles instead of (Ti₁B) particles. In most alloy specifications for Al-Mg-Si alloys (International Alloy Designations and Chemical Composition Limits for Wrought Aluminum and Wrought Aluminum Alloys, The Aluminum Association, Washington DC, USA, April 2004) an upper limit for Ti in the range 0.1 - 0.2 wt.% is specified. However, it is well known that the actual Ti content needed for obtaining grain refinement in Al-Mg-Si alloys is much lower, typically in the range 0.005 - 0.03 wt.%.

The inventors of the present invention found that Ti also has an effect on the ductility of the Al-Mg-Si alloys. This requires a Ti content in excess of what would otherwise be used for grain refinement, and it requires Ti in excess of the (Ti₁B) and/or (Ti₁C) particles from the grain refiner. The amount of Ti required is within the range 0.03 - 0.25 wt.%, and preferably within 0.05 - 0.20 wt.%. As for V, additions of Ti up to about 0.25% will probably not produce any substantial amount of dispersoids. Thus, the mechanism for improving ductility is probably the same for both V and Ti. The improvement in crush performance obtained by adding Ti to an Al-Mg-Si alloy is comparable to the improvements by adding Mn, Cr or V to the alloys. This is substantiated by the following examples:

Example 5

Tests were performed with Al-Mg-Si alloys as specified in Table 8 below, which have essentially equal contents of Mg and Si but different amounts of the elements Mn, Cr, V, Cu and Ti.

Table 8: Composition of the alloys of Examples 5 and 6.

The same tests as that of Example 3 were performed, with cooling of the extrusions in forced air. Specimens of the age-hardened profiles were subjected to axial crush testing, and grades were given according to Table 2. The grades of the individual alloys are shown in Fig. 9.

By comparing BO with alloy variants B2, B5 and B6 one can see that Mn and Cr without any V or Ti have a positive effect on the crush behaviour. By comparing B2 and B5 one can see that increased Mn in B2 as compared to B5 has a positive effect on the behaviour in the crush test. The same is the case with Cr where alloy B2 has a higher Cr content than alloy B6 and a correspondingly higher grade in the crush test. However, the levels of Mn and Cr used in alloys B2, B3 and B4 were in this case found to be too high to get acceptable grain sizes, see examples in Fig. 10. When comparing the grain size in a profile from any of the alloys B2, B3 and B4 with for example the grain size in alloy B1 with only 0.15% Mn and no Cr, it is found that the grain size in for instance alloy B3 is too large and may lead to extensive orange peeling in a subsequent forming operation. Thus, the improvement in crush behaviour that can be achieved by adding Mn and Cr alone is therefore limited by the grain size requirements that apply.

By comparing alloys B1 and B7 one finds that the additions of V and Ti have approximately equal positive effects on the behaviour in the present crush test. By comparing alloys B2 and B3, where only the V content is different, one can see the positive effect of V. The same is the case for Ti, which is seen when comparing alloys B6 and B8. The positive effect of V and Ti thus comes in addition to the positive effect of Mn and Cr. Since V and Ti both are elements that do not form dispersoid particles for the amounts added here, no problems with the grain size is expected by additions of these elements. The best behaviour is obtained when several elements are added in combination, such as Mn, V, Cr and Ti in alloy B9.

The maximum amount of Ti and V that can be added is limited by the amount that can be kept in solid solution during processing of the alloy. Another factor that must be considered is the increase in the deformation resistance that these elements result in. High levels of these elements will reduce the extrudability of the alloy and the improvement in crush behaviour must be seen in the light of the reduction in extrudability. Finally, additions of V and Ti to the alloys represent increased cost due the higher price of V and Ti as compared to aluminium. With the current prices of V, Ti and Al an addition of 0.10% V leads to an increase of approximately 110 € per ton of the aluminium alloy. For the same addition of Ti the increase is only 10 € per ton of the aluminium alloy. Thus, addition of Ti is preferred as compared to V from a cost perspective.

Robustness of the effects of alloying elements on ductility.

The extrusion reduction ratio and extrusion exit speed may vary considerably between different extruded geometries. These variations have an effect of the microstructure of the extruded profile, which in turn may have an impact on the ductility in a crush test. Further, it is known that higher strength in general leads to poorer folding behaviour in an axial crushing test. Thus several tests were performed in order to verify that the findings of Example 5 are valid for variations in processing conditions and for variations in strength.

Example 6

All alloys of Table 8 were extruded to the geometry P3 shown in Fig. 1 b). This gives an extrusion reduction ratio of 48, which is the double of the extrusion ratio in Example 5. Two extrusion speeds were employed for the P3 geometry: 15 m/min and 30 m/min. Air cooling of profiles with the geometry P3 was somewhat faster than that of geometry P1 , due to the smaller size and wall thickness. The time of cooling in the temperature interval 500⁰C - 250⁰C was approx. 1.3 min for extrusions with the geometry P3.

The extruded and cooled profiles were age hardened to maximum hardness and cut to specimens with lengths of 70 mm. The specimens were subjected to axial crushing, whereby the specimen length was reduced to 32mm. Grades were given according to: Table 2, and the grades of the individual alloys under the different conditions are shown: in Figs. 11 and 12. Figure 12 also includes the yield strength of the specimens

Except for unexpected low values for variants B5 and B6 in Figure 11 , the results confirm the findings of Example 5. Again alloys B1 with 0.10% V and B7 with 0.10% Tl have approximately the same grades after crush testing, indicating that V and Ti has about the same effect on the ductility. In Fig. 12 the results are more as expected since alloy variants B5 and B6 have better crush behaviour than alloy BO.

On average, the grades after crush testing are higher in Figure 12 than in Figure 11. This is to some extent related to the grain structure of the material. A high extrusion speed is favourable for avoiding a coarse-grained structure of these types of alloys, and thus also favours an improved performance in crush testing. The alloys that contain Ti and/or V give on average less variation in crush performance when the extrusion speed varies.

Example 7

Further tests were performed with the alloys as specified in Table 9 below. The alloys

GO and G5 through G9 are essentially equal to the alloys BO and B5 through G9 with the exception of the Mg and Si content. The Mg and Si content of the alloys of Table 9 is somewhat higher than that of the alloys of Table 8, meaning that the alloys of Table 9 should have a somewhat higher strength than the corresponding alloys of Table 8. The stronger alloys are expected to have a slightly lower ductility, and therefore also a slightly lower performance in an axial crush test.

Table 9: Com osition of the allo s of Exam le 7.

The alloys were extruded under four different conditions:

Geometry P1 , extrusion exit speed 15m/min, water quenching after extrusion Geometry P1 , extrusion exit speed 15m/min, air cooling after extrusion Geometry P3, extrusion exit speed 15m/min, air cooling after extrusion Geometry P3, extrusion exit speed 30m/min, air cooling after extrusion

The extruded and cooled profiles were age hardened to maximum hardness and cut to specimens with lengths of 100 mm for geometry P1 and 70 mm for geometry P3. The specimens were subjected to axial crushing, whereby the specimen length was reduced to 40mm for geometry P1 and 32mm for geometry P3. Grades were given according to Table 2, and the grades of the individual alloys under the different conditions are shown in Figs. 13 through 16. Figure 13 and Figure 16 also includes the yield strength of the specimens.

Except for alloy G7 in Fig. 14 and G6 in Fig. 16 the results confirm the findings of Example 5 and Example 6, where additions of Mn, Cr, V and Ti give an improvement in the ductility and performance in an axial crush test. Charpy testing

The Charpy V-notch test is a test of a material's ability to absorb energy during failure. It is found that within groups of alloys that are somewhat similar there is a high degree of correlation between the amount of energy absorbed in a Charpy test and the behaviour in an axial crush test. This is substantiated in Fig. 17, which shows the correlation between the grades in the crush test of Fig. 16 and the absorbed energy in Charpy tests of the same material. Except for alloy GO, which shows a too low Charpy energy compared to the grade in the crush test, one finds that there is almost a linear relationship between the Charpy energy and the crush test grade for the alloys.

Further, it is a general trend that the Charpy energy decreases with increasing strength of the alloy. Consider the alloys BO and B5 through B9 of Table 8 and the alloys GO and G5 through G9 of Table 9. The alloys of table 9 have a higher Mg and Si content and thus reach a higher strength after age hardening than the corresponding alloys of Table 8. Charpy testing of all alloys extruded to geometry P3 at an exit speed of 30 m/min and air-cooled prior to ageing has been performed. The differences in yield strength and Charpy energy between the alloys of Table 9 and the corresponding alloys of Table 8 are shown in Figure 18. One finds that within this group of alloys there is almost a linear decrease in Charpy energy with increasing strength.

Given these correlations, one may state that for comparing a set of alloys that are somewhat similar the Charpy testing gives a good indication on their expected relative behaviour in an axial crush test. Such a comparison is given in Example 8 below.

Example 8

Tests were performed with alloys as specified in Table 10, having essentially equal contents of Mg and Si but different amounts of the elements Mn, Cr, V, Cu and Ti. The alloy X1 is the base alloy, meaning that the other alloys consist of alloy X1 with additional alloying elements. The Mg and Si contents are slightly higher than those of the alloys of Table 9, which means that the strength after age hardening of the alloys of Table 10 in general is slightly higher than the strength after age hardening of the alloys of Table 9. Table 10: Composition of the alloys of Example 8.

Flat bars of the alloys were extruded. Two different extrusion exit speeds were used, 10m/min and 40 m/min. Specimens were solution heat-treated and either water- quenched or air cooled before age hardening. Uniaxial tensile tests and Charpy V-notch testes were performed on the age hardened material. Figure 19 and 20 show the Charpy energy vs. the yield strength of the profiles that were water quenched and air- cooled, respectively; prior to age hardening. No distinctions between the two. extrusion speeds are made in Fig. 19 and Fig. 20. The data tables indicate which alloying - elements that were added to the base alloy, in accordance with Table 10. '

For the profiles that were water-quenched prior to age hardening, Fig. 19 indicates that among the tested alloy compositions, the addition of Mn+Cr has the highest positive influence on the Charpy energy whereas the additions of Mn+V and Mn+Ti have the second highest positive influence on the Charpy energy.

The same ranking is largely found in Fig. 20 for the profiles that were air-cooled prior to age hardening. Addition of Mn+Cr had the highest positive influence, addition of Mn+V had the second highest positive influence and addition of Mn+Ti had the third highest positive influence on the Charpy energy as such. However, the yield strength is lowest for the alloy with Mn+Cr addition, second lowest for the alloy with Mn+V addition and third lowest for a group of other alloys, including the alloy with Mn+Ti addition. These differences in strength may partially account for the differences in Charpy energy (see

Fig. 18). If all the three alloys discussed above had had the same yield strength it is likely that the ranking in Charpy energy would have remained unchanged, but that the differences in Charpy energy between them would have been smaller.

From the examples and discussion given in the present text, it is clear that carefully chosen additions of alloying elements such as Mn, Cr, Zr, V and Ti significantly improves the ductility and crush properties of Al-Mg-Si alloys. For optimal combination of properties and processability it is particularly useful to combine alloying elements that form dispersoids (Mn, Cr, Zr) with alloying elements that are predominantly in solid solution (V, Ti). These principles are valid for the whole range of Mg and Si contents of Al-Mg-Si alloys. However, for optimal combination of process ability with properties such as strength and thermal stability it is favourable to choose alloys that have a Si_eff/Mg ratio close to 1.4 as discussed initially. Fig. 21 shows a Mg-Si diagram where the Si/Mg ratio of 1.4 is drawn, and also shows alloy compositions of particular interest for embodiments of the invention as defined in the claims.

Claims

1. An Al-Mg-Si alloy with good ductility and improved crush properties, in particular useful for structural components in crash exposed areas in vehicles, characterised in that the alloy contains in wt %: Mg 0,25 - 1 ,2 Si 0,3-1,4 Ti 0,1 - 0,4, where Ti is present in solid solution and where the alloy contains in addition one or more of the following alloy components:

Mn max 0.6 Cr max 0.3

Zr max 0,25, and incidental impurities, including Fe and Zn, up to 0,5 with balance Al.

2. Alloy according to claim 1 , characterised in that itcontains in addition one or more of the following alloy components in wt%: Cu max 0.4 preferably 0,3

V max 0.25

3. Alloy according to claims 1 and 2, characterised in that the alloy contains by wt% between 0,1 - 0,3 Ti and preferably between 0,1 - 0,2 Ti.

4. Alloy according to claims 1 - 3, characterised in that the composition of the alloy is defined within the following coordinate points of an Mg-Si diagram: a1 - a2 - a3 - a4 - a1 , where in wt% a1=0.25Mg, 0.55Si, a2=0.50Mg,1.0Si , a3=0.75Mg ,0.75Si and a4=0.45Mg, 0.40Si.

5. Alloy according to claim 4, characterised in that the alloy is defined within the coordinate points b1 - b2 - b3 - b4 - b1 , where in wt% b1=0.30Mg, 0.60Si, b2=0.50Mg, 0.90Si , b3=0.65Mg, 0.75Si and b4=0.45Mg, 0.50Si.

6. Alloy according to claim 5, characterised in that more preferably the alloy is defined between the coordinate points d - c2 - c3 - c4 - c1 , where in wt% d=0.33Mg, 0.60Si, c2=0.47Mg, 0.80Si , c3=0.59Mg, 0.70Si and c4=0.45Mg, 0.52Si.

7. Alloy according to claims 1- 6 characterized in that the alloy contains by wt% Mn 0,05 - 0,30 Cr max 0,05

Zr max 0,15

8. Alloy according to claims 1 -7, characterised in that the alloy is cast to billets and then homogenised.

9. Alloy according to claims 1-8, characterised in that the alloy is reheated to a preferred temperature and then extruded.