The effect of frequency and microstructure on corrosion fatigue crack propagation in high strength aluminium alloys

Fatigue crack growth in high strength aluminium alloy 7150 commercial plate material has been studied in both laboratory air and acidified aqueous salt solution. The aggressive aqueous environment enhanced fatigue crack growth rates by up to an order in magnitude compared to laboratory air. The enhancement in fatigue crack growth rate was accompanied by evidence of embrittlement in the crack path, involving both brittle intergranular and transgranular failure modes. Both the enhancement of fatigue crack growth rates and the extent of intergranular growth modes are dependent on cyclic frequency which, along with the absence of a similar frequency effect in a spray-formed version of the material with a significantly different grain structure, supports a mechanism of grain boundary hydrogen diffusion for intergranular corrosion fatigue crack growth. The convergence of corrosion fatigue crack growth rates at high ΔK in both spray-formed and conventional plate materials coincides with the operation of identical transgranular corrosion fatigue modes dependent on strain-controlled hydrogen diffusion ahead of the crack tip. © 1997 Acta Metallurgica Inc.

Fatigue crack growth behaviour in molten-metal processed SiC particle-reinforced aluminium alloys

Fatigue crack initiation and subsequent short crack growth behaviour of 2014-5wt%SiC aluminium alloy composites has been examined in 4-point bend loading using smooth bar specimens. The growth rates of long fatigue cracks have also been measured at different stress ratios using pre-cracked specimens. The distributions of SiC particles and of coarse constituent particles in the matrix (which arise as a result of the molten-metal processing and relatively slow cooling rate) have been investigated. Preferential crack initiation sites were found to be SiC-matrix interfaces, SiC particles associated with constituent particles and the coarse constituent particles themselves. For microstructurally short cracks the dispersed SiC particles also act as temporary crack arresters. In the long crack growth tests, higher fatigue crack growth rates were obtained than for monolithic alloys. This effect is attributed to the contribution of void formation, due to the decohesion of SiC particles, to the fatigue crack growth process in the composite. Above crack depths of about 200 μm 'short' crack growth rates were in good agreement with the long crack data, showing a Pris exponent, m = 4 in both cases. For the long crack and short crack growth tests little effect of specimen orientation and grain size was observed on fatigue crack growth rates, but, specimen orientation affected the toughness. No effect of stress ratio in the range R = 0.2-0.5 was seen for long crack data in the Paris region.

Comparison of crack growth behaviour between spray-formed and commercial plate Al-Zn-Mg-Cu alloy 7150

This paper compares the crack growth resistance of an experimental spray-formed extrusion with that of a commercial aluminium alloy, the two alloys having similar compositions but markedly different grain structures. Tensile and fracture behaviour is similar in both materials and is influenced by inclusion content. The two materials differ in their crack growth resistance, which is shown to be dependent upon grain size and shape. Environmentally-induced crack growth is favoured by aligned grain boundaries and small grain size.

Development and evaluation of a non-destructive adhesion test

The object of this work was to further develop the idea introduced by Muaddi et al (1981) which enables some of the disadvantages of earlier destructive adhesion test methods to be overcome. The test is non-destructive in nature but it does need to be calibrated against a destructive method. Adhesion is determined by measuring the effect of plating on internal friction. This is achieved by determining the damping of vibrations of a resonating specimen before and after plating. The level of adhesion was considered by the above authors to influence the degree of damping. In the major portion of the research work the electrodeposited metal was Watt's nickel, which is ductile in nature and is therefore suitable for peel adhesion testing. The base metals chosen were aluminium alloys S1C and HE9 as it is relatively easy to produce varying levels of adhesion between the substrate and electrodeposited coating by choosing the appropriate process sequence. S1C alloy is the commercially pure aluminium and was used to produce good adhesion. HE9 aluminium alloy is a more difficult to plate alloy and was chosen to produce poorer adhesion. The "Modal Testing" method used for studying vibrations was investigated as a possible means of evaluating adhesion but was not successful and so research was concentrated on the "Q" meter. The method based on the use of a "Q" meter involves the principle of exciting vibrations in a sample, interrupting the driving signal and counting the number of oscillations of the freely decaying vibrations between two known preselected amplitudes of oscillations. It was not possible to reconstruct a working instrument using Muaddi's thesis (1982) as it had either a serious error or the information was incomplete. Hence a modified "Q" meter had to be designed and constructed but it was then difficult to resonate non-magnetic materials, such as aluminium, therefore, a comparison before and after plating could not be made. A new "Q" meter was then developed based on an Impulse Technique. A regulated miniature hammer was used to excite the test piece at the fundamental mode instead of an electronic hammer and test pieces were supported at the two predetermined nodal points using nylon threads. This instrument developed was not very successful at detecting changes due to good and poor pretreatments given before plating, however, it was more sensitive to changes at the surface such as room temperature oxidation. Statistical analysis of test results from untreated aluminium alloys show that the instrument is not always consistent, the variation was even bigger when readings were taken on different days. Although aluminium is said to form protective oxides at room temperature there was evidence that the aluminium surface changes continuously due to film formation, growth and breakdown. Nickel plated and zinc alloy immersion coated samples also showed variation in Q with time. In order to prove that the variations in Q were mainly due to surface oxidation, aluminium samples were lacquered and anodised Such treatments enveloped the active surfaces reacting with the environment and the Q variation with time was almost eliminated especially after hard anodising. This instrument detected major differences between different untreated aluminium substrates.Also Q values decreased progressively as coating thicknesses were increased. This instrument was also able to detect changes in Q due to heat-treatment of aluminium alloys.

The mechanical and wear behaviour of B(SiC) fibre-reinforced composite materials

The mechanical properties and wear behaviour of B(SiC) fibre-reinforced metal matrix composites (MMCs) and aluminium alloy (2014) produced by metal infiltration technique were determined. Tensile tests were peliormed at different conditions on both the alloy matrix and its composite, and the tensile fracture surfaces were also examined by Scanning Electron Microscopy (SEM). Dry wear of the composite materials sliding on hardened steel was studied using a pin-on-disc type machine. The effect of fibre orientation on wear rate was studied to provide wear resistance engineering data on the MMCs. Tests were carried out with the wear surface sliding direction set normal, parallel and anti-parallel to the fibre axis. Experiments were perfonned for sliding speeds of 0.6, 1.0 and 1.6 m/s for a load range from 12 N to 60 N. A number of sensitive techniques were used to examine worn surface and debris, i.e: Scanning Electron Microscopy (SEM), Backscattered Electron Microscopy (BSEM) and X-ray Photoelectron Spectroscopy (XPS). Finally, the effect of fibre orientation on the wear rate of the Borsic-reinforced plastic matrix composites (PMCs) produced by hot pressing technique was also investigated under identical test conditions. It was found that the composite had a markedly increased tensile strength compared with the matrix. The wear results also showed that the composite exhibited extremely low wear rates compared to the matrix material and the wear rate increased with increasing sliding speed and normal load. The effect of fibre orientation was marked, the lowest wear rates were obtained by arranging the fibre perpendicular to the sliding surface, while the highest wear was obtained for the parallel orientation. The coefficient of friction was found to be lowest in the parallel orientation than the others. Wear of PMCs were influenced to the greatest extent by these test parameters although similar findings were obtained for both composites. Based on the results of analyses using SEM, BSED and XPS, possible wear mechanisms are suggested to explain the wear of these materials.

The deoxidation of steel with simple and complex deoxidisers

The deoxidation of steel with complex deoxidisers was studied at 1550°C and compared with silicon, aluminium and silicon/aluminium alloys as standards. The deoxidation alloy systems, Ca/Si/Al, Mg/Si/Al and Mn/Si/Al, were chosen for the low liquidus temperatures of many of their oxide mixtures and the potential deoxidising power of their constituent elements. Product separation rates and compositional relationships following deoxidation were examined. Silicon/aluminium alloy deoxidation resulted in the product compositions and residual oxygen contents expected from equilibrium and stoichiometric considerations, but with the Ca/Si/Al and Mg/Si/Al alloys the volatility of calcium and magnesium prevented them participating in the final solute equilibrium, despite their reported solubility in liquid iron. Electron-probe microanalysis of the products showed various concentrations of lime and magnesia, possibly resulting from reaction between the metal vapours and dissolved oxygen.The consequent reduction of silica activity in the products due to the presence of CaO and hgO produced an indirect effect of calcium and magnesium on the residual oxygen content. Product separation rates, indicated by vacuum fusion analyses, were not significantly influenced by calcium and magnesium but the rapid separation of products having a high Al2O3Si02 ratio was confirmed. Manganese participated in deoxidation, when present either as an alloying element in the steel or as a deoxidation alloy constituent. The compositions of initial oxide products were related to deoxidation alloy compositions. Separated products which were not alumina saturated, dissolved crucible material to achieve saturation. The melt equilibrated with this slag and crucible by diffusion to determine the residual oxygen content. MnO and SiO2 activities were calculated, and the approximate values of MnO deduced for the compositions obtained. Separation rates were greater for products of high interfacial tension. The rates calculated from a model based on Stoke's Law, showed qualitative agreement with experimental data when corrected for coalescence effects.

Crack closure and residual stress effects in fatigue of a particle-reinforced metal matrix composite

A study of the influence of macroscopic quenching stresses on long fatigue crack growth in an aluminium alloy-SiC composite has been made. Direct comparison between quenched plate, where high residual stresses are present, and quenched and stretched plate, where they have been eliminated, has highlighted their rôle in crack closure. Despite similar strength levels and identical crack growth mechanisms, the stretched composite displays faster crack growth rates over the complete range of ΔK, measured at R = 0.1, with threshold being displaced to a lower nominal ΔK value. Closure levels are dependent upon crack length, but are greater in the unstretched composite, due to the effect of surface compressive stresses acting to close the crack tip. These result in lower values of ΔKeff in the unstretched material, explaining the slower crack growth rates. Effective ΔKth values are measured at 1.7 MPa√m, confirmed by constant Kmax testing. In the absence of residual stress, closure levels of approximately 2.5 MPa√m are measured and this is attributed to a roughness mechanism.

The influence of ageing on fatigue crack growth in SiC-particulate reinforced 8090

A study of the influence of SiC-particulate reinforcement on ageing and subsequent fatigue crack growth resistance in a powder metallurgy 8090 aluminium alloy-SiC composite has been made. Macroscopic hardness measurements revealed that ageing at 170°C in the composite is accelerated with respect to the unreinforced alloy, though TEM studies indicate that this is not due to the enhanced precipitation of S′. Fatigue crack growth rates in the naturally aged condition of the composite and unreinforced matrix are similar at low to medium values of ΔK, but diverge above ≈ 8 MPa√m owing to the lower fracture toughness of the composite. As a result of the presence of the reinforcement, planar slip in the composite is suppressed and facetted crack growth is not observed. Ageing at or above 170°C has a deleterious effect on fatigue crack growth. Increased ageing time decreases the roughness of the fracture path at higher growth rates. These effect are though to be due to microstructural changes occurring at or near to the SiC/matrix interfaces, providing sites for static mode failure mechanisms to operate. This suggestion is supported by the observation that as ΔK increases, crack growth rates become Kmax dependent, implying the crack growth rate is strongly influenced by static modes.

Lithium-aluminium casting alloys and their associated metal-mould reactions

Aluminium - lithium alloys are specialist alloys used exclusively by the aerospace industry. They have properties that are favourable to the production of modern military aircraft. The addition of approximately 2.5 percent lithium to aluminium increases the strength characteristics of the new alloys by 10 percent. The same addition has the added advantage of decreasing the density of the resulting alloy by a similar percentage. The disadvantages associated with this alloy are primarily price and castability. The addition of 2.5 weight percent lithium to aluminium results in a price increase of 100% explaining the aerospace exclusivity. The processability of the alloys is restricted to ingot casting and wrought treatment but for complex components precision casting is required. Casting the alloys into sand and investment moulds creates a metal - mould reaction, the consequences of which are intolerable in the production of military hardware. The primary object of this project was to investigate and characterise the reactions occurring between the newly poured metal and surface of the mould and to propose a method of counteracting the metal - mould reaction. The constituents of standard sand and investment moulds were pyrolised with lithium metal in order to simplify the complex in-mould reaction and the products were studied by the solid state techniques of powder X-Ray diffraction and magic angle spinning nuclear magnetic resonance spectroscopy. The results of this study showed that the order of reaction was: Organic reagents> > Silicate reagents> Non silicate reagents Alphaset and Betaset were the two organic binders used to prepare the sand moulds throughout this project. Studies were carried out to characterise these resins in order to determine the factors involved in their reaction with lithium. Analysis revealed that during the curing process the phenolic hydroxide groups are not reacted out and that a redox reaction takes place between these hydroxides and the lithium in the molten alloys. Casting experiments carried out to assess the protection afforded by various hydroxide protecting agents showed that modern effective, protecting chemicals such as bis-trimethyl silyl acetamide and hexamethyldisilazane did not inhibit the metal - mould reaction to a sufficiently high standard and that tri-methylchlorosilane was consistently the best performer. Tri-methyl chlorosilane has a simple functionalizing mechanism compared to other hydroxide protecting reagents and this factor is responsible for its superior inhibiting qualities. Comparative studies of 6Li and 7Li N.M.R. spectra (M.A.S. and `off angle') establish that, for solid state (and even solution) analytical purposes 6Li is the preferred nucleus. 6Li M.A.S.N.M.R. spectra were obtained for thermally treated laponite clay. At temperatures below 800oC both dehydrated and rehydrated samples were considered. The data are consistent with mobility of lithium ions from the trioctahedral clay sites at 600oC. The superior resolution achievable in 6Li M.A.S.N.M.R. is demonstrated in the analysis of a microwave prepared lithium exchanged clay where 6Li spectroscopy revelaed two lithium sites in comparison to 7Li M.A.S.N.M.R. which gave only a single lithium resonance.

Fatigue of commercial aluminium alloys

Fatigue crack propagation has been observed for a number of commercial aluminium alloys. Comparable data was obtained for a variety of crack and specimen geometries over a range of crack lengths for a given alloy. Where crack propagation only was of interest the initiation event has been excluded by pre-cracking the specimen using a fin of material adjacent to the crack face. By this method a controlled defect size is introduced in to the specimen. By modification of the D.C. potential drop method it has been shown possible to measure the growth of cracking from 0.12mm by this method. Crack growth from defects greater than 0.6mm have been shown to give conventional crack propagation deduced by principle of similitude. Fatigue fracture surface analysis has been conducted for cracking from both free surfaces and from blunt notches. A `quasi cleavage' feature has been identified and is shown to be prominent when the fatigue stress intensity range is below 10 MNm-3/2.

The creep behaviour of pressure diecast zinc-aluminium based alloys

The creep behaviour of three pressure diecast commercial zinc-aluminium based alloys: Mazak 3, corresponding to BS 1004A, and the new alloys ZA.8 and ZA.27 with a series of alloys with compositions ranging from 0% to 30% aluminium was investigated. The total creep elongation of commercial alloys was shown to be well correlated using an empirical equation. Based on this a parametrical relationship was derived which allowed the total creep extension to be related to the applied stress, the temperature and the time of test, so that a quantitative assessment of creep of the alloys could be made under different conditions. Deviation from the normal creep kinetics occurred in alloys ZA.8 and ZA.27 at very low stresses, 150°C, due to structural coarsening combined with partial transformation of ε -phase into T' phase. The extent of primary creep was found to increase with aluminium content, but secondary creep rates decreased in the order Mazak 3, ZA.8 and ZA.27. Thus, based on the above equation, ZA.8 was found to have a substantially better total creep resistance than ZA.27, which in turn was marginally better than Mazak 3 for strains higher than 0.5%, but inferior for smaller strains, due to its higher primary creep extension. The superior creep resistance of ZA.8 was found to be due to the presence of strictly-orientated, thin plate-like precipitates of ε(CuZn4) phase in the zinc matrix of the eutectic and the lamellarly decomposed β phase, in which the precipitation morphology and orientation of ε in the zinc matrix was determined. Over broad ranges of temperature and stresses, the stress exponents and activation energies for creep were found to be consistent with some proposed creep rate mechanisms; i.e. viscous glide for Mazak 3, dislocation climb over second phase particles for ZA.8 and dislocation climb for ZA.27, controlled by diffusion in the zinc-rich phase. The morphology of aluminium and copper-rich precipitates formed from the solid solution of zinc was clearly revealed. The former were found to further increase the creep rate of inherently low creep resistant zinc, but the latter contributed significantly to the creep resistance. Excess copper in the composition, however, was not beneficial in improving the creep resistance. Decomposition of β in copper-containing alloys was found to be through a metastable Zn-Al phase which is strongly stabilised by copper, and the final products of the decomposition had a profound effect on the creep strength of the alloys. The poor creep resistance of alloy ZA.27 was due to the presence of particulate products derived from decomposed β-phase and a large volume of fine, equiaxed products of continuously decomposed α-dendrites.

The fatigue properties of pressure diecast zinc-aluminium based alloys

The fatigue behaviour of the cold chamber pressure-die-cast alloys: Mazak3, ZA8, ZA27, M3K, ZA8K, ZA27K, K1, K2 and K3 was investigated at temperature of 20°C. The alloys M3K, ZA8K and ZA27K were also examined at temperatures of 50 and 100°C. The ratio between fatigue strength and tensile strength was established at 20°C at 107 cycles. The fatigue life prediction of the alloys M3K, ZA8K and ZA27K was formulated at 20, 50 and 100°C. The prediction formulae were found to be reasonably accurate. All of the experimental alloys were heterogeneous and contained large but varying amounts of pores. These pores were a major contribution and dominated the alloys fatigue failure. Their effect, however, on tensile failure was negligible. The ZA27K possessed the highest tensile strength but the lowest fatigue strength. The relationship between the fracture topography and the microstructure was also determined by the use of a mixed signal of a secondary electron and a back-scattered electron on the SEM. The tensile strength of the experimental alloys was directly proportional to the aluminium content within the alloys. The effect of copper content was also investigated within the alloys K1, K2, ZA8K and K3 which contained 0%, 0.5%, 1.0% and 2.0% respectively. It was determined that the fatigue and tensile strengths improved with higher copper contents. Upon ageing the alloys Mazak3, ZA8 and ZA27 at an ambient temperature for 5 years, copper was also found to influence and maintain the metastable Zn-Al (αm) phase. The copper free Mazak3 upon ageing lost this metastable phase. The 1.0% copper ZA8 alloy had lost almost 50% of its metastable phase. Finally the 2.0% copper ZA27 had merely lost 10% of its metastable phase. The cph zinc contained a limited number of slip systems, therefore twinning deformation was unavoidable in both fatigue and tensile testing.

Adhesive bonding of aluminium

Adhesive bonding of aluminium is widely used in the aerospace industry. High initial bood strengths can be obtained, but bond failure occurs atter prolonged exposure to humid enviroments. The thesis contains details ot a test procedure which has been designed and developed for the assessment of different alloys, pretreatments, and adhesives, which will give adhesively bonded aluminium joints of high strength coupled with long term durability. The test involves assembly of lap shear specimens in a precision jig using 250 ballotini spacers in the adhesive to control the bond line thickness. The test is modified by drilling three accurately located holes through the bonded area after assembly of the joint and curing of the adhesive. Further important features at the test, such as fillet control, are detailed. The test was assessed, modified and developed to give a reliable and reproducible method which would discriminate amongst different bonding systems after exposure to humid test environments. This is the first test to have achieved the discrimination necessary for short term assessment of bond systems where long term durability is required. Even better discrimination has been obtained by applying stress in a stress humidity test. Having established accurate, reliable and discriminating test methods they were used to study the durability of structural epoxy adhesive bonds to aluminium as a function of alloy, pretreatment, adhesive and environment. It was established that the long term durability or adhesively bonded aluminium was directly related to the infulence of water migrating within the adhesive. Pretreatments differed in their ability to prevent hydration of the aluminium oxide by the water absorbed within the adhesive.

The creep properties of a series of zinc-rich zinc-aluminium alloys

The compressive creep behaviour of six sand cast zinc-rich alloys: No3 and No5, corresponding to BS 1004A and BS 1004B, respectively, alloy No2, ILZRO,.16 and two newer alloys ACuZinc5 and ACuZinc10 was investigated. The total creep contraction of the alloys was found to be well correlated using an empirical equation. On the basis of this equation, a parametrical relationship was derived which allowed the total creep contraction to be related to the applied stress, the temperature and the time of test, so that a quantitative assessment of compressive creep of the alloys could be made under different testing conditions. The primary creep and secondary creep rates were found for the alloys at different temperatures and stresses. Generally, the primary creep contraction was found to increase with copper content, whereas secondary creep rates decreased in the order No3, ACuZinc10, ACuZinc5 and No2. ILZRO.16 was tested only at the highest stress and two higher temperatures. The results showed that ILZRO.16 had higher creep resistance than all the other alloys. Thus, based on the above empirical equation, alloy No2 was found to have a substantially better total creep resistance than alloys No3 and No5, and slightly better than ACuZinc5 and ACuZinc10 for strains up to 1%. Both ACuZinc alloys had higher creep strength than commercial alloys No3 and No5. Alloy No5 had much higher creep resistance than alloy No3 under all conditions. The superior creep resistance of alloy No2 was considered to be due to the presence of small precipitates of -phase in the zinc matrix and a regular eutectic morphology. The stress exponents and activation energies for creep under different testing conditions were found to be consistent with some established creep-controlling mechanisms; i.e. dislocation climb for alloy No3, dislocation climb over second phase particles for alloys No5, No2, ACuZinc10, controlled by lattice diffusion in the zinc-rich phase. The lower creep resistance of alloy No3 was mainly due to the lower creep strength of copper-free primary particles having greater volume than eutectic in the microstructure. Alloys No5, ACuZinc5 and ACuZinc10 showed much better creep resistance than alloy No3, based on the precipitation-hardening due to the presence of small -phase precipitates. The primary dendrites in both ACuZinc alloys however were not of much benefit in improving the creep resistance of the alloys.

Melting and solidification of Zinc-Aluminium alloys

Following a scene-setting introduction are detailed reviews of the relevant scientific principles, thermal analysis as a research tool and the development of the zinc-aluminium family of alloys. A recently introduced simultaneous thermal analyser, the STA 1500, its use for differential thermal analysis (DTA) being central to the investigation, is described, together with the sources of support information, chemical analysis, scanning electron microscopy, ingot cooling curves and fluidity spiral castings. The compositions of alloys tested were from the binary zinc-aluminium system, the ternary zinc-aluminium-silicon system at 30%, 50% and 70% aluminium levels, binary and ternary alloys with additions of copper and magnesium to simulate commercial alloys and five widely used commercial alloys. Each alloy was shotted to provide the smaller, 100mg, representative sample required for DTA. The STA 1500 was characterised and calibrated with commercially pure zinc, and an experimental procedure established for the determination of DTA heating curves at 10°C per minute and cooling curves at 2°C per minute. Phase change temperatures were taken from DTA traces, most importantly, liquidus from a cooling curve and solidus from both heating and cooling curves. The accepted zinc-aluminium binary phase diagram was endorsed with the added detail that the eutectic is at 5.2% aluminium rather than 5.0%. The ternary eutectic trough was found to run through the points, 70% Al, 7.1% Si, 545°C; 50% Al, 3.9% Si, 520°C; 30% Al, 1.4% Si, 482°C. The dendrite arm spacing in samples after DTA increased with increasing aluminium content from 130m at 30% to 220m at 70%. The smallest dendrite arm spacing of 60m was in the 30% aluminium 2% silicon alloy. A 1kg ingot of the 10% aluminium binary alloy, insulated with Kaowool, solidified at the same 2°C per minute rate as the DTA samples. A similar sized sand casting was solidified at 3°C per minute and a chill casting at 27°C per minute. During metallographic examination the following features were observed: heavily cored phase which decomposed into ' and '' on cooling; needles of the intermetallic phase FeAl4; copper containing ternary eutectic and copper rich T phase.