Hydrogen Porosity in Al-Si Foundry Alloys
MetadataVis full innførsel
Aluminium alloys based on the aluminium-silicon system are popular for automotive and aerospace applications, thanks to their high strength to weight ratio, excellent castability, and corrosion resistance. Microporosity is widely acknowledged to affect both static and dynamic properties of structural aluminium alloy castings. Formation of microporosity is a complex phenomenon and depends upon various factors but mainly hydrogen content and melt cleanliness, i.e., oxide films and inclusions. In the past few decades numerous studies on microporosity formation have been reported. However, several aspects of this subject are not fully understood. The motivation of this doctoral thesis has been to improve the knowledge of porosity formation and its affect of the mechanical properties of the cast products. This study aimed at understanding the effect of hydrogen and defects on microporosity of aluminium-silicon based castings. A literature review of the theories of porosity formation, and previously reported results on the factors affecting the microporosity formation in aluminium alloys is reported briefly in order to establish a basis of the present study. The research presented in this thesis is divided in five phases. In the first phase, lab scale directional solidification experiments were carried out with an A 356 alloy. Six hydrogen levels from 0.48 to 0.07 mL/100 g melt were reached by various up-gassing and de-gassing treatments of the melt. The melt quality was assessed by porous disc filtration apparatus (PoDFA) and reduced pressure test (RPT) methods. Microporosity distribution in the castings was characterized by the Archmedian method, image analysis and X-ray radiography. The results show that the porosity distribution is strongly dependent on the cooling rate and hydrogen content in the presence of the oxide films type defects. In the next phase, casting experiments with an A356 melt were carried by using a step mould die and castings were produced with and without filtration. The experiments were divided into series I and series II. In the first series 70 kg of the A356 ingots were melted in an electric resistance furnace. Three hydrogen levels namely 0.1, 0.2, and 0.4 mL/100 g melt were reached first by Ar-degassing, and up-gassing with Ar-10 % H2 and Ar-water vapour mixture, respectively. In the second series of experiments similar procedure was applied to another melt but in the reverse order. The castings were characterized in terms of microporosity and mechanical properties distribution. In addition high pressure die casting of an A380 alloy was carried out with different processing parameters like plunger speed, commutation point between first and second phase and pouring temperature. The results shows that tensile properties in both gravity and high pressure die castings were affected by the amount and distribution of casting defects. A series of casting experiments were carried out with the same type of step mould and with two hydrogen levels, namely 0.1 and 0.2 mL/100 g melt, in the third phase of the project. The main focus of these trials was to carefully control the variables during melt preparation, casting and solidification and keep the variations at minimum. Experiments were performed in two different days in order to check the reproducibility of the measurements. For the low gas level (0.1 mL/100 g melt) higher porosity levels are observed in the thinnest step, while the trend is opposite with the medium gas level (0.2 mL/100 g melt) castings. The data obtained form these trials where casting variables were controlled in a reproducible way form a basis for evaluating the results of simulated microporosity distributions using recent modelling approaches. The reproducibility of the step mould used was calculated to be 10-15 %. In the fourth phase, a study in an automotive casting plant was targeted to address the problem of high rejection rate of a cylinder head casting made from an A354 alloy. In the foundry, the castings under investigation were usually made from the mixture of A 354 alloys from three different suppliers in uncontrolled proportions. In an attempt to solve this problem, a number of systematic experiments were carried out to assess the melt quality with RPT, and pressure filtration (PREFIL) tests. The results from these trials show that the bi-film index data from the RPT has a potential to be used in an industrial environment for routine melt quality control as the cause of high scrap was successfully identified with this technique. In the last phase of this work, state-of-the-art low pressure die casting (LPDC) experiments with an A356 melt were carried out using the same step mould die used in second and third phases. Castings were made with two hydrogen levels, namely 0.1 and 0.2 mL/100 g melt. There is not much reported on microporosity simulation under LPDC conditions and reliable experimental validation is needed. The results from these LPDC trials provided a sound basis to assist and further develop the porosity modelling tools with the microporosity distribution data from carefully controlled LPDC experiments.