The solidification microstructure has a pronounced influence on the mechanical properties of metallic alloys. It can be either columnar, equiaxed or mixed, and the factors influencing its formation and evolution need to be studied. In-situ synchrotron X-ray radioscopy of the solidification process was performed in the past years, for example by groups in Norway and France, which studied the solidification behavior of different alloys [4,5,6,7,8]. They were able to observe directly dendrite fragmentation, porosity formation, directional and equiaxed dendritical solidification, phase separation and even fluctuations in elemental concentration at the solidification front in Al-based alloys. X-ray transparent Bridgman and isothermal furnaces were used.

Until now, ex-situ and post-mortem analyses could not reveal the kinetics of diffusion. Effects such as convection-driven fluid flow could now be observed. An in-situ X-ray technique for tracking concentration profiles by exploiting Beer-Lambert’s law was presented and chemical diffusion measurements in Al-Ni-Ce melts are compared with self-diffusion and thermodynamic data in order to estimate the influence of thermodynamic forces on diffusion [9]. Further X-ray-supported diffusion experiments will be carried out under microgravity conditions in the frame of the XRMON (X-ray monitoring) project conducted with the support of European Space Agency (ESA) and the German Space Agency (DLR).

The flow of the semi-solid aluminum-germanium alloy Al-Ge32 (wt.%) through a constricted opening into a small cavity was monitored by in-situ radioscopy using white hard X-ray synchrotron radiation [10,11]. The acquisition rate was 67 fps. Liquid drainage and de-/re-agglomeration of small aluminum-rich solid particle clusters—associated with the slurry’s thixotropic behavior—could be visualized. Such studies can help in improving the production of both high-quality and light-weight metallic components using semi-solid casting techniques. The information obtained from experimental flow kinetics can be further used as input for simulations, validation of models or optimization of injection molding tools.

The stabilization of metal films—or cell walls—in liquid foams is a key issue in metal foam science but still not fully understood. More stable films imply less coalescence and therefore smaller bubbles with a more uniform size distribution, which is highly desirable. To investigate the nature of rupture, experiments have been performed in which there was a simultaneous demand for both high spatial and high time resolution. It was possible to observe the coalescence of two adjacent bubbles with 40,000 fps and an exposition time of 25 μs for each frame [15]. Recent experiments were conducted at a recording speed of 105,000 fps with a frame interval of 9.5 μs and effective pixel size of 20 μm. Although the contrast of such images is low due to the short exposure time and consequent limited dynamics of the images, it is clearly visible in Figure 2 that the coalescence of two bubbles is completed in ~475 μs and that the rupture of a film lasts for ~380 μs, if we consider the end of the rupture as the point, where the contour of the new bubble becomes straight before it ends as convex.

In this experiments it was possible to demonstrate that the rupture time of a film is dominated by the inertia of the fluid and not by its viscosity as rupture occurred so fast [14]. Therefore, it could be demonstrated that stabilization by an effective viscosity only, which would be as high as η = 0.4 Pa·s (as calculated by Gergely et al. [19]) does not apply to metal foams of the type investigated here. Other factors such as bubble size and alloy composition may influence the rupture time and will be studied in future.

One of the advantages of using X-ray radioscopy is the possibility to observe qualitatively various effects of the foaming process in-situ even inside a foaming mold. This can provide a better process control. For example, one can observe the growth of nucleated bubbles, defect formations and evolution and the evolution of cracks or big pores [18]. Later, the influence of thermocouples acting as heat sinks on the heat transfer from the mold to the precursors, of the arrangement of the precursors, the mold filling process, density variations during process development, etc. were studied [25].

Figure 3 shows the inhomogeneous filling of a cylindrical mold at 700 °C. This was caused by a temperature gradient of ~100 K over 40 mm in the mold. The foam was generated from a rectangular bar of AlSi9 precursor placed in the center of the mold. As expected, the foam started foaming at the hottest part (left side) and continued expanding up to the right. This led to a broader, aged porous structure on the left and partially expanded foam on the right side. The resulting gradient in density and cell size can be directly observed in the X-ray images.

Radioscopy has provided a deeper insight into foam evolution that allowed for the discovery of new effects. One of those was an unexpected behavior during solidification. In contrast to mechanical or laser expandometers, X-ray radioscopy, due to its high sensitivity, allowed us to find the new effect named “solidification expansion” [26]. It consists of an additional temporary expansion of the metal foam during solidification, where one would actually expect shrinkage as both the gas in the bubble and the metallic matrix shrink in the course of temperature reduction and solidification. The explanation of this counterintuitive phenomenon was found as a distortion of the balance between gas production and gas loses in the foam during solidification caused by the change in the gas-diffusion rate through the solidifying outer skin. This effect is usually small, e.g., a few percent of the total volume, but important because it takes place just during foam solidification and can induce, e.g., cell rupture.

Another interesting effect found was an alignment of the first line of bubbles next to the face sheets of aluminum foam sandwiches (AFS). By reducing the sandwich thickness to around three times the average bubble size, we could obtain an AFS with three aligned bubble layers (Figure 4). This alignment reduces the randomness of the pore distribution and makes foam properties more anisotropic.

To study the stability of foams in the liquid state, the coalescence rate calculated from the number of observed cell wall ruptures can be obtained quantitatively from X-ray image sequences. The dedicated software AXIM is able to compare subsequent images and to recognize individual ruptures [20]. The results are their spatial or temporal distributions, which provide information about, e.g., temperature dependence, drainage effects or long term stability [21,23,24].

The influence of foaming atmosphere could be analyzed in-situ with X-ray radioscopy. It was found that in a non-oxidizing atmosphere the diffusion of hydrogen from the foam into the surrounding atmosphere is much faster, demonstrating that the collapse of the foam is mainly due to gas losses and not due to cell wall ruptures [27]. Figure 5 compares X-ray radiographies of two different Al samples at their maximum expansion and after cooling down to 400 °C. They were foamed under air and argon atmosphere. The expansion plot (Figure 5e) shows that the foam produced in air is much more stable than the one produced under argon over the holding time of 600 s, demonstrating the hydrogen retention capability of the oxide layer.

Not only the cover gas plays a role, but also the gas pressure. The effect of the surrounding pressure on foam nucleation and expansion was investigated [28,29,30]. An improved gas nucleation under high pressure was found. If the pressure is released in the liquid state, it leads to an increased expansion [30]. Variations of pressure showed a reversible behavior of liquid metal foams [29]. Radioscopic observations of the foaming of blowing agent-free powder compacts allowed us to demonstrate that the gas contained in the metal powders used for Al-based foams is sufficient for foaming if there are no gas losses during heating. This can be achieved by Pressure Induced Foaming (PIF), where powder compacts without any additional blowing agent are molten under high gas pressure (e.g., 5 bar) and foamed by subsequent pressure release to 1 bar [31].

Foaming of pure magnesium in argon as protective atmosphere was monitored in-situ with X-rays. This enabled us to find the conditions under which pure magnesium foams can be produced in a good quality following the standard PM-route. Mg-based foams are thought to have a potential for application in the field of biocompatible and biodegradable materials [32].

Foaming experiments under microgravity using inline X-ray diagnostics were performed for the first time in 2007 in the 46th parabolic flight campaign of the European Space Agency (ESA). The X-ray radiographic images showed that fast liquid flow through a cellular structure is possible during strong gravity transitions [24]. Imbibition of drained liquid into a foam occurs under microgravity, leading to a homogeneously distributed liquid fraction. This demonstrates that capillary forces can compensate the effect of gravitational drainage. An even more challenging experiment was conducted during the MASER 11 sounding rocket campaign, where a liquid metal foam was analyzed during several minutes under microgravity (Figure 6). A comparison of the coalescence rate of the metal foamed under gravity and microgravity conditions showed that the total coalescence rate is almost the same in both conditions, although there is no drainage effect under microgravity [33]. This proved that gravity-driven drainage is not the main factor inducing cell wall instabilities, in contrast to what is known from aqueous foams.

In 2009, in the frame of the 51st parabolic flight campaign of ESA, it could be confirmed that the blowing agent is the main destabilizing factor, thus shifting the emphasis of research to the analysis of the blowing agent kinetics and eventually to the further development of blowing agent free methods [31,34]. Figure 7a shows an X-ray radiography of a liquid AlMg15Cu10 foam without extra blowing agent obtained in-situ during a microgravity experiment. In this alloy the intrinsic gas sources of the metal powders provide adequate amount of gas for foaming. The homogeneous gas nucleation and the reduced amount of cell wall ruptures lead to a very homogeneous pore structure, as can also be directly observed in the tomography (Figure 6b) or in a photograph of the polished surface structure (Figure 6c).