Experiments with liquid metals could not only lead to exciting insights into geophysical and astrophysical flow phenomena, such as atmospheric disturbances at the rim of the sun or the flow in the Earth’s outer core, but also foster industrial applications, for example, the casting of liquid steel.

However, as liquid metals are non-transparent, suitable measurement techniques to visualize the flow in the entire volume are still lacking. A team of the Helmholtz-Zentrum Dresden-Rossendorf (HZDR) has now, for the first time, obtained a detailed three-dimensional image of a turbulent temperature-driven liquid metal flow using a self-developed method. In the Journal of Fluid Mechanics, they report on the challenges they had to overcome on the way.

Ever since researchers have been investigating the properties of turbulent flows in fluids, they have used an experiment that initially seems quite simple: the fluid is filled into a container/vessel whose base plate is heated and whose lid is cooled at the same time. A team of the Institute of Fluid Dynamics at HZDR is investigating the very details of this process.

The facility will extract the equivalent carbon of 40 million trees annually from Scotland.

The answer may have been hiding in our own bodies all along.

Active shielding was first proposed in the ’60s. We’re finally close to making it work.

Electromagnetic waves in the terahertz frequency range offer many advantages for communications and advanced applications in scanning and imaging, but realizing their potential poses challenges. Researchers at Tohoku University have addressed one of the key challenges by developing a new type of tunable filter for signals in the terahertz wave band. They published their work in the journal Optics Letters.

The company, which filed for bankruptcy in September, had been negotiating a deal for its founders to provide fresh capital and buy it out of Chapter 11.

Pixart-σ weak-to-strong training of diffusion transformer for 4K text-to-image generation.

PixArt-Σ

Weak-to-strong training of diffusion transformer for 4K text-to-image generation.

In this paper, we introduce PixArt-Sigma, a Diffusion Transformer model~(DiT) capable of directly generating images at 4K resolution.

---

Join the discussion on this paper page.

Towards Real-World Vision-Language Understanding.

We present DeepSeek-VL, an open-source Vision-Language (VL) Model designed for real-world vision and language understanding applications.

---

Join the discussion on this paper page.