Video

TechTalk – Rational Approach for Seismic Analysis of Long Underground Tunnels Based on 2.5D Formulation

This lecture presents a rational procedure for the seismic analysis of underground tunnels using recorded free-field earthquakes based on the 2.5D finite/infinite element approach. The near and far fields of the half space are modeled by finite and infinite elements, respectively. Using the 1D wave theory, the nodal force and displacement on the near-field boundary are computed for each spectral frequency of the earthquake. Then, equivalent seismic forces are computed for the near-field boundary for the imposition of earthquake spectrum. By assuming the soil-tunnel system to be uniform along the tunnel axis, the 2.5D approach adopted can duly account for the wave transmission along the tunnel axis, which reduces to the 2D case for infinite transmission velocity. The horizontal and vertical components of the 1999 Chi-Chi Earthquake (TCU068) are adopted as the free-field motions in the numerical analysis. The maximal stresses and distribution patterns of the tunnel section under the P- and SV-waves are thoroughly studied by the 2.5D and 2D approaches, which should prove useful to the design of underground tunnels. Comments on the idea to extend the present approach to include the effect of overlying water, such as the case for the sites below reservoirs, rivers, or sea, will also be pointed out.

TechTalk – The Metaverse Heatwave

It may not be an overstatement that most of us using the internet has heard of metaverse. The term ‘metaverse’ has seen to stir up global hype for business opportunities and fantasy for mankind, if not became the Oxford Word of the Year 2022 – a word reflecting the ethos, mood, or preoccupations, one that has potential of lasting cultural significance. Metaverse describes a virtual reality environment in which users interact with one another’s avatars and their surroundings in an immersive way. We are going to explore what metaverse meant for us, its fantasy and reality, and the development in the current hype. Experience of exploration and creation of the metaverse is shared and lesson learnt, and takeaway is discussed.

TechTalk – Understanding the Turkey-Syria Earthquakes with Methane Gas Refined Fault Theory of Tectonic Earthquakes

At 4:17 am (Turkey time), Feb. 6, 2023, a damaging Mw 7.8 (or 8.0) earthquake struck southern and central Turkey and western Syria and was followed by many aftershocks including an unusually powerful Mw 7.8 (or 7.5) that occurred at 13:24. The earthquakes caused widespread damage including collapsing of many buildings. So far over 11,000 deaths were reported. Figures were projected to rise dramatically by World Health Organization.
In this Teck Talk, Professor Yue will present his understanding of the causes of the earthquakes and the associated building collapses using his methane gas refined fault theory of tectonic earthquakes. Each earthquake involved a rapid release of highly compressed methane gas expansion energy that was previously stored in deep aperture of rock fault zone. The highly compressed gas mass can rapidly expand, rupture, penetrate, and flow from the deep fault zone to shallow ground at a speed of 3 to 1 km/s. The rapid gas flow and expansion in fault rock zone generate massive seismic waves and induce huge concentrated damage to localized grounds and buildings. The earthquake is a cooling process since the gas expansion absorbs heat and cools the surrounding materials in the ground and sky, which can cause local weather changes including the occurrence of air temperature drop-down, rainfall and/or snow.

TechTalk – Adaptable AI-enabled Robots to Create a Vibrant Society – Moonshot R&D Program in Japan –

This talk introduces our Moonshot project which is a project in the National Research and Development (R&D) program in Japan. The Moonshot program promotes high-risk, high-impact R&D aiming to achieve ambitious Moonshot Goals and solve issues facing future society such as super-aging populations. Our project is accepted under the Moonshot Goal 3: Realization of AI robots that autonomously learn, adapt to their environment, evolve in intelligence, and act alongside human beings, by 2050. Our project aims to create adaptable AI-enabled robots available in a variety of places. We are now developing a variety of assistive robots called the Robotic Nimbus which can change their shape and form according to the user’s condition, environment, and the purpose of the task, and provide appropriate assistance to encourage the user to take independent action.

TechTalk – Search and Rescue in Rubble Piles

Disaster response is an important area where robotics has to be applied intensively. Residents are sometimes left in rubble piles in destroyed buildings and soils in many natural disasters like earthquakes and landslides. The search-and-rescue process is slow and inefficient because of high-risk and demanding situations. This talk will introduce the achievement of research and development of serpentine robots led by the speaker. Active Scope Camera (ASC) is a soft serpentine robot that adapts its configuration to the complex shape of debris and moves by ciliary vibration drive. It was used at some disaster sites in the world. The new version of the ASC levitates and moves by adding an air-jet drive. Its vision, auditory and tactile sensing capability supports the teleoperation of its long body. Its performance was tested at first responder’s training sites and actual disasters.

TechTalk – Defect Tolerant Brain-inspired Computing with Memristors

Human brain can perform many tasks much better than classical electronic computers, such as face recognition, reasoning based on vague information, and learning from experience, to name a few. Recently, brain-inspired algorithms have promoted in the rapid development of artificial intelligence, however, they cannot work well in classical computers. In this talk, Dr. Can Li will present his recent works on building brain-inspired computers to fit better with brain-inspired algorithms. Those computers are based on an emerging nanoelectronics device – a memristor – which can store information and compute simultaneously, similar to synapses and neurons in our brain. The built hardware can function similar to human brains, for example, it can tolerate hardware defects, make full use of the nonlinearity of devices, learn from rare samples, and so on.

TechTalk – Non-Fourier Phonon Heat Conduction: Ballistic, Coherent, Localized, Hydrodynamic, and Divergent Modes

Beyond the Fourier diffusion theory on heat conduction, the classical size effects—the Casimir regime—caused by phonon boundary scattering is well known and extensively studied. However, over the last three decades, new regimes beyond the Fourier and the Casimir pictures of heat conduction have been demonstrated. In this talk, I will discuss different phonon heat conduction regimes, including the Knudsen regime, the hydrodynamic regime, the quantization regime, the coherence and localization regimes, and the divergence regime. The Knudsen regime expands Casimir’s picture to many other quasi-ballistic transport geometries, and is being exploited to develop phonon mean free path spectroscopy techniques. Phonon hydrodynamic transport happens when the normal scattering dominates over the resistive scattering, which is a condition difficult to satisfy and only observed at a narrow temperature range less than 20K. However, our recent experiments have observed second sound—a consequence of phonon hydrodynamic transport—at as high at 200K, while simulations point to possibility of observing hydrodynamic heat conduction even at room temperature. Quantized phonon transport was observed at very low temperatures. Signatures of coherent heat conduction, including localization, will be discussed, together with experimental evidences. Divergent thermal conductivity, implying thermal superconductors, is predicted to be possible in low-dimensional materials, although no experiments have provided conclusive evidence. These different phonon heat conduction regimes will be summarized in a regime map, demonstrating the rich phonon transport physics rivaling that of electrons.

TechTalk – (RE)-Ba-Cu-O Single Grain Bulk Superconductors with Improved Superconducting and Mechanical Properties

Extensive research has been carried out over the last three decades, in general, and over last 10 years, in particular, to produce single-grain, high-performance RE-Ba-Cu-O [(RE)BCO bulk superconductors, where RE is a rare earth element or yttrium, for a variety of high field engineering applications. Sample assemblies of bulk (RE)BCO bulk superconductors reinforced under different configurations, remarkably, have enabled trapped fields of more than 17.5 T to be achieved, which is the current world record. More recently, hybrid (RE)BCO bulk superconductors containing Ag, composite and fibre-reinforcements are being developed specifically for both conventional, static devices and more challenging engineering applications where the presence of large electromagnetic stresses has been of concern for the operation of these ceramic-like materials. This seminar will describe the key developments in the processing and properties of high-performance, state-of-the-art (RE)BCO bulk superconductors with a view to develop practical applications over the next 5 years.

TechTalk – A Simple Way of Doing Machine Learning, without Learning

Machine learning and deep neural networks have revolutionized various fields, most obvious examples are computer vision and natural language processing. Apart from the surging sizes of sophisticated models, an emerging trend is to go down the opposite route of deploying lightweight models on the edge (terminal or user end) for relatively simple AI tasks. This is named edge AI which is often constrained to run under restrictive compute and storage resources. In this talk, we will explore the latest theory in neural network modeling that allows the total avoidance of AI training that used to be slow, daunting or even impossible for the edge. Specifically, we will scratch the surface of the neural tangent kernel, and try to establish (well…. qualitatively) the equivalence of data and network, such that once the data are ready, the network is instantly ready, too.

TechTalk – To See a World in a Grain of Sand: A Geotechnical Researcher’s Perspective

Many large earth structures (e.g. slopes, dams, and artificial islands) are made up of sand or sandy soil. The stability of these structures is a major concern of the public as well as the professional. The bitter memories of the deadly slope failures in Hong Kong in 1972 remind us of the importance of proper stability evaluation. The difficulty in predicting the mechanical behavior of sand and sandy soil mainly comes from the granular nature of these materials. A sand or sandy soil is an assembly of numerous small grains of varying size, shape and even mineral composition. It can exist over a spectrum of states that corresponds to a variety of responses, ranging from fluid-like flow to solid-like strain hardening. The groundwater brings additional difficulty and uncertainty. This talk will present some results and findings yielded from our long-term research endeavor at HKU, which is aimed to advance scientific understanding of the complex behaviors of granular earth materials and thereby provide better engineering solutions. Focus will be placed on the fascinating roles played by the small constituent particles. The significance of these findings to engineering practice will be open to discussion.