PhD Researcher · University of Groningen
QS 2027 #157 · U.S. News 2026–2027 #77 ↗韩笑东
Open to academic & industrial opportunities
Railway engineering · Track materials and component mechanics · Concrete and smart sleepers · Composite materials and timber strengthening · Vehicle–track dynamics · Vibration measurement and smart sensing · Track irregularity and support-stiffness estimation · Rail corrugation and ballast assessment · Physics-constrained machine learning · Non-destructive testing and structural health monitoring · Predictive maintenance
My work spans railway materials and structural components, vehicle–track dynamics, vibration measurement and inverse condition estimation.
Research
From structural materials and track components to sensing, dynamic models and inverse condition estimation.
UHPC, steel-fibre and recycled-material concrete sleepers, CNT self-sensing and PU-tube self-healing designs, alongside steel/composite sleepers, fastenings and timber strengthening.
Full-scale testing · FEM · FRP strengtheningAxle-box and trackside vibration, camera-based measurement and smart rail sensing, with ongoing work on GPR and vibration for ballast assessment.
Time–frequency analysis · video tracking · sensor fusionVehicle–track models and Kalman filtering for track geometry and stiffness estimation, alongside research on corrugation and physics-constrained learning.
Vehicle–track coupling · system identification · physics constraintsPublished work and current investigations, grouped by subject. Expand a topic to see its scope and related papers.
Vehicle–track models, model updating and dual/extended Kalman filtering are used to estimate track geometry and support condition from vehicle- and track-side responses. The trackside EKF study evaluates its method through numerical examples.
Vehicle–track dynamics · model updating · dual KF / EKF
Vehicle–track simulation and field-response assessment are used to examine how changes in track support relate to wavelength-selective dynamic behaviour and rail corrugation.
Wheel–rail interaction · variable support · wavelength selection
See current research ↓A critical review connects corrugation mechanisms, measurement and diagnosis with grinding and other interventions. It distinguishes immediate surface restoration from evidence that subsequent growth or recurrence is suppressed.
Critical review · monitoring evidence · maintenance decisions
See current research ↓Current work combines neural representations of spatial irregularity with vehicle–track dynamics to reconstruct a continuous track profile from vibration observations. Model development and comparative studies are ongoing; independent validation remains to be completed.
Neural representations · dynamics constraints · continuous-window inversion
See current research ↓Experimental axle-box accelerations are analysed to compare time–frequency response characteristics on ballasted tracks and across a bridge transition zone.
Field measurement · axle-box acceleration · time–frequency analysis
Transformer-based CoTracker and dynamic mode decomposition are combined to investigate structural vibration measurement and dynamic characteristics from video tracking.
Video tracking · CoTracker · dynamic mode decomposition
A published review compares rail-sensing technologies, monitoring targets and application conditions, including the limits of measurement interpretation and practical use.
Rail sensing · optical-fibre sensing · technology review
Ground-penetrating radar features are investigated together with axle-box vibration and auxiliary inspection information to interpret ballast structure and condition.
Ground-penetrating radar · vibration features · multimodal assessment
See current research ↓Experimental and numerical studies examine sleeper mechanical response using ultra-high-performance concrete, non-prestressed steel-fibre concrete and recycled materials.
UHPC · steel-fibre concrete · recycled materials · mechanical testing
Collaborative experimental and numerical work on CNT self-sensing concrete sleepers examines response changes associated with hanging-sleeper defects and support-condition identification.
CNT concrete · self-sensing · hanging-sleeper defects
Self-healing concrete sleepers incorporating different-size polyurethane tubes are investigated through development and performance evaluation.
Self-healing concrete · polyurethane tubes · performance evaluation
Tests and finite-element models examine the mechanical behaviour of steel turnout sleepers and fastening systems for fibre-reinforced foamed-urethane (FFU) sleepers.
Steel turnout sleepers · FFU · fastening systems · FEM
Work on GFRP- and carbon-fabric-strengthened timber sleepers is complemented by earlier collaborative research on the eccentric-compression behaviour of hybrid-strengthened square timber columns.
GFRP / CFRP · timber sleepers · strengthened timber columns
Academic and industrial relevance
Connections between material behaviour, component response, system dynamics and the quantities that can be measured.
Measurement, modelling and diagnostic methods for track inspection, component assessment and maintenance research.
Publications
17 peer-reviewed publications spanning railway condition estimation, vibration measurement and track-component mechanics: 15 journal articles and 2 conference papers. Journal versions replace earlier preprints.
Experimental setups, field measurements, estimation workflows and sensing concepts from work across the research portfolio.
Instrumented three-point bending tests assess how GFRP wrapping changes the flexural response of timber sleepers.
Author manuscript · Fig. 6
An extended Kalman filter estimates track irregularity from fixed rail or sleeper responses. Performance is evaluated numerically.
Author manuscript · Fig. 1
A review schematic relates optical-fibre placement to the rail cross-section, neutral axis and bending response under a passing wheel.
Author manuscript · Fig. 3
A dual-filter architecture couples track-geometry estimation with resonance-based support-stiffness updating.
Video-tracked points on a laboratory beam provide motion information for vibration analysis using CoTracker and dynamic mode decomposition.
Published article · tracking example
An instrumented wheelset-drop setup measures the dynamic response of rails and sleepers in a ballasted track panel.
Published article · test setup
A jack, force sensor and LVDTs record the load–displacement response during lateral-resistance testing of a steel turnout sleeper.
Published article · lateral-resistance test
Electrode installation, concrete casting, curing and the finished sleepers illustrate the fabrication process in this collaborative study.
Siahkouhi et al. · Fig. 4
The schematic locates the axle-box sensor and shows the two experimental cases: ballasted track on an embankment and an embankment–bridge transition zone.
Han et al. · Fig. 9 · CC BY 4.0
Two patents related to railway turnout sleepers.
Current work
Simulation and field-response assessment examine variable support and wavelength selection. A parallel review connects physical mechanisms, monitoring, maintenance interventions and recurrence evidence.
GPR, vibration and auxiliary inspection information are investigated together to interpret ballast structure and condition.
Neural representations and vehicle–track dynamics are combined for continuous spatial reconstruction from vibration observations. Model development and comparative studies are ongoing; independent validation remains to be completed.
The ongoing studies above are separate from the published results. Validation scope is stated for each project, and no claim of routine operational deployment is made.
Experience
University of Groningen
Beijing Jiaotong University
Beijing Jiaotong University
Outstanding Master’s Graduate, Beijing Jiaotong University, 2023
Vehicle–track coupling, model updating, dual/extended Kalman filtering
Axle-box and trackside vibration, GPR, time–frequency analysis, CoTracker and DMD
Sleeper testing, FRP strengthening, self-sensing and self-healing concepts
ABAQUS, ANSYS, Simpack, RecurDyn, EDEM, Rocky DEM
Physics constraints, neural representations, parameter studies and model–measurement comparisons
Academic service
Verified peer-review contributions recorded publicly on ORCID.
Verify on ORCID ↗Contact
Open to collaboration and research-focused opportunities in railway engineering, structural health monitoring and predictive maintenance.