氷-構造相互作用
Theory and Physics
Overview of the Phenomenon
What kind of problem is ice-structure interaction?
It's the load evaluation when Arctic offshore oil platforms or icebreakers collide with sea ice. Ice exhibits complex failure behaviors such as crushing, buckling, and creep, so simple load models are sometimes insufficient.
Governing Equations
What kind of mechanical models are there for ice?
Ice is often modeled as a visco-elasto-plastic body. In Sinha's (1978) model,
the strain rate is decomposed into three components: elastic strain, delayed elastic strain (grain boundary slip), and viscous creep strain. For high strain rates (during impact), Maxwell-Brittle type models are also used.
The structure side uses standard elasto-plastic FEM. The ice-structure contact problem is handled as an interface using the penalty method or Augmented Lagrangian method. Ice crushing is represented by element erosion or CZM (Cohesive Zone Model).
Are there empirical formulas for ice load?
ISO 19906 (Arctic offshore structures) gives the ice pressure relative to the contact area $A$ as,
where $C_R$ is the Ice Reference Strength and $h$ is the ice thickness. Comparison with this empirical formula is useful for validating numerical simulations.
Ice is "More Complex Than Metal" – The Mechanical Properties of Ice Change with Temperature, Speed, and Salinity
When dealing with ice in structural analysis, the first wall you face is the problem of "what material model should I use for ice?" Ice's strength more than doubles between -2°C and -20°C, it creeps (viscous flow) at low strain rates, and fails in a brittle manner at high strain rates. Furthermore, sea ice is weaker than pure ice because it contains salt, so even the same "ice" can have vastly different properties. In engineering, the boundary where "the failure mode transitions with strain rate" is said to be around 10⁻³/s. Brittle fracture dominates in ship collisions (high speed), while creep dominates in ice pressure on bridge piers (low speed). Representing these two failure modes with a single material model is the theoretical core of ice-structure interaction simulation.
Physical Meaning of Each Term
- Structural-Thermal Coupling Term: Thermal expansion due to temperature change induces structural deformation, and the deformation affects the temperature field. $\sigma = D(\varepsilon - \alpha \Delta T)$. 【Everyday Example】Railroad tracks in summer where the rails expand and the gap narrows – temperature rise → thermal expansion → stress generation is a typical example. Warping of electronic circuit boards after soldering is also due to differences in thermal expansion coefficients between materials. Thermal stress occurs in engine cylinder blocks due to temperature differences between hot and cold parts, potentially leading to cracks.
- Fluid-Structure Interaction (FSI) Term: Fluid pressure and shear forces deform the structure, and structural deformation changes the fluid domain – a bidirectional interaction. 【Everyday Example】Suspension bridge cables vibrating in strong winds (vortex-induced vibration) – wind forces shake the structure, the shaken structure changes the airflow, further amplifying the vibration. Blood flow in the heart and elastic deformation of blood vessel walls, and aircraft wing flutter (aeroelastic instability) are also typical FSI problems. One-way coupling may suffice in some cases, but bidirectional coupling is essential when deformations are large.
- Electromagnetic-Thermal Coupling Term: Joule heating $Q = J^2/\sigma$ causes temperature rise, and temperature change alters electrical resistance, creating a feedback loop. 【Everyday Example】The nichrome wire in an electric stove heats up (Joule heat) and glows red when current flows – as temperature increases, resistance changes, altering current distribution. Eddy current heating in IH cooking heaters and increased sag in power lines due to temperature rise are also examples of this coupling.
- Data Transfer Term: Interpolation resolves mesh mismatches between different physical fields. 【Everyday Example】When calculating "feels-like temperature" in weather forecasting by combining "air temperature data" and "wind data," interpolation is needed if the observation points differ – in CAE coupled analysis, structural and CFD meshes generally don't match, so the accuracy of data transfer (interpolation) at the interface directly affects result reliability.
Assumptions and Applicability Limits
- Weak coupling assumption (one-way coupling): Effective when one physical field affects the other but the reverse is negligible.
- Cases requiring strong coupling: Large deformations in FSI, cases with strong temperature dependence in electromagnetic-thermal coupling.
- Time scale separation: When characteristic times of each physical field differ greatly, efficiency can be improved via subcycling.
- Interface condition consistency: Ensure energy and momentum conservation at the coupling interface is satisfied numerically.
- Non-applicable cases: When three or more physical fields are strongly coupled simultaneously, monolithic methods may be necessary.
Dimensional Analysis and Unit Systems
| Variable | SI Unit | Notes / Conversion Memo |
|---|---|---|
| Thermal expansion coefficient $\alpha$ | 1/K | Steel: ~12×10⁻⁶, Aluminum: ~23×10⁻⁶ |
| Coupled interface force | N/m² (pressure) or N (concentrated force) | Check force balance on fluid and structure sides |
| Data transfer error | Dimensionless (%) | Interpolation accuracy depends on mesh density ratio. Under 5% is a guideline. |
Numerical Methods and Implementation
Discretization Methods
How is ice crushing handled numerically?
There are three main approaches.
| Method | Characteristics | Application |
|---|---|---|
| FEM + Element Erosion | Element deletion upon reaching failure criteria | LS-DYNA, Abaqus/Explicit |
| DEM (Discrete Element Method) | Represents ice as an aggregate of particles | PFC, YADE |
| SPH | Mesh-free. Easy to track crushing. | LS-DYNA SPH |
| Peridynamics | Non-local model. Cracks occur naturally. | Peridigm |
Using DEM for ice sounds interesting.
In DEM, ice is represented as an aggregate of many disks (2D) or spheres (3D), and bonds between particles break when they exceed failure criteria. It has the advantage of naturally reproducing ice crushing patterns (radial crack, circumferential crack).
Contact Algorithm
How is contact between ice and structure handled?
In LS-DYNA, frictional contact is defined using *CONTACT_AUTOMATIC_SURFACE_TO_SURFACE. The friction coefficient of ice strongly depends on temperature, varying in the range $\mu = 0.01$ to $0.3$.
How is contact between ice and structure handled?
In LS-DYNA, frictional contact is defined using *CONTACT_AUTOMATIC_SURFACE_TO_SURFACE. The friction coefficient of ice strongly depends on temperature, varying in the range $\mu = 0.01$ to $0.3$.
The explicit method time step follows the Courant condition determined by the minimum element size and sound speed.
The speed of sound in ice is about 3,000 m/s, so for an element size of 0.01 m, $\Delta t \approx 3 \times 10^{-6}$ s, which is very short.
Calculating "Ice Floes" with DEM (Discrete Element Method) – Large-Scale Icebreaking Impossible with FEM
In actual icebreaker navigation, countless broken ice fragments continue to flow around the hull. Modeling each of these "ice fragment floes" individually with FEM (Finite Element Method) would result in millions to tens of millions of elements, making computation impossible. This is where DEM (Discrete Element Method) is used. In DEM, each ice fragment is represented by a simple "rigid body + spring + dashpot" model, efficiently calculating contact forces between ice fragments and between fragments and the hull. Norway's SINTEF research institute has conducted simulations containing over 1 million ice fragments, matching the resistance of icebreakers with measurements within ±15%. DEM is widely used not only for ice but also for analyzing sand, rock, and granular materials, and its development for ice-structure problems contributes to the overall technological advancement of granular mechanics.
Monolithic Method
Solves all physical fields simultaneously as a single system of equations. Stable for strong coupling but complex to implement and memory-intensive.
Partitioned Method (Separated Iterative Method)
Solves each physical field independently and exchanges data at the interface. Easy to implement and can utilize existing solvers. Suitable for weak coupling.
Interface Data Transfer
Nearest neighbor method (simplest but low accuracy), projection method (conservative), RBF interpolation (robust to mesh mismatch). Balance between conservation and accuracy is important.
Sub-iteration
Performs sufficient iterations within each coupling step to ensure interface condition consistency. Residual criteria are scaled based on typical values for each physical field.
Aitken Relaxation
Automatically adjusts the relaxation coefficient for coupling iterations. An adaptive method that prevents divergence from over-relaxation and accelerates convergence.
Stability Condition
Beware of added mass effect (in fluid-structure coupling when structural density ≈ fluid density). If unstable, apply Robin-type interface conditions or IQN-ILS method.
Analogy for Aitken Relaxation
Aitken relaxation is like "balancing a seesaw." If one side pushes too hard, the other side flies up, and the recoil causes it to push too hard again – Aitken relaxation automatically adjusts the pushing force to suppress this oscillation. It's an adaptive method that automatically adjusts the next correction amount based on the previous correction when coupling iterations oscillate and fail to converge.
Practical Guide
Model Construction Procedure
Please tell me the steps to start an ice-structure interaction simulation.
1. Create a 3D FE model of the structure (for steel structures, use shell elements)
2. Create an ice plate model (solid elements. Add erosion settings if handling failure)
3. Define contact (surface-to-surface contact, set friction coefficient)
4. Set initial velocity / drift velocity of ice
5. Define material models (ice: e.g., Tsai-Wu failure criterion; structure: elasto-plastic)
6. Run with explicit method
Ice Material Parameters
How are the material properties of ice determined?
Sea ice properties strongly depend on temperature, salinity, and strain rate.
| Parameter | First-year ice (-10°C) | Multi-year ice (-10°C) |
|---|---|---|
| Young's modulus | 3–9 GPa | 5–10 GPa |
| Compressive strength | 2–10 MPa | 5–15 MPa |
| Tensile strength | 0.5–2 MPa | 1–3 MPa |
| Poisson's ratio | 0.33 | 0.33 |
| Density | 900 kg/m³ | 910 kg/m³ |
The property variation is large.
Therefore, parametric studies are essential. ISO 19906 specifies using characteristic values with a 50-year return period. Probabilistic evaluation using Monte Carlo simulation is also sometimes performed.
Icebreaker Design – Optimizing the "Ice Breaking Process" with Simulation
Icebreakers don't simply "collide and break" ice; they induce "bending failure" in the ice through the inclination angle of the bow to break it efficiently. If the angle is too shallow, the ice doesn't break and slides under the hull; if it's too steep, it pushes the ice in compression, generating extremely high loads. The optimal angle varies with ice thickness and strength, so in practice, simulations are run under design conditions like "continuously breaking sea ice of 1.5m thickness and 2MPa compressive strength" to optimize the combination of bow shape and propulsion. In the design of Russia's Arktika-class (nuclear-powered icebreakers), simulation results directly led to changes in bow shape, reportedly improving icebreaking capability by 20% over the initial design.
Analogy for Analysis Flow
Have you ever blown up a balloon? At that moment, a sophisticated fluid-structure interaction is actually occurring. Internal air pressure (fluid) pushes and expands the rubber wall (structure) → the expanded wall changes the internal pressure distribution → the changed pressure further deforms the wall... Repeating this catchball at each computational step is FSI analysis.
Common Pitfalls for Beginners
"One-way coupling should be enough, right?" – This misjudgment is the most dangerous in coupled analysis. One-way coupling may indeed suffice if structural deformation is minimal, but in cases like heart valve opening/closing where deformation significantly alters the flow path, one-way coupling is completely inadequate. A rule of thumb is "does the deformation exceed 1% of the characteristic length?" If it does, bidirectional coupling is essential. If you settle for one-way coupling, the result can be "plausible but actually completely wrong" – this is the scariest pattern.
Way of Thinking About Boundary Conditions
Data exchange at the coupling interface is like "border control." Each country (physical field) has its own laws (governing equations), but if the exchange of people and goods (force, temperature, displacement) at the border (interface) is not managed accurately, the economies (energy balance) of both countries collapse. Interpolation when meshes don't match is like a "translator" – the smaller the mistranslation (interpolation error), the better the result.
Related Topics
なった
詳しく
報告