科研 Skill

连续尺度有限元力学建模

finite-element-modeling

Covers continuum-scale mechanical, thermal, and multiphysics simulation with Abaqus/COMSOL conventions, mesh generation. Use for macroscale FEM consuming measured or computed material properties.

发布者 Codar安全审核通过v0.1.0更新于 2026-11-18

SKILL.md

Finite Element Modeling

Overview

Finite element analysis (FEA) solves continuum mechanics, heat transfer, and coupled multiphysics problems over a discretized (meshed) geometry — the macroscale end of this repo's modeling spectrum, consuming material properties (elastic modulus, thermal conductivity, CTE) that mechanical-testing and thermal-analysis measure, or that computational-materials/phase-field-calphad predict, as model inputs rather than computing them from first principles.

When to use this skill

Activate when the request involves:

  • FEA, finite element analysis, Abaqus, COMSOL, ANSYS, mesh generation, boundary conditions, multiphysics simulation
  • Terms: mesh convergence, element type, constitutive model, von Mises stress, thermal-stress coupling
  • "Set up an FEA model of...," "check mesh convergence," "what boundary conditions for..."

Core usage

Mesh convergence checking (the first validity check for any FEA result)

def mesh_convergence_check(mesh_sizes, result_values, tolerance_percent=2):
    """Run the same model at progressively finer mesh sizes; the result should
    plateau. Report as converged only once consecutive refinements change the
    result by less than the tolerance."""
    percent_changes = [
        abs((result_values[i] - result_values[i - 1]) / result_values[i - 1]) * 100
        for i in range(1, len(result_values))
    ]
    converged_at_index = next((i for i, pc in enumerate(percent_changes) if pc < tolerance_percent), None)
    return {
        "converged_mesh_size": mesh_sizes[converged_at_index + 1] if converged_at_index is not None else None,
        "percent_changes": percent_changes,
    }

Material property input from measured/computed data

def elastic_properties_from_testing(elastic_modulus_gpa, poisson_ratio):
    """The FEA model's constitutive behavior is only as good as these inputs —
    typically sourced from mechanical-testing's stress-strain analysis, or
    computational-materials' elastic-constant calculations, not assumed generic
    handbook values unless the specific material/condition is genuinely
    well-represented by a handbook value."""
    lame_lambda = (elastic_modulus_gpa * poisson_ratio) / ((1 + poisson_ratio) * (1 - 2 * poisson_ratio))
    shear_modulus = elastic_modulus_gpa / (2 * (1 + poisson_ratio))
    return {"lame_lambda_gpa": lame_lambda, "shear_modulus_gpa": shear_modulus}

Thermal-stress coupling setup considerations

def thermal_stress_estimate(cte_per_k, delta_T, elastic_modulus_gpa, constrained_fraction=1.0):
    """First-order estimate of thermal stress under full constraint — a sanity
    check before a full coupled simulation, not a substitute for one when the
    actual constraint condition is more complex than fully fixed."""
    thermal_strain = cte_per_k * delta_T
    return thermal_strain * elastic_modulus_gpa * constrained_fraction  # GPa

Validation & Pitfalls

Canonical reference: Zienkiewicz, Taylor & Zhu, The Finite Element Method: Its Basis and Fundamentals (7th ed., 2013), for FEA methodology generally.

  • Mesh convergence must be checked for every new geometry/loading condition, not assumed from a prior similar model — a mesh density adequate for one geometry or stress concentration doesn't automatically transfer to a different one, especially near geometric features (sharp corners, holes) where stress gradients are steep.
  • Boundary condition choice is often the single largest source of discrepancy between simulation and reality, more than mesh resolution or material model refinement — an overly idealized boundary condition (perfectly fixed, perfectly frictionless) can dominate model error; justify boundary conditions against the actual physical constraint being modeled, and check sensitivity to reasonable alternative choices.
  • Material property inputs sourced from a different material state/condition than what's being modeled produce a precise but wrong answer — e.g. using room-temperature elastic modulus for a high-temperature simulation, or bulk properties for a thin-film/small-scale structure where size-dependent effects matter. Match input data conditions to the model's actual conditions.
  • A converged, non-crashing simulation is not the same as a validated one — numerical convergence confirms the solver found a stable solution to the equations as posed, not that those equations (material model, boundary conditions, geometry simplifications) correctly represent reality. Validate against at least one independent data point (analytical solution for a simplified case, or experimental measurement) before trusting predictions for conditions without such a check.
  • Element type choice (linear vs. quadratic, reduced vs. full integration) affects both accuracy and characteristic numerical artifacts (e.g. shear locking in linear elements for bending-dominated problems, hourglassing in reduced-integration elements) — match element type to the problem physics, and watch for these specific artifact signatures rather than assuming a default element type is always appropriate.

示例提示词

我想在Abaqus里给这个零件设置网格划分和边界条件做热力耦合分析。

所属插件

材料表征、性能与材料信息学面向实验与数据驱动的材料研究:XRD/电镜/谱学/热分析/力学测试等表征,合成与加工,电化学与半导体表征,相场、CALPHAD 与连续介质力学,材料数据库、特征工程、贝叶斯优化与主动学习,选材与生命周期评价。

安装该插件会一并获得同领域的 42 个 Skill / MCP。

同插件的其他科研 Skill

常见问题

连续尺度有限元力学建模能做什么?

Covers continuum-scale mechanical, thermal, and multiphysics simulation with Abaqus/COMSOL conventions, mesh generation. Use for macroscale FEM consuming measured or computed material properties。

如何安装连续尺度有限元力学建模?

在科搭 Codar 左侧点击「插件」,在「组件」页签搜索「连续尺度有限元力学建模」并安装;在 Claude Code 或 Codex 中可以发送本页的安装提示词,或下载 Skill 包后解压到 ~/.claude/skills 或 ~/.codex/skills。

连续尺度有限元力学建模需要哪些权限?

该科研 Skill没有声明额外权限;执行具体操作时仍由 Agent 按自身规则请求确认。

连续尺度有限元力学建模经过官方审核吗?

是。SpeedSkills 收录的每个版本都会记录内容摘要与权限声明,当前版本已通过安全审核。