HHT-α法（Hilber-Hughes-Taylor）
Theory and Physics
What is the HHT-α Method?
Professor, is the HHT-α method an improved version of the Newmark method?
Yes. It's an improved version by Hilber, Hughes, Taylor (1977) that added a numerical damping parameter $\alpha$ to the Newmark method. It's the default time integration method in Abaqus and Ansys.
Problems with the Newmark Method
The Newmark method ($\beta=1/4, \gamma=1/2$) is second-order accurate and unconditionally stable, but has zero numerical dissipation. High-frequency numerical noise, once introduced, never disappears. High-frequency noise is prone to occur with contact impacts or sudden load changes.
It's problematic that high-frequency noise doesn't disappear.
The HHT-α method solves this problem. It preserves low-frequency accuracy while selectively damping only high frequencies.
HHT-α Method Algorithm
Modified equation of motion:
Parameter relationships:
So it matches the Newmark method when $\alpha = 0$, right?
The range for $\alpha$ is $-1/3 \leq \alpha \leq 0$. $\alpha = 0$: No damping (Newmark method). $\alpha = -0.05$: Gentle high-frequency damping. $\alpha = -1/3$: Maximum high-frequency damping (but accuracy degrades).
Practical recommendation: Around $\alpha = -0.05$. This effectively damps high-frequency noise while maintaining second-order accuracy.
Abaqus
```
*DYNAMIC, ALPHA=-0.05 $ HHT-α α value
0.001, 1.0
```
Abaqus default is equivalent to $\alpha = -0.05$ (APPLICATION=MODERATE DISSIPATION).
Ansys
```
TINTP, , , , , 0.05 $ γ = 1/2 + 0.05 → equivalent α
```
In Ansys, set Newmark parameters with the TINTP command. γ > 1/2 gives numerical damping.
Nastran
```
PARAM, NDAMP, 0.01 $ Numerical damping parameter
```
Summary
Key points:
- Newmark method + numerical damping $\alpha$ — Selective damping of high-frequency noise
- $\alpha = -0.05$ is recommended — Suppresses high frequencies while maintaining 2nd-order accuracy
- Default in Abaqus/Ansys — Often used without realizing it
- Degenerates to Newmark method at $\alpha = 0$ — No numerical damping
- HHT-α is effective when high-frequency noise appears from contact or sudden load changes
So many engineers are "using the HHT-α method without knowing it."
The default for *DYNAMIC in Abaqus is the HHT-α method. Appropriate numerical damping is applied even without changing settings. However, understanding the $\alpha$ value allows you to adjust it when noise appears.
HHT-α: A Numerical Damping Scheme Born in 1977
The HHT-α algorithm, published by Hilber, Hughes, and Taylor in 1977, extends Newmark-β to selectively damp only high-frequency components. The α parameter ranges from −1/3 ≤ α ≤ 0, where α = 0 matches the Newmark method, and α ≈ −0.1 can suppress high-frequency noise while maintaining second-order accuracy and unconditional stability. Abaqus's *DYNAMIC procedure adopts α = −0.05 as default, making it a practical standard for seismic response calculations in building structures.
Physical Meaning of Each Term
- Inertia Term (Mass Term): $\rho \ddot{u}$, i.e., "mass × acceleration". Have you ever experienced being thrown forward when slamming on the brakes? That "feeling of being carried away" is precisely the inertial force. Heavier objects are harder to set in motion and harder to stop once moving. Buildings shake during earthquakes because the ground moves suddenly while the building's mass "gets left behind". In static analysis, this term is set to zero, assuming "forces are applied slowly enough that acceleration is negligible". It absolutely cannot be omitted for impact loads or vibration problems.
- Stiffness Term (Elastic Restoring Force): $Ku$ or $\nabla \cdot \sigma$. When you stretch a spring, you feel a "force trying to return it", right? That's Hooke's law $F=kx$, the essence of the stiffness term. Now a question—an iron rod and a rubber band, which stretches more under the same force? Obviously the rubber. This "resistance to stretching" is the Young's modulus $E$, which determines stiffness. A common misconception: "High stiffness ≠ strong". Stiffness is "resistance to deformation", strength is "resistance to failure"—they are different concepts.
- External Force Term (Load Term): Body force $f_b$ (e.g., gravity) and surface force $f_s$ (pressure, contact force). Think of it this way—the weight of a truck on a bridge is a "force acting on the entire volume" (body force), while the force of the tires pushing on the road is a "force acting only on the surface" (surface force). Wind pressure, water pressure, bolt tightening force... all are external forces. A common pitfall here: Getting the load direction wrong. Intending "tension" but applying "compression"—sounds like a joke, but it actually happens when coordinate systems rotate in 3D space.
- Damping Term: Rayleigh damping $C\dot{u} = (\alpha M + \beta K)\dot{u}$. Try plucking a guitar string. Does the sound continue forever? No, it gradually fades. That's because vibration energy is converted to heat by air resistance and internal friction in the string. Car shock absorbers work on the same principle—they intentionally absorb vibration energy to improve ride comfort. What if damping were zero? Buildings would keep swaying forever after an earthquake. Since that doesn't happen in reality, setting appropriate damping is crucial.
Assumptions and Applicability Limits
- Continuum assumption: Treats material as a continuous medium, ignoring microscopic heterogeneity
- Small deformation assumption (for linear analysis): Deformation is sufficiently small compared to initial dimensions, and stress-strain relationship is linear
- Isotropic material (unless specified otherwise): Material properties are independent of direction (anisotropic materials require separate tensor definitions)
- Quasi-static assumption (for static analysis): Ignores inertial and damping forces, considering only equilibrium between external and internal forces
- Non-applicable cases: Geometric nonlinearity is required for large deformation/large rotation problems. Constitutive law extensions are needed for nonlinear material behavior like plasticity and creep
Dimensional Analysis and Unit Systems
| Variable | SI Unit | Notes / Conversion Memo |
|---|---|---|
| Displacement $u$ | m (meter) | When inputting in mm, unify loads and elastic modulus to MPa/N system |
| Stress $\sigma$ | Pa (Pascal) = N/m² | MPa = 10⁶ Pa. Beware of unit system inconsistency when comparing with yield stress |
| Strain $\varepsilon$ | Dimensionless (m/m) | Note the distinction between engineering strain and logarithmic strain (for large deformation) |
| Elastic modulus $E$ | Pa | Steel: ~210 GPa, Aluminum: ~70 GPa. Note temperature dependence |
| Density $\rho$ | kg/m³ | In mm system: tonne/mm³ (= 10⁻⁹ tonne/mm³ for steel) |
| Force $F$ | N (Newton) | Unify as N in mm system, N in m system |
Numerical Methods and Implementation
Numerical Characteristics of HHT-α Method
Please explain the numerical characteristics of the HHT-α method in detail.
| $\alpha$ | Numerical Damping | Accuracy | Application |
|---|---|---|---|
| 0 | None | 2nd order | Pure Newmark method |
| -0.05 | Gentle | Nearly 2nd order | Recommended (Standard) |
| -0.1 | Moderate | Slightly degraded | Problems with much noise |
| -0.33 | Maximum | Close to 1st order | Special applications only |
So accuracy drops if you make $\alpha$ larger (more negative) than -0.1.
Numerical damping also affects low frequencies. Larger $|\alpha|$ damps low-frequency response more. $\alpha = -0.05$ strikes a good balance of "damping high frequencies while hardly affecting low frequencies".
Relationship with Generalized-α Method
Chung-Hulbert's (1993) Generalized-α method further generalizes the HHT-α method. It allows independent control of low-frequency accuracy and high-frequency damping. Abaqus's APPLICATION=MODERATE DISSIPATION is based on the Generalized-α method.
Summary
Analysis Accuracy Changes with α Selection
Setting HHT-α's α between −0.05 and −0.10 results in a numerical damping ratio ξnum of a few percent to about 10% for the highest mode. Making α too small (e.g., α = −0.3) damps even physical low-order modes, so for structural engineering, it's desirable to have it affect modes with natural periods below about 0.01s. In MSC Nastran SOL 109 direct transient response, HHT-α can be set with the DTI,DIRECTT,ALPHA card.
Linear Elements (1st-order Elements)
Linear interpolation between nodes. Low computational cost but low stress accuracy. Beware of shear locking (mitigated with reduced integration or B-bar method).
Quadratic Elements (with Midside Nodes)
Can represent curved deformation. Stress accuracy improves significantly, but degrees of freedom increase by about 2-3x. Recommendation: When stress evaluation is important.
Full Integration vs Reduced Integration
Full Integration: Risk of over-constraint (locking). Reduced Integration: Risk of hourglass modes (zero-energy modes). Choose appropriately for the situation.
Adaptive Mesh
Automatic refinement based on error indicators (e.g., ZZ estimator). Efficiently improves accuracy in stress concentration areas. Includes h-method (element subdivision) and p-method (order increase).
Newton-Raphson Method
Standard method for nonlinear analysis. Updates tangent stiffness matrix every iteration. Quadratic convergence within convergence radius, but high computational cost.
Modified Newton-Raphson Method
Updates tangent stiffness matrix at initial value or every few iterations. Cost per iteration is low, but convergence speed is linear.
Convergence Criteria
Force residual norm: $||R|| / ||F_{ext}|| < \epsilon$ (typically $\epsilon = 10^{-3}$ to $10^{-6}$). Displacement increment norm: $||\Delta u|| / ||u|| < \epsilon$. Energy norm: $\Delta u \cdot R < \epsilon$
Load Increment Method
Applies total load in small increments rather than all at once. The arc-length method (Riks method) can trace beyond limit points on the load-displacement curve.
Analogy: Direct Method vs Iterative Method
The direct method is like "solving simultaneous equations accurately with pen and paper"—reliable but takes too long for large-scale problems. The iterative method is like "repeatedly guessing to approach the correct answer"—starts with a rough answer but improves accuracy with each iteration. It's the same principle as looking up a word in a dictionary: it's more efficient to open it at an estimated location and adjust forward/backward (iterative) than to search sequentially from the first page (direct).
Relationship Between Mesh Order and Accuracy
1st-order elements are like "approximating a curve with a ruler"—represented by straight line segments, so accuracy is limited. 2nd-order elements are like a "flexible curve"—can represent curved changes, dramatically improving accuracy even at the same mesh density. However, computational cost per element increases, so judge based on total cost-effectiveness.
Practical Guide
HHT-α Method in Practice
In practice, situations where you "consciously use the HHT-α method" are limited. It's fine to rely on the solver's default for most cases.
When to Adjust $\alpha$
| Situation | $\alpha$ Adjustment |
|---|---|
| High-frequency noise present in response | Strengthen to $\alpha = -0.1$ |
| Low-frequency accuracy is critical (flutter, etc.) | Revert to $\alpha = 0$ (Newmark method) |
| Spike noise from contact impact | $\alpha = -0.05 \sim -0.1$ |
| No issues with default | No change needed |
Practical Checklist
Related Topics
なった
詳しく
報告