Spongy Steel: How Ultrasonic Physics Reveals Hidden Hydrogen Damage

Internal Methane Bubble Formation in Hydrogen-Damaged Steel
Methane bubbles formed by High-Temperature Hydrogen Attack inside steel
Methane-filled cavities formed during High-Temperature Hydrogen Attack (HTHA). These microscopic voids strongly reflect ultrasonic waves and can eventually lead to catastrophic material failure.

The Invisible Breach: Unmasking Hydrogen Damage with Ultrasonic Physics

Industrial Physics • Materials Science • Non-Destructive Testing

⚠ The Anacortes Tragedy

On April 2, 2010, a catastrophic failure occurred at the Tesoro Anacortes Refinery in Washington State. A heat exchanger ruptured violently, killing seven workers.

Investigators later identified the likely culprit as High-Temperature Hydrogen Attack (HTHA), a dangerous form of material degradation that can remain hidden deep within steel components for years before suddenly causing catastrophic failure.

The most dangerous defects are often the ones that remain invisible until the moment they become fatal.

⚛ What is High-Temperature Hydrogen Attack?

At elevated temperatures and hydrogen pressures, atomic hydrogen can diffuse into carbon steel. Once inside the metal, hydrogen reacts with cementite (iron carbide), an important strengthening constituent of steel.

$$Fe_3C + 4H \rightleftharpoons 3Fe + CH_4$$

The reaction generates methane gas \((CH_4)\) inside the material. Unlike atomic hydrogen, methane molecules are too large to diffuse through the crystal lattice.

Consequently, methane accumulates at grain boundaries, microscopic interfaces separating individual crystals within the steel.

๐Ÿ”ฌ Physics Insight

Atomic hydrogen is extraordinarily small and can diffuse through steel relatively easily. Methane molecules, however, are much larger.

Once methane forms, it becomes trapped. The trapped gas builds pressure inside the material, creating microscopic cavities and fissures that gradually weaken the steel structure.

๐Ÿ” From Microscopic Voids to Macroscopic Failure

As methane accumulates, internal pressure rises. Tiny voids begin forming along grain boundaries.

Over time these voids grow, connect, and form networks of microcracks. What initially appears to be healthy, load-bearing steel gradually transforms into a brittle, weakened structure.

The danger lies in the fact that the damage develops internally. External inspection may reveal little or no evidence until failure becomes imminent.

๐Ÿšจ Why HTHA is Dangerous

Traditional visual inspections are often incapable of detecting hydrogen attack because the damage begins deep inside the material.

A component may appear perfectly sound while extensive internal cracking is already developing beneath the surface.

๐Ÿ“ก Ultrasonic Physics to the Rescue

Engineers use ultrasonic testing to detect these hidden defects. Ultrasonic waves propagate through steel and interact with internal features such as cracks, voids, and methane-filled cavities.

The effectiveness of ultrasonic testing depends upon differences in acoustic impedance.

Acoustic impedance is defined as

$$Z = \rho c$$

where:

  • \(\rho\) = density of the medium
  • \(c\) = speed of sound in the medium

๐Ÿ“ˆ Why Gas Voids Reflect Sound So Strongly

When an ultrasonic wave encounters a boundary between two materials with different acoustic impedances, part of the wave is reflected.

The intensity reflection coefficient is

$$R_I= \left( \frac{Z_2-Z_1} {Z_2+Z_1} \right)^2$$

In hydrogen attack, one medium is steel and the other is methane gas.

Because the acoustic impedance of methane is extremely small compared to steel,

$$Z_{gas} \ll Z_{steel}$$

the reflection coefficient approaches unity:

$$R_I \approx 1$$

This means methane-filled cavities behave almost like perfect acoustic mirrors, reflecting most of the incident ultrasonic energy.

๐Ÿ’ก Physical Interpretation

Imagine shouting at a concrete wall and hearing a strong echo. A methane-filled void inside steel produces a similar effect for ultrasonic waves.

The enormous impedance mismatch creates powerful reflections that reveal otherwise invisible internal damage.

๐Ÿงช Must Experiment: Interactive Ultrasonic Inspection Lab

Theory becomes much clearer when you can see ultrasonic waves interact with defects in real time.

Try the Interactive Ultrasonic NDT Laboratory developed for Passion of Physics. Create artificial defects, switch between Pulse-Echo and Phased Array modes, and observe how internal discontinuities generate reflections, A-Scans, and B-Scans.

The same physical principles are used to detect hydrogen attack, methane-filled cavities, and grain-boundary cracking in industrial pressure vessels and refinery equipment.

๐Ÿ”ฌ Launch Ultrasonic NDT Simulation

๐Ÿ“Š The Characteristic "Grass" Signal

Experienced inspectors often identify hydrogen attack through the appearance of ultrasonic backscatter known as grass.

Instead of receiving a clean signal from the back wall of a component, the instrument displays numerous small echoes generated by methane-filled voids distributed throughout the material.

The denser the damage, the stronger and more widespread this backscatter becomes.

✅ Key Takeaways

  • Hydrogen diffuses into steel at elevated temperatures.
  • Hydrogen reacts with cementite to produce methane gas.
  • Methane becomes trapped inside the material.
  • Internal pressure creates voids and grain-boundary cracking.
  • Gas-filled cavities strongly reflect ultrasonic waves.
  • Ultrasonic backscatter provides an early warning of HTHA.
  • Physics allows engineers to detect invisible damage before catastrophic failure occurs.

๐ŸŽฏ Challenge

Open the Ultrasonic NDT Lab and place several defects of different sizes. Observe how the A-Scan changes as defect density increases. Can you identify the point where the signal begins to resemble the "grass" pattern associated with hydrogen damage?

๐ŸŽฅ Related Video

Watch an excellent explanation of hydrogen damage and ultrasonic inspection by Steve Mould:

▶ Watch on YouTube

Non Destructive Testing - A Great Tool in Intustrial Physics

ULTRASONIC LAB

Initializing FDTD Engine...

Drag on canvas to place defects • Right-click/Long-press to erase

Simulation Mode

Wave Parameters

Frequency5.0 MHz
Pulse Width2.0
Velocity5900 m/s

Phased Array Setup

Elements16
Steering Angle0°
Focus Depth50 mm

Test Block

Brush Size4 mm
A-SCAN (Amplitude)
B-SCAN (History)
AUTO-GATE TOF
Scanning...

General Theory of Relativity - A Simple Visualization of Space Time Curvature - The The Schwarzschild Metric

Curvature of Space-Time

Spacetime Fabric

According to General Relativity, gravity is not a force, but rather a curvature of spacetime caused by mass and energy.

Massive bodies like stars and planets create "potential wells." The denser the mass, the deeper the metric deformation.

Telemetry Data

Sun Mass (M) 1.50
Earth Mass (m) 0.15
Orbit Radius (r) 25.0 AU
Time Dilation (ฮ”t') 0.982
Status: STABLE ORBIT
Sun Mass 1.5
Earth Mass 0.15
Curvature 30
Orbit Radius 25
Orbital Velocity 1.0
Sun Brightness 1.0

Euler-Lagrange Equation — Derivation & Live Simulation

Euler–Lagrange Lab

An interactive environment exploring analytical mechanics and the Principle of Stationary Action. Discover the fundamental equations that govern the paths of nature.

$$S[q] = \int_{t_1}^{t_2} L(q, \dot{q}, t) \, dt$$

$$\delta S = 0$$

The Derivation Pathway

Step 1: Path Variation +
$$q(t) \to q(t) + \epsilon \eta(t)$$
Assume a true path \(q(t)\) that minimizes action. We introduce a small variation \(\eta(t)\) scaled by parameter \(\epsilon\). The endpoints are fixed.
Step 2: Action Functional +
$$S(\epsilon) = \int L(q + \epsilon\eta, \dot{q} + \epsilon\dot{\eta}) \, dt$$
Substitute the varied path into the action integral. The action \(S\) is now a function of the scalar parameter \(\epsilon\).
Step 3: First Order Expansion +
$$\frac{dS}{d\epsilon} \bigg|_{\epsilon=0} = \int \left( \frac{\partial L}{\partial q}\eta + \frac{\partial L}{\partial \dot{q}}\dot{\eta} \right) dt = 0$$
For stationary action, the derivative with respect to \(\epsilon\) evaluated at \(\epsilon=0\) must vanish. We use the chain rule to expand the Lagrangian.
Step 4: Integration by Parts +
$$\int \frac{\partial L}{\partial \dot{q}} \dot{\eta} \, dt = \left[ \frac{\partial L}{\partial \dot{q}} \eta \right] - \int \frac{d}{dt}\left(\frac{\partial L}{\partial \dot{q}}\right) \eta \, dt$$
Apply integration by parts to the second term to factor out \(\eta(t)\). This shifts the time derivative onto the momentum term.
Step 5: Euler-Lagrange Equation +
$$\frac{\partial L}{\partial q} - \frac{d}{dt}\left(\frac{\partial L}{\partial \dot{q}}\right) = 0$$
Since \(\eta(t)\) is zero at the boundaries, the surface term vanishes. Because the remaining integral must be zero for any arbitrary variation \(\eta(t)\), the term in brackets must be identically zero.

Action Minimization (Gravity Field)

ACTION PHASE SPACE (S vs ฮต)
VARIATION ฮต →
← VARIATION -ฮต
TRUE PATH (CYAN) / VARIED PATH (VIOLET)
ACTION S = 0.000

Observe a particle in a uniform gravitational field. Modify the variation amplitude \(\epsilon\) to see how arbitrary wavy paths increase the total Action compared to the natural parabolic path.

Classical Systems

Lagrangian: $$L = \frac{1}{2} m \dot{x}^2$$

E-L Eq: $$\frac{d}{dt}(m \dot{x}) - 0 = 0$$

EOM: $$\ddot{x} = 0 \implies v = \text{const}$$

Lagrangian: $$L = \frac{1}{2} m \dot{x}^2 - \frac{1}{2} k x^2$$

E-L Eq: $$\frac{d}{dt}(m \dot{x}) - (-kx) = 0$$

EOM: $$\ddot{x} = -\frac{k}{m}x$$

Lagrangian: $$L = \frac{1}{2} m l^2 \dot{\theta}^2 + m g l \cos(\theta)$$

E-L Eq: $$\frac{d}{dt}(m l^2 \dot{\theta}) - (-m g l \sin(\theta)) = 0$$

EOM: $$\ddot{\theta} = -\frac{g}{l}\sin(\theta)$$

Numerical Workspace

Runge-Kutta 4 (JS)
Euler Method (Python)
// RK4 Integrator for 2nd Order ODE: x'' = f(x, v, t)
function rk4_step(x, v, t, dt, force_func) {
    const k1_v = force_func(x, v, t);
    const k1_x = v;

    const k2_v = force_func(x + 0.5*dt*k1_x, v + 0.5*dt*k1_v, t + 0.5*dt);
    const k2_x = v + 0.5*dt*k1_v;

    const k3_v = force_func(x + 0.5*dt*k2_x, v + 0.5*dt*k2_v, t + 0.5*dt);
    const k3_x = v + 0.5*dt*k2_v;

    const k4_v = force_func(x + dt*k3_x, v + dt*k3_v, t + dt);
    const k4_x = v + dt*k3_v;

    const new_x = x + (dt / 6.0) * (k1_x + 2*k2_x + 2*k3_x + k4_x);
    const new_v = v + (dt / 6.0) * (k1_v + 2*k2_v + 2*k3_v + k4_v);
    
    return { x: new_x, v: new_v };
}
# Basic Euler Integration for Harmonic Oscillator
def simulate_ho(k, m, x0, v0, dt, steps):
    x = x0
    v = v0
    
    history = [(x, v)]
    
    for _ in range(steps):
        # EOM: a = - (k/m) * x
        a = -(k / m) * x
        
        # Update state
        v_new = v + a * dt
        x_new = x + v * dt
        
        x = x_new
        v = v_new
        history.append((x, v))
        
    return history

Phase Space Geometry

The state of a system is fully described by coordinates \(q\) and momenta \(p\). For a conservative system like the Harmonic Oscillator, energy conservation manifests as closed orbits in phase space.

INITIATIVE FOR PHYSICS VISUALIZATION NOTEBOOKS // V2.0 // KAZ
NOTEBOOK 01: CLASSICAL MECHANICS - EULER - LAGRANGE EQNS FROM VARIATIONAL PRINCIPLE
Subscribe to: Posts (Atom)