Photothermal & Photomechanical Ablation
Two coupled mechanisms — instantaneous thermal expansion and an acoustic shockwave — eject contaminants from the substrate without dumping bulk heat into the underlying metal.
What's actually happening
A short, high-peak-power pulse from the fiber laser deposits energy into the contaminant layer faster than the layer can conduct it away. The local temperature climbs by hundreds of kelvins in tens of nanoseconds — far faster than the thermal-diffusion timescale of the underlying nickel-based superalloy.
Two things happen at once. First, the contaminant expands — explosively. Second, a pressure wave (an acoustic shock) propagates through the contaminant–substrate interface. The contaminant fragments and lifts off; the substrate barely warms because the bulk thermal-diffusion timescale is orders of magnitude longer than the pulse.
This is why laser cleaning can remove a 50-µm CMAS deposit from a turbine blade without measurably altering the blade's microstructure. The energy never has time to thermalize into the substrate.
The two coupled mechanisms
"Photothermal" is the heating term: the contaminant absorbs photons, electrons heat the lattice, the lattice expands. "Photomechanical" is the resulting impulsive stress: rapid expansion launches a stress wave that travels at the speed of sound in the contaminant and reflects off the substrate interface.
When the reflected tensile wave exceeds the contaminant's adhesion strength to the substrate, the contaminant spalls. For brittle CMAS deposits this happens at remarkably modest fluences because the deposit is already stressed by service-induced microcracks.
Key constraints
- Pulse must be shorter than thermal diffusion time across the contaminant layer (typically <200 ns for ceramic-on-metal systems)
- Peak power density of 10⁶–10⁹ W/cm² is the working window for turbine-blade contaminants
- Spallation threshold scales with √(adhesion strength) — heavily-stressed deposits clean at lower fluence
Working parameters for turbofan use
These are the operating envelopes our automation runs within for typical hot-section components. Exact values are tuned per part and per contaminant via the closed-loop reflectivity feedback in Step 2.
| Parameter | Value | Note |
|---|---|---|
| Pulse duration | 20–100 ns | Short enough to confine heat, long enough to build mechanical impulse |
| Pulse energy | 0.5–5 mJ | |
| Repetition rate | 20–500 kHz | |
| Peak power density | 10⁶–10⁸ W/cm² | Below substrate ablation threshold for Inconel 718 / CMSX-4 |
| Spot diameter | 30–120 µm | |
| Scan overlap | 40–70% | Higher overlap on cracked CMAS, lower on smooth oxide layers |
Why this matters for turbofan cleaning
The hot section of a turbofan accumulates two contaminant classes that are notoriously hard to remove without damage: CMAS (calcium-magnesium-aluminosilicate glass from ingested dust) and high-temperature oxides. Conventional methods — abrasive blasting, chemical stripping, ultrasonic baths — either erode the thermal barrier coating, leave residues, or change blade geometry enough to detune aerodynamics.
Photothermal/photomechanical ablation is the only mechanism known to remove these deposits without measurably altering the substrate. Because the energy never thermalizes into the blade, even the TBC's strain-tolerant columnar structure is preserved.
Common pitfalls
- → Below threshold: the deposit absorbs heat but doesn't spall — it sinters and becomes harder to remove on the next pass
- → Above substrate ablation threshold: surface melts. Recast layer forms. Geometry changes. Bad.
- → Pulse too long (>500 ns): heat thermalizes into the substrate, defeating the entire mechanism
Further reading
- ScienceDirect — Pulsed laser ablation of CMAS deposits on TBCsQuantifies the threshold-fluence window for clean spallation vs. recast
- SSRN — Photomechanical removal of high-temperature oxides from Ni-based superalloysDemonstrates substrate-preserving cleaning across CMSX-4 and Inconel 718
- Journal of Laser Applications — Acoustic-impulse modeling of laser cleaning