Deep Hole Drilling of Ceramic Matrix Composites: Ultrasonic, Laser and Abrasive Methods

A manufacturer of SiC/SiC turbine vane components (requiring Ø3 mm × 30 mm cooling holes, L/D = 10:1, 50 holes per vane) was using conventional diamond core drilling (5,000 rpm, 0.5 mm/min feed) producing exit delamination > 1.5 mm, fibre pull-out, and tool life of only 8 holes per drill. Switching to high-frequency ultrasonic vibration-assisted drilling (HFUAD) — brazed diamond core drill, 35 kHz, 8 µm amplitude, 8,000 rpm, 1.0 mm/min feed — achieved: drilling forces reduced by 72%, exit delamination < 0.3 mm, surface roughness Ra 1.2–1.8 µm, and tool life of 120 holes per drill. Annual tooling cost was reduced by 55% and hole quality acceptance rate increased from 65% to 97%.

CMC Types and Drilling Characteristics ​

Ceramic Matrix Composite Types and Key Properties ​

Process Comparison for Deep Hole Drilling of CMCs ​

HFUAD Parameters for CMC Deep Hole Drilling ​

FAQ ​

What makes ceramic matrix composites fundamentally different from other materials for deep hole drilling? ​

Ceramic matrix composites (CMCs) are fundamentally different from metals, polymers, or even conventional composites (CFRP) for deep hole drilling because of their unique combination of high hardness, heterogeneous structure, notch sensitivity, and low fracture toughness. The differences manifest in four critical ways: (1) Tool engagement mechanism — in metal drilling, the cutting edge shears material by plastic deformation, producing a continuous chip. In CMC drilling, the material does not undergo plastic deformation — it fractures brittlely ahead of the cutting edge. The diamond abrasive grains on a core drill or the laser pulse must individually fracture the ceramic fibres and matrix, producing fine, dust-like debris rather than chips. The mechanism is closer to grinding or erosion than to conventional cutting. This means that conventional gun drills or twist drills — regardless of coating or geometry — cannot cut CMCs. The material simply fractures or delaminates rather than shearing. (2) Anisotropic damage propagation — CMCs have a heterogeneous fibre-matrix structure with an interphase coating (typically boron nitride or pyrolytic carbon) that provides toughness by deflecting cracks along the fibre-matrix interface. This crack deflection mechanism, while beneficial for structural performance, is highly problematic for drilling. As the drill advances, cracks propagate preferentially along the fibre-matrix interface rather than perpendicular to it, causing interlaminar delamination that extends 2–10× the hole diameter beyond the hole edge. The delamination is worst at the drill exit, where the remaining material thickness cannot resist the axial thrust force, causing the material to spall or pop out. (3) Extreme tool wear — CMC fibres have hardness of 20–35 GPa (SiC fibres), comparable to many tool materials. Diamond (hardness 70–100 GPa) is the only viable abrasive for mechanical drilling of CMCs. However, even diamond wears rapidly in CMC drilling because the hard SiC matrix abrades the diamond grits and because the cyclic loading from the ultrasonic vibration causes micro-chipping of diamond particles. Tool wear is measured in micrometres per hole rather than per hundred holes. (4) Scale of the deep hole problem — CMC components in aerospace applications typically require many small-diameter cooling holes (Ø0.5–5 mm) over complex curved surfaces. A single turbine vane may have 50–200 holes at angles of 15–30° to the surface — requiring the drill to enter at a compound angle through a curved, sometimes coated surface. The shallow entry angle causes asymmetric loading on the drill, increasing the tendency for delamination and drill wander. The practical consequence is that CMC deep hole drilling requires entirely different processes than conventional deep hole drilling: ultrasonic vibration-assisted drilling, laser drilling, or abrasive waterjet cutting — each with its own parameter optimisation requirements.

How does ultrasonic vibration improve deep hole drilling of CMCs? ​

Ultrasonic vibration improves deep hole drilling of CMCs through several distinct mechanisms that collectively enable hole quality and tool life unattainable with conventional rotary drilling. The ultrasonic vibration (typically 20–40 kHz, 5–15 µm amplitude) is superimposed on the diamond core drill's rotation and feed motion, causing the abrasive grits on the tool face to impact the workpiece at ultrasonic frequency — effectively creating a microscopic hammering action. The primary mechanisms are: (1) Reduced average cutting force — the intermittent contact between the abrasive grits and the workpiece (the tool contacts only during the downward stroke of each vibration cycle) reduces the average axial thrust force by 60–80% compared to conventional drilling at the same feed rate. For CMCs, lower thrust force directly translates to reduced delamination at both entry and exit. The maximum instantaneous force in HFUAD may be similar to conventional drilling, but the impulse loading (force applied for a very short duration) allows the brittle CMC to fracture locally without propagating cracks into the surrounding material. (2) Improved coolant access — the ultrasonic vibration creates a pumping action at the tool-workpiece interface, drawing coolant (typically water mist) into the cutting zone during the retraction phase of each vibration cycle. In conventional diamond drilling of CMCs, the tight clearance between the core drill and the bore wall blocks coolant from reaching the cutting face, leading to thermal damage of both the workpiece and the diamond bond. Ultrasonic vibration provides intermittent clearance that allows coolant penetration, reducing the cutting zone temperature by 40–60% compared to conventional drilling. (3) Self-sharpening of diamond grits — the ultrasonic impact causes micro-fracture of worn diamond grains, exposing fresh cutting edges. This self-sharpening effect extends tool life by 3–5× compared to conventional diamond core drilling. The effect is most pronounced with brazed diamond tools (where individual diamond crystals are chemically bonded to the tool shank) rather than electroplated or sintered tools. (4) Reduced fibre pull-out — the high-frequency impact shears fibres cleanly at the hole edge rather than tearing them from the matrix. Fibre pull-out depth in HFUAD-drilled holes is typically 20–50 µm, compared to 200–1,000 µm in conventional drilling. The clean edge finish is critical for CMC components because exposed fibre ends can oxidise at high temperature, initiating crack propagation that reduces component life. (5) Lower exit delamination — the intermittent tool-workpiece contact means that when the drill breaks through the exit surface, the material is loaded impulsively rather than continuously. The impulse loading causes local fracture at the breakthrough point without propagating the crack radially. Exit delamination in HFUAD is typically 0.2–0.5 mm compared to 1.0–5.0 mm in conventional drilling. The key process parameters for HFUAD are: frequency (higher frequencies > 30 kHz produce better surface finish but require more power); amplitude (higher amplitudes > 10 µm reduce cutting forces but increase tool wear and noise); and the ratio of rotational speed to vibration frequency, which determines the overlap of abrasive grit paths per revolution and thus the surface finish.

What laser drilling methods are suitable for deep hole drilling of CMCs? ​

Three laser drilling methods are suitable for deep hole drilling of CMCs, each with different capabilities and limitations: water-jet-guided laser drilling, millisecond fibre laser percussion drilling, and picosecond/femtosecond laser trepanning. Water-jet-guided laser (WJGL) drilling is the most promising method for deep cooling holes in CMC aerospace components. A pulsed laser beam is focused into a water jet nozzle, where it is guided by total internal reflection within the water jet (typically Ø30–100 µm). The laser pulse ablates the material while the water jet cools the surface, removes debris, and re-solidifies any molten material. The advantages for CMCs are: (1) high aspect ratio — depth-to-diameter ratios of up to 40:1 are achievable in SiC/SiC and C/SiC, compared to 10–15:1 for conventional laser drilling in air; (2) minimal thermal damage — the water jet continuously cools the hole wall, limiting the heat-affected zone (HAZ) to < 10 µm, compared to 50–200 µm for dry laser drilling; (3) no mechanical loading — as a non-contact process, there is no delamination from thrust force. The limitations of WJGL are: limited maximum diameter (typically < 3 mm); reduced depth capability for complex geometries; and high equipment cost ($150,000–500,000). Millisecond fibre laser percussion drilling is an established method for drilling cooling holes in metal turbine components and is being adapted for CMCs. A millisecond-duration laser pulse (0.5–20 ms, up to 20 kW peak power) melts and vaporises the material, with material ejection assisted by a coaxial gas jet. For CMCs, the key challenge is differential ablation rates — the SiC matrix (vaporisation temperature ~2,700°C) and carbon fibres (sublimation temperature ~3,600°C) ablate at different rates, creating uneven hole surfaces. The recast layer in laser-drilled CMCs is typically 20–60 µm thick and contains re-solidified SiC and amorphous carbon, which has different thermal expansion behaviour than the base material. The advantages of millisecond laser drilling are: high speed (1–5 seconds per hole); capability to drill at shallow angles (15–30°); and moderate equipment cost ($50,000–200,000). The limitations are: recast layer formation (must be removed by post-processing for critical applications); limited L/D ratio (< 20:1 for reliable quality); and hole taper (entry is 10–30% larger than exit). Picosecond and femtosecond (ultrashort pulse) laser trepanning uses pulses of 1–100 picoseconds or < 1 picosecond duration to ablate material by non-thermal mechanisms (multiphoton absorption and Coulomb explosion). The ultra-short pulses vaporise material without heat conduction to the surrounding area, producing HAZ < 1 µm — essentially no thermal damage. For CMCs, ultrashort pulse lasers produce the highest quality holes (surface roughness Ra 0.5–1.0 µm, no delamination, no recast layer) but at very low material removal rates (0.01–0.5 mm³/min). A Ø1 mm × 10 mm hole in SiC/SiC requires 5–20 minutes of drilling time — acceptable for prototype or low-volume production but uneconomical for high-volume applications. The practical strategy for CMC production drilling is to use water-jet-guided laser drilling for holes requiring the best combination of depth and quality, millisecond laser percussion for high-speed drilling where moderate HAZ is acceptable, and HFUAD for larger-diameter holes (> 3 mm) that require the best mechanical properties.

What quality control methods are specific to CMC deep-drilled holes? ​

Quality control for CMC deep-drilled holes must address defects that are specific to brittle, fibre-reinforced materials — delamination, fibre pull-out, matrix cracking, and thermal damage — in addition to conventional dimensional and surface finish inspection. The key methods are: (1) Delamination measurement — the most critical quality parameter for CMC holes. Delamination is measured as the maximum radial extent of interlaminar cracking from the hole edge, typically at the exit face where delamination is most severe. Measurement is performed by optical microscopy at 50–100× magnification. The delamination factor (Fd = maximum delamination diameter / hole diameter) should be < 1.3 for aerospace CMC components. For comparison, conventional drilling produces Fd of 1.5–3.0, while HFUAD typically achieves Fd of 1.05–1.2. For WJGL-drilled holes, there is no mechanical delamination, but the HAZ must be assessed instead. (2) Microscopy-based defect assessment — a sectioned hole (cut axially through the hole centre) is examined by scanning electron microscopy (SEM) at 200–2,000× to identify: fibre pull-out depth (acceptable < 100 µm for mechanical drilling, < 20 µm for laser); matrix micro-cracking (no radial cracks > 50 µm from the hole edge); interphase degradation (the BN or PyC interphase coating should remain intact for at least 90% of the hole circumference); and HAZ thickness (for laser-drilled holes, acceptable < 30 µm for WJGL, < 100 µm for millisecond laser). (3) Computed tomography (CT) — for critical components, industrial CT scanning at 5–20 µm voxel resolution provides 3D visualisation of internal defects: delamination planes, matrix cracks that are invisible from the surface, and density variations from thermal damage. CT scanning is typically performed on a sample basis (1 per 50–200 holes for process validation) due to the scan time (10–30 minutes per hole). (4) Surface finish — for mechanical-drilled holes, surface finish is measured by non-contact optical profilometry with a 90° borescope probe. Ra values of 1.0–2.5 µm are typical for HFUAD-drilled CMCs. For laser-drilled holes, the recast layer surface is typically measured by SEM rather than profilometry because the surface may have glassy features that are not well characterised by Ra. (5) Mechanical property verification — for first-article qualification, a sample of CMC panels with drilled holes is subjected to tensile or fatigue testing to verify that the drilling process has not reduced the material's mechanical properties below design allowables. The typical test is tension-tension fatigue (R = 0.1, 10⁵ cycles at 75% of ultimate tensile strength) on a coupon with a centrally located hole. The acceptance criterion is no reduction in fatigue life compared to an undrilled coupon. (6) Process monitoring — for production quality assurance, the most effective method is spindle power monitoring during HFUAD. Spindle power increases by 20–40% as the diamond tool wears, providing a reliable indicator for tool change. For laser drilling, in-process optical emission spectroscopy can detect changes in the plasma composition that indicate the transition from fibre to matrix ablation, enabling real-time parameter adjustment. (7) Dimensional inspection — hole diameter is measured by air gauging (for holes > 2 mm) or by optical measurement (borescope with calibrated optics). The critical dimensional parameter for CMC cooling holes is not absolute diameter but the minimum cross-sectional area, which determines the cooling air flow rate. For film cooling holes, the flow coefficient (C_d = actual flow / theoretical flow) should be measured by flow testing on a sample basis and should be within ±5% of the design value.

When should ultrasonic drilling, laser drilling, or abrasive waterjet be selected for CMC deep holes? ​

The selection between ultrasonic drilling, laser drilling, and abrasive waterjet for CMC deep holes depends on hole geometry, material type, production volume, and quality requirements. Decision criteria: (1) Hole diameter — for bore diameters < 1 mm, laser drilling (water-jet-guided or ultrashort pulse) is the only viable method. Diamond core drills below Ø1 mm are too fragile for CMC drilling, and ultrasonic vibration accelerates their wear rather than improving it. For bore diameters 1–3 mm, either HFUAD or laser drilling is feasible. HFUAD provides better hole quality (no recast layer, no HAZ) but requires mechanical access to both sides of the component (for core drill through-hole). Laser drilling can produce holes at compound angles on complex surfaces but leaves a recast layer that may require post-processing. For bore diameters > 3 mm, HFUAD is the preferred method because it provides the best combination of hole quality, material removal rate, and cost per hole. (2) Depth-to-diameter ratio — for L/D < 10:1, all three methods are feasible. HFUAD is most economical at L/D of 5–20:1. For L/D of 10–40:1, water-jet-guided laser drilling is the best choice because the water jet provides continuous debris removal and cooling along the full hole depth. Abrasive waterjet is limited to L/D < 10:1 because the water jet loses cutting energy with depth. (3) CMC type — for SiC/SiC MI (melt-infiltrated, the hardest CMC), HFUAD is preferred because the hard SiC matrix is efficiently removed by diamond abrasion but poorly removed by laser ablation (the dense SiC matrix conducts heat away rapidly, reducing laser efficiency). For C/SiC, both HFUAD and laser drilling are effective — the carbon fibres are more easily laser-ablated than SiC fibres. For OX-Ox (oxide-oxide, the lowest toughness CMC), HFUAD is preferred because OX-Ox is prone to micro-cracking from thermal shock in laser drilling. (4) Production volume — for high-volume production (> 10,000 holes per year), HFUAD provides the lowest cost per hole ($0.50–2.00 per hole for Ø3 mm × 30 mm in SiC/SiC, including tooling) because the equipment cost is moderate and the per-hole cycle time is 30–120 seconds. Laser drilling has lower per-hole cost at very high volumes (the non-contact tool has no consumable wear) but higher capital cost, making it economical at volumes > 50,000 holes per year. Abrasive waterjet has the highest per-hole cost for deep holes because the slow feed rate (2–10 mm/min for L/D > 5:1) increases cycle time. (5) Hole angle — for holes drilled at shallow angles (< 30° to the surface), laser drilling has an advantage because the laser beam can be oriented at any angle relative to the surface without mechanical interference. HFUAD at shallow angles requires special tooling (angled bushings, custom fixtures) and the asymmetric loading increases tool wear and delamination risk. Abrasive waterjet at shallow angles causes excessive delamination at the hole exit because the water jet penetrates the thin edge of the CMC before it has cut through the full thickness. (6) Quality requirements — if the application requires zero HAZ and no recast layer (e.g., high-cycle fatigue-critical components), HFUAD is required. If the application can accept a recast layer of < 50 µm (e.g., combustor liner cooling holes that will be coated after drilling), laser drilling is acceptable. If delamination must be < 0.2 mm, laser drilling or HFUAD with optimised parameters are both capable; abrasive waterjet typically cannot meet this requirement.

This article provides an overview of deep hole drilling of ceramic matrix composites. Process selection, parameters, and quality control methods depend on the specific CMC type (SiC/SiC, C/SiC, OX-Ox), component geometry, and operating requirements. The technical data presented here reflects published research and documented industrial applications as of 2026.