High Shear Squeeze Flow Analysis in Microvia Substrate Lamination

Melt

Thermosetting Polymer Rheology under Dynamic Strain Rates

Thermosetting prepreg matrix material undergoes rapid fluidity changes during substrate lamination as thermal energy breaks physical cross-links before chemical gelation locks the polymer network. In high-density interconnect and IC substrate lamination, the prepreg resin or build-up film does not experience simple parallel sliding. Instead, hydraulic force applied normal to the platen surfaces drives resin laterally outward through narrow channels formed by inner layer copper features and core dielectrics.

This physical displacement represents classic squeeze flow between parallel plates, where thin gap geometry converts platen closure speed into extreme fluid velocity gradients. The local strain rate scales inversely with the square or cube of the dielectric gap height depending on edge boundary conditions. When dielectric clearance narrows below 50 µm over inner layer copper planes, local shear rates routinely exceed 1,000 s⁻¹ and can reach 10,000 s⁻¹ near copper feature transitions.

Under these high strain rates, thermosetting resin systems like high-Tg FR-4 epoxies, polyphenylene ether formulations, and Ajinomoto Build-up Films exhibit non-Newtonian shear-thinning behavior. Standard zero-shear viscosity figures published on supplier material datasheets fail to capture resin movement under processing conditions. Zero-shear viscosity measured via low-strain oscillatory rheometry reflects material response during unconstrained thermal expansion rather than forced hydraulic squeeze.

As high strain rates develop inside narrow lamination channels, long polymer chains align parallel to the direction of flow, reducing intermolecular entanglement resistance. Effective viscosity drops sharply below the zero-shear baseline, altering both the lateral flow front velocity and the filling dynamics into laser-drilled microvias.

At a shear rate of 2500 inverse seconds and a temperature of 145 degrees Celsius, high-Tg epoxy prepreg resin exhibits a shear-thinning viscosity drop from 120 Pascal-seconds down to 8.4 Pascal-seconds.

Non-Newtonian Shear Thinning in Microvia Channels

Mathematical modeling of squeeze flow in substrate lamination requires extending the classical Stefan equation to accommodate power-law fluid mechanics. For an incompressible power-law fluid, shear stress depends on strain rate according to the fluid consistency index and the power-law exponent. High shear rates force the effective viscosity down according to the expression:

η_eff = m · γ˙^{n − 1}

where η_eff is the effective dynamic viscosity, m represents the fluid consistency index in Pa·sⁿ, γ˙ is the local shear rate in s⁻¹, and n is the non-Newtonian power-law exponent. For high-density substrate epoxies and build-up resins, the power-law exponent n typically ranges between 0.35 and 0.55 during the melt window prior to gelation. Values of n significantly below unity confirm pronounced pseudoplastic behavior.

Low viscosity aids cavity filling.

Resin pressure drops rapidly. When platen closure forces resin through a 30 µm gap over an inner layer copper pad, local shear rate γ˙ is given by:

γ˙ = (2 + 1/n) · (v_closure / h)

where v_closure is the hydraulic platen closure velocity and h is the instantaneous dielectric gap height. A closure speed of 5 µm/s across a 25 µm gap with n = 0.4 yields local shear rates exceeding 1,100 s⁻¹. The resulting viscosity drop enables resin to enter microvia cavities before thermal cross-linking halts matrix motion, but the localized shear stress creates non-uniform drag across inner layer copper features.

How the polymer chain length distribution controls shear-thinning stability across sequential press cycles remains an open question in material characterization.

Cavity

Recess Penetration and Hydrodynamic Pressure Drops

Blind laser-drilled microvias present narrow cylindrical blind holes into which flowing prepreg resin moves during the vacuum press cycle. Microvia filling operates under a coupled pressure field: the global lateral squeeze pressure generated between the press platens forces resin sideways, while local hydrostatic pressure differentials drive resin downward into the microvia barrel. Microvia aspect ratio, defined as depth divided by target pad opening diameter, governs flow resistance.

Aspect ratios exceeding 0.8:1 generate steep entrance head losses that require higher local squeeze pressures or lower matrix viscosity to achieve complete filling without air entrapment.

Hydrodynamic pressure profiles across an inner layer pattern vary based on copper pattern density and feature clearance. Resin flowing over high-density line-and-space routing encounters localized flow obstructions, producing localized high-pressure zones on the upstream side of copper traces and pressure shadows on the downstream side. When a microvia sits inside a pressure shadow zone, the local squeeze pressure drops below the threshold required to force resin into the via corners.

Liquid resin fills the void. If resin gelation occurs before local pressure equalizes, microvia dimpling or internal voiding develops at the target pad interface.

Inner layer copper surface profile depth alters local boundary layer drag and forces shear rates upward along the pad edges.

Substrate reinforcement style influences squeeze flow dynamics. Woven glass fabrics in standard prepregs introduce macro-scale flow channels and structural boundaries that modify local shear fields during lamination. Table 1 summarizes rheological parameters and squeeze flow thresholds across standard substrate laminate grades and build-up dielectric films.

Influence of Glass Weave Architecture on Resin Inflow

Woven glass reinforcements like 1035 or 1067 glass style act as porous barriers during lamination squeeze flow. The glass filaments constrain lateral resin movement, forcing the resin matrix to shear through micro-cavities between warp and weft yarn bundles. Glass bundles resist lateral displacement.

In prepregs with heavy glass styles such as 2116 or 7628, glass bundle stiffness prevents uniform squeeze flow into microvias, limiting their application to low-density outer layers. Spread-glass fabrics like 1078 or 1035 offer uniform filament distribution, reducing local permeability variations and permitting predictable shear-thinning flow across microvia arrays.

Selecting prepreg materials for sequential microvia lamination demands careful alignment between resin volume fraction and microvia cavity volume. Insufficient resin content results in resin starvation over copper-dense regions, while excessive resin content promotes severe layer misregistration and inner layer copper distortion. Shear thinning accelerates lateral flow.

• Consistency Index Stability – Prepreg lot acceptance requires verification of consistency index m within +/- 8 percent of target baseline to ensure repeatable squeeze flow velocities.
• Power-Law Exponent Window – Substrates specifying power-law exponents between 0.40 and 0.50 maintain shear-thinning capability without risking matrix breakdown under localized pressure spikes.
• Dynamic Gelation Window – Polymer formulation must provide a minimum 150-second working window where dynamic viscosity remains below 15 Pa·s under processing press pressures.
• Glass Bundle Permeability – Spread-glass prepreg configurations eliminate resin flow stagnation zones over microvia capture pads, ensuring consistent cavity fill metrics.

Failing to match prepreg rheology to microvia aspect ratios results in structural voids that trigger catastrophic inner layer delamination during assembly reflow thermal cycles.

Clamp

Hydraulic Force Profiling and Isothermal Windows

Multi-platen hydraulic presses deliver precise force trajectories to balance matrix consolidation against excessive lateral squeeze-out. The lamination cycle consists of three distinct phases: vacuum evacuation, controlled thermal ramp under low pressure, and high-pressure consolidation. During initial heating, prepreg temperature rises toward the melt transition, causing dynamic viscosity to drop rapidly.

Applied squeeze pressure during this low-viscosity window forces resin to flow laterally, filling microvias and enveloping inner layer copper traces. Squeeze flow determines final dielectric thickness.

Can Hydraulic Vacuum Ramp Rates Prevent Microvia Voiding?

Pressure escalation before thermal ramp-up evacuates trapped gas from microvia structures prior to fluid movement initiation. Vacuum levels below 15 mbar must be established before platen temperature crosses 80 °C. If hydraulic pressure is applied before vacuum drawdown completes, ambient gas gets trapped inside blind microvia barrels. trapped gas forms high-pressure pockets that resist resin entry during the minimum viscosity window. Vacuum pressure suppresses bubble formation.

When hydraulic press pressure ramps too quickly while resin viscosity sits at its absolute minimum, the lateral squeeze velocity exceeds the microvia filling rate. Resin bypasses the via opening, trapping micro-voids at the target pad base. Conversely, delayed pressure application allows polymer cross-linking to advance, raising matrix viscosity above the threshold required for complete cavity penetration.

Mathematical Model of Microvia Matrix Inflow

Fluid movement into a blind microvia during lamination follows a modified Hagen-Poiseuille relationship driven by the local hydrodynamic pressure gradient. Consider a laser-drilled microvia with top radius r_t = 50 µm, bottom radius r_b = 35 µm, and depth h_v = 60 µm. The total volume V_v of this truncated conical cavity is calculated as:

V_v = (π · h_v / 3) · (r_t² + r_t · r_b + r_b²) = (π · 60 × 10−6 / 3) · (2500 + 1750 + 1225) × 10−12 = 3.44 × 10−13 m³

Assume an applied lamination press pressure P = 2.4 MPa, resulting in a local hydro-static driving pressure ΔP = 2.0 MPa at the microvia entrance after accounting for lateral squeeze pressure losses. For a non-Newtonian prepreg resin exhibiting power-law parameters m = 45 Pa·sⁿ and n = 0.45 at 145 °C, the local shear rate γ˙ inside the microvia entrance channel reaches 1,600 s⁻¹. The effective dynamic viscosity η_eff is:

η_eff = 45 · (1600)^{0.45 − 1} = 45 · (1600)^−0.55 = 0.782 Pa·s

The volumetric flow rate Q into the microvia cavity under this localized pressure gradient is governed by:

Q = (π · ΔP · r_avg&sup4;) / (8 · η_eff · h_v)

Using average radius r_avg = 42.5 µm (4.25 × 10⁻⁵ m):

Q = (π · 2.0 × 10&sup6; · (4.25 × 10−5)&sup4;) / (8 · 0.782 · 60 × 10−6) = 5.41 × 10−17 / 3.75 × 10−4 = 1.44 × 10−13 m³/s

The calculated fill time t_fill is:

t_fill = V_v / Q = 3.44 × 10−13 / 1.44 × 10−13 = 2.39 seconds

Flow timing governs microvia filling. If lower press pressure or insufficient shear rate leaves the resin viscosity at its zero-shear value of 120 Pa·s, the volumetric flow rate drops to 9.38 × 10⁻16 m³/s, extending the required fill time to 366 seconds. Because the thermal processing window near minimum viscosity lasts only 180 seconds before gelation elevates viscosity exponentially, incomplete fill occurs under low-shear conditions.

Over-pressing starves central microvias.

Material suppliers claim that standard vacuum lamination cycles automatically compensate for resin viscosity variations across batch lots.

Deformation

Shear-Induced Copper Distortion and Fiber Wash

Excessive hydrodynamic forces during squeeze flow induce lateral forces on inner layer traces and underlying target pads. When high-viscosity resin moves horizontally across thin copper features, shear stress τ = η_eff · γ˙ exerts mechanical drag on copper sidewalls. For 12 µm thick copper traces sitting on unreinforced dielectric layers, lateral shear stress exceeding 3.5 kPa causes copper line bending, trace stretching, and pitch variation.

Pad displacement causes layer misregistration.

In fine-line sequential build-up layers, high squeeze flow velocities pull glass filaments out of alignment, creating a defect known as glass fiber wash. Displaced glass fibers bunch together in routing channels while leaving adjacent areas unreinforced. Fiber bundle displacement alters localized dielectric constants, causing severe differential impedance mismatches across signal pairs.

Copper thickness governs flow resistance.

Microvia dimpling reduces assembly yield. High shear stress strips resin. Unfilled microvias cause electrical opens.

1. Perform differential scanning calorimetry on incoming prepreg lots to determine exact minimum viscosity temperature window.
2. Adjust vacuum press pressure ramp timing to reach maximum hydraulic clamp force before resin gelation initiates.
3. Calibrate platen temperature uniformity to within two degrees Celsius across the entire working panel area.
4. Monitor inner layer pattern copper density to rebalance flow barriers in low-density signal regions.
5. Inspect trial coupon microsections to verify full microvia cavity fill and measure final dielectric thickness.

IPC-6012 Class 3 specification mandates zero microvia voids within the target pad interface, causing full lot rejection when squeeze flow entrapment leaves unfilled microvia voids.

Defect Prevention Procedures for Sequential Build Substrates

Process technicians adjust press parameters through systematic evaluation steps to maintain structural integrity across high-density layers. Controlling initial pressure rise rates limits fluid velocity spikes, protecting sensitive fine-line traces from mechanical sweep. Balancing copper coverage across internal layers reduces asymmetric flow resistance, preventing panel bow and twist induced by differential squeeze dynamics.

Contract specifications incorporating IPC-6012 Section 3.4.2 enforce strict limits on dielectric separation variations resulting from resin squeeze-out. Under this specification, final dielectric clearance after lamination cannot vary by more than +/- 10 percent from nominal drawing values across any 500 mm span.

Outlay

Panel Area Efficiency and Materials Pricing

Substrate panel layout decisions dictate the net count of high-density interconnect units harvested from raw sheet stock. Standard panel sizes including 457 mm x 610 mm and 508 mm x 610 mm require edge clearance margins reserved for press tool pins, vacuum sealing gaskets, and resin flow retention barriers. Squeeze flow behavior dictates margin width: high-bleed prepregs require wider peripheral waste zones to contain extruded matrix resin, reducing usable board area per panel.

Higher pressure increases resin bleed. Table 2 illustrates how scrap rates driven by squeeze flow defect modes escalate finished unit costs across different substrate architectures.

Excessive hydraulic pressing force during prepreg gelation displaces functional resin into array borders and reduces central dielectric thickness.

Commercial Trade-Offs in Laminate Slash Sheet Selection

Specifying IPC-4101 slash sheet parameters balances base material thermal capability against purchasing expenditure. Standard high-Tg FR-4 materials compliant with IPC-4101/126 cost substantially less than advanced low-loss polyphenylene ether laminates specified under IPC-4101/102. Selecting higher-grade laminates solely for improved thermal rating without considering rheological flow compatibility increases material expenditure while failing to resolve microvia fill defect rates.

• Slash Sheet Equivalency Rules – Material substitution requests must demonstrate matching viscosity profiles under identical heating rates to preserve process window validity.
• Resin Content Tolerance Bands – Purchasing specifications requiring prepreg resin content within +/- 1.5 percent prevent batch-to-batch dielectric thickness drift.
• Working Panel Size Optimization – Squeeze flow containment margins must scale with panel surface area to optimize usable substrate yield per press sheet.
• Target Pad Clearance Allowances – Fabrication artwork clearances surrounding laser microvias must accommodate calculated lateral resin drag displacement tolerances.

Matching prepreg flow scale parameters to target copper foil weights yields reliable cavity filling without incurring premium laminate surcharges.

Compliance

Fabrication Drawing Notes and Inspection Coupons

Engineering drawings specify microvia fill targets, dielectric clearance limits, and microsection acceptance criteria. Fabrication notes must explicitly state acceptable microvia dimple depth, typically capping dimple depression at 10 µm or 15 percent of total microvia depth, whichever is smaller. Notes should also define acceptable resin fill percentages inside blind microvias, demanding 100 percent fill for Class 3 high-reliability aerospace and medical applications.

Quality conformance inspection requires specialized microsection test coupons placed at panel corners and panel center. Center coupons capture worst-case squeeze pressure drop conditions, where resin flow distance reaches its maximum. Corner coupons evaluate edge-bleed effects, where excessive squeeze flow can starve surrounding dielectric layers.

Coupon design must mirror the functional board microvia aspect ratios and copper plane coverage to ensure representative flow testing during the press cycle.

Cross-Sectional Verification under Industrial Standards

Quality assurance teams perform destructive microsection analysis on quality conformance test coupons to evaluate microvia fill percentages. IPC-A-600 Acceptability of Printed Boards defines visual criteria for microvia structural integrity, identifying internal voids, dimpling, and target pad separation under optical magnification up to 200X. Scanning electron microscopy evaluation provides higher resolution verification for microvia entrance wetting and resin-to-copper interface adhesion.

Verification protocols under IPC-6012 Section 3.6.2 require microsection coupons to undergo three simulated assembly reflow cycles at 260 °C before metallographic preparation. Thermal exposure induces mechanical strain along the z-axis, exposing latent voids or micro-cracks generated by incomplete squeeze flow fill. Substrate lots passing thermal stress testing without delamination or target pad lift demonstrate robust hydro-mechanical consolidation during lamination, verifying that the press parameter profile achieved complete cavity filling within the resin fluidity window.