Heat Staking vs. Ultrasonic Welding for PCB Assemblies
Both processes form a permanent thermoplastic joint without fasteners or adhesives. For a populated board, the deciding factor is usually not which forms a stronger head – it is where the process energy ends up.
by Alex Spurgeon
Mechanically fastening plastic to a printed circuit assembly is a routine but consequential step. Boards are retained in plastic housings, light pipes and bezels are located to displays, EMI shields and connectors are captured, and standoffs are formed in a single operation. When designers want to avoid screws, snap fits and adhesives, two thermoplastic joining methods dominate the conversation: heat staking and ultrasonic welding. Both melt and reform plastic to lock parts together, and both are fast and free of consumables. They are not, however, interchangeable on a populated board.
This article compares the two processes specifically for PCB assembly work. The thesis is stated up front, because it should drive the decision: the most important difference for a board is not joint strength but how each process delivers energy. Heat staking conducts heat into a localized boss; ultrasonic welding couples high-frequency mechanical vibration through the assembly. On a board carrying vibration-sensitive components and mixed materials, that distinction usually decides the process before throughput or cost enter the discussion.
How each process forms a joint. A heat-staked joint is formed by an external heat source. A temperature-controlled tool, a stream of hot air, or an impulse-heated thermode raises a plastic boss above its softening point. The softened boss is reformed into a head that captures the mating part, then cooled – ideally under pressure – so it solidifies in the formed shape before the tool retracts. The heat is conductive and stays largely within the boss. Thermoplastic staking can be performed with thermal tooling, a hot punch, hot-air/cold-upset tooling or cold forming, and the process readily joins plastics to dissimilar materials, including metal and PCBs1.
Ultrasonic welding forms a joint by a different physical mechanism. A transducer converts high-frequency electrical energy into mechanical vibration, typically at 20 to 40kHz, which a horn (sonotrode) delivers to the part under pressure. The vibration produces intermolecular and surface friction at a designed contact point, melting the plastic in a fraction of a second; the melt then solidifies under a hold force2. When two thermoplastics are joined, the melt is concentrated at a small triangular feature called an energy director. When a single stud is reformed over a board – the closer analog to staking – the operation is often called ultrasonic staking. The defining characteristic is that the energy arrives as vibration, and that vibration does not stop at the joint.
Figure 1 contrasts the two mechanisms on the same task: a plastic boss retaining a PCB in a housing.
The practical consequence is visible in the figure. In heat staking, the energy is conductive and localized. In ultrasonic welding, the energy is vibratory and travels through the boss into the board, its solder joints, and any components in the acoustic path.
The crux for populated boards: vibration coupling. A printed circuit assembly is a mechanical structure with its own stiffness, resonances, and fragile features. When ultrasonic energy passes through it, those features are loaded. Solder joints, in particular, are brittle interconnects that do not tolerate repeated high-frequency strain well, and ceramic capacitors, crystals, MEMS devices and fine-pitch packages can be sensitive to vibratory input.
This is not a theoretical concern. A finite-element study of ultrasonic plastic assembly modeled the vibration that reaches the solder joints of a cell phone assembly during welding and identified solder-joint failure driven by the process itself. The analysis showed the failures depended on three structural factors – the location of the chip, the shape of the PCB, and the structure of the housing – rather than on the weld joint in isolation. After the design was revised based on the simulation, the reported reject ratio in mass production fell by roughly 90 percent3.
Two conclusions follow, and both matter for process selection. First, ultrasonic energy can damage a board even when the weld itself is sound, and the location of the damage is governed by the board and housing structure, not by weld parameters alone. Second, the risk is manageable: with the horn kept away from sensitive devices, the board supported and fixtured, amplitude controlled, and energy directors used to localize the melt, ultrasonic welding runs reliably on many electronic assemblies. The point is that the risk has to be engineered out, and that engineering is specific to each board.
Heat staking sidesteps this class of risk. Because the energy is conductive rather than vibratory, there is no high-frequency input to couple into solder joints or fragile parts. The thermal load is real and must be managed – standoff height, clearance from traces and components, and clamp force that does not flex the board all require attention – but the failure mode of vibration-stressed interconnects is largely absent. This is a primary reason heat staking is frequently specified when bosses are located close to populated areas on a board.
Materials and joint design. Material behavior separates the two processes as much as energy delivery does.
Heat staking is tolerant of a wide range of thermoplastics – ABS, polycarbonate, nylon and polypropylene among them – and the tool temperature is dictated by the polymer rather than by a single universal setpoint. Amorphous resins such as ABS and polycarbonate soften over a broad range and are forgiving; semicrystalline resins such as nylon and polypropylene have sharper melt transitions and a narrower process window. As rough heated-tip starting points, ABS forms at roughly 200 to 230°C and polycarbonate at roughly 260 to 300°C, with the exact window set by the grade. Nylon should be dried before staking because absorbed moisture flashes to steam in the molten head, leaving a porous, weak stake; acetal has a narrow window and degrades if overheated. Because there is no vibration, heat staking is also comparatively forgiving of glass- and mineral-filled grades and of higher-temperature resins, which can be prone to micro-cracking under vibratory input. It also joins dissimilar materials and mechanically secures metal to plastic, which is useful when a board is captured between a plastic boss and a metal chassis1.
Ultrasonic welding is more sensitive to material selection and joint design. Welding is most reliable between identical or closely compatible resins, and amorphous polymers generally transmit and weld more predictably than semicrystalline ones, which absorb and redistribute vibratory energy differently. Fillers and additives alter how the material conducts ultrasonic energy, and the joint must be designed with an appropriate energy director or shear geometry to concentrate the melt. For stud reforming specifically, melt that flows freely under vibration can yield heads of varying diameter and shape unless the tooling and parameters are well controlled, whereas a thermally formed head cooled under pressure tends to be geometrically consistent.
Throughput, cost and process control. Speed is ultrasonic welding’s signature advantage. A weld or stake is typically made in well under a second, which makes the process attractive on high-volume lines where a single joint sets the cycle time2. The trade-offs are a higher equipment investment – a generator, converter, booster, and a horn tuned to the operating frequency – a more involved setup, and a narrower process window that is sensitive to part and material variation.
Heat staking inverts several of those terms. A single cycle is longer because it includes both a heat phase and a cooling phase, often a few seconds in total. But multi-point thermal tooling forms many bosses in one stroke, so per-board throughput can be competitive even when per-joint time is longer. Equipment costs are generally lower, maintenance is simpler and modern systems use distance- or energy-based control with position feedback to hold head geometry within tolerance. Two process details carry most of the quality on a board: cooling the head under pressure so it does not relax back and a clamp-and-standoff scheme that locates the board without flexing it. On the ultrasonic side, the equivalent variables are amplitude – set by the booster and horn gain – trigger and weld force, weld and hold time, and rigid support of the part; these govern both joint quality and the amount of energy reaching the board. Figure 2 summarizes the qualitative comparison across the attributes that most often determine a PCB assembly application.
The figure is a qualitative synthesis intended for orientation rather than a set of measured values; the right answer for a specific assembly depends on its components, materials, geometry and volume.
A selection framework. For PCB assemblies, the decision can be reduced to a short sequence of questions, shown in Figure 3.
Work through it in order. If vibration-sensitive components are near the joint, heat staking avoids the energy path that would put them at risk. If the assembly joins dissimilar materials or metals to plastics, or uses glass-filled or high-temperature resins, heat staking is the more tolerant process. If neither of those applies and per-joint cycle time is the genuine bottleneck at high volume on a mechanically robust assembly, ultrasonic welding’s speed becomes the deciding advantage. Absent a strong cycle-time argument, heat staking is a reasonable default on cost and simplicity grounds. Whichever process is chosen, the joint should be validated with pull tests and at least one cross-section before it is committed to a production line; neither a hollow head, a cracked boss base, nor a vibration-stressed solder joint is visible from the outside.
Validation and the defects each process hides. Neither process should be released on appearance alone, because the failures that matter are internal. A heat-staked head can look complete while hiding a cold, unfused core that fails a pull test, an over-formed head that has cracked the boss base, or thermal exposure to a component that sat too close to the tool. An ultrasonic joint can look clean while concealing a partially welded interface, a marked or diaphragmed area beneath the horn, or – as the finite-element work showed – solder-joint damage at a location remote from the weld itself3. A short validation routine catches all of these. Pull-test a sample of joints to failure, cross-section at least one to inspect head fill and the boss base, and, where the assembly is vibration-sensitive, functionally test or inspect the solder joints nearest the tool after ultrasonic processing. Building that routine into first-article approval is inexpensive relative to a field return.
Conclusion
Heat staking and ultrasonic welding both produce strong, fastener-free thermoplastic joints and belong in a PCB assembly engineer’s toolbox. The choice between them for a populated board should be based on energy delivery first. Ultrasonic welding is faster per joint and well suited to high-volume work, but its vibration couples into the assembly and can stress solder joints and fragile components unless the board, housing, and horn placement are engineered to prevent it – a risk demonstrated in finite-element analysis and resolved only by structural redesign3. Heat staking delivers localized, conductive heat that avoids the input, tolerates dissimilar and filled materials, and reduces cycle time through multi-point tooling, at the cost of a longer individual cycle. Match the process to the assembly: choose ultrasonic when per-joint speed on a robust board dominates, and heat staking when the board carries sensitive components, mixes materials, or runs filled or high-temperature resins. In every case, validate the joint before the line depends on it.End of article content
References
[1.] Troughton, M. J. (Ed.). (2008). Handbook of plastics joining: A practical guide (2nd ed.). William Andrew / Plastics Design Library. (Chapters on staking and ultrasonic welding.)
[2.] Grewell, D., & Benatar, A. (2007). Welding of plastics: Fundamentals and new developments. International Polymer Processing, 22(1), 43–60.
[3.] Structural improvement for solder joint failure in ultrasonic plastic assembly. (2008). Simulation Modelling Practice and Theory. Elsevier. ScienceDirect article identifier S1569190X08001159. (Finite-element analysis of ultrasonic-induced solder-joint failure; ~90% reject reduction after structural redesign.)
Alex Spurgeon is the founder of U.S. HeatStake (usheatstake.com), a U.S. manufacturer of precision impulse heat-staking and brass-insert installation systems based in Roanoke, Indiana. He works with assembly teams on thermoplastic joining process development. He can be reached at info@usheatstake.com or (614) 549-0627.