PES Laser Welding: A Practical Guide to Polyethersulfone

A medical-device engineer clamps two amber-coloured polyethersulfone housings together, fires a diode laser through the clear wall, and watches a leak test that failed three times last week suddenly pass. Nothing was glued, no fumes came off the line, and the cycle took under two seconds. That is PES laser welding, and the strange part is that it only works because of a few millimetres of polymer that is almost transparent to the very beam doing the melting.

PES Laser Welding: What the Term Actually Means

PES laser welding is the laser joining of polyethersulfone (PES, also sold under the designation PESU) thermoplastic parts. In the most common setup, a laser beam passes through a transparent upper component and is absorbed at the joint interface, where it melts a thin film of polymer and lets the chains interdiffuse into a single continuous bond. Heat, placed precisely, for a precisely controlled amount of time.

One clarification before we go further, because the acronym trips people up. Searching for “PES laser welding” can land you on three different things. This article is about polyethersulfone. It is not about polyester, which covers polyesters such as PET and PBT, and it is not a process parameter acronym. Polyesters behave differently enough to matter: polyesters are used with the same through-transmission approach, but their absorption behaviour, lower softening temperatures and, in the case of PBT, semi-crystalline structure mean parameters do not transfer between the two families. If your part is moulded from PET rather than PESU, start from scratch rather than adjusting these numbers.

Where PES sits in the polysulfone family

Polyethersulfone belongs to a small family of polysulfones, and there are three industrially used members: PSU (polysulfone), PES/PESU (polyethersulfone) and PPSU (polyphenylsulfone). All three are rigid, amorphous, high-temperature thermoplastics with excellent chemical and hydrolytic resistance. They differ in cost, in maximum temperature and in how readily they absorb laser energy, which is why the family shows up together in aerospace, medical and electronics work but not interchangeably in a welding process.

How this differs from welding metal

If you normally arrive at laser welding from steel or aluminium, the vocabulary will mislead you. There is no filler wire, no keyhole, no metallurgical grain and no heat-affected zone with a distinct microstructural property. The joint is a diffusion bond across a melted polymer interface: the two surfaces melt, the polymer chains interdiffuse, and on cooling you have one continuous mass of material. If a supplier tells you the weld strength equals the base material, they mean something different than they would for steel, because the ceiling is set by how well the chains interdiffuse, not by a deposit chemistry. Our guide to alpha laser welding systems covers the metal side in more depth.

The three configurations

Two dominate practice, and a third is a variation on them:

• Through-transmission laser welding (TTLW). The beam travels through a clear upper part and is absorbed at the interface, usually in a lower part containing an absorber. This is the workhorse for clear medical and fluid-handling components.
• Direct or surface laser welding. The beam hits the joint directly and the surface absorbs the energy. Faster and simpler, but it heats the visible face and can discolour it.
• Film or interlayer variant. A thin absorber-bearing film is placed between the parts and welded through. It adds a step but also lets you join polymers that would not otherwise bond.

Why the amorphous nature matters

PES has no crystalline phase, so a weld does not undergo the post-weld shrinkage seen in polypropylene or polyethylene. That is a practical advantage: the joint does not pull out of flat, and the finished part stays optically clear. Laser welding of polymers depends on this chain-interdiffusion mechanism, and the wider family picture is covered by TWI’s polymer welding work.

Thermally, the polysulfones have a glass transition temperature between 190 and 230 degrees Celsius and are usable across roughly -100 to 200 degrees Celsius. Dimension change after exposure to boiling water or 150 degrees Celsius air or steam generally stays below 0.1 percent. The welding window is therefore high but narrow compared with polyolefins, which is why absorption and energy control decide the outcome.

Where it is used

• Sterilizable medical fluid-handling components and device housings
• Filtration membranes, cartridges and filter housings
• Electrical and electronic enclosures where heat must stay local
• Automotive lighting and fuel-system components
• Analytical and laboratory instruments

Now that the terminology is clear, the next question is the one that decides every project: what happens when the beam actually meets the polymer.

Why Polyethersulfone Is Both a Gift and a Challenge

Ask anyone who runs a plastic welding cell what makes a polymer hard to weld, and the answer is never strength, cleanliness or cycle time. It is absorption. Everything hinges on one question: what happens when the beam reaches the plastic?

Light has three fates in a polymer: reflected, transmitted or absorbed. Reflection is the least interesting, because reflected laser energy in most polymers is typically on the order of 5 to 10 percent, low compared with absorption and transmission. That makes it an optics problem rather than a process one, and a clean anti-reflective coating on the beam delivery beats most process tweaks.

Absorption length decides the process

How deep light travels before it is absorbed follows the Bouger-Lambert law of absorption: intensity at depth z equals intensity at the surface multiplied by e to the power of negative K times z, where K is the absorption coefficient. Intensity falls off exponentially with depth, so the deeper light goes before absorption, the more transparent the material behaves at your wavelength.

The number engineers use is the absorption length, L-alpha, simply one divided by the absorption coefficient. It is a distance, usually in micrometres or millimetres. Because it depends on laser wavelength, it dictates which of three processes you have:

• Direct welding, when the absorption length matches the thickness of the top layer, so nearly all the energy lands where you want it.
• Through-transmission welding, when the top layer transmits and the layer beneath absorbs, heating the interface directly.
• No welding at all, when both parts transmit and the beam exits the far side having done nothing.

PES is an amber, optically transparent high-performance polymer that transmits much of the near-infrared band, so an unmodified PES-to-PES joint usually does not melt at all. The beam crosses the upper part, passes the interface and leaves through the bottom, leaving a joint that looks clean and fails a leak test. This is not a power problem. More watts simply give you a brighter exit.

What poor absorption looks like in practice

The symptom is unmistakable: a joint that looks heated, browned or smoked but pulls apart with a dull, chalky surface. The beam scoured the surface without ever heating the interface, leaving oxidation on top and a cold joint underneath. Operators call it “the part looks welded but it is not.”

The aromatic penalty

PES is aromatic, and that chemistry is both why it survives the process and why it can be ruined by it. Aromatic backbones brown and degrade before they flow properly if overheated. The resin is tough: polysulfones have outstanding resistance to heat, oxidation, hydrolysis, aqueous media and alkaline media, along with good electrical properties. The failure mode is process abuse, not weak plastic.

The paradox: the glass transition of polysulfones sits in the region of 200 degrees C, so reaching a flowing melt takes a lot of energy. Too little gives you a cold joint; enough to guarantee flow risks charring the aromatic backbone before the chains can interdiffuse. The window between “nothing happened” and “the part is ruined” is narrow, sitting near the top of the material’s thermal tolerance. Properly made, the joint keeps its strength across that same elevated range, exactly what sterilisation and automotive under-bonnet service demand.

The amorphous advantage

PES is amorphous, and that is why it welds far more cleanly than polypropylene or polyethylene. There is no crystalline phase to reject weld-line impurities, no shrinkage on cooling, and no residual stress between competing morphologies. A good PES weld has a clear interface and often looks untouched. Semi-crystalline polymers are the opposite: crystals nucleate and grow as the weld cools, the material contracts, and the interface can be left with voids or a weak, stressed boundary layer.

Chemically, PES resists pH 2 to 13 and shrugs off hydrocarbon oils and surfactants. The weakness is low-polarity organic solvents such as ketones and chlorinated hydrocarbons, which attack the surface, so keep degreasers and solvent wipes away from the joint face.

Three levers

Every fix for PES laser welding moves one of three levers: absorption (additive, dye or coating), geometry (joint design and gap), or energy delivery (wavelength, power, speed, clamping).

For readers coming from metal fabrication the shift is large: a metal weld makes a keyhole and melts filler, while a polymer weld must deposit energy in a thin, precise volume. Transmission and absorption matter far more than beam shape, the opposite of what guides like ours on alpha laser welding systems teach.

Choosing a Laser Source for Polyethersulfone

Choosing a laser for polyethersulfone is an optical decision, not a brand decision. A source that melts a black-pigmented PES flange cleanly at 940 nm may barely warm a clear one at the same setting, and the same source can brown one grade and seal another. Start from your actual parts: how they absorb, how thick they are, and what the joint geometry demands. Beam quality, delivery method and pulse control only matter once that physics is settled.

Wavelength options and what each one does to PES

Source	Typical wavelength	Effect on PES
Diode (GaAs)	780-980 nm (808, 940 and 980 nm common)	Best energy efficiency of the plastic-welding options and cheap per watt, but PES transmits much of the near-infrared, so an absorber is normally required
Fibre / Nd:YAG	~1064 nm	High beam quality and fibre delivery for fine focus and three-dimensional seams; same absorber logic as diodes
1470 nm diode	1470 nm	Clear polymers absorb noticeably better here than at 940 nm, so less additive is needed and less light scatters around
Thulium fibre	~1940 nm	Sits in the band where many clear thermoplastics begin to absorb, giving subsurface heating on modest thicknesses
Mid-infrared fibre	2.7-3.0 micrometres	Strongly absorbed by clear polymers; short absorption lengths that match typical part wall thickness
CO2	10.6 micrometres	Absorbed within roughly 0.5 mm, so it is a surface-heating and cutting tool rather than a bulk welding tool

Diode lasers in the 780-980 nm range are noted for superior energy efficiency compared with Nd:YAG and CO2 sources, which is why they dominate cost-sensitive plastic-welding cells. Fibre and Nd:YAG systems around 1064 nm buy beam quality and fibre delivery instead, which matters when you need a small spot or a robot-mounted head. Moving up the spectrum changes the physics: many optically clear thermoplastics are partially absorbing in the 1500-2000 nm range, which is what makes subsurface heating and clear-to-clear welding possible without a pigment. A 2018 overview of laser welding methods for optically clear plastics documents this band well, and mid-infrared sources push it further.

Absorption length decides which regime you land in

Absorption length, as covered earlier, is the distance over which the beam’s intensity decays to about 37 percent of its entry value. Compare it with your part geometry and only three outcomes are possible:

• The top layer absorbs too much. Energy is spent at the surface, so you get ablation and surface damage before a weld can form.
• The top layer transmits and the bottom absorbs. This is through-transmission welding, the classic PES configuration with a pigmented, dyed or coated lower component.
• Both transmit. Nothing happens. This is unmodified clear PES welded to unmodified clear PES, and it is the single most common reason a first trial fails.

For direct, single-surface welding the optimum sits elsewhere: choose a wavelength whose absorption length closely matches the thickness of the top layer. Almost nothing is wasted heating the material above the joint, and the melt forms right where you want it.

The mid-infrared argument

Lasers beyond roughly 2.5 micrometres are strongly absorbed by most clear polymers. Mid-infrared fibre lasers around 3 micrometres deliver absorption lengths for most polymers in the 10-500 micrometre range, which lines up neatly with typical medical and electronic part thicknesses. As an illustration of the physics, one study measured more than 70 percent absorption through 180 micrometres of PET at 2800 nm (MDPI, 2020).

The practical payoff is that an absorbing additive may not be needed at all, which removes the colour, the regulatory question and the extra moulding step. It also explains why mid-infrared polymer processing has become the focus of so much development work.

Continuous versus pulsed, and why peak power matters

Continuous-wave sources accumulate heat along the seam, which is efficient for long welds but raises the risk of browning an aromatic polymer like PES. Pulsed mid-infrared sources keep the heat-affected zone tight and largely avoid the discoloration associated with CO2 surface heating, because energy is delivered in short bursts and the surrounding material never dwells in the degradation window (precision polymer welding with pulsed mid-infrared fibre lasers).

Peak power is the lever that controls melt depth. A short, high-peak pulse melts locally and deeply into the interface without raising the average heat load, so the rest of the part stays below its degradation temperature.

Delivery, integration and the medical precedent

Beyond wavelength, think about how the beam reaches the part: a fixed head with clamping fixture, a scanner for simultaneous illumination, or fibre delivery on a robot for three-dimensional seams (TRUMPF on laser plastic welding). If you are new to laser equipment in general, the Alpha Laser Welding guide to systems, uses and safety is a useful orientation before comparing quotations.

There is medical precedent for near-infrared bonding: a Fraunhofer ILT feasibility study on a sutureless wound-closure device used 1470 nm laser power to achieve higher tensile strength than the reference closure method.

Joint Design and Fixturing for PES Parts

A PES laser weld is won or lost at the drawing stage, not the parameter screen. You get a hermetic, optically clear joint or a cold, brittle seam depending on how the parts are shaped, how closely they meet, and how they are held while the polymer melts.

Metal readers sometimes assume the principles are identical to alpha laser welding on aluminium or steel. They are not. A polymer joint has no keyhole, no filler and no metallurgical heat-affected zone. You are creating intimate contact between molten surfaces, and geometry alone holds that contact together.

The classic through-transmission flange joint

For a through-transmission laser weld (TTLW), the standard geometry is a stepped or rebated lip on the transparent upper component landing on a flat, absorbing lower surface. The beam enters through the clear wall and is absorbed in a thin layer at or just beneath the interface, so the melt forms at the joint line, not on the visible face.

Keep the absorbing surface directly under the laser path with a consistent stand-off, and keep the transparent member thin: one to two millimetres is common, so the beam still has energy on arrival. A good overview of laser welding optically clear plastics shows the same principles applied to polycarbonate, copolyester and acrylic housings.

Gap control: the 150 micron rule

For through-transmission welding, the joint gap should be as small as possible, ideally below 150 micrometres. This one number decides more outcomes than any laser setting.

A gap ruins the process three ways at once: the beam no longer reaches the interface in a controlled way, so the melt pool lands off the joint; heat conducts away through the air gap instead of flowing into the lower part; and the faces never make intimate contact, so polymer chains cannot interdiffuse. TWI’s published work on through-transmission laser welding makes the point: without firm, uniform contact you produce a weld that looks plausible and leaks.

Temperature, time and pressure

Every plastics welding process, whether hot plate, ultrasonic, vibration or laser, is governed by the same three parameters: temperature, time and pressure. Laser welding delivers the first two with extreme precision, which leaves pressure, and pressure is the fixture’s job.

Typical clamping pressure for through-transmission laser welding sits between 0.1 and 1 N/mm², according to TWI’s polymer welding guidance. Too little and the faces are not in true contact, which limits chain interdiffusion and drops weld strength. Too much and you squeeze molten material out of the joint or distort a thin-walled housing. TWI’s rule is simple: no intimate contact, no weld strength.

Fixtures: fixed clamps, moving clamps and compliance

Two broad fixture strategies exist. Fixed clamps squeeze the parts together and stay put, suiting short, simple welds on small components. Moving clamps travel along the seam, so force lands exactly where the beam works. For large parts that is the advantage: a shoe or roller follows the seam, so the far end is never pushed out of flat.

Compliance matters too. A rigid clamp can distort a moulded housing, and a distorted housing changes the gap in the next part. An elastomeric element or spring-loaded foot absorbs that variation without adding a hard constraint, and rollers reduce scratching on clear surfaces near the weld.

Transparent clamping elements are common because operators need to see the part under the beam. Acrylic and plain plate glass are the everyday choices. Borosilicate is worth the extra cost where thermal shock is a risk, and quartz suits high-temperature polymers and hot fixtures.

Motion options and their trade-offs

• Contour welding: the part moves under a fixed beam. Round parts turn at up to about 25 m/min, fast and easy to fixture, but the seam must be clean and continuous.
• Scanning (quasi-simultaneous): the beam is steered along the seam. It runs at up to 15 m/s (900 m/min) and handles complex outlines, including closed and non-planar seam geometries.
• Robot-mounted fibre delivery: suits three-dimensional seams on larger assemblies, at the cost of programming effort and floor space.
• Curtain laser with a mask: one shot illuminates a small, complex geometry through a patterned mask. Fast on tiny parts, limited by mask size.
• Simultaneous illumination: the whole joint is exposed at once, as described in TRUMPF’s laser plastic welding overview. Very short cycle times, but it needs a designed optical system and heat sinking.

Part variation, warp and seam access

Moulded parts are not identical. Cavity fill, cooling and fibre orientation shift flatness, and a housing out of flat by 200 micrometres gives an inconsistent gap on the line. Give the fixture somewhere to push the part flat instead of clamping over the seam, and make sure the beam reaches the last millimetres at a start-stop point.

Corners need their own thought: a restarted beam leaves a start-up mark, and a sharp turn needs enough dwell to build a full melt pool. Either mask and jump, or scan continuously, and accept that corners run hotter.

The limits of reinforcement

Glass fibre changes the rules. Above roughly 40 percent reinforcement, laser welds tend to become somewhat brittle, because the weld interface no longer offers a continuous polymer matrix. A laser-transparent reinforced part should not be thicker than 2 mm, since a thick reinforced upper layer will not transmit the beam cleanly.

Absorbers, Additives and Surface Preparation for PES

Most polyethersulfone laser welding projects go wrong for reasons unrelated to the laser itself. Amber PES transmits so much of the near-infrared band that an unmodified joint can come out of a full exposure looking untouched. Decide where the energy is absorbed, because that decision lives in the material, not the machine.

Strategy 1: Bulk additive in the moulded part

Carbon black is the classic solution. At roughly 0.1 to 1 percent by weight it turns a transparent PES part into one that soaks up near-infrared laser radiation instead of passing it through. Across that narrow range the behaviour changes sharply: loading is the main control on where the melt forms.

Carbon black loading	What the beam does	Weld result
0.1 to 0.15 percent	Partial transmission, absorption deeper in the wall	Subsurface heating, thicker melt layer, less risk of surface degradation
0.5 percent and above	Strong absorption at the entry surface	Thinner melt layers, higher risk of charring or bubbling

Subsurface heating is why the low-loading end is worth chasing. If you need a deep, generous melt without discolouring the visible surface, a lightly loaded part keeps the energy travelling a short distance into the wall before it converts to heat.

Strategy 2: Infrared-absorbing dyes and pigments

Carbon black works but makes the part opaque and dark, disqualifying it for a clear medical housing or an optical fluid path. The better answer is a dye engineered for the near-infrared. Work at TWI and Gentex produced infrared dyes nearly transparent in the visible range yet absorbing strongly between 800 and 1100 nm, exactly where most diode and fibre sources operate, leading to the commercial Clearweld grades.

These grades can be introduced into the bulk during moulding of the absorbing component, so the part welds cleanly with no visible coating, or applied as a coating onto the weld surface only. Bulk loading follows the absorber everywhere, while a surface grade keeps the resin virgin and limits the additive to the seam. TWI’s polymer welding research covers the detail.

Strategy 3: Coated absorbers and intermediate films

The third route puts the absorber where you want it. A dispersion of dark pigment or carbon black in solvent can be brushed, sprayed or inkjet-printed onto the joint face and welded once dry. It is slower than a moulded-in additive and adds a handling step, but the resin keeps its clarity and it works on parts you already have.

A thin extruded film containing an absorber works differently. Laid between the two parts, it melts first and then fuses the interface on either side, opening the door to joining polymer pairs that would not normally interdiffuse because the film provides a wetting layer. New to the process? A general primer on alpha laser welding systems and safety will get you oriented on the equipment side.

The colourant trap

Not every pigment helps: titanium dioxide, which gives polymers their white colour, scatters laser energy rather than absorbing it, so white PES grades are notoriously awkward to weld. Carbon black differs: its concentration can be dialled to control the effective area of the weld, since a tighter footprint concentrates the heat into a smaller weld line. Glass fibre and other fillers scatter and divert light too, reducing the energy that reaches the interface.

Surface preparation

Even a tuned absorber will not save a contaminated joint. Clean the faces, isopropanol being a common choice for PES, and make sure they are completely dry and free of release agents, mould oils, dust and fingerprints. Parts sit bagged for weeks, so the joint face is the most contaminated surface in the assembly. The prep discipline metalworkers expect before striking an arc applies here, with a cleaner that will not attack the polymer, as outlined in our guide to prepping aluminum for welding.

PES is hygroscopic. Moisture in the bulk polymer or on the joint face turns to steam at weld temperatures, the usual root cause of bubbles and porosity in otherwise sound parts. Condition components to a controlled moisture level before assembly and keep them warm and dry up to the cell.

The regulatory layer for medical work

Any additive, dye or film in a patient-contact or fluid-path device needs a biocompatibility and leachables assessment, because a moulded-in carbon black or infrared dye is part of the finished article, not a processing consumable. Validate the final welded assembly rather than the raw resin, since the weld line changes what the patient and the fluid see. Laser welding of polymers covers the process background, and absorber selection belongs in the quality system as a documented design decision reviewed alongside joint geometry and parameters.

I’ll rewrite this section, tightening it into the 600–900 word range while preserving every fact and link.

Process Parameters and the PES Weld Quality Window

Polyethersulfone gives you a narrow target: melt the interface without degrading the polymer. Between a cold joint on one side and brown, brittle char on the other sits a working window that is wide enough to hit but narrow enough to miss if you guess. The good news is that the window is not a mystery. It is built from a handful of parameters you can measure, log and hold.

Energy density is the governing variable

Almost everything in plastic laser welding reduces to one number: energy density. It folds temperature and time into a single figure, because the same joules delivered slowly and the same joules delivered quickly produce completely different melts.

• Moving processes (contour or raster welding): energy density = laser power ÷ (welding speed × spot width at the joint).
• Fixed processes (stationary or simultaneous heating): energy density = (power × irradiation time) ÷ irradiated area.

Most laser welding applications land between roughly 0.1 and 2 J/mm², with the value scaling to the melt depth you need rather than to the thickness of the part. A shallow interface bond wants the low end. A deeper melt through a thick wall wants more, and this is where the window starts to feel tight.

Now the caveat that matters more than the formula. Energy density describes energy delivered, not energy retained. Heat conducts away from the joint into the surrounding polymer and into the clamp, so two joints with identical energy density can produce very different weld quality purely because their thermal surroundings differ. This is a well-documented limitation in the research literature on laser welding of polymers, and it explains why parameter tables borrowed from another part number rarely transfer cleanly (review of laser polymer welding).

The practical parameter set

• Power (W) — the primary driver of melt depth and pool size.
• Speed or dwell — travel speed for contour welding; dwell time for stationary or simultaneous heating.
• Spot size at the joint — set by the optical setup, not just the beam diameter at the lens.
• Focal position and focus offset — where the beam waist sits relative to the interface. A negative offset (defocus) spreads the energy and limits depth; a positive offset concentrates it. This is your main control over how deep the melt grows without raising power.
• Pulse duration, repetition rate and duty cycle — in pulsed mode these set the peak power and the cooling gap between pulses, which is what keeps the heat-affected zone tight.
• Clamp force — typically in the region of 0.1 to 1 N/mm², enough to hold the joint faces in intimate contact without squeezing molten polymer out of the seam.
• Hold or cooling time — the period after the beam switches off while the fixture keeps the parts pressed together, allowing the molten film to solidify without drawing back into a void.

For general laser system fundamentals, including how the source is chosen and delivered, see our Alpha Laser Welding guide. The physics of beam delivery is shared with metals; the interaction with the polymer is not.

Typical scale of a production process

• Laser plastic welding is routinely performed with sources of up to approximately 300 W.
• Contour welding of round parts has been demonstrated at speeds of up to 25 m/min.
• Simultaneous heating can deliver process times of only around 100 ms, which suits simple high-volume geometries.
• Short cycle times of 1 to 3 seconds have been demonstrated for through-transmission laser welding processes (TRUMPF on laser plastic welding).

Building the process window experimentally

Do not trust a borrowed parameter sheet. Build your own window on coupons cut from production parts, with the real absorber strategy and the real fixture in place.

1. Run a power-versus-speed matrix across several power levels and several speeds, one weld per combination.
2. Inspect every sample for the cold-joint limit (lowest combination that produced no bond) and the degradation limit (highest combination that produced discolouration, char or flash).
3. Plot both limits and operate in the middle of the band, not at either edge.
4. Log every sample. A parameter log from day one is the only evidence you will have later when a batch shifts.

In-line monitoring and quality verification

Closed-loop monitoring is what separates a prototype from a validated production process. The options worth evaluating are back-reflection monitoring of the melt pool, melt-pool temperature sensing, pyrometry and acoustic emission. TWI’s programme on real-time temperature monitoring for polymer welding addresses exactly this class of problem, where a small shift in absorption or part geometry would otherwise push a good weld out of the window.

Verification should combine two streams. Non-destructive, on every part: dye-penetrant and helium leak testing for hermetic joints. Destructive, on a defined sampling rate: cross-sections and micro-slicing to check melt penetration, plus tensile, shear and peel testing, and burst and hoop tests for pressure-containing parts. Fix the sampling rate in the control plan and adjust it only through documented change control.

Common PES Laser Welding Failures and How to Fix Them

There is no comfortable margin in PES laser welding. The usable process window sits between a cold joint that peels apart in your hand and a burnt, brittle weld that smells of decomposition. Almost every production headache traces back to three culprits: energy density, fit-up, or contamination. The good news is that most failures announce themselves clearly, and most are recoverable without scrapping the part.

Symptom	Most likely cause	First thing to check
No bond at all	Energy below threshold, gap too wide, absorber misplaced, focus mis-set	Cut a cross-section
Intermittent welds	Mould variation, release agent, part warp, loose fixture	Gap at every station
Yellowing or browning	Excess energy, slow scan, long hold time, contamination	Reduce power or speed up
Bubbles or porosity	Moisture in or on the joint, trapped air, degradation gases	Dry parts and joint face
Flash or molten expulsion	Too much energy plus wide gap and heavy clamp force	Back off energy density
Weak or brittle joint	Under-melt, contamination, degraded polymer, excess reinforcement	Cross-section the joint
Warpage	Unbalanced heating, rigid fixture, non-simultaneous contour	Add fixture compliance
Leak at corner or start point	Clamp variation, poor seam termination, mask versus scanning	Leak test in production setup

No bond at all

The classic cold joint. Energy density may simply be under threshold, but check the physical causes too: a joint gap wider than about 150 micrometres, the transparent member sitting in the beam path when it should be off-axis, a missing absorber or one on the wrong face, focus set for a different part height, or a top component too thick for the chosen wavelength. Poor fit-up and insufficient clamping pressure cut heat conduction into the upper workpiece and limit interdiffusion of polymer chains, which is exactly what a strong PES bond depends on.

Intermittent welds

Intermittent failures usually mean the process sits on the edge of the window, not that it is broken. Usual culprits are mould-to-mould variation in part dimensions, a film of mould release agent or fingerprint oil on the joint face, a gap that changes because the part is warped, a fixture that loosens over a shift, or focus drifting as the tool heats up. Log the gap at every station rather than trusting the setup sheet.

Yellowing, browning and charring

PES is aromatic, so it browns and degrades before it flows if you give it too much heat for too long. Excess energy density, a scan that is too slow, long hold time after the beam switches off, surface contamination, or a hot shop that soaks the parts before welding will all push the polymer into degradation. Darkening at the weld line is the earliest warning that you are burning chemistry rather than melting polymer.

Bubbles and porosity

PES is hygroscopic, so moisture in the resin or on the joint face boils off during the weld and releases on cool-down. Trapped air at the interface and degradation gases produce the same symptom. Dry components before welding, keep the joint face free of condensation, and check the melt pool temperature profile rather than guessing.

Flash and molten expulsion

Flash at the seam edge means energy density is too high, especially when paired with excessive clamp force and a wide seam gap. The polymer is being squeezed out rather than fused. A degraded melt behaves the same way: material that has already browned will spit. Lower the energy, and check whether your aluminum MIG welding habits are creeping in here, since arc processes tolerate heat soak and clamping that plastics do not.

Low or brittle joint strength

If the weld peels apart, look for insufficient melt, contamination trapped at the interface, polymer degraded during the cycle, or excessive reinforcement. Glass fibre content above roughly 40 percent tends to give brittle laser welds. A subtler issue is absorber loading: at 0.1 to 0.15 percent carbon black you get partial transmission and subsurface heating, which builds a thicker melt layer without charring the surface. At higher loadings, absorption concentrates near the surface and you get thin, weak layers.

Warpage and residual stress

Contour heating on bulky components generates stress because the weld contour is not heated simultaneously, and TRUMPF flags this as the biggest problem with the process. Rigid fixtures make it worse. Elastomeric elements, rollers and sliding shoes let the part expand as it heats.

Leak paths at corners and start-stop points

Leaks concentrate where the seam changes direction or where the beam starts and stops, because the melt pool does not get the same dwell time there. Clamping variation at a corner changes the gap mid-weld. Mask-based systems hold the seam geometry but freeze one energy profile along the whole contour, while scanning systems vary the dwell to suit the path. Test with the exact production geometry.

Edge cracking

Cracks at the edge of a moulded part usually mean stress concentration plus a degraded or under-melted local region. Increase the edge radius, confirm the beam is not clipping the edge, and rule out contaminated regrind material in the joint face.

A diagnostic sequence that works

1. Cut a cross-section first. Every other step is guesswork without it.
2. Look at the melt pool. No melt, a thin melt, a wide melt or a charred melt each point to a different cause.
3. Check energy density against your process window.
4. Check fixture and geometry: gap, clamping force, part flatness.
5. Check absorber and optics: loading, placement, focus, window cleanliness.
6. Check material condition: moisture, contamination, regrind content, degradation history.

Fume, and why monitoring matters

Thermal degradation of an aromatic high-temperature polymer such as PES releases decomposition products, so local exhaust ventilation at the cell is the control, not general room ventilation. There is no arc and no spatter as there would be in a process such as alpha laser welding, but the smoke is still a real exposure and a quality signal. Charring you can smell at the cell is polymer you are destroying in the part.

This is where closed-loop monitoring earns its keep. TWI’s programme on real-time temperature monitoring for polymer welding targeted exactly this class of problem, joining difficult-to-weld polymers by controlling interface temperatures rather than relying on a fixed power setting. Back-reflection monitoring and pyrometry can catch a cold joint before it leaves the cell.

PES Laser Welding Compared With Other Joining Methods

Choosing a joining method for polyethersulfone is rarely about which weld is strongest. It is about which process fits the part, the volume, the tolerance and the leak requirement without creating a downstream problem. PES laser welding fits more cases than fabricators expect, but it is not for everything and rarely the cheapest.

Method	Strength vs parent material	Cycle time	Cosmetics and heat effect	Leak tight and dissimilar materials	Capital, consumables and best fit
Laser welding (TTLW)	Approx. base material, optimised joints	30 m/min and above, 1 to 3 s cycles	Clean seam, no flash; very localised heat	Excellent; dissimilar limited	High capital, no consumables, absorber needed; thin, clear, hermetic parts
Ultrasonic welding	High, often near base material	Fraction of a second to a few seconds	Can mark or stress-whiten; cracking risk	Good, gasket dependent; dissimilar limited	Moderate capital, per-part tooling; high-volume flat flanges
Vibration / hot-friction welding	High	Several seconds	Fine seam, flash possible; bulk joint heat	Good; dissimilar limited	Moderate to high capital; large parts, sealing rings
Hot-plate welding	High	Long, tens of seconds	Visible flash or fin; whole part heated, distortion risk	Good; poor with dissimilar polymers	Low capital, fixture wear; large housings, low volume
Hot-gas / extrusion welding	Lower to moderate, filler dependent	Slow, continuous	Visible bead; continuous heat into the part	Moderate; good with compatible filler	Low to moderate capital, filler rod used; sheet and structural work
Resistance / spiral welding	High in the seal zone	Fast	Clean sealing face; very localised heat	Excellent for seals; dissimilar poor	Moderate capital, wire or electrodes; containers, filters, fluid paths
Adhesive bonding	Well below parent material	Slow, cure time included	Visible bond line; no heat but long dwell	Good if designed for it; dissimilar good	Low capital, ongoing adhesive cost; stress relief, low volume
Solvent bonding	Moderate, bond-line dependent	Slow plus drying	May cloud or distort; no heat	Variable; poor for PES with most solvents	Low capital, solvent handling; acrylic and PS parts, rarely PES
Mechanical fastening / overmoulding	Not a weld-strength joint	Fast	Visible parting line; no heat or full part heat	Seal geometry dependent; dissimilar excellent	Mould cost, no consumables; snap fits, medical disposables

Where laser wins outright

Much of the advantage is what laser welding does not do. It is contactless, so there is no flash, no plastic flakes and no adhesive residue. In cleanroom and medical assembly that matters more than the numbers suggest, since flakes are a particulate problem and adhesive outgassing is a validation headache. TWI’s polymer welding research makes the same point, noting that adhesive, vibration and ultrasonic welding can deposit particles where laser welding does not.

Heat input is severely limited locally, so delicate electronics survive. A pump controller with a PCB ten millimetres from the seam can be assembled without cooking it, while hot-plate welding heats the whole component.

Speed closes the case: processing speeds of at least 30 m/min, cycle times of 1 to 3 seconds and no consumables to manage. TRUMPF notes that seam strength in optimised joints corresponds approximately to the base material, a best-case figure. Scanners and robots also follow three-dimensional seams no ultrasonic horn or flat hot plate can reach.

Where laser falls short

The entry cost is high and it sits on the line for years. Every joint also needs an absorber strategy: carbon black in the bulk, an infrared dye, a coated layer or an intermediate film, an extra design and validation task adhesive bonding does not have.

Geometry must be under control. Gap, flatness and part variation become your problem rather than the process’s, because the process does not tolerate them. And joining genuinely incompatible polymers is no easier with a laser than with any other plastics welding technique, the conclusion in TWI’s polymer welding research.

The comparison metal readers will care about

A metal laser weld is a keyhole fusion weld: it melts, optionally adds filler, and leaves a metallurgical heat-affected zone that can be inspected and qualified against a code. A PES laser weld is a diffusion bond across a molten polymer interface, with no keyhole, no filler wire and no metallurgy. Strength depends on interdiffusion of polymer chains and on how clean the molten surfaces were. If you came from alpha laser welding systems or TIG work, swap that mental model before setting parameters.

A decision rule of thumb

If the joint is a simple flat flange on a high-volume part, ultrasonic welding is usually cheaper: lower capital, faster cycle, no absorber.

If the part is clear, thin, hermetic or holds sensitive electronics, laser is usually right. So is the case for a cosmetic joint with no flash, an automated line with no consumables or a three-dimensional seam.

And if you are joining materials that do not want to be joined, no plastic welding process will save you. Adhesive bonding or a mechanical redesign is the honest answer. For the physics behind the laser route, see how optically clear plastics are laser welded in practice.

Safety, Standards and Validation for PES Laser Welding

Welding a clear polymer inside a medical device is regulated in a way that hot-plate welding a plastic tray is not. Three drivers: a Class 4 laser beam, a hot tool, and an aromatic high-temperature polymer that decomposes rather than simply melts. Get all three under control and PES welding becomes one of the cleanest joining operations you can run in a regulated plant.

Class 4 laser safety: the controls that do the actual work

Industrial polymer-welding sources, whether diode, fibre or mid-infrared, are almost always Class 4: the process needs high peak power delivered to a spot you cannot see. A 1064 nm fibre source is invisible, and near-infrared diodes look deceptively dim, which is what makes them dangerous. Read our guide to laser welding systems and laser safety before the first power-on.

Eye protection with the correct optical density for the wavelength is essential during alignment, but it is the last line of defence. No goggles protect against a Class 4 beam at full power. The primary controls are engineered:

• Enclosure. An interlocked chamber containing the beam, fixtures and parts.
• Interlocks and key switch. Opening the door removes power; the key switch prevents accidental operation.
• Beam dump and beam stop. A rated termination so the beam never ends on a reflective fixture.
• Controlled area, signage and training. Trained operators only, with warning labels and a documented procedure.

Two consensus standards carry most of the weight. ANSI Z136.1 covers safe use of lasers in the workplace: training, procedures and area management. IEC 60825-1 covers laser classification and safety requirements for laser equipment. In the US, OSHA laser safety requirements for industrial exposure sit on top of that. Builders usually supply a compliant cell, but the operating procedure, training records and area management stay yours to own.

Physical hazards specific to plastics welding

Stored energy in a transparent cover. A clamped glass or acrylic cover holds significant energy and does not fail gracefully if it fractures under load. Borosilicate glass tolerates welding thermal shock better than acrylic or plate glass, and quartz glass is used for high-temperature polymers such as PES. Any transparent cover needs eye protection, a fragment handling plan and a check on clamp load.

Heat. Tooling and parts come off the machine hot, even when the cycle is under two seconds. Guarding, a cool-down time and a hot-parts area are the sensible controls.

Fume. Fume is underestimated. PES is aromatic and does not evaporate cleanly; it degrades, releasing decomposition products that concern operator exposure and part quality. General room ventilation is not adequate; local exhaust captured at the weld and sized for the cell is.

Validating the welded article, not just the weld

A strong weld is the start. For a device that touches a patient or carries a fluid, you are validating a joint that must survive sterilization, shelf life and cycling.

• Biocompatibility and leachables. The absorber, dye or pigment that catches the beam is part of the device, not a consumable. Anything in patient-contact or fluid-path position needs its own evaluation, tested as the welded assembly.
• Sterile barrier and packaging integrity. PES membranes can be made with reproducible pore sizes down to 40 nanometres, which is why filtration and sterile barrier applications dominate PES use. Welding adds a failure mode to package validation.
• Accelerated ageing. Polysulfones suit medical applications needing autoclave and steam sterilization because of their high hydrolysis stability. Demonstrate that your weld survives that exposure and has not embrittled or crept.
• Process validation. Installation, operational and performance qualification on the production cell, not a good sample.

Where the qualification language comes from

TWI sits on BSI WEE/3 (Welding Thermoplastics), CEN TC249/WG16 (Thermoplastics Welding) and ISO TC61/SC11/WG14 (Plastic Joining), which produce the qualification and specification wording your quality system will be audited against. TWI also runs plastic welder training and qualification programmes covering process selection, joint design, procedure development and validation, the fastest route to competent staff for a material as unforgiving as PES. Its polymer welding technical pages and the through-transmission laser welding application study are the best public grounding available, while real-time temperature monitoring for polymer welding points at where the evidence is heading: measured, logged weld temperature instead of inferred quality.

A Practical Workflow for Taking PES Welding to Production

Most PES laser welding projects do not fail on the physics. They fail in the gap between the bench sample that worked and the line that has to make good parts every hour of every shift. This order of operations is what keeps that gap from opening.

Ten Steps From Coupon to Production Line

1. Define the joint requirements in numbers. Strength as a percentage of parent material, hermeticity or a pressure rating, cycle time, sterilisation exposure, cosmetic standard. A joint described as “strong enough” fails qualification.
2. Choose the geometry and put the absorber where the beam will find it. Design the flange, rebate and sealing surface, then decide whether the absorbing layer sits in the moulding, in a dye, or on the interface. This overview of laser welding methods for optically clear plastics is a useful sanity check on joint shapes.
3. Select the source from absorption data measured on your own parts. Transmission figures for “clear PES” on a data sheet will not match your part with its colour package, wall thickness and surface finish. Mid-infrared sources change the calculation entirely, because clear polymers absorb strongly above roughly 2.5 micrometres, as this mid-infrared polymer processing overview sets out.
4. Design the fixture around gap control. Hold the gap within about 150 micrometres, apply load only where the beam lands, and add a little compliance so rigid tooling does not distort a bulky part. Glass, borosilicate and quartz clamps let the beam through, but treat them as a stored-energy hazard: a transparent cover under load can release dangerously if it fractures.
5. Run a coupon matrix. Sweep power against speed on simple test bars, find the cold-joint limit and the degradation limit, then work in the middle. Control energy density, spot size and speed together, because heat conducted away from the joint changes the outcome even at identical nominal energy density. TWI member studies of transmission laser welding, including a thermal model for the process, are a useful reference point for building that window.
6. Build the quality plan before the cell is installed. Non-destructive leak testing, in-process monitoring and a defined destructive sampling rate. TWI’s current work on real-time temperature monitoring for polymer welding reflects why closed-loop control separates a prototype from a validated process.
7. Validate for the regulatory context. Accelerated ageing through sterilisation, biocompatibility and leachables for patient-contact parts, and documentation that the process, not just the first article, holds.
8. Train operators and write the parameters down. A short recipe card beats a memory. Keep a parameter log from the first shift onward.
9. Plan maintenance. Optics, chiller, focus position and beam alignment are consumables in everything but name.
10. Scale in stages. Prove the window on one part, then transfer it to the family.

Choosing the Equipment Setup

Setup	Best fit	Watch-outs
Single-head diode with clamping head	Simple flat flanges, short cycles, moderate volume	Fixed geometry only
Scanner-based quasi-simultaneous	Complex contours, varying seams	Higher capital cost, more optics to maintain
Robot-mounted fibre delivery	Three-dimensional parts and multiple seams	Programming time and guarding
Simultaneous illumination	Very high volume, simple geometry such as filter housings	Limited to parts that fit the illuminated area

One structural advantage is worth knowing before you spec a cell: because laser plastic welding melts only a thin film, flash-trap requirements largely disappear. Most PES jobs run without one, which simplifies layout and housekeeping. Broader context on these systems sits in our guide to alpha laser welding systems, uses and safety.

The Economics, Honestly

The case for laser is straightforward: no consumables, no fixture wear, fast cycles and no rework from flash or adhesive residue. Against that sits a high entry cost and a process that punishes sloppy optical maintenance. Compare it with ultrasonic and hot-plate alternatives using your own numbers: cycle time per part, labour minutes loaded, scrap and rework rate, changeover time. Vendor claims about cost per part are marketing, so build the model yourself and stress it with your worst-case yield.

Four Tips Worth More Than Most of the Advice Above

• Start with a simple straight test bar, not a production part.
• Keep the parameter log from day one, including rejects.
• Measure the near-infrared transmission of your actual parts if you can.
• Treat the joint face finish as a designed feature, not a moulding afterthought.

Conclusion

PES laser welding is a clean, fast, hermetic joining route for a demanding high-temperature polymer, and it works when you control three things: absorption, gap and clamping. Get those right and the joint matches the parent material, the seal holds, and the part looks untouched.

It is also the wrong choice in a few honest situations. Thick sections beyond a few millimetres outrun what the beam can do evenly. High-volume simple joints are usually cheaper on ultrasonic. Incompatible polymer pairs will not bond, whatever the absorber does. And a team without the capital or the process discipline to run a laser cell will not get a hermetic joint out of one, no matter how good the sample was.

Frequently Asked Questions

What does PES mean in laser welding?

In this context PES means polyethersulfone, a rigid, amorphous, high-temperature thermoplastic also known as PESU. It belongs to the polysulfone family with PSU and PPSU. Laser welding of polyethersulfone joins parts by melting a thin film at the interface, usually with the beam passing through a transparent upper component and being absorbed at the joint.

Do you need carbon black to laser weld PES?

In most cases, yes. PES transmits much of the near-infrared band, so an unmodified joint usually will not melt at all. The usual solutions are a small carbon black loading in the moulded part (roughly 0.1 to 1 percent by weight), an infrared-absorbing dye, or a coated or printed absorbing layer at the interface. Longer wavelengths around 1470 nm and mid-infrared sources near 3 micrometres can reduce or eliminate the need for additives, and that must be verified by coupon trials on your own parts.

What laser wavelength works best for polyethersulfone?

It depends on the grade, colour, filler content and thickness, so it has to be measured rather than assumed. 808-980 nm diode and 1064 nm fibre systems need an absorber at the joint, while 1470 nm, 1940 nm and mid-infrared sources around 3 micrometres are absorbed more strongly by clear polymers and can weld without additives. Run a coupon matrix at several wavelengths before committing to capital equipment.

How strong is a laser-welded PES joint compared with the parent material?

In optimised homogeneous joints, suppliers report seam strength close to that of the base material. In practice the strength is governed by the melt film, surface cleanliness, clamping pressure and whether the polymer degraded, so dissimilar or reinforced assemblies commonly land lower than the parent material. Always qualify with tensile, shear or peel testing on your own geometry.

What gap should a PES laser joint have?

As close to zero as you can hold it, and ideally under 150 micrometres. A wider gap lets heat conduct away before the interface melts and prevents intimate contact between the two surfaces, which is where the strength of the joint is created. That is why fixture and tolerance stack design matters more than any single laser parameter.

What are the most common causes of a failed PES laser weld?

In order of frequency: energy density below threshold, a gap or surface mismatch, the absorber on the wrong component or missing, focus set incorrectly, and contamination on the joint face. Yellowing or charring points to the opposite problem – too much energy, a scan that is too slow, or too long a hold time. Cutting a cross-section is the fastest way to tell which failure you actually have.

Is laser welding of PES safe for operators?

It can be, with the right controls. Industrial plastic-welding sources are typically Class 4 lasers, so the primary safeguards are enclosure and interlocks, key control, beam dumps and stops, and eyewear with the correct optical density, in line with ANSI Z136.1 and IEC 60825-1 practice. Add local exhaust ventilation because thermally degraded PES releases decomposition products, and use fragmentation-resistant clamping glass where a cover is under load.

How does PES laser welding compare with ultrasonic welding?

Ultrasonic welding is usually cheaper for simple flat flange joints in high volume and does not need an absorber, but it cannot weld many clear, thin or very rigid parts and it leaves flash or a visible joint line. Laser welding costs more in capital and requires careful fixture design, yet it is residue-free, produces a clean cosmetic seam, welds delicate electronics without damaging them, and handles three-dimensional contours and hermetic sealing far better.

Can PES be laser welded to other plastics?

Same-material and similar-polymer pairs weld most reliably. Joining PES to something like polycarbonate, ABS or PP can work but is limited by thermal compatibility and crystallisation behaviour, and research from TWI confirms laser welding carries the same dissimilar-material limitations as other plastics welding processes. An intermediate film with an absorber or adhesion promoter is often the practical route when the pair must be joined.

When is laser welding the wrong choice for PES?

Avoid it for thick sections beyond a few millimetres, for high-volume simple joints where ultrasonic will cost far less, for polymer pairs that do not interdiffuse, and for teams without the capital and the optical maintenance discipline to run a laser cell. It is also the wrong tool if you need a full-strength structural load path, since a polymer weld is a diffusion bond rather than a fusion weld.