EOS PA 640-GSL Nylon 12, Glass Bead Filled, Carbon Reinforced is a dark-coloured polyamide 12 powder formulated for powder bed fusion–laser beam (PBF-LB/P) platforms. The material combines spherical glass bead filler with carbon reinforcement in a nylon 12 matrix to yield a grade that is stiffer and more dimensionally stable than unfilled EOS PA 2200 while retaining the hydrocarbon and oil resistance associated with polyamide 12. Supplier data sheets report mechanical values under standard test methods including ISO 527-1/-2, ISO 178, ISO 179-1/1eA, and ISO 75-1/-2. The glass bead phase is included to reduce anisotropic shrinkage and improve powder bed flow, while the carbon reinforcement contributes higher tensile and flexural modulus, darker surface appearance, and more uniform heat distribution during laser scanning. The material is used for functional jigs, fixtures, housings, and underhood brackets where unfilled nylon 12 grades exhibit excessive creep or heat sag at elevated service temperatures.

The combination of glass beads and carbon reinforcement alters the crystallisation behaviour of the nylon 12 matrix. Glass beads act as hard nucleation sites that promote finer crystal domains, and the carbon phase increases the effective thermal conductivity of the powder bed. In laser-sintered parts, this can reduce the Z-direction property drop caused by interlayer melting defects. However, the benefits are condition-dependent: below the correct energy density, sidewall porosity increases because the filled melt cannot fully wet adjacent particles; above the optimum, the polyamide matrix begins to degrade, producing resin-coloured fumes and reducing Charpy impact values. Production parameter sets therefore differ from PA 2200 and PA 3200 GF and should not be interchanged.

Why Powder Feed and Recoater Dynamics Shift with Combined Glass Bead and Carbon Fillers

On EOS P 396 and FORMIGA P 110 systems, the operating window for PA 640-GSL is approximately 172–176°C chamber temperature with a 0.12 mm layer thickness. The EOS P 396 uses a 70 W CO₂ laser; the scan speed and fill spacing are controlled by the material parameter set rather than manually entered. Carbon-filled powder surfaces absorb more laser energy than unfilled nylon 12, so laser power is reduced or scan spacing is widened to avoid over-sintering. Build chamber heating must be uniform because carbon-filled powder retains more heat in thick sections; users report warpage at the transition from dense cross-sections to thin walls when chamber temperatures are below the set tolerance.

The carbon reinforcement increases absorptivity in the 10.6 µm CO₂ wavelength, which improves energy coupling but can produce local overheated zones if scan spacing is too tight. Thermal runaway in the powder bed appears as dark brown or black patches and a sharp reduction in Charpy impact strength. On P 396 systems, the operator can detect this by monitoring part density coupons and by comparing sleeve weight against part volume. The appropriate response is to reduce fill laser power or increase scan spacing within the parameter set, not to alter chamber temperature. Incorrect chamber temperature often produces curl at the part base, which is a different failure mode.

Moisture control is critical. Powder from open containers at ambient relative humidity above 60 % absorbs moisture on the particle surface. The recommended drying condition is 80°C for 12–24 h before use. Drying at higher temperatures or longer times may oxidise the carbon surface and shift the melt viscosity. Recycled powder must be sieved through a 150 µm mesh, blended with virgin powder at a refresh ratio of 30–50 %, and sampled for particle size distribution. The combined filler system can segregate during extended recoating; glass beads tend to flow more readily, while carbon-rich fines remain in the bed. This produces local changes in mechanical properties and is detectable as black streaking on part surfaces.

Powder flowability for PA 640-GSL is not fully captured by Hausner ratio alone. The glass beads increase bulk density and reduce interparticle friction; carbon reinforcement decreases the effective flowability at high recycle ratios. Production lines can use a granulation drum or avalanche angle tester to compare fresh blends against a validated baseline. A Hausner ratio below 1.25 is often used as a pass criterion for SLS polyamide blends, but the exact value must be established on the supplier’s powder tonnage.

Recoater blade wear is a more visible failure mode on hardened metal recoater blades than with unfilled PA 2200. Glass beads are harder than polyamide, and carbon reinforcement increases the abrasive character of the powder bed. Maintenance intervals on P 396 systems may need to be shortened when PA 640-GSL is run continuously. The recoater speed is sometimes reduced to 120–150 mm/s to prevent powder dragging and short feeds. The build bed density is lower for carbon-reinforced powders, so the feed chamber setting may need a higher refill depth to maintain uniform layer density.

Tensile Modulus, Heat Deflection, and Moisture Uptake Under Standardized Test Protocols

Mechanical performance for PA 640-GSL is generally reported on laser-sintered test specimens conditioned at 23°C and 50 % relative humidity according to ISO 291. The following table summarises the typical literature and supplier datasheet ranges for this filled grade.

Because PBF-LB/P parts are anisotropic, values measured in the XY build plane are higher than those measured in the Z orientation. Internal studies and published property sheets frequently report Z-direction tensile elongation at 60–80 % of XY values, with Charpy impact showing a larger directional drop. Published data for specific Z-direction fracture toughness of PA 640-GSL are limited; therefore, any design using out-of-plane mechanical loading should use production-run coupons and failure analysis rather than supplier datasheet values alone. Dimensional tolerance capability for well-designed parts is often cited as ±0.3 mm for features below 100 mm, but this must be established on the exact machine and powder blend.

When comparing XY and Z specimens on a P 396, the ratio of Z tensile modulus to XY tensile modulus is commonly 0.85–0.95; the ratio of elongation at break is lower and machine-dependent. The precise anisotropy is not a fixed material constant because it changes with layer thickness and energy density. Thin layers of 0.10 mm improve interlayer bonding but increase build time; thicker layers of 0.15 mm reduce build time but can lower Z-direction tensile strength. Published data for PA 640-GSL at 0.15 mm layer thickness are limited, so the standard 0.12 mm parameter set remains the most documented baseline.

Heat deflection testing applies a three-point flexural stress at 1.80 MPa according to ISO 75-1/-2. The resulting HDT value is a comparative ranking tool, not a maximum continuous use temperature. Long-term exposure at temperatures slightly below HDT can still produce creep in the nylon 12 matrix. The glass bead and carbon fillers reduce creep strain but do not convert the polymer into a thermoset. Applications that require continuous load at 120°C should not be specified with this material without mandatory creep-rupture testing.

Water uptake after saturation is lower than unfilled PA 2200 because the filler fraction reduces the mass of hygroscopic polyamide. Nonetheless, conditioned moisture equilibrium at 50 % RH is approximately 0.7–1.0 wt%. This water uptake reduces tensile modulus by a few percent and increases elongation at break slightly. Dimensional changes from moisture absorption must be considered for parts used in high-humidity enclosures or outdoor equipment. The glass bead and carbon fillers reduce the absolute linear expansion due to moisture but do not eliminate it.

When Dimensional Stability and Stiffness Outweigh Unfilled PA 2200 Ductility

Compared with EOS PA 2200, PA 640-GSL is specified when the loading mode is flexural or compressive and when the part must maintain flatness under a thermal gradient. The tensile modulus of PA 640-GSL is roughly 70–90 % higher than that of unfilled PA 2200, while elongation at break is lower by approximately 50–70 %. This is a deliberate trade-off. Unfilled nylon 12 remains the better candidate for snap-fit clips, living hinges, and parts that require repeated high-strain flexing. PA 640-GSL is not suitable for those applications because the filled matrix fractures earlier and cannot absorb the same elastic energy.

Against PA 3200 GF, the addition of carbon reinforcement in PA 640-GSL improves stiffness and reduces the coefficient of linear thermal expansion in the build plane. The linear CTE of PA 640-GSL is typically in the range of 60–80 µm/m·°C, whereas unfilled PA 2200 is closer to 110–120 µm/m·°C. This matters in assemblies with metal inserts or in parts that must remain stable across a 20–80°C service cycle. The carbon also provides a dark surface that hides some post-process wear marks, but it makes colour dyeing impractical; PA 640-GSL is supplied dark grey to black and is usually left in its natural laser-sintered state or painted rather than dyed.

Chemical exposure follows the nylon 12 matrix. The grade resists many oils, greases, fuels, and aliphatic hydrocarbons. It should not be exposed to strong acids, phenols, boiling water, or strong oxidising agents. Filled grades can be more susceptible to environmental stress cracking at filler–matrix boundaries when exposed to aggressive high-polarity fluids. Any chemical compatibility claim should be validated under ISO 22088-3 or an application-specific stress cracking test because the glass beads and carbon reinforcement introduce interfaces that can accelerate solvent attack.

Surface finishing options for PA 640-GSL are more limited than for unfilled PA 2200. The exposed glass bead filler creates a matte, slightly gritty surface that can be improved by bead blasting or vibratory finishing. Machining of as-built surfaces may cause filler pull-out unless sharp tooling and low feed rates are used. Sealing with a low-viscosity epoxy or cyanoacrylate is possible; the high modulus of the filled substrate reduces deflection under load, but the sealant must withstand the same thermal environment. Dimensional calibration is sometimes performed after a thermal conditioning cycle at 80°C for 2 h to relieve residual stresses before final machining.

Drying at 80°C Does Not Compensate for Segregated Recycled Powder

Production-scale experience indicates that severe property drift occurs when recycled PA 640-GSL is returned to the feed silo without blend ratio control. The glass bead and carbon phases do not pack uniformly, and their distribution changes after multiple layers of recoating. A powder management procedure should include loss-on-ignition measurement to track filler content, bulk density checks, and sieve residue checks. Dry powder can still produce weak parts if the filler fraction has dropped below the specification band. Pre-drying at 80°C addresses only moisture; it does not restore glass bead spacing or redisperse carbon-rich fines.

Storage should follow the same rules as other nylon 12 powders: sealed plastic containers or nitrogen-purged silos, no direct sunlight, and no introduction of incompatible polymer powders. Nylon 12 powders should not be mixed with PA11 or PA6 powders because melt viscosity differences create local delamination. Any recovered powder that has been exposed to high humidity above 70 % RH should be isolated, dried once, and tested before returning to production. The drying oven must not have residual lubricants or plastisol vapours because nylon 12 absorbs low-molecular-weight organic compounds that later volatilise in the laser chamber.

For compliance documentation, request current statements from EOS or the material distributor. Standard commercial PA 640-GSL may be covered under REACH and RoHS supplier declarations, but aerospace or medical certifications are lot-specific and require additional traceability. The grade is not automatically equivalent to food-contact nylon 12; it should not be specified for applications requiring FDA 21 CFR 177.1500 without formal confirmation. For any application involving high-voltage insulation or sensitive electronics, the electrical behaviour of the carbon phase must be tested at the specific wall thickness and humidity.

On automotive manufacturing lines, PA 640-GSL is used where unfilled PA 2200 jigs creep at paint-oven temperatures. It appears in robotic gripper inserts that must maintain clamping force at temperatures where unfilled nylon 12 softens. In industrial equipment, carbon-reinforced glass bead PA12 is specified for locating fixtures, inspection gauges, and assembly nests that require high bending modulus and dimensional stability. Typical low-volume production parts include underhood brackets, ducting with thick flanges, sensor mounts, and automation components where metal replacement is constrained by weight. The material is less suitable for high-impact housings, snap-fit closures, or parts that require a bright dye finish. For any new design, the production process should include a minimum of three tensile specimens built in XY and Z orientations per build, because the property range across machines is wider for filled PA12 than for unfilled PA12. No universal substitute for process validation exists when the application lies near the 85°C HDT boundary.