Polybutylene Terephthalate in Relays: Why an Insulation Housing Is a Safety Part

Polybutylene terephthalate (PBT) is selected for relay insulation housings because the moulded housing functions as both the mechanical carrier and the electrical separation system for live contact elements. A relay housing is not an inert enclosure; it is a safety part under IEC 61810-1, IEC 60664-1, and the end-product standards that invoke them. The component must retain metallic contact springs and terminal pins within tolerance while preventing leakage current, surface tracking, and arc flash between circuits. When the housing loses dimensional stability, a contact gap can change. When the polymer carbonises under electrical stress, the creepage path becomes a conductive track. These failure modes place PBT in a different technical class from general enclosures because the insulation housing is part of the relay’s clearance and creepage system, not merely packaging around it.

Material selection therefore begins with dielectric and thermal requirements rather than tensile strength. The polymer must tolerate the continuous heat of an energised coil, repeated contact-breaking arcs, soldering heat during PCB mounting, and high-humidity exposure without allowing creepage distances to shrink or insulation resistance to collapse. IEC 60664-1 links creepage allocation to the comparative tracking index measured according to IEC 60112. For relay housings, this means the selected PBT grade must be characterised for CTI, dielectric strength, flammability, and thermal endurance in the exact wall thickness and colour used in production, because fillers, flame retardants, and pigments alter each property independently.

What Causes an Insulation Housing to Function as a Safety Part?

The safety function of a relay housing appears only when the relay operates under fault conditions. During contact breaking, an electric arc forms between the separating contacts. The housing must not support combustion from arc spatter; it must also resist the deposition of vaporised metal and carbonaceous particles that can reduce surface resistivity. In miniature power relays, the contact gap is small, and the housing wall is often the only barrier between a live contact spring and a metallic bracket, an adjacent circuit trace, or an accessible conductive part. A crack at a terminal insert, a warp-induced change in air gap, or a carbon track along the housing surface can convert a switching circuit into an ignition source.

The relay standard IEC 61810-1 defines dielectric withstand tests between coil and contact circuits, between contact sets, and across open contacts. The housing must not flash over or allow breakdown during these tests. In addition, IEC 60664-1 assigns material groups based on CTI: group I is 600 V or greater, group II is 400 V to 599 V, group IIIa is 175 V to 399 V, and group IIIb is 100 V to 174 V. Many flame-retarded glass-reinforced PBT grades fall into group IIIa, while unfilled PBT grades can achieve group II in some formulations. The practical effect is that a flame-retardant package may lower CTI enough to require a larger creepage distance for the same working voltage, and the housing geometry must be designed around the final material grade, not around generic PBT data.

Comparative tracking data for PBT grades is obtained under IEC 60112 using a platinum electrode test on the moulded surface, not on raw resin. The test result is sensitive to surface finish, colour, moisture conditioning, and the distribution of glass fibres. For this reason, supplier datasheet values for CTI are not transferable from one colour to another. A relay manufacturer is required to verify CTI on production-representative parts because a flame-retardant package may produce a carbonaceous residue under arcing that accelerates tracking even if the unfilled resin shows a higher value. This is an operational boundary: PBT should not be used in high-pollution outdoor relay applications where continuous surface contamination and wetting are expected unless well-sealed or used with larger creepage margins.

These ranges are drawn from supplier technical datasheets and UL Yellow Card listings; the exact values for a specific relay grade must be taken from the production lot because flame-retardant loading and glass content shift each row. The table confirms a central trade-off: glass reinforcement raises the heat deflection temperature and strength but reduces dielectric strength and CTI. Relay housings with thin walls and tightly spaced terminals therefore cannot simply select the stiffest PBT grade; the grade must be balanced against tracking and breakdown requirements.

Arc resistance is a separate property from CTI. ASTM D495 measures the time a material withstands high-voltage low-current arc on the surface. PBT is not inherently an arc-resistant thermoset; glass-filled flame-retarded grades can show shorter arcs than some phenolic or melamine materials. In sealed relays, arc extinction is controlled by the switching atmosphere and contact design rather than the housing alone, but the housing still receives ejected metal. If the housing is too close to the contact gap, the arc can erode the polymer surface and deposit conductive film. Supplier datasheets for PBT relay grades rarely publish arc-resistance values for finished housings; published data for this specific configuration is limited. A safe design therefore does not rely on PBT arc resistance as a primary protective layer; it uses contact-gap shielding, sufficient wall thickness, and arc suppression circuits where needed.

Thermal ageing limits, contact-force retention, and material substitution thresholds

Contact pressure in a relay is generated by a preloaded moving spring, and the housing must keep the spring anchor points fixed. If the polymer loses stiffness at elevated temperature, contact force decays, contact resistance increases, and the relay can fail in service despite a mechanically intact housing. For PBT, the heat deflection temperature of unreinforced grades is only 50–60 °C at 1.82 MPa, which is insufficient for a relay carrying contact and coil heat. A 30% glass-reinforced grade moves HDT/A to 200–220 °C, but this does not mean the material is suitable for continuous exposure at that temperature. Long-term ageing is governed by UL 746B relative thermal index values. Many 30% glass-reinforced PBT grades carry an RTI electrical of 130 °C and an RTI mechanical-with-impact of 120 °C. For a relay specified to 125 °C continuous operation, the margin is therefore narrow; for a coil surface exceeding 135 °C, grade substitution to a higher-temperature polymer such as PPS or PEI is required unless forced cooling or lower coil power is applied.

Thermal ageing also affects the physical integrity of terminal insert areas. PBT can embrittle when aged above its RTI for extended periods, and the resulting microcracks around moulded-in metal pins reduce the dielectric separation path. A mechanical impact during handling or a relay drop can then open a crack that becomes a moisture retaining path. In production, this is not only a resin issue: moulding conditions that create high residual stress around terminals will accelerate the same failure. The safety limitation is therefore not a single temperature number but the combination of continuous coil temperature, hot-spot temperature at the contact, ambient operating temperature, and the stress state produced in the moulded part.

Moisture uptake in a relay housing is not a cosmetic variable. A polymer that swells in humid environments changes the physical distance between terminal pins and slightly alters the effective creepage surface. PBT shows an equilibrium moisture uptake of about 0.2 % at 23 °C and 50 % relative humidity, roughly an order of magnitude below polyamide 66, which can take up 2.5 % under the same conditions. This low uptake preserves moulded dimensions and prevents the insulation resistance from collapsing in moist environments. For relay applications where the housing is exposed to condensing humidity after soldering, PBT can maintain a stable creepage geometry more reliably than polyamide grades, but it is not hydrophobic enough for continuous condensation without surface sealing. The material carries its own wet limitation: prolonged hot-water exposure or steam sterilisation is outside the practical operating envelope for standard PBT relay housings.

When the Moulding Process Determines Dielectric Safety Instead of the Resin Supplier

The safety properties of PBT exist only if the resin is processed inside a narrow thermal and moisture window. PBT is a polyester and undergoes hydrolysis when melt-processed with moisture. Before injection moulding, pellets must be dried with a closed-loop desiccant dryer to a moisture content at or below 0.02 %, commonly achieved at 120 °C for 4 h with a dew point of -30 °C or lower. Drying at higher temperatures above 140 °C risks pellet clumping and discolouration. If the material enters the barrel above the moisture limit, the melt degrades in terms of molecular weight, producing short shots, low weld strength, and a fall in elongation at break. A housing moulded from hydrolysed PBT may still meet a dimension check but fail the insulation withstand test after thermal cycling because microcracks form at reduced molecular weight.

Melt temperature for glass-reinforced PBT is typically set between 250 °C and 270 °C. Barrel residence time should be kept below 8–10 min at these settings, with shorter residence at the upper limit. Mould temperature is normally 80–100 °C for 30% glass-reinforced grades to allow the PBT to crystallise fully. A low mould temperature can save cycle time but leaves underdeveloped crystallites and increases post-mould shrinkage, which changes terminal pin orientation after assembly. On a multi-cavity relay base tool, this produces a safety-relevant drift: the nominal creepage distance measured immediately after moulding is not the same as the creepage distance after the part has equilibrated overnight. Post-mould shrinkage of unfilled PBT is about 1.5–2.0 %, while glass-reinforced PBT shrinks substantially less but anisotropically, often below 0.5 % in the flow direction and up to 1.0 % transverse. Tool design must compensate for this anisotropy around terminal pins or the final part may violate the minimum creepage distance.

Compounding and moulding also influence tracking and arc resistance. Glass fibres near the surface can create local field concentrations. Weld lines formed around moulded-in metal pins are structurally and dielectrically distinct from the bulk material because the fibres orient parallel to the melt front rather than across it. Notched impact strength measured by ISO 179-1/1eA at a weld line may be less than half the unknit value. If the weld line sits between two live terminals, it can become the lowest-energy pathway for breakdown or crack propagation. For this reason, production tools for relay housings position gates so that weld lines are moved away from high-voltage gaps, and cavity-pressure sensors are used to confirm that the melt reaches the last-filling regions before freeze-off. Published data for weld-line-specific CTI or dielectric strength of PBT relay housings is limited; verification on production-representative parts is required.

Production-scale relay housings are commonly moulded from flame-retarded PBT in multi-cavity tools with clamp force from 350 kN to 1000 kN depending on cavity count and flow length. The compound is generally produced on a co-rotating twin-screw extruder with an L/D ratio of 40:1 or greater; glass fibre is side-fed downstream to limit fibre attrition. This process affects batch-to-batch variation in fibre length distribution, which in turn shifts mechanical and electrical properties across lots. Incoming material for safety housings should be checked for melt volume-flow rate under ISO 1133-1:2022, moisture content, and UL Yellow Card recognition for the exact colour. Without those controls, a nominally identical PBT grade can deliver different creepage and clearance performance after moulding.

Flame retardancy is evaluated according to UL 94. Flame-retarded PBT relay grades are commonly available with V-0 ratings at 0.75 mm or 1.5 mm thickness; some grades achieve V-0 at 0.4 mm, but the thinnest wall ratings require careful mould temperature control because fast cooling can create a flame-retardant-rich skin with a less flame-resistant core. For relay housings, the relevant flammability test is not a raw-pellet check; the rating applies to the moulded thickness of the final part. Any change in wall thickness, regrind content, or colour requires re-evaluation under the relevant UL program. In addition, brominated flame-retardant packages can lower CTI, while some halogen-free packages increase melt viscosity and narrow the processing window. The final grade selection therefore must lock not only the polymer family but also the specific flame-retardant system and the regrind ratio used in production.

For glass-reinforced PBT relay housings, the process boundary is the hotter barrel zone and the moisture limit. If the barrel is kept at 270 °C without reducing residence time, thermal degradation will depress tensile elongation and create volatile by-products that can stain the tool and reduce surface tracking resistance. The material cannot be purged with ordinary high-moisture air; dry-air conveying is required from the desiccant hopper to the machine throat. When regrind is used at levels above 20 %, the decrease in viscosity and possible contamination from dust and metal fines must be monitored because relay housings with insert-moulded terminals are sensitive to metal-particle contamination. The instructions associated with the safety approval require the moulder to keep lot traceability because an unreported grade or colour change can invalidate the UL listing, the IEC 60664-1 creepage assumptions, and the end-product relay certification.

An operational boundary exists in high-density relay designs where the coil bobbin and housing are combined into a single PBT moulding. In these parts, the coil winding is in direct contact with the polymer, and any moisture trapped at the interface can become a leakage path through the winding slot. Insulation resistance measured at 500 V DC under ASTM D257 after humidity conditioning must remain above the relay end-product acceptance limit; a drop below that limit may require coil encapsulation or a separate bobbin material. This limitation is not visible in the raw material certification and is only observed when the complete relay is tested under IEC 61810-1 humidity and dielectric sequences.