Specifying Polycarbonate Glazing: Span Tables, Deflection and Thermal Movement

Specifying polycarbonate glazing requires coordinated evaluation of short-term deflection, long-term creep, and reversible thermal expansion. Unlike glass, solid and multiwall PC sheets have a relatively low flexural modulus of approximately 2300 MPa when measured to ASTM D790-17 and a linear coefficient of thermal expansion near 70 × 10⁻⁶ m/(m·K) under ASTM D696-16. The consequence is that serviceability limit states—rather than ultimate tensile strength—usually control support spacing. Typical PC sheet products are specified under EN 16240:2014 for solid sheet or EN 16153:2013+A1:2015 for multiwall sheet, while model codes such as IBC Section 2608 impose material, edge engagement and fastener requirements for plastic glazing. A span table is therefore not a single material property but a calculation output that includes load magnitude, edge fixity, allowable deflection ratio, creep correction and thermal edge clearance.

What governs deflection in simply supported polycarbonate glazing?

For a simply supported rectangular plate under uniform pressure, maximum deflection is described by δ = k q a⁴ / D, where D = E t³ / [12(1-ν²)], q is the design pressure, a is the short span, E is the flexural modulus from ASTM D790-17, t is sheet thickness, and ν is Poisson’s ratio. For solid PC, ν is commonly taken as 0.38. The coefficient k depends on aspect ratio and edge support; for a square plate with simply supported edges, a plate coefficient of 0.00406 is used. In service, published polycarbonate span tables frequently adopt an allowable short-duration deflection of L/20 for wind loads and a more restrictive L/50 or L/60 where glazing is overhead and drainage requires positive slope. Tensile yield strength of PC is approximately 60 MPa per ASTM D638-14, but at L/20 the calculated bending stress under typical wind pressures often remains below 10 MPa, indicating that deflection controls before material strength becomes limiting.

Span selection for solid sheet can be estimated from plate theory, but the values must be adjusted for edge support continuity, aspect ratio and long-duration loading. The table below is calculated for a simply supported square solid PC sheet using E = 2300 MPa, ν = 0.38, a uniform pressure of 1.0 kPa, and a deflection limit of L/20. The results are representative for a square plate with four edges supported; they are not a substitute for product-specific technical approvals, and profiled or multiwall sheets require manufacturer test data.

Thermal movement is not a secondary design check for solid sheet

Linear thermal movement is calculated as ΔL = L₀ α ΔT. For PC, α is approximately 70 × 10⁻⁶ m/(m·K) per ASTM D696-16; some product data sheets report values from 65 × 10⁻⁶ m/(m·K) to 72 × 10⁻⁶ m/(m·K) depending on grade and test method. A 3000 mm clear sheet undergoing a 50 K temperature differential therefore requires 10.5 mm of combined edge clearance if both ends are free to move. When one edge is fixed, movement occurs at the free edge; when both edges are restrained, the panel cannot expand without compression or bowing. Thermal movement must be evaluated separately from deflection because deflection is a bending displacement under transverse load, while thermal expansion is an in-plane displacement that can accumulate across the full panel width and length.

Frame detail should permit the larger of thermal movement and fabrication tolerance. For a 3 m multiwall sheet at ΔT = 50 K, the 10.5 mm movement is usually divided between two edges, approximately 5 mm per side. Fastener holes should be oversized or slotted, with EPDM or silicone gaskets specified to isolate the PC from direct metal contact; contact with plasticized PVC or certain sealants can cause stress cracking. In multiwall sheets, edge closure tapes must permit drainage and pressure equalization while excluding dirt, and the lower edge should be kept free to expand. Avoid contact with amine-cured sealants and solvent-based adhesives containing ketones or aromatic hydrocarbons unless explicitly validated for PC glazing.

When edge clearance is less than calculated thermal movement

When edge clearance is less than calculated thermal movement, the panel is subjected to compressive in-plane strain. Fully restrained thermal stress can be estimated from σ = E α ΔT. At 2300 MPa flexural modulus and 70 × 10⁻⁶ m/(m·K) expansion coefficient, a 50 K temperature rise produces approximately 8.05 MPa of compressive stress. Although this is below the short-term tensile yield of 60 MPa, it is significant when combined with external wind suction or snow load and can cause buckling in thin sheets. Repeated daily thermal cycles can induce fatigue at fastener holes and microcracks at saw-cut edges if the hole-to-edge distance is below the minimum recommended by the sheet manufacturer. Thermal bowing is also observed when the exterior surface temperature exceeds the interior surface temperature; for dark-colored multiwall sheets, published manufacturer calculators commonly use a surface temperature elevation of 20 K to 30 K above ambient under 1000 W/m² solar irradiance. This bowing changes the apparent deflection and can increase mid-span movement beyond the value calculated from uniform transverse load alone.

Multiwall polycarbonate span tables are product-specific because web geometry, rib spacing, and sheet weight vary even at the same overall thickness. A 10 mm multiwall sheet from one manufacturer may have a different flexural stiffness and deflection-limited span than a 10 mm solid sheet or a 10 mm multiwall sheet with a different number of walls. Published span tables under EN 16153:2013+A1:2015 are based on type tests and often state whether the listed span includes a creep correction for continuous load. Where roof glazing carries snow for more than 30 days per year or where dead load is continuous, the short-term flexural modulus is not sufficient; long-term deflection should be checked using creep modulus data generated to ISO 899-2:2003 or product-specific technical approvals. Published data for the specific configuration of a given multiwall PC sheet under simultaneous sustained snow load and thermal cycling is limited, so the safer design approach is to reduce the span or increase thickness rather than extrapolate short-term tables.

Wind load and snow load combinations in multiwall polycarbonate span selection

Design wind actions should be determined in accordance with EN 1991-1-4:2005 or ASCE 7-22, and snow loads in accordance with EN 1991-1-3:2003 or the applicable national annex. For plastic glazing in roof applications, the governing load combination may be wind uplift on the leeward slope combined with internal pressure, rather than the maximum downward snow load. The span table should therefore be entered with the larger of the two serviceability pressures, not with the ultimate limit state pressure, because deflection is the controlling parameter. When using manufacturer span tables, the applicable deflection limit must be confirmed; a table prepared for L/20 will yield larger spans than one prepared for L/50, and neither is interchangeable with glass deflection criteria under ASTM E1300-23, which applies to soda-lime silicate and not to polycarbonate.

At the sheet extrusion stage, melt temperature and roll stack temperature control affect residual stress and thickness tolerance that later alter deflection performance. Production-scale solid PC sheet is typically extruded on single-screw extruders with L/D ratios of 32:1 to 38:1, using melt temperatures of 280 °C to 320 °C and polished roll stack temperatures of 110 °C to 140 °C. Residual stress retained from uneven cooling can produce anisotropic flexural behavior; incoming sheets should therefore be inspected for uniform optical retardation before cutting. On production lines, edge cracking around punched holes in multiwall sheets has been traced to undersized holes and rigid fasteners that prevent expansion; field inspections of overhead glazing show tensile crazing adjacent to aluminum frames when edge clearance is fully consumed. On site, verification of actual sheet thickness, support width, and fastener location is necessary because a ±1 mm thickness deviation on a 6 mm solid sheet changes flexural stiffness by approximately −42% to +59% relative to the nominal value, directly affecting the deflection-limited span.