What the short-term test misses

Nearly every load table a manufacturer publishes comes from a short-term test: load for a few minutes, read the deflection, unload. That figure describes instantaneous stiffness with no time in it, and FRP is viscoelastic, so time is exactly the variable most easily left out.

A panel that passes on the test rig will keep deflecting if it carries the same load all year, and after several years the deflection can reach twice the initial value or more. The growth is not instability and does not happen suddenly, but it changes drainage fall, step heights and the load on fixings. Long-term behaviour is discussed under FRP grating service life

What viscoelastic means

The resin matrix in FRP is a polymer, and under load its molecular chains gradually slide and rearrange, converting part of the instantaneous elastic strain into permanent deformation. The glass fibre itself does not creep; creep comes mainly from the resin and from the fibre-resin interface.

Creep therefore depends on resin content, cure and fibre direction. A product with more resin and incomplete cure creeps more; a pultruded profile with continuous fibre along the load direction generally resists creep better than a molded panel with random fibre.

Moisture accelerates the process. Resin absorbs water and plasticises, its glass transition temperature falls, and creep runs faster at the same temperature and stress, so hot and wet duty needs a more severe coefficient.

Reading a creep curve

A typical creep curve has three parts on log-log axes: a fast initial stage, a long, nearly straight steady stage, and a final acceleration close to failure. Design only uses the first two; the third means failure is near.

Some makers publish isochronous curves, plotting stress against strain at different times with one curve each for a year, ten years and fifty years. Those curves are more useful than a single creep factor because they show the effect of stress and time together.

Without isochronous curves, at least ask how the creep factor was measured and extrapolated. A 1,000-hour test extrapolated to ten or fifty years under ASTM D2990 or ISO 899 is common practice, and how much to trust the result depends on the number of test points and the consistency of the material.

Using the creep factor

The creep factor is the most practical tool: long-term deflection equals short-term deflection times the factor. Values commonly fall between two and four, with room-temperature, low-stress vinyl ester near the low end and hot, high-stress orthophthalic polyester near the high end; the maker's data governs.

The factor is not the same for every load case. A sustained uniform load develops creep fully, while intermittent and short-term live loads barely creep, so loads have to be classified by duration before any factor is applied.

One combination is easily missed: when short-term live load and long-term dead load act together, the dead load takes the creep factor, the live load does not, and the two results are added. Multiplying the total by one factor overestimates, and multiplying nothing underestimates.

Sustained and intermittent load

Dead load on a platform includes the grating itself, permanent equipment, piping and insulation, and it stays for years, so creep develops fully. Live load includes people, tools, temporary material and maintenance gear, acts briefly, and is handled with short-term stiffness.

Deciding which class a load belongs to means asking how much of a working day it is present. Pallets stored for months and pumps that never move are dead load; a few hours of maintenance work a day is live load; anything in between goes on the conservative side.

Snow and dust are seasonal dead load and can persist for months in some regions, so their creep cannot be ignored. Local weather conditions belong in the load combination rather than being treated as zero.

Temperature accelerates creep

Temperature affects creep non-linearly. As the resin approaches its glass transition temperature the modulus falls quickly and creep rate multiplies; polyester transition temperatures typically lie between 60 and 100 degrees Celsius, and the exact grade matters a great deal.

Hot duty needs two steps together: reduce short-term strength for temperature, then calculate long-term deflection with a higher creep factor. Doing only one of the two leaves the result optimistic.

A surface in direct sun can run twenty or thirty degrees above air temperature, and dark grating shows it most. The design temperature for an outdoor platform is the surface temperature, not the weather station reading.

Resin and fibre direction

Isophthalic polyester, vinyl ester and phenolic behave differently in creep. Vinyl ester shrinks less on cure and resists hydrolysis better, so it usually retains more modulus under sustained load; phenolic is stiff but brittle, suited to fire duty rather than necessarily to creep-critical work.

Fibre direction decides whether the fibre carries the load directly. Pultruded grating has continuous roving along the bearing bar, so sustained load acts mostly on the fibre and creep stays smaller; molded grating has chopped fibre at random, so more load passes through the resin and creep is more visible.

The selection logic and long-term behaviour of the two processes are compared in molded versus pultruded grating

How to run the design check

The check has two steps. First select the panel on short-term load against the short-term deflection limit, commonly span divided by 160 or 6.4 millimetres. Then multiply the dead-load share by the creep factor, recalculate long-term deflection and compare it with a looser long-term limit.

There is no universal long-term limit. A common approach is span divided by 100, or whatever keeps drainage working and avoids a visible step. The figure should be agreed with the owner and the process discipline and written into the specification.

If long-term deflection fails, add supports before adding panel depth. Support spacing affects deflection to the third or fourth power while extra depth raises section stiffness only linearly, so closer supports are usually the cheaper fix. The effect of spacing is covered in support spacing and deflection for FRP grating

Writing the specification

The specification should state the design life, the split between dead and live load, the long-term deflection limit, the design temperature and the resin type. A requirement that capacity is adequate leaves the supplier unable to tell whether to select on short-term or long-term data.

Ask the supplier for creep data or isochronous curves, with the test standard, duration and extrapolation method stated. Where no measured data exists, at least require the basis for the creep factor and have it written into the calculation. Full parameter wording is in the FRP grating specification guide

• Separate dead from live load, and apply the creep factor to dead load only.
• Put the long-term deflection limit into the specification.
• At high temperature, reduce strength and raise the creep factor together.
• Ask the supplier where the creep data comes from.