Charring rate

Charring rate is the rate at which heated wood is converted into char, usually expressed as a depth per unit time. It describes the inward progress of the char boundary rather than the movement of flames across a surface. In structural calculations, this boundary identifies material that has largely lost its load-bearing capacity.

Charring is one part of the behaviour of timber in fire. Heat decomposes wood into gases and a carbonaceous residue, while heat also penetrates the remaining uncharred section. The resulting loss of section and changes in material properties determine how the member's resistance changes. Fire exposure, moisture, density and grain direction influence the rate used in a calculation.

A constant rate is a useful approximation under specified conditions, particularly for sufficiently large members during standard fire exposure. It is not an exposure-independent property of every timber product. Protected members, smaller sections and changing fire conditions require additional consideration. The calculated char depth must also be distinguished from the effective section used to assess structural capacity.

Wood decomposition and the char boundary

The underlying process is pyrolysis, the thermal decomposition of wood into gases and char. Gases released through the heated surface can burn in the surrounding air. The remaining solid residue has a lower density than the original wood. Its outside surface can subsequently recede through glowing combustion, physical deterioration or loss of char.

White describes a reasonably distinct boundary between char and uncharred wood. The base of the char layer is associated with a temperature of approximately 300 °C. This temperature provides a useful engineering marker for char development, although it does not make the wood immediately behind the boundary unaffected by fire. A temperature gradient continues into the remaining section.

The depth of wood converted into char differs from the thickness of char remaining after exposure. Char contracts as decomposition proceeds, and its outside surface may lose material. White defines a char-contraction factor using the remaining char thickness divided by the original thickness of wood converted. Measuring only the visible residue therefore does not necessarily recover the original depth of section lost to charring.

Char contraction and surface loss distinguish remaining char thickness from original wood depth converted. Based on White (2002), pp. 4-260–4-261.

The thermal gradient explains why a large timber member can retain a cooler interior while its surface chars. The IFEG discussion of heavy timber identifies progressive section reduction as the main structural consequence of this process. It does not treat the charred material as having the original timber strength. Assessment therefore concerns the heated residual member and the load it must continue to support.

Conditions behind a quoted rate

Under standard exposure, the rate can be higher initially before becoming approximately constant. White reports a commonly used empirical approximation of 0.6 mm/min transverse to the grain for the ASTM E119 exposure. This is an example from the 2002 handbook's treatment of that test. It is not a statement that every timber product chars at that rate under every fire.

Species differences are associated with density, chemical composition and permeability. Moisture content also affects charring, and grain direction changes the route through which heat and decomposition products interact with the material. White reports that charring parallel to the grain can be approximately twice the transverse rate. A quoted value therefore requires its direction and material conditions to be retained.

The imposed heating conditions are equally relevant. White describes experiments in which the charring rate varied with external radiant heat flux, including tests on spruce and southern pine. A value derived during one exposure cannot automatically represent a weaker, stronger or changing exposure. The source distinguishes empirical standard-test rates from methods developed for nonstandard fires.

The standard fire curve establishes a particular heating history for a resistance test. A real compartment fire has a history affected by available fuel and ventilation. White discusses separate charring relationships for such exposures and specifies limits on their application. Their coefficients and assumptions cannot be exchanged freely with a rate calibrated against a standard test.

The dimensions of the member also limit the constant-rate approximation. It is most defensible when the remaining section is sufficiently deep that its centre or rear face remains little heated. White notes that the charring rate can increase once those regions warm. Continuing a large-member approximation after the entire remaining section becomes thermally affected can therefore miss a change in behaviour.

Rate conventions also require attention when using published equations. Some relationships express distance charred per unit time, while others use the reciprocal, time required per unit depth. White presents both forms when reviewing empirical methods. A coefficient in minutes per millimetre cannot be inserted directly into an expression expecting millimetres per minute. The distinction also matters when describing a larger coefficient as faster or slower charring: increasing the reciprocal coefficient means more time for the same depth, rather than more depth in the same time.

Calculating char depth and remaining geometry

For a constant linear rate, char depth equals the rate multiplied by the period of charring. This expression is the geometric basis of the constant-rate approach described by White. Its time variable starts with the assumed onset of charring at the exposed wood. A protective lining can alter that onset, so elapsed fire time and charring duration need not be identical.

x_char = βt

For a rectangular beam exposed on both sides and its underside, the simplified residual width loses two char depths. The residual depth loses one because the upper face is unexposed in this idealisation. White's rectangular-section expression then gives the section modulus about the horizontal bending axis. This geometric quantity describes how the remaining section resists bending stress for a given bending moment.

B_r = B − 2βt; D_r = D − βt

S_r = B_rD_r²/6

Symbol	Description	Unit
xchar	Depth of wood converted into char	mm
β	Constant linear charring rate in the example	mm/min
t	Duration of charring	min
B, D	Original breadth and depth	mm
Br, Dr	Simplified residual breadth and depth	mm
Sr	Residual rectangular section modulus	mm³

These expressions assume the same constant rate on the three exposed faces and a rectangular remaining section. Actual corners become rounded as heating arrives from adjacent faces. White discusses incorporating that effect into design rates or the geometrical treatment. The simple equations consequently illustrate section loss without reproducing every correction in a structural fire-design method.

The section modulus changes differently from cross-sectional area because the depth enters its expression squared. The loss of the outer material therefore affects bending geometry as well as total area. White's structural method subsequently compares the applied bending moment with the residual section and the appropriate material resistance. A char-depth calculation alone does not perform that load-bearing check.

A worked geometric example

White includes a beam example with dimensions 130 mm by 533 mm and three-sided exposure. Those dimensions are used here for a separate illustration of section loss. Applying the handbook's 0.6 mm/min constant-rate approximation for a selected 60-minute charring period gives a depth of 36 mm. This arithmetic example is not a reproduction of White's separate fire-rating calculation.

Removing 36 mm from both sides reduces the calculated width from 130 mm to 58 mm. Removing the same depth from the underside reduces the beam depth from 533 mm to 497 mm. The simplified residual area is approximately 28,800 mm², compared with an original 69,300 mm². About 42% of the original cross-sectional area therefore remains in this idealised geometry.

Three-sided geometric section loss for the illustrative constant-rate calculation. Based on White (2002), pp. 4-260, 4-266 and 4-268.

The corresponding section modulus decreases from approximately 6.16 million mm³ to 2.39 million mm³. Its residual value is about 39% of the original value, calculated using the rectangular-section equation. This ratio is a statement about section geometry. It is not a prediction that exactly 39% of the original structural capacity remains, because heated-wood properties and other failure modes also matter.

The example deliberately isolates the geometric effect of a chosen rate. It does not include a zero-strength allowance, corner rounding, protection performance or changes in the rate as the section becomes smaller. Nor does it specify the member's applied load or restraint. These omissions prevent the arithmetic from establishing a sixty-minute fire-resistance rating for an installed beam.

Char depth and the effective structural section

Uncharred wood close to the char boundary can be hot enough to lose strength and stiffness. Its moisture distribution also changes during heating. White therefore distinguishes simple geometrical removal of char from evaluation of the remaining material's mechanical properties. Treating every uncharred part as if it remained at room temperature would omit this heated region.

One approach retains the uncharred geometry but reduces the strength and stiffness assigned to it. Another removes an additional effective layer and evaluates the smaller section using the corresponding room-temperature properties. The latter underlies the reduced cross-section method. The extra removed layer represents mechanical degradation; it is not additional measured char.

Alternative treatments of the heated uncharred region must use compatible design assumptions. Based on White (2002), pp. 4-265–4-267.

White also describes more detailed models that divide the residual section into layers or elements. Each part receives properties associated with its local temperature and moisture condition. Such a model can represent variations that a single reduction factor approximates. Its additional detail also increases the need for reliable thermal and mechanical material data.

The chosen rate, geometrical correction and strength model must belong to a compatible calculation method. A design rate may already include allowances beyond the physical linear recession observed in a test. Adding unrelated corrections can therefore change the meaning of the calculation. White explicitly cautions that design values and structural methodologies account for performance factors in different ways.

For mass timber, the large-member relationships discussed here support questions about residual sections in sawn and glued-laminated members. Their application still depends on the particular construction and exposure represented by the source method. The IFEG discussion specifically addresses heavy structural members, while White separates member response from protective layers and connections. A member calculation does not by itself establish the performance of an entire assembly.

Protection and assembly details

Protective linings can postpone the onset of charring by reducing heat transfer to the timber. They can also affect the rate after charring starts. White describes gypsum-protected beam experiments in which the reduction in char depth exceeded that attributable only to delayed onset. A calculation that merely shifts an unprotected charring curve in time may therefore miss part of the protection's behaviour.

The protective layer's continued attachment and physical integrity also matter. White identifies deterioration of panel products as a difficulty in numerical analysis of protected timber assemblies. The model must describe the boundary conditions actually experienced by the timber as protection heats and changes. Assuming an unchanged protective layer throughout exposure is a substantive assumption about assembly performance.

Surface fire-retardant treatment and structural fire protection are distinct questions. Some treatments change ignition and surface flame spread without producing a corresponding improvement in charring resistance. White reports that certain treatments can even increase charring while reducing flammability. A favourable surface-burning result therefore does not automatically provide a reduced structural charring rate.

Connections and fasteners introduce further conditions beyond the exposed timber section. White discusses their protection alongside the calculation of member fire resistance. The ability of a beam to retain material is insufficient if its support or connection cannot carry the load. This is why an assembly assessment includes details that are absent from the simple depth-times-rate calculation.

Evidence, modelling and uncertainty

Charring relationships can be derived from experiments or from coupled models of heat transfer and decomposition. White describes models that account for wood, moisture, gases and char as separate components. Some also include char contraction and recession of the outside surface. These mechanisms help explain why a visible char thickness, a thermal boundary and an effective structural section are different quantities.

A more detailed calculation remains limited by its input data. White identifies the adequacy of thermophysical properties as a major issue for sophisticated pyrolysis models. An uncertain material description does not become certain when represented by more computational elements. The selected model must still be supported for the material, exposure and geometry that the analysis concerns.

The ABCB 2020 Data Sheets place char-rate methods within the final stage of structural fire assessment. Fire exposure is established first, thermal response follows, and the structure's ability to perform its function is then evaluated. The guidance also discusses thermal stresses, deformations and interactions between members. Char depth is therefore an intermediate result used within a wider structural assessment.

Experimental resistance data have their own limits of transfer. White notes that listed ratings generally concern the tested assembly and specified construction details. The ABCB guidance similarly relates direct use of test data to representative heating and loading, or to justified treatment of differences. Neither a nominal rate nor a test duration establishes equivalent performance for every changed member arrangement.

References

• Australian Building Codes Board. (2020). Fire safety verification method data sheets: Handbook annex (Version 1.1).
• Australian Building Codes Board, National Research Council of Canada, International Code Council, & Department of Building and Housing. (2005). International fire engineering guidelines.
• White, R. H. (2002). Analytical methods for determining fire resistance of timber members. In P. J. DiNenno (Ed.), SFPE handbook of fire protection engineering (3rd ed., Chapter 4-11). National Fire Protection Association.