Failure of spruce under compressive low-cycle fatigue loading parallel to grain

Structural wood elements are frequently subjected to cyclic loads in service. These loads can cause micro-structural changes (damage) in the wood cells. Under extreme conditions such as earthquake and hurricane, this invisible minute damage may lead to fatigue failure in a relatively short time. This phenomenon is called low-cycle fatigue with relatively few load cycles at a high stress level causing the damage.; This thesis elucidates the fatigue behaviour and failure mechanisms of clear spruce wood under parallel-to-grain compressive low-cycle fatigue loads. The failure mechanisms under sustained and static loads are also considered.; Spruce was the research material. A ‘density-matched’ sampling strategy was used to group small clear specimens. ‘End-to-end’ pair matching was used to estimate the reference compressive strength of specimens subjected to fatigue or creep (sustained) loading. In fatigue tests, three types of waveforms (triangular, sinusoidal, and square) were used. The square waveform employed several duty ratios. High-to-low and low-to-high step-loading sequences were performed. ‘Dark-field incident’ light, polarised-light and scanning electron microscopy were used to investigate the progression of failure mechanisms. The changes in engineering properties during cyclic loads were considered. To quantify damage, a non-dimensional damage index was introduced based on the continuum damage mechanics theory. A damage model was proposed to describe the low-cycle fatigue behaviour.; It was found that: (1) The ‘density-matched’ sampling design is a good means of matching properties between groups of specimens. The ‘end-to-end’ pair matching method is an effective and accurate means of matching properties of small specimens; (2) The load cycling effect strongly influences the fatigue behaviour of wood under low-cycle fatigue conditions. The effect of load waveform on the damage accumulation is dependent on the loading/unloading rate rather than the rate of change in loading/unloading rate. The square waveform is the most damaging due to its fast loading/unloading rate; (3) The nature of failure in small clear spruce specimens can be attributed to the formation of kinks in the latewood tracheid walls under parallel-to-grain compression regardless of the type of loading. A kink is the permanent micro-structural change in cell walls. The number of kinks is a ‘direct’ damage indicator and strongly linked to the accumulated, non-recovered, strain. The damage accumulation due to fatigue and creep is different. The damage loci in creep specimens are mainly kinks formed during the initial loading. For fatigue specimens, new kinks can be generated due to load cycling. The failure mechanisms are different between the low-to-high and high-to-low step-loading sequences; (4) Cyclic creep strain (deformation), accumulated strain (deformation), effective modulus, and hysteresis loop area are good indicators for describing the fatigue behaviour and accumulation of damage; (5) It is feasible to use Continuum Damage Mechanics to describe the damage accumulation in spruce.