Compression test methods

Compression testing is particularly difficult for PMCs due to the occurrence of macro- and micro-buckling modes. The rectangular prism specimen in ISO 604 is not suitable for laminated composites as these will split vertically between the laminate layers when loaded axially in the plane of the laminations. Most composites are available as relatively thin laminated sheet, so that, tensile-type specimens, 2 mm to 4 mm thick, are not suitable for compression tests as they fail by Euler column buckling. Three approaches have been used to avoid specimen buckling;

• increased specimen thickness - limited by machine capacity
• short test span (ie column height equal to the "gauge" length) - most popular,
• support jig on faces of specimen - for example as used in ASTM D695, however, the level of support and load carried by the jig is unknown.

This led to several varieties of the test with a range of aspect ratios, both height to thickness and height to width.

A framework has been developed in BS EN ISO 14126 in order to rationalise these many test methods and even greater number of loading fixtures/jigs. The approach taken was to concentrate on the quality of the test by requiring the strain to be measured on both faces of the specimen and for the test to be completed, or terminated, within a bending (buckling) strain of less than 10% of the axial strain. This limit was also applied in the equivalent ASTM standard, D 3410.

Concentrating on the quality of the test allows for a range of jigs to be used, for the continuing development of jigs and ensures that established tests are used correctly. Both shear face loaded and end loaded jigs, as in the discontinued ISO 8515, are allowed. The jig used must be quoted, either an established jig from those listed in the annex or any other that is used. The revision of ASTM D 3410 enabled greater harmonisation with the CEN/ISO standard (i.e. 10 mm widths and gauge lengths rather than 12.5 mm (~ 0.5 inch) when moving to metric dimensions.

A disposition was requested by the USA, so that providing the first five specimens fail correctly within the 10% bending strain criteria, further specimens of the same batch, using the same operator, equipment etc., can be tested without, or with a single, strain gauge(s).

A second specimen size, 25 mm wide by 25 mm gauge length was included for fabrics, mats and multidirectional lay-ups but uses a 4 mm or greater thickness based on ASTM D3410 guidelines to avoid premature failure by Euler buckling. There was no general support for including the machined specimen in prEN 2850. This specimen requires a separate special laminate lay-up, careful machining and polishing of the cut faces and radii. This is not a preferred situation for general engineering use when a finished test panel is supplied to obtain a set of properties. The results for this specimen are inconclusive, as some materials have shown improved results, while others have recorded lower strengths.