Fracture tests - shock tests

In-service failures are extremely costly: when they do occur, if fortunately they do not lead to loss of human life, in addition to the replacement of damaged equipment, you have to count the hours and production lost, the damaged brand image, the lost markets... For example, the broken bolts in the rudder control system on the Amoco Cadiz sank on March 16, 1978, causing the oil slick we all remember; the broken tail bulkhead on Japan Airlines flight 123, a Boeing 747, led to the loss of control of the aircraft on August 12, 1985, and the deaths of 120 people; on June 3, 1998, a ruptured wheel rim caused the derailment of the ICE Wilhem Conrad Röntgen high-speed train near Eshede, Germany, killing 101 people and injuring a hundred more.

These failures occur as a result of the loads borne by the parts exceeding the breaking strength of the materials of which they are made. It is obviously important to know as much as possible about this latter property, which depends on various factors: temperature, deformation rate, environment.

Numerous tests have therefore been devised to assess the fracture resistance of materials, some of which have long been standard practice in industry. To fully appreciate their scope and limits, it is necessary to understand the fracture mechanisms involved. This is why notched impact tests, developed by Georges Charpy a hundred years ago [1], are so useful. These tests provide invaluable data for assessing the fracture toughness of materials in terms of resilience, the energy absorbed during impact. In particular, they can be used to determine the risk of brittle fracture of steels at temperatures below the brittle-ductile transition temperature. They are particularly useful for welded constructions. Simple to implement and inexpensive, Charpy tests are therefore extremely widespread and still extremely useful (see "Resilience-toughness relationship. Contributions of numerical modeling" [M 4 168] ).

Resilience measured in this way is only a datum that cannot be transposed to a real structure. It gives only a relative indication. More recently, Charpy tests have been instrumented to provide more quantitative data by recording the variation in force exerted by the hammer during impact (ISO 148-1 standard).

The main difficulty in transposing the resilience results of Charpy tests lies in the size effect, as thick structures are more brittle than small Charpy specimens. This has led...