~m~~arbon and graphite materials stressed by external or internal forces release energy in the form of stress waves. These waves which characterize the behavior of the material, have been analyzed by means of acoustic emission analysis. Graphite will emit acoustic energy already under small bending and tension stresses (observed frequency range O&O,3 MHz). The acoustic emission rate is higher during the first loading than during the following load cycles where emission can only be observed when the stress level exceeds the preceding one. When multiple loading is repeated after some stress-free time, acoustic emission will show up again during the first new load cycles. This (relaxation) effect is the higher the longer the load-free interval has been. It can also be achieved by heating the prestressed graphite, e.g. up to 1ooO"C. The approaching failure of graphite (under a constant or alternating load) can be predicted from the acoustic emission curve. Thus a critical crack growth may be stopped by diminishing the load. During heat treatment of baked material (petroleum coke base) up to 2500°C acoustic emission can be observed in the temperature range from MOO to 17OO"C, as the formation and evolution of sulphur compounds creates increased microporosity and an irreversible elongation of the carbon body. On cooling down a graphitized body very strong acoustic emission will occur in the temperature range from 2200" to 1SOOT. By means of distribution analysis of the acoustic emission events with regard to counts per event, to pulse widths, and to amplitudes, one can distinguish between low and high loading phases of graphite components. By applying dist~bution analysis one can also determine the difference in time of arrival of the stress waves at two different observation points and thus locate the origin of the acoustic emission. During a 3 point-bend-test critical crack growth begins at about 80% of the fracture load. 1. RINLRITUNG * Belastung bis zum Bruch (680 Nl 0 Belastung bis 550 N Schaltemissionssumme AuJnehmerI I , , A.,..$ Biegeprobe Abb. 14. HPufigkeitsverteilung der Lage der Schallquellen iiber die LHnge einer 3-Punk&Biegeprobe.