Damage Identification for a Spatial Steel Frame

Tbe paper reports tbe resùlts of an experimental investigation carried out in tbe laboratory of Civil and Environmental Engineering at tbe University ofFlorence wbere a reduced scale (2.5 m width x 2.5 m deptb x 4.8 m beigbt), tbree storey steel spatial frame was instrumented and progressively damaged by cutting one of its pillars just above tbe first leve!. Tbe deepness of tbe cut was measured and was taken as tbe entity of tbe damage. A seri es of 14 accelerometers was installed on tbe spatial steel frame, and acceleration measurements induced by ambient vibrations were recorded as tbe frame was progressively damaged. A'cceleration signals were subsequently analyzed in botb time and frequency domain (SSI, EFDD, neural networks): a frequency sbift and a cbange in tbe corresponding modal sbapes were observed as tbe damage increased.