Examination of a cracked Wave Fan casing. 1. Introduction 1.1 A request was received for an examination to be carried out on the subject wave fan casing to determine the mechanism that had resulted in cracks that were reported to be present down two welded joints. 1.2 The wave fan size was recorded as being 800CBL'N. 2. Examination Results 2.1 The wave fan casing, as examined is shown on Plate 01. Magnetic tests revealed the material from which the fan casing had been manufactured was a ferromagnetic steel and brittle cracks were seen to be present at positions A & B (see Plates 02 & 03). Macroscopic examination of the exposed fracture face of the crack at position A showed that the separation had taken place as a result of a fatigue mechanism. The fatigue cracks had initiated in the root of the weld bead. Penetration of the weld was estimated to average 90% with a minimum of 70% (see Plate 04). 2.2 In order to investigate the fracture mechanism further and the quality of the welding, small mating pieces were cut out from what was considered to be the region in crack A that had separated first (see Plate 05). Macroscopic and microscopic examination of these pieces confirmed that a fatigue mechanism had been responsible for separation and it was estimated that many hundreds of thousands of stress cycles would have been required to produce the type of damage present. 2.3 The pieces from crack A were sectioned and one half of each was encased in polymethylmethacrylate so that the cut edges could be prepared for metallographic examination after suitable etching. On examination it was seen that welding had been carried out by a TIG or MIG process. The weld was clean and sound and approximately 95% penetration had been achieved. The etched structure indicated that the parent metal was probably a low carbon steel and the weld metal a carbon/manganese steel. The parent metal thickness was measured as being 2.8mm. 2.4 Hardness tests were carried out on the prepared micro section (Para 2.3) in order to determine the approximate tensile strengths of the parent metal and the weld bead. The following results were recorded: Parent metal: 126 & 127 HV(10) Weld bead: 224 HV(10) These results indicate that the parent metal tensile strength approximated to 390N/mm² (25 tonf/in²) and the weld metal 690N/mm² (45 tonf/in²). 3. Conclusions 3.1 It is considered that the cracks present in the subject wave fan casing were the result of a simple bending fatigue mechanism resulting from in service cyclic loads generated by pulsing air pressure within the casing. It is also considered that the weld quality, although not perfect, was satisfactory for this kind of construction. 4. Comment 4.1 During the examination the fan was run under normal service conditions. Crack A opened to a maximum of approximately 38mm and the free edge of the side panel was seen to be vibrating laterally. 4.2 From the overall consideration of the above findings it is evident that a design deficiency existed in the subject wave fan casing which resulted in the cracking experienced (This opinion is only valid if the material thickness of the casing is as called up by the designers and the fan is being operated within its performance envelope). 4.3 In order to repair the cracked wave fan casing it will be necessary to cap over the cracks with 40mm steel angle attached by welding. In order to prevent further cracking it will also be necessary to stiffen the side panels to prevent them drumming under the action of the pulsating air pressure. This can be achieved by the addition of an external waffle iron pattern of stiffeners formed by an angle iron attached externally by welding and extending round the sides and ends of the casing. It will probably also be necessary to fit a tie bar across the side panels at their approximate centre of pressure. It is recommended that agreement should be reached with those responsible for the design of the casing before the repair/modification proposed is carried out.