Page 201 - Fiber Fracture
P. 201

186                                                             H.U. Kiinzi















                                                ___ -
                      Fig. 1. A bnttle inclusion  at the beginning of a channel of debns In  a 25 wrn Au wire.


                 A great variety of various drawing defects are known to occur (Catalogue of Drawing
              Defects, 1985). Some defects may already be present in the rod prior to drawing the wire
              and others appear during the drawing process. The former often produce catastrophic
              failures during  the  drawing  and  fortunately  the  latter  are  mostly  due  to  inadequate
              drawing tools  and  drawing parameters. These  defects are  readily  detectable  and  are
              therefore of little or no concern in commercial products. In some cases, however, hidden
              defects such as voids along the wire centerline (central bursts) and inclusions may occur
              even under sound drawing conditions.
                 Non-metallic inclusions such as oxide particles are often trapped in the metal during
              remelting. Even though  in  composite materials hard  particles  are  deliberately added
              to  reinforce  ductile  metals,  the  presence  of  such  particles  is  incompatible with  the
              drawing  process.  The  difference in  their  yield  and  flow  properties  rapidly  leads  to
              complete  decohesion  along  the  interface  and  even  voids  may  be  formed  along  the
              drawing direction. Bigger particles may also break up and produce a channel of debris.
              Similarly, metallic inclusions with yield properties that are different from the matrix
              may produce substantial drawing defects. The effect of inclusions is particularly feasible
              in micro-wires where they may cover an important fraction of the cross-section. Fig. 1
              shows a brittle inclusion appearing at the surface in a 25 km thick Au wire.
                 Murr  et  al.  (1997)  and  Murr  and  Flores  (1998)  describe  an  interesting  case  of
              contaminated submicron Cu particles in Cu wires. During the fabrication of precursor
              rods (later used for drawing the wire) bursting vapor bubbles may produce a spray or
              mist of  liquid Cu particles. The vapor in the bubbles results from the reaction of  H2,
              dissolved in  Cu  from  the  reducing furnace  atmosphere, with  02 from  the  air  when
              molten Cu is cast into open molds. The emanating particles solidify rapidly but remain
              hot enough to react with air, slagelements or carbon from the graphite blanket covering
              the liquid Cu. After reintegration in the melt the reaction layer acts as a diffusion barrier
              and prevents complete dissolution in the Cu melt. During drawing the surface-reacted
              Cu particles behave somewhat like second-phase particles. The surface layer smears out
              in the drawing direction and tiny voids form in front and behind the particles.
                 Cup-shaped  voids  along  the  center  line  of  the  wire  are  known  to  result  from  a
              phenomenon called central burst or chevroning. This phenomenon has its primary origin
              in the flow pattern of the yielding metal. The zone where this occurs is located at the end
   196   197   198   199   200   201   202   203   204   205   206