English: While casting directionally solidified crystals in the late 1960s, AMRDL researcher Barry Piearcey found that if a right-angle bend occurred in the casting mold, a short distance above the knurled chill plate surface (called the “starter” chamber), the number of columnar crystals exiting the bend would be reduced. Two such bends reduced the number even more. Later, while investigating the properties of single-crystal springs, Giamei found that a helical channel with smooth continuous turning was a natural substructure filter, admitting columnar crystals from the starter and emitting one single crystal to start the single- crystal structure of the turbine blade.