The definition of acceptable flaws in structures is a science unto itself. Typical design specifications do not allow voids in "stressed" areas of parts (at critical fillets, notches, holes, etc.) The next level of sophistication for flaw acceptance criteria comes from the specifications that accompany the inspection method - in other words, flaws are defined within the limits of what a particular method can find. That allows designers to specify "no flaws" when the true interpretation is "no detectable flaws."<br>
<br>
Having said that, the approach to defining acceptable flaws can be done by 1) determining the level of stress within the part, 2) using fracture toughness or J-integrals to define the material's strength and 3) creating a "map" of the part that defines critical flaw size for various locations. Once this is known, a suitable inspeciton method (ultrasonic, eddy current, die penetrant, magnetic particle, radiographic, n-radiographic, etc.) can be used. Each has its own positive and negative aspects and an optimal inspeciton plan (one that finds critical flaws within reasonable cost constraints) can be established.<br>
<br>
Voids typically cause problems because less material exists for strength in areas where strength is needed. They can be mitigated by using larger part section areas. Voids indicate potential problems with part fabrication processes and, although acceptable from a strength point of view, may be symptoms of other problems. They can also contain foreign material which can have other affects on a part (corrosion resistance, etc).