Gas Flow Method For Detection Of Local Preform Defects Based On Statistical Analysis

Quality Control of Preforms in the Resin Transfer Molding (RTM) Process for Composite Structures NU 2002-052 and NU 2004-068 Inventors Isaac Daniel Sun Kyoung Kim Jeremy Opperer Abstract Northwestern researchers have developed a method to test individual preforms used in composite parts. This particular invention injects low-pressure gas into the preform and uses statistics to characterize the quality of the preform quality. Because only low-pressure gas is used, in-line production testing can be run on a low cost fixture capable of sustaining only low mechanical stress. The savings potential for certain applications can be great, particularly if detection of defects is made early or if a defect is one of a multi-part die. Early detection would enable some preforms to be repaired before they are processed into composite parts, at which point the part needs to be scrapped or reworked. Each variation of the invention injects gas through the preform and takes pressure readings to identify the location of preform defects before resin is injected into the mold. A software tool has been used to statistically evaluate pressure readings and to characterize the severity, location and nature of a preform defect. A 'normal' part is identified from a base of data generated for parts identified as normal. The differences between actual and normal pressures form a basis for quality control criteria. The software employs statistical methods to classify preforms into normal or defective groups, and to answer four basic questions: (1) Is there a defect?; (2) How severe is the defect?; (3) Where is the defect?; and (4) What is the nature of the defect? Applications o Composite Molding Advantages o Lowers fixture cost o Exerts low mechanical stress o Saves time o Functions with various materials (fabric to semi-rigid/rigid) IP Status Issued US Patent No. 6,853,925 Issued US Patent No. 7,305,308