R,E&D Program Manager for Advanced Materials/Structures AAR-430, Federal Aviation Administration, NJ
Associate Professor, Wichita State University, Wichita, KS
Senior Statistician, Dana-Farber Cancer Institute, Boston, MA
Pages: 17 Published: Jan 2002
The development of a database and guidelines for its generation are described. The developed common database guidelines include types and number of mechanical tests, test methods, and environmental conditions. Methods to reduce the test data to the level of obtaining A- and/or B-basis values include statistical procedures that allow pooling across different environments for the same failure modes in order to increase sample sizes to estimate material variability. Either Weibull or normal distribution statistics can be used. For the normal distribution, knockdown factors from mean value to A-and B-basis material (lamina) allowables were developed that are based on a large sample size for variability and a small sample size for the mean. The concept of establishing large sample (population) variability for particular failure modes has implications for the statistical significance of design allowables for notched and unnotched laminates where the test samples are historically small. In order to utilize the database, each company must demonstrate that their processing of the identical composite material results in mechanical properties that are the same as in the database. Equivalency tests, along with statistical acceptance criteria, were developed that would perform that function. In addition, the developed acceptance criteria may be used by the manufacturing company to control the quality of the incoming material. This paper will also contain an example of a database for a carbon/epoxy material along with example equivalency tests and statistics.
composite materials, databases, material allowables, statistical methods, pooling, equivalency tests, acceptance criteria
Paper ID: STP10633S