THE PRIMARY OBJECTIVES OF THE PROPOSED WORK ARE: 1. CAREFULLY CHARACTERIZE THE TOOL/PART INTERFACE FOR FRICTION STIR WELDING OF HIGH STRENGTH ALUMINUM ALLOYS USING A SIMPLE EXPERIMENT. 2. DEVELOP A MODEL OF LOCAL FLOW STRESSES AT THE TOOL/PART INTERFACE USING THE DATA FROM OBJECTIVE #1. 3. INTEGRATE THE INTERFACE MODEL INTO A PREVIOUSLY DEVELOPED FINITE ELEMENT MODEL OF THE FSW PROCESS. 4. VALIDATE THE FSW MODEL RESULTS FOR TORQUE REACTIVE LOAD AND TEMPERATURES FOR STANDARD FSW TOOLS AND FOR A SELF-REACTING FSW TOOL. IN PRIOR EFFORTS THE FRICTION LEVEL HAS BEEN AN ADJUSTABLE PARAMETER USED TO "TUNE" A FINITE ELEMENT MODEL FOR FSW. HOWEVER THIS SIMPLE TUNING APPROACH NEGLECTS THE UNDERLYING PHYSICS AT THE INTERFACE AND DOES NOT PROVIDE ACCURATE PREDICTIONS FOR ALL OF THE CRITICAL PHENOMENA THAT WOULD RENDER THE MODEL TRULY PREDICTIVE AND USEFUL FOR DEVELOPMENT. FOR EXAMPLE IF THE MODEL IS "TUNED" FOR AN ACCURATE PREDICTION OF TEMPERATURES THEN THE REACTIVE LOAD IS OFTEN OVERESTIMATED. THIS RESULTS IN INACCURATE STRESS PREDICTIONS WITHIN THE PART. THE PHYSICS AND MECHANICS OF HEAT GENERATION AND MATERIAL DEFORMATION AT THE SCALES AND TEMPERATURES AT WHICH THEY OCCUR IN FSW HAS NOT BEEN ADEQUATELY STUDIED AND UNDERSTOOD. WHILE THE FLOW STRESS IN THE BULK MATERIAL CAN BE CHARACTERIZED BY TENSION TESTING COMPRESSION TESTING OR TORSION TESTING THE INTERFACE MATERIAL FLOW STRESS UNDERGOES LARGE AMOUNTS OF SHEAR AND THEREFORE RECRYSTALLIZATION WHICH CAN DRAMATICALLY INFLUENCE THE LOCAL FLOW STRESS. RATHER THAN TRYING TO FULLY UNDERSTAND THE MICROSTRUCTURE EVOLUTION AT THE INTERFACE OUR APPROACH WILL BE TO DEVELOP A SIMPLE EXPERIMENT THAT WILL ALLOW FOR CHARACTERIZING THE LOCAL FLOW STRESSES AT THE INTERFACE WHICH ARE THE PRIMARY DATA NEEDED FOR ACCURATE MODELING WITHOUT THE NEED FOR "TUNING" BY FREQUENT ADJUSTMENT OF A FRICTION COEFFICIENT. DATA FROM THE EXPERIMENT WILL BE USED AS INPUT TO AN INVERSE PARAMETER IDENTIFICATION SCHEME THAT WILL PROVIDE LOCAL FLOW STRESSES IN WHAT WE ARE CALLING THE "HEAT GENERATION ZONE" (HGZ). IN ORDER TO ACCOMPLISH THE OBJECTIVES OF THIS PROPOSED RESEARCH THE FOLLOWING TASKS WILL BE CARRIED OUT: 1. FLAT PINLESS FSW TOOLS WILL BE ROTATED AT DIFFERENT RPM AND REACTIVE LOADS IN AA 2219. TWO DIFFERENT TOOLING MATERIALS WILL BE EMPLOYED IN ORDER TO INTRODUCE DIFFERENT LEVELS OF FRICTION AT THE INTERFACE. THERMOCOUPLES WILL BE EMBEDDED IN THE AL PLATE TO MEASURE TEMPERATURES NEAR THE INTERFACE. THEY WILL ALSO BE EMBEDDED IN THE TOOL. DATA COMING OUT OF THE EXPERIMENT WILL BE TORQUE REACTIVE LOAD AND TEMPERATURES IN THE PLATE AND THE TOOL. 2. A FINITE ELEMENT MODEL OF THE FLAT PINLESS TOOL EXPERIMENT WILL BE DEVELOPED AND COUPLED WITH AN INVERSE PARAMETER ANALYSIS ALGORITHM TO MODEL THE LOCAL FLOW STRESSES IN THE HGZ. 3. THE LOCAL FLOW STRESS RELATIONSHIP FROM TASK #2 WILL BE INCORPORATED INTO AN EULERIAN FINITE ELEMENT MODEL OF FSW DEVELOPED IN PRIOR EFFORTS BY THE PIS. 4. THE FSW MODEL WILL BE VALIDATED BY EXPERIMENTS SUCH THAT TEMPERATURES REACTIVE WELDING LOAD AND WELDING TORQUE ARE ALL PREDICTED SIMULTANEOUSLY WITH REASONABLE ACCURACY (5%). INITIAL VALIDATION WILL BE WITH A SMOOTH TOOL. AFTER APPROPRIATE AGREEMENT ON A SMOOTH TOOL A TOOL WITH A THREADED PIN WILL BE EMPLOYED IN THE MODEL AND VALIDATED BY EXPERIMENT. FINALLY A SELF-REACTING FSW TOOL WILL BE MODELED AND VALIDATED BY EXPERIMENT. MODEL RESULTS WILL BE USED TO PREDICT MECHANICAL PROPERTIES IN THE HEAT AFFECTED ZONE OF THE WELD. THE PROPOSED APPROACH WOULD PROVIDE NASA WITH A METHOD FOR CHARACTERIZING INTERFACE BEHAVIOR FOR A GIVEN ALLOY AND TOOL MATERIAL. THIS INTERFACE BEHAVIOR IS ESSENTIAL TO ACCURATE MODELING. FSW MODELS USING ACCURATE INTERFACE BEHAVIOR WILL BE MORE PREDICTIVE AND USEFUL FOR THE RAPID OPTIMIZATION OF PROCESS PARAMETERS AND TOOL DESIGNS.