COMBUSTION RESEARCH&FLOW TECHNOLOGY, INC. (CRAFT TECH) SBIR PHASE III PROJECT STATEMENT OF WORK TITLE: SPEE…

Public award description

COMBUSTION RESEARCH&FLOW TECHNOLOGY, INC. (CRAFT TECH) SBIR PHASE III PROJECT STATEMENT OF WORK TITLE: SPEED UP OF THERMAL/HEAT TRANSFER ANALYSIS USING A GPU ENABLED COMPUTING ENVIRONMENT PURPOSE: TO AGGRESSIVELY SPEED-UP THE TIME FOR ANALYSES OF HEAT TRANSFER AND THERMAL EFFECTS THROUGH THE UTILIZATION OF GPU (GRAPHICAL PROCESSOR UNITS) ENABLED COMPUTING. INTRODUCTION: THE DESIGN AND ANALYSIS DIVISION AT NASA SSC IS RESPONSIBLE FOR CONDUCTING RELIABLE AND MEANINGFUL ENGINEERING SIMULATIONS IN SUPPORT OF POTENTIAL AND ACTIVE ROCKET TEST PROGRAMS. GENERAL ANALYSIS OR DATA PROCESSING TASKS HAVE BEEN CONDUCTED USING IN-HOUSE OR COMMERCIAL ENGINEERING SOFTWARE ON STANDARD WORKSTATIONS. WHEN THE NEED ARISES FOR HIGHER-FIDELITY MODELING, SUCH AS 3D TRANSIENT COMPUTATIONAL FLUID/THERMAL/STRUCTURAL ANALYSES, SUPERCOMPUTING FACILITIES ARE USED TO TAKE ADVANTAGE OF MULTIPROCESSING AND THEREBY ENABLE RAPID TURN-AROUND TIMES. HOWEVER, THESE INFRASTRUCTURES ARE LIMITED IN AVAILABILITY AND CAN BE IMPRACTICAL TO USE FOR DAY-TO-DAY DESIGN AND ANALYSIS TASKS. HOWEVER, WITH THE ADVENT OF GRAPHICAL PROCESSING UNIT (GPU) COMPUTING TECHNOLOGY, MODELING TASKS THAT WERE RESTRICTED TO HIGH-END COMPUTATIONAL FACILITIES ARE NOW TANGIBLE ON LOWER COST HIGH-END WORKSTATIONS. GPU COMPUTING HAS THE POTENTIAL OF REDUCING MODELING TIME BY ONE TO TWO ORDERS OF MAGNITUDE COMPARED TO CURRENT CPU WORKSTATIONS. FOR NASA SSC, ONE OF THE MOST POTENTIAL ADVANTAGES OF GPU COMPUTING IS PERHAPS THE OPPORTUNITY TO PERFORM REAL-TIME COMPUTING, VISUALIZATION OR DATA PROCESSING DURING ROCKET ENGINE TESTING THEREBY PROVIDING FEEDBACK DIRECTLY TO THE TEST OPERATIONS PERSONNEL AND CUSTOMERS. IN A SBIR PHASE II CONTRACT (NNX08CA36C), CRAFT TECH USED THEIR NUMERICAL SIMULATION SOFTWARE (CRUNCH CFD CODE) TO DEVELOP A MODEL THAT CAN PREDICT PERFORMANCE OF KEY SYSTEMS IN HIGH ALTITUDE TEST FACILITIES FOR ROCKET AND SATELLITE PROPULSION ENGINES, INCLUDING COLLOIDAL/ION THRUSTERS AND NUCLEAR PROPULSION ENGINES, PROVIDE DESIGN SUPPORT BY SUPPLEMENTING CURRENT EMPIRICAL RULES, AND DIAGNOSE SYSTEM ANOMALIES AS WELL AS EVALUATE TRANSIENT PERFORMANCE. IN A SUBSEQUENTLY FUNDED PHASE III PROJECT, THEY INCLUDED THERMAL/HEAT TRANSFER ANALYSIS CAPABILITY TO THIS MODEL OF THE NEW SSC A3 TEST STAND. SCOPE: THE OBJECTIVE OF THIS TASK WILL BE TO AGGRESSIVELY SPEED-UP THE TIME FOR ANALYSES OF HEAT TRANSFER AND THERMAL EFFECTS THROUGH THE UTILIZATION OF GPU (GRAPHICAL PROCESSOR UNITS) ENABLED COMPUTING. THIS EFFORT WILL SIGNIFICANTLY ENHANCE THE CURRENT ANALYSIS FRAMEWORK BY FACILITATING ANALYSES OF COMPLEX HEAT TRANSFER PROBLEMS ON DESKTOP WORKSTATIONS WITH RAPID TURNAROUND THAT CAN SUPPORT A TESTING ENVIRONMENT. TRADITIONALLY SUCH ANALYSES ARE CARRIED OUT ON MULTIPLE PROCESSORS OR PC CLUSTERS THAT REQUIRE SIGNIFICANT COSTS FOR ESTABLISHING AND MAINTAINING A COMPUTING ENVIRONMENT. IN THIS PROGRAM, THE ALGORITHMS WILL BE INNOVATIVELY RE-ENGINEERED FOR DESKTOP COMPUTERS WITH GPU ENABLED ARCHITECTURES LEADING TO SUBSTANTIAL REDUCTION IN THE COST OF CARRYING OUT ANALYSES AND SPEEDUP IN TURNAROUND TIME. THE WORK UNDER THIS TASK MUST BE PERFORMED BEFORE SEPTEMBER 30, 2012 WITH THE FINAL COMPLETED DELIVERABLE DUE WITHIN 30 DAYS OF TASK COMPLETION. HOWEVER, DUE TO THE NATURE OF THE FUNDING SOURCE AND HQ EXPECTATIONS COMPLETION OF THIS ACTIVITY WITHIN SIX MONTHS OF RECEIPT OF THE AWARD IS GREATLY DESIRED. DELIVERABLES: PROGRESS REPORTS WILL BE SUBMITTED AS APPROPRIATE. THE FINAL DELIVERABLE WILL PROVIDE THE FOLLOWING: WORKING GPU THERMAL/HEAT TRANSFER ANALYSIS ALGORITHMS OPERATING INSTRUCTIONS FOR THE GPU ALGORITHMS FINAL REPORT TO INCLUDE PROBLEMS AND ISSUES IN THE CONVERSION TO THE GPU ENABLED ARCHITECTURE. SHIP TO: ALL REQUIREMENTS HEREIN SHALL BE DELIVERED TO THE FOLLOWING ADDRESS FOB