Computational harmonic analysis on plate-like base panel of air conditioning outdoor unit using direct method / Choon Wei Feng

• Main Author: Choon, Wei Feng
• Format: Thesis
• Published: 2022
• Subjects: TK Electrical engineering. Electronics Nuclear engineering
• Online Access: http://studentsrepo.um.edu.my/14365/1/Choon_Wei_Feng.jpg; http://studentsrepo.um.edu.my/14365/3/wei_feng.pdf; http://studentsrepo.um.edu.my/14365/

Harmonic response analysis is a useful tool to check for the structural response (acceleration) when subjected to constant harmonic excitation. By extracting the response of the desired spectrum from the structure by using finite element method, the potential resonance or peaks with high vibration can be observed. In HVAC industries, the source of vibration are from unit transportation, fan motor vibration and compressor vibration, engineers will have to design and test the components of the unit to ensure the component can withstand the vibration. This study focusses on the effect of transportation vibration against the base panel of an air-conditioning outdoor unit. To ensure all component can withstand road and ship transportation vibration, Daikin utilizes hydraulic shaker with a prefix harmonic vibration profile to represent actual transportation condition. FEA can be used during design stage of component to check the vibration integrity. Commercial FEA solver is usually expensive, with its licensing cost. The alternatives would be using open�source FEA solver such as Code_Aster®. Usage on Code_Aster® will lead to extensive saving in licensing cost. To ensure the method of an open-source solver is valid, simulation results from Code_Aster® will be compared against commercial solver (Altair Optistruct®) currently use in Daikin. Modal and harmonic response analysis will be constructed on both solvers, with similar boundary condition and properties and proper keyword. To further enhance the validity, two experiment will be done to verify the results from the simulations. EMA are used to check the modal parameters such as natural frequency and mode shapes of the base panel against the result from FEA. Shaker with harmonic excitation from 5Hz-100Hz will be used to excite the base panel, accelerometers will be used to collect the response such as acceleration from different location. Response from simulation and experiment can be use as comparison and verification. After the validation process through experiment, Code_Aster® modeling technique can be integrated into design workflow of DRDM with confidence. When comparing Code_Aster® with Altair Optistruct®, the frequency difference are only within 1.89% and the MAC plot in all 8 modes are near to 1. This shows that the settings used are correct in Code_Aster®. The results validation for normal mode in Code_Aster® with EMAshows that mode 1, 2, 3, 5 and 7 are present and comparable, mode 4 and mode 6 are missing from EMA. This might be due to localized mode which is not captured by accelerometer. The maximum error on frequency against EMA is only 12.6%. Experiment harmonic response analysis can capture the peaks of resonance and the shapes are comparable to FEA although the magnitude in FEA is higher. Improvement can be made in FEA by using element with actual material data in weld and screw area to reduceoverall stiffness. More accelerometers can be used to measure the response on the panel to capture localized mode in multiple DOF, actual damping coefficient from EMA can beinputted into FEA to generate an accurate response.