Design Validation and Improvement Study of HVAC Plumbing Line Assembly under Random Loading Condition Rakesh Jakhwal Senior Engineer Chrysler Group LLC RMZ Millenia II, Perungudi Chennai 600096, India Yogesh Kumar Senior Engineer Chrysler Group LLC RMZ Millenia II, Perungudi Chennai 600096, India Muthukumar Arunachalam Manager Chrysler Group LLC RMZ Millenia II, Perungudi Chennai 600096, India Abstract Automotive OEMs are currently addressing faster product development cycle needs with increasingly robust and efficient virtual validation processes facilitated by improved CAE tools and techniques. This includes developing reliable virtual methods for complex problems. The objective of this paper is to present a validation process for HVAC Plumbing line Assembly under random vibration loading condition for a specific test setup. The process is validated by correlating virtual prediction of potential problems locations on the plumbing line to the corresponding observations in physical test. The method is then used to develop design modifications in the virtual models to address and mitigate the problem which is then verified in the physical re-test. The virtual methodology has been currently developed on HyperWorks platform (version 11.0). Introduction: The purpose of the HVAC plumbing assembly is to provide a reliable leak-free and efficient method of transferring refrigerant and oil between air conditioning components at high and low temperatures and pressures. HVAC Plumbing shall be routed and supported in a manner which ensures functionality during life cycle of the product. HVAC Plumbing must be routed so as to either be sufficiently constrained or have sufficient clearance with surrounding components to prevent harmful contact. Suction Line-Tube that convey refrigerant gas or vapor from evaporator outlet to the suction inlet compressor. Discharge Line Tube that convey the compressed refrigerant vapor from compressor to the condenser inlet. Liquid Line Tube used to convey liquid refrigerant from the condenser outlet to the refrigerant control device of evaporator. 1
Figure 1: Plumbing line routing sketch diagram An automobile traveling down a rough road is subjected to random excitation. The Random vibration test verifies the components capability to meet applicable functional/performance requirement during or after exposure to the service life random vibration environments. Real world vibrations are usually of the random type. Figure 2: (a) sine wave (b) Random profile (c) PSD curve The difference between sine vibration and random vibration is that for random vibration, at any point in time, any frequency can be happening that will excite, or resonate, multiple bodies at once. What this means is although one member of the system may have a low frequency, and another high, they can both be resonating at the same instant in a random vibration environment. The sine environment will run the input curve over the frequency band from 20 to 2000 hertz that can be experienced in the vehicle and only one member will resonate at a time, assuming of course two members do not share a natural frequency. Most of the time a sine scan is run to determine the resonances of the vehicle. A 1G input over the entire frequency band is used, and is not meant to induce fatigue damage. This data is used to validate the FEA before certification or qualification testing occurs. 2
Sigma (σ) & Sigma Clipping: Figure 3: sigma clipping Sigma is the standard deviation of a statistical PDF. A Gaussian PDF distribution is assumed for random vibration which takes the shape of a bell shaped curve. Since the amplitude or intensity of the random vibration will change over time, the time spent at different amplitude excursions is measured using a PDF. Figure 3 shows a Gaussian PDF. The vertical axis would be 1/G, the horizontal axis would be sigma and µ is the mean which is equal to zero for a shaker control. For a Gaussian distribution, 68.2% of the peak G excursions occur between ± 1 sigma, 95.4% between ± 2 sigma, 99.7% between ± 3 sigma. Process Methodology: Figure 4: HVAC Plumbing line Validation Process 3
Physical Test Setup: Test set up placed in a group of 4-6 samples to be tested in tri - axial shaker table with bolted with the location as per vehicle condition. Generic random vibration profile as per standard provided to run specific hours for each individual axis. Figure 5: Controlled input PSD profiles for shaker table Performance Criteria: Vibration testing shall be performed with plumbing line assembly installed in bench test condition. The acceptance criteria are: - No issues, Leak Test prior to the vibration test for each axis for specifichrs. cycle run. - No issues s (cracks, deformation, etc.) of add-on components such as brackets, clips, isolators, transducer, service valves, etc., following evaluation, Florescent Liquid Penetrate Inspection. CAE process and correlation: Meshing -AC plumbing line FE modeling is done as per standard guidelines. Boundary Condition -Plumbing Line is constrained as per Physical Test shown above. Analysis -Modal analysis is performed to check the natural frequency of the plumbing line assembly, followed by random vibration analysis using vehicle road profile. Problem could be predicted when the 3 sigma stress is higher than the material yield stress. Example of which is shown below. Figure 6: CAE Model Set Up 4
Results & Discussions: Two- Piece bracket design Baseline Physical Test Problem observed beforespecific hours in first axis at marked location Virtual Test Maximum 3-sigma RMS stress isobserved more than material yieldstress at marked location Figure 7: Physical and CAE results for baseline bracket CAE simulation correlates well with the Test. 3-sigma RMS stress are higher than the Yield stress of the bracket. Design Iterations were carried out to avoid problems with material and gauge change. Design Modification Baseline - 2 pieces Bracket material aluminum. Proposed - single piece Bracket Material aluminum with bead. Figure 8: Bracket design modification 5
CAE proposed- single piece bracket design with bead Physical Test No issue observed tillspecific hours in the test. Virtual Test Maximum 3-sigma RMS stress isobserved less than material yieldstress. Figure 9: CAE and Physical Test Results for Modified Bracket Benefits Summary: The random vibration analysis tool reduces the product development phase as virtual tools consumes less time. It helps to evaluate the performance of a design before cutting tools or physically building prototypes. The cost of the design verification in the different stages eliminates by reducing number of physical test or sample size. Challenges: The major challenge with random vibration analysis to solving time and file size is more. Nonlinearity affect need to take care for this type of analysis. Extra care need to be carried for averaging and filtering the data from road profiles to make PSD curve. Future Plans: Current CAE validation using random vibration analysis can be extended to fatigue life prediction which would yield the information of number of cycles. Conclusions: It is necessary to take structural loads due to sine and random environment to define the frequency domain and reach a conclusion. Even though there have been many advances in CAE, and it is widely used in the engineering field, physical testing is still used as a final confirmation for subsystems due to the fact that CAE cannot predict all variables in complex assemblies. 6
ACKNOWLEDGEMENTS We would like to thank the Chrysler Management for their continuous support in accomplishing this task and also like to thanks to design and testing team for continuous support to execute the process. REFERENCES [1] Francis J. Andrews: Random Vibration An Overview [2] Tom Irvine : Power Spectral Density units:[g^2/hz] Revision A [3] Tom Irvine: An Introduction to Random Vibration Revision B [4] OptiStruct - 11.0 reference guide [5] NWM Bishop: Vibration Fatigue Analysis In Finite Element Environment, 1999 7