A Highly Accurate Flow Measurement by Ultrasonic Flow Meter Based Simulation and Experiment
|Keywords:||超音波流量計;裝置效應;計算流體力學;校正因子;辛普森1/3規則;求積法;ultrasonic flowmeter;installation effect;computational fluid dynamics;calibration factor;quadrature||Issue Date:||2009||Abstract:||
精確的流量量測對於許多工業上的應用非常重要，以天然氣的交易而言，台灣目前每年的交易額約一千億，若流量計的器差平均為1.0%，且不修正這個器差，則每年交易的差額高達10億。因此，正確挑選及使用流量計，並定期執行流量計校正，對達到交易的公平性，實為重要的一大課題。較其他型式的流量計，超音波流量計擁有一些優點，例如：可自我診斷、無可動件、結構簡單、低維護成本。本論文的目的在於借著使用數值模擬及實驗，闡明流場效應對超音波流量計計量準確性的影響。超音波流量計被裝設於互處不同平面150 mm肘管及400 mm匯管的下游，流量計的平均速度借著應用適當的數值積分技巧來積分超音波流量計各軌的平均速度，並進而得到流量。數值積分技巧分別是適用於等間距的修正辛普森1/3規則，及適用於非等間距的高斯積分法及闕比闕弗的積分法。每軌超音波軌道的流體平均速度都已模擬及考慮到橫向流(cross flow)對平均速度的影響。論文成功的探討管流場對超音波流量計定性與定量的影響，流場因為90度彎頭、T型管、…等影響流場穩定的因素存在，若不裝整流器，所產生流場的不對稱性及渦漩，將會持續到43.5D以上。整流器可以迅速且有效的消除渦漩流，超音波流量計最好裝設於整流器下游至少10D以上。對於整流器裝設於肘管下游3.5D，計量的預測差值都小於0.36%。對於實驗，一具英斯托梅特(Instromet) Q sonic-3超音波流量計置放於肘管3.5D – 43.5D下游，不論有無整流器，對於每一種測試條件，器差都在0.0%~-0.8%之內，這器差與數值模擬，流量積分技巧採用辛普森1/3規則乘以修正係數法、高斯積分法及闕比闕弗的積分法，4軌以上之超音波流量計，位於整流器下游10D以上之預測差值，在相同的級數之內，隱含流場模擬的正確性，並提供更多管線設計的指引。
A highly accuracy of flow measurement is very important for many industrial applications. In Taiwan, a turnover of natural gas custody transfer reaches 100 billion NTD per year. If an error goes up to 1% for a flowmeter with no correction, the amount of loss in trade achieves one billion per year. As a result, selection, operation and calibration of a flowmeter is of great importance in trade business.everal advantages of an ultrasonic flowmeter over others are self-diagnosis, no moving parts, simple structure and low maintenance cost. The thesis investigates the flow-field effect on ultrasonic flowmeter, and demonstrates the superior results in accuracy by using numerical and experimental approaches. The ultrasonic flowmeter was experimentally installed downstream in a pipeline with an out-of-plane 150 mm single elbow and thus following along a 400 mm header. Averaged velocity of the flowmeter was integrated over averaged velocities of all paths by applying suitable integration techniques, and then obtained the flow characteristics of the flowmeter. The specific numerical integration method for equal interval was developed based on the Simpsons’ 1/3 rule method in accordance with the configuration of the experimental ultrasonic flowmeter, but Gaussian and Tchebychev methods for unequal interval methods generally. his paper explored successfully the qualitative and quantitative analysis of flow field in pipeline imposed on ultrasonic flowmeter. With the installation of the flow conditioner, the non-symmetry flow and swirl flow of disturbant factors were effectively reduced. Meanwhile, the flow field was nearly stable from a distance of 43.5D. The optimum position of flow meter was located above 10D downstream of flow conditioner. With a typical case of flow conditioner installed 3.5D downstream of the 90° elbow, the flow rate metering deviated was less than 0.36%. The Instromet Q.sonic-3 meter that was placed at 3.5D–43.5D downstream of an elbow showed the error of within 0.0%~-0.8% in a condition with and with no a flow conditioner. These experimental results were in an agreement with the predicted errors of simulation of the flow integration technique for the ultrasonic flowmeter of beyond 4 paths. This indicates the validation of the simulation approach to the experiment practically, and provides a guide of pipelines for further design in modular package.
|Appears in Collections:||應用力學研究所|
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.