選擇性雷射熔融之單線軌跡建模與驗證

Abstract

本論文以選擇性雷射熔融系統，透過雷射功率與掃描速度等調整下，進行單線熔融軌跡之連續性結構的實驗探討，並與ANSYS 14.0有限元素軟體的模擬數據進行比對，以達到模擬分析與實驗驗證之目的。在雷射波長1070 nm、雷射聚焦光斑直徑40 μm與雷射功率分別為60 W、79 W及調整掃描速度10 ~ 150 mm/s (每10 mm/s遞增)等條件下，獲得共30條單線熔融軌跡，並依光學顯微鏡觀測其單線軌跡之表面形貌，判斷出最佳穩定連續成型速度區間分別為50 ~ 90 mm/s與40 ~ 100 mm/s。透過橫切雷射功率79 W之15條單線熔融軌跡，而取得其斷面形貌，並以掃描式電子顯微鏡量測軌跡與底板間的界面熔寬，及使用能量分散式X -ray元素分析儀量測軌跡熔入底板的底板熔深，可得隨掃描速度變化的界面熔寬及底板熔深等兩關係折線圖。從兩折線圖可發現，皆因掃描速度提高而有緩慢下降的趨勢。但掃描速度過高達到臨界狀態(130 mm/s)時，從底板熔深的折線圖可發現驟降的現象。 在模擬部分，考量實際雷射與粉床的吸收作用深度有別於雷射與塊材的情形，將分別採取不同的方式施加熱源。從溫度場分析數據，可得雷射功率79 W及掃描速度40 ~ 90 mm/s時，其界面熔寬之模擬與實驗結果的百分誤差為10 ~ 15 %。而雷射功率79 W及掃描速度40與60 ~ 100 mm/s時，其底板熔深之模擬與實驗結果的百分誤差為15 ~ 30 %。

In this thesis, effects of the processing parameters such as laser power and scanning speed on single molten tracks formed are investigated during the selective laser melting (SLM) Process. The simulation results solved from commercial finite element software (ANSYS 14.0) are validated by the experimental results. Thirty single molten tracks were formed by the laser wavelength 1070 nm, laser focused beam diameter 40 μm, and laser power of 60 W and 79 W with scanning speeds in the range of 10 ~ 150 mm/s. The best formed continuous tracks are determined by measuring the surface morphology of single tracks from the optical microscopy. The experimental results show that the best continuous tracks are formed at a scanning speed from 50 to 90 mm/s at the laser power of 60 W and from 40 to 100 mm/s at the laser power of 79 W. At the laser power of 79 W, we obtain two line charts with variation of scanning speed from measuring the interface width from the cross section morphology of single tracks by the Scanning Electron Microscope and analyzing the remelted depth from the cross section morphology of single tracks by the Energy Dispersive X-ray Analyzer. Both of the two line charts decrease slowly with the increasing of scanning speed. When the scanning speed reaches a critical state (130 mm/s), the line chart of the remelted depth drops drastically. In the numerical simulation, this study taking account of the absorption penetration depth in the powder bed is different from that obtained in an opaque solid material. This model is applied to the heat source by taking different methods. From analyzing between the simulation and experimental results for the laser power of 79 W and scanning speed regime 40 - 90 mm/s, the percentage error of the interface width is 10 ~ 15 %. The percentage error of the remelted depth between simulation and experimental results for the laser power of 79 W and the scanning speed of 40 or 60 ~ 100 mm/s is 15 ~ 30 %.