在鋁合金的製造過程中，熱處理是強化鋁合金的關鍵步驟，然而，淬火過程中強烈的熱梯度會導致較高的殘留應力，消除鋁板的殘留應力可以避免在後續加工操作中產生變形。本研究的目的是採用有限元方法 (FEM) 來模擬商用尺寸鋁板在淬火過程中的溫度場以及伴隨產生的殘留應力，以ANSYS APDL 軟體來建立有限元模型，也模擬拉伸消除殘留應力的過程；在模擬中，使用了隨溫度變化的熱對流係數及材料性質。 另外，使用尺寸為90 mm x 90 mm x 16 mm的試片進行淬火實驗，與模擬結果比對，以驗證FEM模型的有效性。實驗過程，整塊試片加熱到475 °C之後，在很短的時間內浸入水中進行水淬，並藉由黏貼於試片表面中心的K-type熱電偶來量得淬火過程中的溫度變化。鋁塊與水之間的熱對流係數藉由逆向熱傳導分析來求得，隨後用於模擬計算實驗試片的殘留應力。模擬計算的殘留應力結果與實驗量測具有合理的一致性，確認FEM模型的有效性及應用可行性。 對於厚度為20-80 mm的商用尺寸鋁板，淬火後的殘留應力隨著厚度的增加而增加，並且會趨近於一個飽和值。在沒有夾持效應下的理想拉伸狀態，施加1.5% 至4.5%的拉伸量來釋放殘留應力，結果顯示，在拉伸3.5%之後，殘留應力之最大張應力以及最大壓應力皆可以被消除達95%。隨著拉伸量增加至4.5%，殘留應力的消除效果並沒有進一步的增加。端面的夾持效應會導致拉伸過後的應力分布區分為夾持區、過渡區及均勻區。在夾持區的應力並沒有明顯被消除，而消除後的殘存應力是從過渡區向均勻區逐漸減小。在均勻區內的應力可以得到顯著的消除效果，但是在過渡區則還會有較大的殘留應力存在。 以應變的觀點來看，在理想拉伸狀態下拉伸量為3.5% 時，為了消除殘留應力所需要的真實應變為0.035。對端面夾持的拉伸狀況而言，在夾持區域內會有較小的真實應變值存在，但是真實應變會從過渡區向均勻區逐漸增加；當施以2.5% 的公稱拉伸量時，3.3% 的真實應變值會在夾持區長度為150 mm時發生在均勻區，而3.9% 的真實應變則會在夾持區長度為300 mm時發生在均勻區，也因此讓均勻區內的應力得以有效的消除。 ;In the fabrication process of aluminum alloys, heat treatment process is a vital step to strengthen the aluminum alloys. However, severe thermal gradients during the immersion quenching process results in high magnitude of thermally-induced residual stresses. Reducing the residual stress in the aluminum plate is able to avoid distortions in the following machining operations. The aim of this study was to employ the finite element method (FEM) to simulate the temperature field during the quenching process and the resulted residual stresses in commercially-sized aluminum plates. An FEM model with ANSYS codes was developed and used. Furthermore, stress relief by tensile stretching was also simulated. Temperature-dependent heat transfer coefficients (HTCs) and material properties were considered. A specimen size of 90 mm x 90 mm x 16 mm was used to conduct the quenching experiment for validation of the FEM model. The block was heated to 475 °C and then immersed in a bucket of water within a short period of time. The center on one side of the specimen was attached with a K-type thermocouple to measure the temperature history during the quenching process. The HTCs were calculated by an inverse heat transfer analysis and applied in the calculation of residual stress in the specimen. The simulation results showed a reasonable agreement with the experimental results. For commercially-sized aluminum plates with a thickness of 20-80 mm, the residual stress after quenching increased and reached a saturated level with increasing thicknesses. Further, tensile stretching ratios of 1.5% to 4.5% were employed to relieve the residual stress under ideal stretching without end-clamping. With 3.5% stretching, both maximum tensile and compressive parts of residual stress were reduced by 95%. No significantly further improvement was found with the increase of stretching ratio to 4.5%. The stress distribution on the plate with end-clamping after tensile stretching was divided into clamping region, transition region, and uniform region. The stress in the clamping region was not relieved significantly while it was increasingly reduced from the transition region to the uniform region. In the uniform region, the stress was relieved to a significant extent while a higher magnitude of residual stress still existed in the transition region. In terms of mechanical strain, the selected optimal stretching ratio of 3.5% in ideal stretching indicated the true strain needed to effectively relieve the residual stress was 0.035. For stretching with end-clamping, a rather small true strain existed in the clamping region while the true strain increased from the transition region to the uniform region. For a nominal stretching ratio of 2.5%, a true strain of 3.3% and 3.9% in the uniform region was achieved with 150-mm and 300-mm clamping length, respectively, which resulted in an effective stress relief in the uniform region.
