Crack line tooth的問題，透過圖書和論文來找解法和答案更準確安心。 我們找到下列推薦必買和特價產品懶人包

熱擠製鋁合金6061三階齒輪之研究：有限體積模擬與實驗的比較

為了解決Crack line tooth的問題，作者PAVAN SAI PAGADALA 這樣論述：

The study introduces a novel technique, bidirectional hot extrusion and uses a selective laser melting (SLM) machine with Inconel 718 alloy to additive manufacturing of a helical-spur-helical three-stage gear. An experimental investigation is carried out on Aluminum 6061 alloy by performing the Ten

sile test on the cylindrical specimen machined based on ASTM E-8M standard to compare the estimated results with the available library data. To determine the effectiveness of the introduced techniques, firstly a computer-based design followed by analysis is carried out with the help of MSC Simufact

forming and MSC Simufact additive. The analysis includes the Finite Volume Method (FVM) for the triple gear and the Powder Bed Fusion (PBF) technique for gear die. Convergence test has been performed using the FVM with Aluminum 6061 as the material based on varying mesh element size. Simulation is e

xecuted at three different temperatures 350°C, 400°C, 450°C, and numerous attempts were made to determine the optimal time and velocity for obtaining a more favorable gear profile and the computation time. With a constant optimal time of 10s and varying the mesh element size between 0.5 to 1.6 mm wi

th the input velocity of 1.75, 2.0 & 2.25 mm/s, it is observed that the die filling rate is very accurate, and crack formation on the tooth bed is minimum. The input parameters of PBF in Simufact additive such as scan velocity, laser power, beam width, and layer thickness are optimized to evaluate t

he residual stresses and distortions formed in the fabrication of gear dies which yielded satisfactory results. IFUM (Institute for forming technology and forming machines) model is employed to evaluate the material flow rate and underfilling of the gear tooth. Considering the simulation results, ex

periments are performed to fabricate the gear die using the SLM technique and the triple-gear using monodirectional hot extrusion (due to unavailability of required bidirectional hot extrusion machine setup). Surface machining is performed to the gear die to obtain a favorable gear profile closely m

atching the simulation results.

基於多輸出時間卷積神經網路之齒輪箱振動缺陷分析系統研究

為了解決Crack line tooth的問題，作者鄭睿閎 這樣論述：

迴轉機械透過齒輪箱將動力傳送至生產設備在各類的機械設備中已被廣泛的使用，但是因為經常性的碰撞及振動導致零件的損耗破壞，進而導致整個機械設備出現故障，影響生產效率的機會提高。因此本研究利用深度學習方式，以直接通過時序振動資料來開發一套齒輪缺陷檢測系統用於機械故障的診斷，希望能透過振動訊號持續監測，提前發現故障訊號並進行故障判斷以提供使用者進行預防保養。 本文首先利用更改後的時間卷積網路(TCN)架構結合改良後的異常檢測算法(TCN-AE)並使用無監督學習的方式進行缺陷檢測，以檢測時序齒輪振動資料中的異常值來確認目前機械系統是否正常。接著以更改的時間卷積網路架構，在其後添加壓縮及激

勵網路(SE-Net)再加上自動編碼器的架構完成本研究之SE-TCN-AE網路模型結構。最後分別將SE-TCN-AE網路模型結合分類損失函數交叉熵(Cross-Entropy)進行故障分類；結合模型回歸損失函數中的均方誤差(MSE)函數進行磨耗程度判斷。 實驗結果顯示使用無監督學習的異常檢測算法及使用SE-TCN-AE網路模型結合交叉熵進行故障分類之系統均能在有限樣本條件下就達到100%的準確度；使用SE-TCN-AE網路模型結合均方誤差函數進行磨耗程度判斷在三種磨耗程度訓練後之判斷MSE值約為 1.4×10^(-7)，若以輕重兩種磨耗程度訓練後去判斷中度磨耗齒輪組，其MSE為9.8×1

0^(-6)。證明了此方法在齒輪缺陷異常診斷以及缺陷分類上使用時序資料有著優異的結果。