【正文】 3226015443703056表8 輪罩沖壓的L27(313)正交表和有限元仿真結(jié)果 試用號工藝參數(shù)有限元仿真壓邊力坯料偏移距離拉延高度拉延半徑斷裂（Ff）皺紋（Fw）111112112231133412125122361231713138132191332102111112122122133132212142223152231162313172321182332193111203122213133223212233223243231253313263321273332表9 輪罩沖壓斷裂特性的方差分析表參數(shù)平方和自由度均方F 比率壓邊力2 b坯料偏移距離2 b拉延高度2 b拉延半徑2交互作用壓邊力 x ?a4壓邊力 x ?a4錯誤錯誤10Ep.20總計26注：a表示輪流檢測b這是重要的結(jié)論，特別是對于實際的沖壓流程設(shè)計，其中的張力受制于坯料尺寸，壓邊力和拉延筋的形狀。該可行的成形性圖被主要工藝參數(shù)，如初始坯料形狀，壓邊力，拉延筋的形狀同時考慮，這些都與沖壓過程有關(guān)。Definition of characteristic values on the basis of the forming limit diagramView Within Article. Procedure of process designThe procedure for process design of stamping to produce sound products without defects on the basis of the feasible formability diagram is shown in Fig. 4. The procedure includes the following steps. FEsimulation and design of experiment (DOE) as well as ANN are applied to the description of the feasible formability diagram. For ANN training, the bination of design parameters in OA table are used as inputs and the corresponding characteristic value for facture and wrinkle are used as objective value. The structure of the neural network consists of an input layer, output layer and four hidden layers having 20, 20, 10 and 5 neurons, respectively. ANN is trained until root mean square (rms) error is less than 10?7.(i) Determine the target contour with a uniform trimming width at the outline of the final product.(ii) Perform FEsimulation by using the assumed initial shape of blank.(iii) Modify the initial shape of blank through an inverse approach.(iv) Perform step (iii) repeatedly until the shape error of Eq. (v)Perform FEsimulation using the shape of blank obtained from step (iv) under the condition of maximum offset distance of blank and maximum blank holding force within the capacity of the given press.(vi) Using。（3）可在可行的成形性圖看出，隨著坯料偏移距離，壓邊力，拉延筋高度和拉延半徑的增加，由于過度拉伸應(yīng)力發(fā)生在坯料的邊緣，斷裂區(qū)域增大，皺紋區(qū)域減少。較高的壓力增加了主要應(yīng)變，減少了產(chǎn)品邊緣的起皺。x ? a4壓邊力 Ra4錯誤錯誤10Ep.20總計26注：a表示輪流檢測b如圖2所示，一個臺肩的拉延筋受制于這項研究中，并且各種不同高度（h）和肩半徑（r）的拉延筋是多種多樣的。= ， ? 90對于整個工藝參數(shù)的范圍，應(yīng)用人工神經(jīng)網(wǎng)絡(luò)找到未在表3中列出的組合特征值。 一個長方形的560 500毫米的坯料是用來設(shè)計初始坯料的。懸架在不對稱的深度為110毫米的大圖中，在沖壓時可能會導(dǎo)致產(chǎn)品的邊圍斷裂。 （三）通過逆算法修改初始板料形狀。各種形狀的拉延筋被賦予不同參數(shù)，如拉延筋的高度和肩部半徑，如圖2。 傳統(tǒng)上把板料最佳形狀稱為毛坯初始形成生產(chǎn)所需要的形狀，其中無論是完全消除或削減焊縫過程。然而，如果改變工藝條件之一，上述被提及的大部分的優(yōu)化設(shè)計程序可能一再的需要重復(fù)有限元與不同的工藝參數(shù)組合。郭及其他人。作 者 簽 名：　　 　 　　日　 期：　　 　 　　 指導(dǎo)教師簽名：　　 　 　　 日　　期：　　 　 　　 使用授權(quán)說明本人完全了解 大學(xué)關(guān)于收集、保存、使用畢業(yè)設(shè)計（論文）的規(guī)定，即：按照學(xué)校要求提交畢業(yè)設(shè)計（論文）的印刷本和電子版本；學(xué)校有權(quán)保存畢業(yè)設(shè)計（論文）的印刷本和電子版，并提供目錄檢索與閱覽服務(wù)；學(xué)校可以采用影印、縮印、數(shù)字化或其它復(fù)制手段保存論文；在不以贏利為目的前提下，學(xué)?？梢怨颊撐牡牟糠只蛉績?nèi)容。汽車板件的沖壓過程，包括支持諸如暫停臺架懸置和輪罩模塊等，已被作為實例來驗證了成形工藝流程設(shè)計圖中的效果。沖壓流程的設(shè)計是非常重要的，因為它可能產(chǎn)生沒有缺陷產(chǎn)品，如開裂和皺紋。魏等人[7]提出了一種工藝參數(shù)的優(yōu)化方法，并預(yù)測關(guān)于覆蓋件外板的沖壓在性能方面的公差。從有限元模擬的結(jié)果與實驗結(jié)果對比看出，汽車板沖壓流程，支持懸掛模塊諸如臺架懸置和駕駛室模塊，作為實例驗證了可行的成形工藝設(shè)計圖中措施的效果。板料的上下界偏移距離最大值是板料沖壓分別成為目標(biāo)輪廓和板料形狀被放大到最終沖壓模面后。對于神經(jīng)網(wǎng)絡(luò)訓(xùn)練，對于表面加工和皺紋，把設(shè)計參數(shù)組合在OA表格中的輸入和相應(yīng)的特征值作為目標(biāo)值。壓邊力，板料偏移距離和拉延筋的形狀，是與薄板壓緊力有關(guān)的主要參數(shù)并且在可行的成形性圖上作為工藝參數(shù)決定的。通過一般的50噸容量的金屬片實驗機(jī)拉伸板材，[17]，直到標(biāo)本發(fā)生破裂。在實際工業(yè)應(yīng)用中，對于沖壓能力和模面1和4是密切相關(guān)的。這項研究使用的材料是冷軋鋼板。在圖16a中可以看出拉延筋被用來提供給坯料一個附加約束力，因為由于嚴(yán)重扭曲坯料元素在邊緣皺紋?？尚械某尚涡詧D可在神經(jīng)網(wǎng)絡(luò)培訓(xùn)結(jié)果的基礎(chǔ)上確定，如圖17。從這些數(shù)字中可以看出，有限元仿真的結(jié)果與試驗得出的結(jié)果高度一致。在這項研究中一個通過可行的成形性圖的沖壓流程設(shè)計已經(jīng)被提出了。 Automotive panelArticle Outline1. Introduction2. Process design through feasible formability diagram. Process variables. Estimation of characteristic values. Procedure of process design3. Process design of automotive panels. Turret suspension. Wheel house4. Experimental verification5. Discussions6. ConclusionsAcknowledgementsReferences1. IntroductionIn metal forming technologies, the stamping process for sheet metal is one of the significant manufacturing processes in the production of sheet metal ponents. Stamping technology has been extensively applied in the automotive industry. The formability of stamping products is generally influenced by various process variables such as the shape of the die, material properties, the shape of the initial blank, the blank holding force, the layout of the draw bead, lubrication. It is very important to design stamping processes that can produce sound products without defects, such as fracture and wrinkle. The design of stamping processes has been mainly performed by either a trialanderror approach, which is both time and costintensive, or Finite Element analysis (FEanalysis) bined with optimal design procedure, which poses some problems in actual industrial applications [1], [2], [3], [4], [5], [6] and [7].Since the formability and product quality in stamping processes depend on the initial blank shape, the optimal design of blanks has been investigated by many researchers. Lee and Huh [8] suggested an inverse finite element approach for the prediction of the blank shape. Guo et al. [1] conducted the optimal design of blanks on the basis of the variation in the thickness of the sheet material. A method for the design of optimal blank shape that uses the initial nodal velocity was proposed by Son and Shim [9]. Yeh et al. [10] suggested a forwardinverse prediction scheme to determine the optimal blank shape. Although the methods mentioned above are excellent, there still remain problems when the methods are applied for the optimal design of blanks in actual industrial problems with various process variables.In recent years, much research has focused on the bination of FEanalysis and optimization technology to optimize stam