壓縮包內(nèi)含有CAD圖紙和說(shuō)明書(shū),均可直接下載獲得文件，所見(jiàn)所得，電腦查看更方便。Q 197216396 或 11970985
任 務(wù) 書(shū)
院（系）： 專業(yè)：
班 級(jí)： 學(xué)生： 學(xué)號(hào)：
1、 畢業(yè)論文課題 直角多孔板沖壓連續(xù)模設(shè)計(jì)
2、 畢業(yè)論文工作自 20xx 年 3 月 12 日起至 20xx 年 6 月 15 日止
三、畢業(yè)設(shè)計(jì)進(jìn)行地點(diǎn) 學(xué)院
四、畢業(yè)設(shè)計(jì)的內(nèi)容要求
(一) 設(shè)計(jì)之原始數(shù)據(jù)：
原始資料： “直角多孔板”實(shí)物一件
(二) 設(shè)計(jì)計(jì)算及說(shuō)明部分內(nèi)容：
1.計(jì)算內(nèi)容與方案確定：
（1）成形零件設(shè)計(jì)：凹、凸模尺寸的計(jì)算和布置。
（2）壓力機(jī)的選擇
（3）結(jié)構(gòu)系統(tǒng)設(shè)計(jì)計(jì)算：卸料力、推件力、頂件力的計(jì)算，排樣圖和工序件圖的設(shè)計(jì)計(jì)算，各工序沖裁力的計(jì)算，沖裁、拉深和壓制成形各工序間尺寸的計(jì)算，各工序沖裁力的計(jì)算 。
（4）強(qiáng)度設(shè)計(jì)和結(jié)構(gòu)草圖設(shè)計(jì)：各部件的強(qiáng)度校核。
2. 設(shè)計(jì)內(nèi)容：
（1）落料、沖孔、拉深多工位級(jí)進(jìn)模模具設(shè)計(jì)（用計(jì)算機(jī)完成繪制裝配圖）；
（2）各工序的凹、凸模機(jī)構(gòu)、卸料機(jī)構(gòu)、導(dǎo)料機(jī)構(gòu)和排樣固定等機(jī)構(gòu)的設(shè)計(jì)；
（3）編寫(xiě)設(shè)計(jì)（論文）說(shuō)明書(shū)（不少于2.0萬(wàn)字，全部用計(jì)算機(jī)輸出）；
（4）綜述文獻(xiàn)（要求書(shū)寫(xiě)6000～8000字與畢業(yè)設(shè)計(jì)內(nèi)容相關(guān)的綜述文章）
(三) 主要參考資料
1、薛啟翔《沖裁模具設(shè)計(jì)結(jié)構(gòu)圖冊(cè)》，化學(xué)工業(yè)出版社。2005.3
2、郝濱?！稕_壓模具簡(jiǎn)明設(shè)計(jì)手冊(cè)》，化學(xué)工業(yè)出版社 2004.11
3、《材料力學(xué)》，高等教育出版社。
4、高軍《沖壓模具標(biāo)準(zhǔn)件選用與設(shè)計(jì)指南》，化學(xué)工業(yè)出版社2007
5、林楓英 《沖壓模具設(shè)計(jì)基礎(chǔ)》，電子工業(yè)出版社 2005.11
6、《沖壓模具設(shè)計(jì)與制造過(guò)程仿真》, 化學(xué)工業(yè)出版社2007
(四)附屬專題
1、專題外文翻譯
檢索與閱讀和設(shè)計(jì)題目相關(guān)的外文資料，并書(shū)面翻譯4篇(3000～5000漢字)外文資料。
指導(dǎo)教師
接受畢業(yè)設(shè)計(jì)任務(wù)開(kāi)始執(zhí)行日期 年 月 日
學(xué)生簽名
附件一 英文文獻(xiàn)翻譯
譯文：
板料沖壓成形工藝與模具設(shè)計(jì)制造中的若干前沿技術(shù)
摘要
板料成形技術(shù)是最重要的金屬成形方法之一,廣泛應(yīng)用于工業(yè)領(lǐng)域。因此,成形制造技術(shù)及裝備水平已成為衡量一個(gè)國(guó)家制造水平的重要標(biāo)志,在很大程度上決定產(chǎn)品的質(zhì)量、效益和開(kāi)發(fā)能力,決定著一個(gè)國(guó)家制造業(yè)的國(guó)際競(jìng)爭(zhēng)力。結(jié)合相關(guān)研究領(lǐng)域的發(fā)展趨勢(shì)和最新研究成果及其工業(yè)應(yīng)用,針對(duì)目前板料成形技術(shù)發(fā)展現(xiàn)狀,提出板料沖壓成形工藝與模具技術(shù)的若干發(fā)展前沿問(wèn)題。
2 板料沖壓工藝優(yōu)化設(shè)計(jì)技術(shù)中的若干前沿問(wèn)題
板料成形過(guò)程是一個(gè)非常復(fù)雜的塑性成形過(guò)程，許多因素都直接或間接地影響到成形件的質(zhì)量。如果工藝參數(shù)、幾何參數(shù)以及材料等選擇不當(dāng)，易產(chǎn)生起皺、破裂和回彈等缺陷。傳統(tǒng)的工藝分析和模具設(shè)計(jì)主要依靠設(shè)計(jì)者的經(jīng)驗(yàn)，反復(fù)修改各種參數(shù)，并進(jìn)行修模，耗時(shí)長(zhǎng)，成本高，難以適應(yīng)現(xiàn)代工業(yè)的要求。隨著計(jì)算機(jī)技術(shù)和數(shù)值計(jì)算方法的發(fā)展，數(shù)值模擬技術(shù)逐漸成為板料成形分析的一種重要手段。有限元模擬分析已廣泛應(yīng)用于板料成形優(yōu)化設(shè)計(jì)中，然而，對(duì)有限元分析結(jié)果的判斷和相關(guān)參數(shù)的選擇依然依賴于經(jīng)驗(yàn)。而最優(yōu)化技術(shù)的引入，使人們從試錯(cuò)(Trial-error)的工作中解脫出來(lái)。在沖壓成形優(yōu)化領(lǐng)域，優(yōu)化方法可以分為三類：傳統(tǒng)優(yōu)化，啟發(fā)式優(yōu)化和近似模型優(yōu)化。由于板料沖壓成形過(guò)程是典型的強(qiáng)非線性問(wèn)題，而傳統(tǒng)的優(yōu)化方法是基于梯度的，對(duì)于Hessian 矩陣為奇異的優(yōu)化問(wèn)題，顯然難以處理沖壓成形優(yōu)化問(wèn)題；啟發(fā)式算法包括遺傳算法、粒子群、擬退火等算法，這類算法可以找到優(yōu)化問(wèn)題的全局優(yōu)化點(diǎn)，可以保證優(yōu)化的精度和效果。但是，優(yōu)化求解中需要反復(fù)修正CAE模型并調(diào)用相應(yīng)的有限元程序，這需要耗費(fèi)大量的計(jì)算資源。隨著設(shè)計(jì)參數(shù)數(shù)量和復(fù)雜度的增加，不僅計(jì)算成本會(huì)成倍增加，而且計(jì)算時(shí)間會(huì)大幅度增加。基于近似模型的優(yōu)化方法是求解大型問(wèn)題最有希望的方法之一，鑒于其高效性，廣泛應(yīng)用于工程優(yōu)化領(lǐng)域，如果能夠高效建立可靠的近似模型，很多大規(guī)模的工程問(wèn)題可以迎刃而解。近年來(lái)，隨著研究的深入，該技術(shù)也大量應(yīng)用于沖壓成形問(wèn)題的優(yōu)化，成為目前沖壓優(yōu)化領(lǐng)域的研究熱點(diǎn)。
2.1 試驗(yàn)設(shè)計(jì)方法
通過(guò)試驗(yàn)發(fā)現(xiàn)，樣本點(diǎn)的布置對(duì)具體的工程問(wèn)題以及相應(yīng)近似模型的構(gòu)建具有較強(qiáng)的敏感性，在有限的樣本點(diǎn)范圍內(nèi)，采用不同方法生成的樣本點(diǎn)會(huì)導(dǎo)致優(yōu)化結(jié)果的差異。針對(duì)這些問(wèn)題，國(guó)內(nèi)外很多學(xué)者提出了縮減空間的概念。空間縮減技術(shù)分為兩類，一類是建立在敏感性分析基礎(chǔ)上的設(shè)計(jì)變量篩選方法[23]。BROWNE 等[24]通過(guò)分析凸凹模幾何形狀、壓邊力、凸模壓力、摩擦和拉伸速度等對(duì)板厚為0.19 mm 的筒形件成形的影響，采用試驗(yàn)設(shè)計(jì)的方法進(jìn)行了大量的試驗(yàn)，結(jié)果表明：除了壓邊力和拉深速度對(duì)最終成形性能影響不大之外，其他參數(shù)對(duì)成形性能均有較大影響。NAKAMURA 等[25]采用敏感性分析的方法將影響目標(biāo)函數(shù)的7 個(gè)參數(shù)減少到2 個(gè)，有利于優(yōu)化的進(jìn)行。另一類是基于設(shè)計(jì)空間的縮減方法。前者主要是通過(guò)敏感性分析，篩選出與目標(biāo)函數(shù)相關(guān)性大的參數(shù)作為設(shè)計(jì)變量。后者則是建立在工程經(jīng)驗(yàn)和空間分類技術(shù)之上，通過(guò)對(duì)設(shè)計(jì)變量的分析，將設(shè)計(jì)變量控制在合理的范圍內(nèi)，并建立更為精確的近似模型。目前已經(jīng)發(fā)展的技術(shù)包括：基于分層技術(shù)的小設(shè)計(jì)空間方法[26-27]，移動(dòng)控制布點(diǎn)技術(shù)[28]、基于信任域的布點(diǎn)方法[29]、自適應(yīng)試驗(yàn)設(shè)計(jì)方法[30]、多層次搜尋方法[31-32]等。同傳統(tǒng)的試驗(yàn)設(shè)計(jì)方法相比，這類方法將設(shè)計(jì)空間控制在比較小的范圍內(nèi)，在減少樣本點(diǎn)數(shù)目的同時(shí)，避免了全局?jǐn)M合，使近似模型的精度大為提高。這些方法通常建立在某些假設(shè)之上，如凸集合連續(xù)性要求等等。對(duì)于實(shí)際工程問(wèn)題的優(yōu)化，“在線”布點(diǎn)技術(shù)依然存在以下瓶頸。首先，初始設(shè)計(jì)參數(shù)區(qū)間的確定是以工程經(jīng)驗(yàn)為依據(jù)的。對(duì)于缺乏先驗(yàn)性的工程問(wèn)題，當(dāng)設(shè)計(jì)空間過(guò)小時(shí)，可能丟失最優(yōu)解；反之，近似模型難以構(gòu)造，精度也得不到保證。其次，對(duì)于復(fù)雜的非線性問(wèn)題，隨著設(shè)計(jì)參數(shù)的增加，迭代次數(shù)和樣本的數(shù)目都會(huì)大幅度增加。而每次試驗(yàn)的開(kāi)銷達(dá)到一定量級(jí)時(shí)，算法的可行性難以保證。近似模型的精度是影響優(yōu)化結(jié)果的最直接因素,在優(yōu)化領(lǐng)域，國(guó)內(nèi)外采用的近似模型構(gòu)造技術(shù)種類繁多，常用的包括基于多項(xiàng)式的擬合技術(shù)，Kriging插值，徑向基函數(shù)(Radial basis function，RBF)插值，自適應(yīng)響應(yīng)面技術(shù)等。近年來(lái)，近似模型構(gòu)造技術(shù)發(fā)展很快，各種逼近和擬合算法層出不窮。但實(shí)質(zhì)上，這些方法并沒(méi)有取得突破性進(jìn)展。總體來(lái)說(shuō)，大多數(shù)近似模型構(gòu)造方法的性能均是建立在回歸分析、方差分析以及無(wú)偏估計(jì)等以統(tǒng)計(jì)學(xué)理論為依據(jù)的基礎(chǔ)之上，判斷這類近似模型優(yōu)劣的標(biāo)準(zhǔn)可以歸結(jié)為所謂經(jīng)驗(yàn)風(fēng)險(xiǎn)最小化準(zhǔn)則，但經(jīng)驗(yàn)風(fēng)險(xiǎn)最小并不一定意味著期望風(fēng)險(xiǎn)最小。簡(jiǎn)而言之，這類技術(shù)試圖用十分復(fù)雜的模型去擬合有限的樣本。因此，導(dǎo)致喪失了推廣能力，難以反映研究目標(biāo)的實(shí)質(zhì)和特性。因此，泛化性能出色的基于概率的支持矢量機(jī)回歸近似模型構(gòu)造技術(shù)提供了一種能夠應(yīng)用于沖壓成形優(yōu)化的選擇。同其他主流方法相比，支持向量機(jī)回歸技術(shù)的缺陷在于其計(jì)算效率，因此需要建立更為高效的試驗(yàn)設(shè)計(jì)方法。
2.2 并行優(yōu)化技術(shù)
同串行算法相比，并行算法具有以下優(yōu)勢(shì)。由于采用并行構(gòu)架，計(jì)算效率大幅度提高，可以擴(kuò)大樣本數(shù)量，從而導(dǎo)致近似模型精度提高；每個(gè)迭代步中的樣本信息遠(yuǎn)遠(yuǎn)大于串行算法，因此更容易收斂。JAKUMEIT 等[33]建立了并行Kriging 迭代算法，在優(yōu)化板金成形參數(shù)時(shí)的結(jié)果表明，它確實(shí)能更快、更穩(wěn)地收斂?jī)?yōu)化過(guò)程；WANG 等[34]建立了基于并行智能布點(diǎn)算法的Kriging 近似模型體系和支持矢量機(jī)回歸優(yōu)化體系，并成功應(yīng)用于拉延筋的優(yōu)化設(shè)計(jì)。隨著計(jì)算機(jī)技術(shù)的發(fā)展，基于圖形處理器的通36 機(jī) 械 工 程 學(xué) 報(bào) 第46卷第10期期用計(jì)算技術(shù)(General purpose graphic process unit，GPGPU)發(fā)展很快，它不僅能夠進(jìn)行圖形處理，而且能完成CPU 的運(yùn)算工作，更適合高性能計(jì)算，并能使用更高級(jí)別的編程語(yǔ)言，在性能和通用性上更加強(qiáng)大。從狹義的GPGPU 應(yīng)用來(lái)說(shuō)，GPGPU 就是功能強(qiáng)化的GPU，彌補(bǔ)了CPU 浮點(diǎn)運(yùn)算能力的嚴(yán)重不足。目前湖南大學(xué)汽車(chē)車(chē)身先進(jìn)設(shè)計(jì)制造國(guó)家重點(diǎn)實(shí)驗(yàn)室已將該技術(shù)成功應(yīng)用于并行有限元仿真中，已取得初步成效，并已著手建立基于GPGPU技術(shù)的板料成形工藝與模具優(yōu)化設(shè)計(jì)體系。
3 基于CAE 技術(shù)的沖壓工藝與模具
制造的若干前沿問(wèn)題板料沖壓成形工藝發(fā)展歷史悠久，隨著信息技術(shù)和其他相關(guān)技術(shù)的發(fā)展和應(yīng)用，基于CAE 的沖壓分析越來(lái)越多地應(yīng)用于板料成形工藝的設(shè)計(jì)中，不僅為工藝設(shè)計(jì)提供了更加精確的設(shè)計(jì)方法，同時(shí)也為工藝的創(chuàng)新提供了新的途徑，如通過(guò)CAE 技術(shù)的計(jì)算和模擬，在沖壓分析中可以實(shí)現(xiàn)基于原理變革的工藝創(chuàng)新和基于計(jì)算方法變革的工藝創(chuàng)新。
3.1 拉延筋技術(shù)
在現(xiàn)有沖壓工藝技術(shù)中，拉延模具通常由凹模、凸模、壓邊圈三部分組成。壓邊圈的壓料面上通常設(shè)置拉延筋以控制板料在沖壓中所受的流動(dòng)阻力，防止工件出現(xiàn)起皺或拉裂等成形缺陷。原有拉延模具的拉延筋的設(shè)置一般為垂直材料自然流動(dòng)方向的直線或環(huán)線。由于拉延筋處的板料變形通常比其他部位的變形更為復(fù)雜，傳統(tǒng)的設(shè)計(jì)計(jì)算方法難以定量計(jì)算給定形狀和尺寸的拉延筋能提供的流動(dòng)阻力，這就使得傳統(tǒng)的拉延筋布置和參數(shù)設(shè)計(jì)主要憑經(jīng)驗(yàn)和直覺(jué)進(jìn)行，再通過(guò)反復(fù)的打磨和調(diào)試來(lái)獲得符合要求的拉延筋形狀和尺寸。這個(gè)過(guò)程通常是導(dǎo)致傳統(tǒng)模具調(diào)試時(shí)間較長(zhǎng)的重要原因。同時(shí)對(duì)于深拉延件，傳統(tǒng)拉延筋布置和設(shè)計(jì)用來(lái)克服角部起皺或拉裂效果不佳，成形缺陷經(jīng)常出現(xiàn)。為解決上述工藝方案中常出現(xiàn)的成形缺陷，湖南大學(xué)汽車(chē)車(chē)身先進(jìn)設(shè)計(jì)制造國(guó)家重點(diǎn)實(shí)驗(yàn)室應(yīng)用CAE 技術(shù)，對(duì)拉延筋機(jī)理進(jìn)行了深入的研究，發(fā)明了斜拉延筋工藝。新的斜拉延筋技術(shù)除在壓邊圈的壓料面上設(shè)置有傳統(tǒng)的拉延筋外，還在壓邊圈的壓料面的角部創(chuàng)新性地設(shè)置有斜拉延筋。傳統(tǒng)拉延筋主要提供板材在沖壓中的具有被動(dòng)性質(zhì)的流動(dòng)阻力，而斜拉延筋則除提供流動(dòng)阻力外，還可提供具有主動(dòng)性質(zhì)的引導(dǎo)材料流動(dòng)的作用力。這就使得斜拉延筋對(duì)材料流動(dòng)具有很好的控制作用，在拉延件沖壓特別是深拉延件沖壓中能有效克服拉延件的起皺或拉裂缺陷。斜拉延筋工藝對(duì)設(shè)計(jì)提出了更高的要求。因?yàn)樾崩咏钅芴峁┚哂兄鲃?dòng)性質(zhì)的引導(dǎo)材料流動(dòng)的作用力，其形狀、位置和尺寸參數(shù)必須足夠精確，否則可能導(dǎo)致相反的效果。由于斜拉延筋的作用機(jī)理不同于傳統(tǒng)拉延筋，現(xiàn)有可借鑒的設(shè)計(jì)理論和經(jīng)驗(yàn)很少，因此必須發(fā)展能準(zhǔn)確模擬真實(shí)拉延筋的計(jì)算機(jī)仿真技術(shù)。
3.2 熱成形技術(shù)
先進(jìn)高強(qiáng)度鋼板由于具有強(qiáng)度高、較好的應(yīng)變硬化能力、較強(qiáng)的均勻變形能力、更高的疲勞特性而得到廣泛應(yīng)用，特別是在汽車(chē)工業(yè)，由于這些優(yōu)越的力學(xué)性能而具有更多的降低板厚的可能，通過(guò)減少鋼板的厚度，在保證其他性能相當(dāng)?shù)臈l件下可達(dá)到減輕汽車(chē)重量的目的。 因此，先進(jìn)高強(qiáng)度鋼目前已成為較理想的汽車(chē)輕量化材料之一。國(guó)際上已研制出超輕鋼車(chē)身，在滿足2004 年美國(guó)碰撞法規(guī)前提下，應(yīng)用先進(jìn)高強(qiáng)度可在不增加成本的條件下使車(chē)身質(zhì)量減少20%。因此，先進(jìn)高強(qiáng)度鋼在汽車(chē)減重、節(jié)能、提高安全性、降低排放等方面展現(xiàn)了廣闊前景。在高強(qiáng)度鋼板的應(yīng)用中，由于強(qiáng)度的大幅度提高，給成形工藝和模具設(shè)計(jì)制造帶來(lái)了許多新的困難，如強(qiáng)度增大和厚度減薄導(dǎo)致沖壓成形中的回彈增加，強(qiáng)度和硬度提高導(dǎo)致成形力增大和模具容易磨損等。和普通鋼板相比，高強(qiáng)度鋼板在沖壓過(guò)程中不僅會(huì)造成較大的彎曲回彈和扭曲回彈，而且還會(huì)產(chǎn)生嚴(yán)重的側(cè)壁卷曲，這將嚴(yán)重影響沖壓件的形狀尺寸精度和整車(chē)裝配。超高強(qiáng)度鋼的回彈問(wèn)題對(duì)當(dāng)今沖壓工藝提出了新的挑戰(zhàn)，成為汽車(chē)制造行業(yè)和學(xué)術(shù)界研究的一大熱點(diǎn)。熱沖壓成形技術(shù)是解決高強(qiáng)度鋼板難成形問(wèn)題和減少回彈的有效途徑。熱沖壓是將沖壓板材在奧氏體溫度區(qū)加熱，在高溫下沖壓成形并使其在模具內(nèi)進(jìn)行馬氏體相變后，在保證高強(qiáng)度的前提下獲得需要的形狀。目前，高強(qiáng)度鋼板采用熱沖壓技術(shù)具有以下特點(diǎn)：成形性能得到優(yōu)化，回彈極小，產(chǎn)品尺寸精度高，成形負(fù)荷小，不必對(duì)沖壓機(jī)進(jìn)行升級(jí)；熱成形過(guò)程工件可以獲得高延展率，并可以大幅度增加零件的強(qiáng)度。目前的研究關(guān)鍵是要建立快速加熱和冷卻過(guò)程中的高強(qiáng)度鋼板本構(gòu)關(guān)系變化規(guī)律及對(duì)材料強(qiáng)度影響規(guī)律，發(fā)展能準(zhǔn)確模擬熱成形過(guò)程的計(jì)算機(jī)仿真技術(shù)，通過(guò)數(shù)值仿真與模擬，探明高強(qiáng)度鋼板在受熱和受力雙重作用下材料性能的變化規(guī)律和塑性成形機(jī)理以及冷卻速度對(duì)成形性能的影響機(jī)理，從而確定熱成形工藝參數(shù)的變化對(duì)成形件質(zhì)量的影響特性。
原文：Some New Topics on Process Design and Mould Manufacture for Sheet Metal Forming
LI Guangyao WANG Hu YANG Xujing ZHENG Gang
（State Key Laboratory of Advanced Design and Manufacturing for Vehicle Body; Hunan University; Changsha 410082）;
機(jī)械工程學(xué)報(bào), Journal of Mechanical Engineering, 編輯部郵箱 2010年 10期??
summary
Sheet metal forming technique is one of the most important forming methods and is widely applied in industrial fields.Therefore,forming manufacturing techniques represent a country s manufacturing level.Furthermore,the quality,profit and development capability of the product also depend on forming techniques.Some new topics on process design and mould manufacture for sheet metal forming are proposed according to the recent development and applications in this field.
The 2sheet metal stamping process design optimization technique of cutting edge issues in
Sheet metal forming process is a very complex plastic forming process, many factors which directly or indirectly affects forming quality. If the process parameters, geometry parameters and the material choice is undeserved, easy wrinkling, cracking and springback defect. The traditional process analysis and die design for the mainly rely on the experience of the designer, repeated modification of various parameters, and to repair mould, time-consuming, high cost, difficult to adapt to the requirements of modern industry. With the development of computer technology and numerical calculation method, numerical simulation technology of sheet metal forming analysis has gradually become an important means of. Finite element simulation analysis has been widely used in sheet metal forming process of optimal design, however, the result of finite element analysis judgment and selection of relevant parameters is still dependent on experience. Optimization of technology introduction, make people from trial and error ( Trial-error ) work out. In forming optimization, optimization methods can be divided into three categories: traditional optimization, heuristic optimization and approximation model optimization. Due to the sheet metal forming process is a typical nonlinear problem, and the traditional optimization method is based on the gradient, for Hessian matrix is singular optimization problem, it is obviously difficult to processing stamping optimization problem; heuristic algorithm including genetic algorithms, particle swarm, quasi annealing algorithm, this algorithm can find the global optimization problem point, can guarantee the precision and performance optimization. However, optimization requires repeated modified CAE model and calls the corresponding program of the finite element method, which requires a lot of computing resources. With the design parameters of quantity and the complexity increases, not only calculate the cost will be multiplied, and the computing time will increase substantially. Based on the approximation model for optimization method for solving large problems is the most promising method, in view of its high efficiency, widely used in the field of engineering optimization, if can efficiently establish reliable approximation model, many large-scale engineering problems can be smoothly done or easily solved. In recent years, with the deepening of the research, the technology is also applied to a large number of stamping forming of the optimization problem, become the research hotspot in the field of stamping optimization.
2.1methods of experimental design
Through the experiment, the sample points layout of the specific engineering problem and the corresponding approximate model has strong sensitivity, the sample points in a limited range, using different methods to generate samples can lead to optimal result differences. In view of these problems, many domestic and foreign scholars put forward the concept of reduced space. Space reduction technology is divided into two categories, one category is established on the basis of sensitivity analysis of design variables screening method [23]. BROWNE [24] through analysis of punch and die geometry, BHF, convex die pressure, friction and stretching speed on the plate thickness of 0.19mm cylinder forming effect, using the experimental design method for a large number of trials, the results show that: in addition to the blank-holder force and drawing speed on the final forming performance little, other parameters on forming properties of. NAKAMURA [25] by using sensitivity analysis method will affect the target function with 7parameters are reduced to 2, is conducive to optimizing the. The other is based on the design of space reduction method. The former mainly through sensitivity analysis, screening out of objective function and correlation with large parameters as design variables. The latter is based on the engineering experience and spatial classification techniques, through the design variables in the analysis, the design variables are controlled in a reasonable range, and build a more accurate approximation model. Has been the development of the technology include: Based on hierarchical technology small space design method [26-27], mobile control layout technique [28], trust domain based layout method [29], the adaptive test design method [30], multi-level search method [31-32]. With the traditional test method compared to, this kind of method to design space control in a relatively small range, the reduced sample number at the same time, to avoid the global fitting, so that the approximation accuracy is greatly improved. These methods are usually based on some assumptions, such as convex set continuity requirements etc.. For the practical engineering optimization problems," online" layout technique is below the bottleneck. First of all, the initial design parameters interval to determine the engineering experience as the basis. For lack of a priori knowledge of engineering problems, when the design space of hours, may lose the optimal approximation model; instead, it is difficult to build, the precision can not be guaranteed. Secondly, for the complex nonlinear problems, with the design parameter increases, the number of iterations and the number of samples will increase. Each time the test spending reaches a certain level, it is difficult to guarantee the feasibility of the algorithm. Approximation accuracy is influenced by the optimization results of the most direct factor, in the field of optimization at home and abroad, the approximate model construction technology of variety, commonly used include based on polynomial fitting technique, Kriging interpolation, radial basis function ( Radial basis function, RBF ) interpolation, adaptive response surface technique. In recent years, approximate model technology develops very fast, and various approximation algorithm fitting emerge in an endless stream. But in fact, these methods have not achieved breakthrough progress. In general, most approximate model construction method performance are based on regression analysis, analysis of variance and unbiased estimation based on the statistics theory based on this kind of approximate model, judging criterions can be attributed to the so-called empirical risk minimization rule, but empirical risk minimization does not necessarily mean that the expected risk minimum. In short, this kind of technology is tried to use very complex model to fitting Limited sample. Therefore, lead to a loss of generalization ability, cannot reflect the research target of the essence and character. Therefore, the generalization performance of the excellent probability based on support vector machine regression approximation model technology provides a can be applied to the optimal selection of stamping forming. Compared with other mainstream method, support vector machine SVM regression technique flaw lies in its computational efficiency, so it is necessary to build more efficient methods of experimental design.
2.2parallel optimization technique
Compared with the serial algorithm, parallel algorithm has the following advantages. Due to the adoption of parallel framework, computational efficiency is greatly improved, can expand the number of samples, thus leading to improve the accuracy of approximate model; each iteration of the sample information is far greater than the serial algorithm, thus more easily convergent. JAKUMEIT [33] to establish a parallel Kriging algorithm in the optimization of sheet metal forming parameters, the results show that, it can be faster, more stable convergence of the optimization process; WANG [34] was established based on the parallel intelligent layout algorithm of approximate Kriging model system and support vector machine regression optimization system, and successfully applied to the drawbead optimization design. With the development of computer technology, based on the graphics processor through36machine machinery engineering process study reported forty-sixth volume tenth period calculation technique ( General purpose graphic process unit, GPGPU ) is developing very fast, it can not only for graphics processing, and can complete the CPU operation work, is more suitable for high performance computing, and can use more high level programming language, in performance and versatility of a more powerful. From the narrow GPGPU application, GPGPU is strengthening the function of GPU, for the CPU floating point arithmetic ability shortage. Advanced design and manufacturing for vehicle at Hunan University State Key Laboratory of the technology has been successfully applied in parallel in the finite element simulation, has achieved initial results, and has started to build the GPGPU technology based on sheet metal forming process and die design system.
3 based on the CAE technology of the stamping process and die
Manufacture of some frontier problems in sheet metal forming process has a long history of development, with the development of information technology and other related technology development and application, based on the CAE analysis is more and more used in stamping of sheet metal forming process design, not only for the process design provides more precise design method, but also for the technological innovation provides new ways, such as through the use of CAE technology calculation and simulation, stamping in the analysis can be achieved based on the principle of change of process innovation and based on the calculation method of transformation process innovation.
3.1bead technique
In the existing stamping technology of drawing die, usually by concave die, punch, a pressing edge ring is composed of three parts. Blank holder pressure surface usually setting drawbead to control sheet in stamping the flow resistance, prevents the workpiece wrinkling or cracking forming defects. The original drawing mould drawbead setting is generally perpendicular to the flow direction of the straight line or natural materials link. As a result of drawbead of sheet deformation usually than in other parts of the deformation is more complex, the traditional design and calculation methods are difficult to quantify the given shape and size of the drawbead can provide resistance to flow, which makes the traditional drawbead layout and parameter design mainly by experience and intuition, and through repeated grinding and debugging to meet the requirements of the drawing rib shape and size. This process is usually lead to the traditional mold debugging time longer important reason. At the same time for deep drawing, traditional drawbead layout and design to overcome the corner wrinkling or cracking effect, forming defect often appears. In order to solve the above technical scheme often appears in the forming defects of automobile body, Hunan University State Key Laboratory of advanced design and manufacture technology for application of CAE, drawbead mechanism of in-depth research, the invention of oblique drawbead. New slant on drawbead technology except in the pressing ring is arranged on the surface of the binder traditional drawbead, still pressing edge ring binder corner innovatively with inclined drawbead. Traditional drawbead mainly provide sheet in stamping with the passive properties of the resistance to flow, and the oblique drawbead except to provide resistance to flow, can also provide the active nature of the guide material flow force. This makes the oblique Drawbead on material flow and has good control effect, in the drawing stamping especially deep drawing stamping can effectively overcome the drawing wrinkling or crack. Oblique drawbead design put forward higher requirements. Because of an oblique drawbead can provide active nature of the guide material flow force, its shape, position and size parameters must be sufficiently accurate, otherwise may cause the opposite effect. Because of the oblique drawbead mechanism is different from the traditional drawbead, existing reference design theory and experience little, therefore must develop accurate simulation of real drawbead computer simulation technology.
3.2 hot forming technology
Advanced high strength steel plate with high strength, good strain hardening ability, strong uniform deformation capacity, higher fatigue properties and widely used, especially in the automotive industry, due to the excellent mechanical performance and more reduce plate thickness may, by reducing the thickness of the sheet, the other performance conditions can reach the purpose of reducing vehicle weight. Therefore, advanced high strength steel has become the ideal lightweight automobile materials. Ultra light steel auto body international has developed in the United States, meet the 2004crash regulations premise, the application of advanced high strength without increasing the cost of conditions to reduce the 20% body quality. Therefore, advanced high strength steel in vehicle weight, energy conservation, improve safety, reduce emissions and so on shows broad prospects. In high strength steel application, due to the intensity of the increase, to forming process and mold design and manufacturing has brought many new difficulties, such as strength increases and the thickness of lead in stamping forming springback increase, strength and hardness increase in forming force and easy mould wear. And ordinary steel, high strength steel sheet in stamping process can not only cause the larger bending and torsion springback, but also will produce the serious side wall curl, which will seriously affect the shape of the stamping parts size precision and the whole assembly. Ultra high strength steel rebound problem of the stamping process and put forward the new challenge, as the automobile manufacturing industry and academic research area. Hot stamping technology is to solve the difficult problem of high strength steel sheet forming and efficient way reduced bounce. Hot stamping is a sheet metal stamping in the austenitic temperature zone heating in high temperature, stamping and make it in the mold of martensitic transformation, in ensure the premise of high strength required shape. At present, the high strength steel plate by hot stamping technology has the following characteristics: the forming performance is optimized, rebound small, product size and high precision, the forming load is small, do not need to stamping machine upgrade; hot forming process of the workpiece can be obtained with high elongation, and can greatly increase the strength of the parts. The current study is the key to the establishment of rapid heating and cooling process of high strength steel constitutive relation change rule a