0073-滾針軸承自動裝針機設(shè)計,軸承,自動,裝針機,設(shè)計 本科生畢業(yè)設(shè)計（論文） 開題報告 學(xué)生姓名： 學(xué) 號： 班 級： 專 業(yè)： 指導(dǎo)教師： 一 課題介紹： 1課題名稱：滾針軸承自動裝針機設(shè)計 2課題背景： ⑴課題來源： 本課題來源于中國第一汽車集團公司（FAW）,專為CA141汽車傳動軸中萬向節(jié)滾針軸承的裝配過程設(shè)計的自動裝針機。自動滾針軸承裝針機能有效解決人工裝針速度慢，合格率低等問題，該機可裝配各種規(guī)格的滾針軸承。 ⑵機床和機床自動線： 機床和機床自動線是一種專用高效自動化技術(shù)裝備，目前，由于它仍是大批量機械產(chǎn)品實現(xiàn)高效、 高質(zhì)量和經(jīng)濟性生產(chǎn)的關(guān)鍵裝備，因而被廣泛應(yīng)用于汽車、拖拉機、內(nèi)燃機和壓縮機等許多工業(yè)生產(chǎn)領(lǐng)域。其中 ，特別是汽車工業(yè)，是機床和自動線最大的用戶。如德國大眾汽車廠在Salzgitter的發(fā)動機工廠，90年代初所采用的金屬切削機床主要是自動線(60%)、組合機床(20%)和加工中心(20%)。顯然，在大批量生產(chǎn)的機械工業(yè)部門，大量采用的設(shè)備是機床和自動線。因此，機床及其自動線的技術(shù)性能和綜合自動化水平，在很大程度上決定了這些工業(yè)部門產(chǎn)品的生產(chǎn)效率、產(chǎn)品質(zhì)量和企業(yè)生產(chǎn)組織的結(jié)構(gòu)，也在很大程度上決定了企業(yè)產(chǎn)品的競爭力。 使用專用自動化機床是大批量生產(chǎn)提高生產(chǎn)率降低成本的重要途徑。專用自動化機床往往具有投資省，見效快等特點，因而在大批量生產(chǎn)中被廣泛采用。自動機床或半自動機床主要用于軸類和盤套類零件的加工和裝配自動化。這類機床的最大特點是可以根據(jù)生產(chǎn)裝配需要，在更換或調(diào)整部分零部件（例如凸輪或靠模等）后可加工不同的零件，適合于大批量少品種生產(chǎn)裝配。因此，這類機慶使用比較廣泛。 專用機床是按一種零件（或一組相似的零件）的一個加工工序而專門設(shè)計制造的自動化機床。專用機床的結(jié)構(gòu)和部件大都是專門和單獨制造的這類機床的，這類機床的設(shè)計制造往往時間較長，投資也較大，因此采用這類機床時，必須考慮以下基本原則：被加工的工件除具有大批量的特點外，還必須結(jié)構(gòu)定型；工件的加工工藝必須是合理和可靠的。在大多數(shù)情況下，需要進行必要的工藝試驗，保證專用機床所采用的加工工藝先進可靠，所完成的工序加工精度穩(wěn)定；在機床上采用一些新的結(jié)構(gòu)方案時，必須進行結(jié)構(gòu)性能試驗，待取得較好的結(jié)果后，方能在機床上采用；必須進行技術(shù)經(jīng)濟分析。只有在技術(shù)經(jīng)濟上效果明顯，才能采用專用機床實現(xiàn)單機自動化。 自動線是由流水生產(chǎn)線方式發(fā)展而來的。20 年代美國Henry Ford 創(chuàng)立了汽車工業(yè)的流水線，由此揭開了現(xiàn)代流水生產(chǎn)的序幕。福特流水線的主要內(nèi)容可以包括以下兩個方面： ①實施零件和產(chǎn)品的標準化，設(shè)備和工具的專用化以及工場專業(yè)化。為了追求高效率和低成本，福特認為首先要將生產(chǎn)集中于唯一最佳的產(chǎn)品型號，提出了所謂的“單一產(chǎn)品原則”，福特汽車公司曾在20 年間連續(xù)生產(chǎn)T型汽車，由此而奠定了現(xiàn)代流水生產(chǎn)線的基礎(chǔ)。這種“單一產(chǎn)品原則”在當(dāng)今市場需求日益多樣化的環(huán)境下也許已不再適用，但在當(dāng)時的經(jīng)濟條件下卻適應(yīng)了美國的國情，福特汽車公司也由此而迅速發(fā)展起來。 零件標準化是產(chǎn)品標準化的進一步發(fā)展，目的在于提高零部件的互換性，減少零件種數(shù)和擴大生產(chǎn)批量，零件標準化后，便于分別組織專業(yè)化工廠或車間制造，這樣可以采用高度專門化的設(shè)備和工具，從而達到生產(chǎn)的高效率。由于工人的作業(yè)活動是不斷地重復(fù)同一作業(yè)，所以作業(yè)和操作也可以實現(xiàn)標準化。 ②創(chuàng)造了流水作業(yè)的生產(chǎn)方法，建立了傳送帶式的流水生產(chǎn)線。由于傳送帶的廣泛應(yīng)用，使得原材料均可在使用機械裝置搬運的移動中，加工成為各種零件。而部件裝配和汽車總裝配，則采用移動裝配法完成。由于把生產(chǎn)工序分細，大大提高了操作熟練程度和勞動生產(chǎn)率。作業(yè)的速度也為傳送帶的速度所規(guī)定，借助于傳送帶的應(yīng)用，使生產(chǎn)過程的各項作業(yè)能在同一時間進行，并且各種零部件在各條流水線的投入和產(chǎn)出互相銜接配合，不至于發(fā)生在制品過多或不足的現(xiàn)象，恰能保證總裝配線的需要，形成同步化的流水生產(chǎn)體系。 自動線生產(chǎn)是指工件按照一定的工藝路線，順序地通過各個工作地，并按照一定的生產(chǎn)速度(節(jié)拍)完成工藝作業(yè)的連續(xù)重復(fù)自動生產(chǎn)的一種生產(chǎn)組織形式。自動線生產(chǎn)的基本特征如下： ①工作地專業(yè)化程度高，在自動線上固定地生產(chǎn)一種或幾種工件，而在每個工作地上固定完成一道或幾道工序。 ②生產(chǎn)具有明顯的節(jié)奏性，即按照節(jié)拍進行生產(chǎn)。所謂節(jié)拍，是指自動線上出產(chǎn)相鄰兩件制品的時間間隔。 ③各道工序的工作地設(shè)備數(shù)量與該工序單件工時的比值相一致( 若不一致則需要臨時緩沖庫) 。 ④工藝過程是封閉的，并且工作地設(shè)備按工藝順序排列成鏈索形式，工件在工序間作單向。 ⑤工件如同流水般從一個工序轉(zhuǎn)到下個工序，消除或最大限度地減少了工件的因等待加工而耽擱的時間和機床設(shè)備加工的間斷時間，生產(chǎn)過程具有高度的連續(xù)性。 ⑥工件從進入加工工位到所有工位操作完成，均在主控制器的控制之下自動完成，這種自動過程不僅包括工件的流動自動化控制，還包括機床的加工全過程的自動化控制。 僅滿足上述前面6項特征的生產(chǎn)線，只能稱之為流水生產(chǎn)線，還不能稱之為自動生產(chǎn)線。加工設(shè)備由專用自動化機床或組合機床組成的，能加工固定一種或少數(shù)幾種相似零件的自動線稱之為剛性自動線。加工設(shè)備由數(shù)控機床或加工中心等組成的可加工多品種少批量零件的自動線稱之為柔性自動線。 在剛性自動化生產(chǎn)條件下，生產(chǎn)過程的連續(xù)性、平行性、比例性、節(jié)奏性都很高，所以它具有可以提高加工設(shè)備專業(yè)化水平、提高勞動生產(chǎn)率、增加產(chǎn)量、降低產(chǎn)品成本、提高生產(chǎn)的自動化水平等一系列優(yōu)越性。但是反過來也有不少不利的地方，例如：由于設(shè)備高度專用化，對產(chǎn)品的變化缺乏適應(yīng)力；一旦在某處發(fā)生設(shè)備故障，就有可能導(dǎo)致全線停車，帶來較大的損失；生產(chǎn)率的調(diào)整幅度不可能很大；技術(shù)改造困難較大等等。 ⑶我國軸承制造業(yè)現(xiàn)狀： 軸承是關(guān)系國民經(jīng)濟發(fā)展的關(guān)鍵機械基礎(chǔ)件，其技術(shù)水平和產(chǎn)品質(zhì)量對主機的性能和質(zhì)量有著重要的影響，被譽為機械的“關(guān)節(jié)”。改革開放以來，我國軸承產(chǎn)品水平和制造技術(shù)水平有了長足進步，與重點主機配套能力有了很大提高。但與工業(yè)發(fā)達國家相比，我國軸承制造業(yè)的整體水平還存在著相當(dāng)大的差距——組織結(jié)構(gòu)散亂差，重復(fù)建設(shè)嚴重，生產(chǎn)集中度低；產(chǎn)品設(shè)計水平低，新產(chǎn)品開發(fā)跟不上主機發(fā)展需求；制造技術(shù)落后，尺寸散差大，振動噪聲大，性能一致性差，壽命可靠性低；產(chǎn)品檔次低，價格亂，競相壓價爭市場，國際競爭能力差。為主機配套和維修的一些高技術(shù)含量的軸承主要依靠進口，而出口軸承則主要是低檔通用軸承。軸承制造業(yè)的這種落后狀況，已經(jīng)成為制約機械工業(yè)發(fā)展的重要因素之一。 ⑷軸承裝配自動線： 國家已將汽車制造業(yè)列為國民經(jīng)濟的支柱產(chǎn)業(yè)，我國目前的汽車保有量為2500～3000萬輛，年產(chǎn)量170萬輛，軸承是汽車的主要配套件，軸承制造業(yè)的發(fā)展必須與汽車制造業(yè)同步甚至超前。因此，提高軸承裝配的技術(shù)水平和自動化程度已經(jīng)迫在眉睫。軸承屬于量大面廣通用性強的機械基礎(chǔ)件，軸承套圈一般以軸對稱的多個回轉(zhuǎn)面、環(huán)面的相互組合為其主要的幾何特征，特別適合自動化裝配。 軸承裝配（含軸承成品的自動檢測、自動包裝）自動線一直是困擾著技術(shù)進步的難題。在國外著名的軸承公司，這項技術(shù)已實際成功應(yīng)用多年。而國內(nèi)靠國外引進，而使用效果仍不能令人滿意。因而研發(fā)軸承裝配自動生產(chǎn)線，并與軸承的磨削自動生產(chǎn)線有機連接，對于提高軸承的生產(chǎn)效率和產(chǎn)品質(zhì)量，減少工序流動中的軸承零件的數(shù)量，減少人工干預(yù)的影響、降低成本，意義是顯著的。由于軸承的品種很多，不同的品種，自動裝配線各有特點，因而軸承行業(yè)對軸承裝配的數(shù)控設(shè)備的多樣性提出了要求。 ⑸萬向節(jié)滾針軸承特點： 目前后驅(qū)動汽車上應(yīng)用最廣的一種普通萬向節(jié)由萬向節(jié)叉、十字軸等基本零件構(gòu)成。十字軸裝配在萬向節(jié)叉上做連接，十字軸的軸頭上裝有滾針軸承，當(dāng)軸頭接入萬向節(jié)叉時，十字軸與萬向節(jié)叉之間就可以有相對旋轉(zhuǎn)，也就產(chǎn)生了多角度變化。萬向節(jié)叉上的花鍵連接又可以做小許的軸向移動，這樣就適應(yīng)了夾角和距離同時變化的需要。十字軸的軸頭上裝有的這種滾針軸承，就是我們此次設(shè)計的裝配對象。 滾針軸承可分為具內(nèi)圈（NA）和不具內(nèi)圈（RNA）兩種結(jié)構(gòu)。如果機軸上的滾道可以淬硬及研磨，則宜用無內(nèi)圈的滾針軸承，由于不需內(nèi)圈，因此機軸直徑可以加大，剛性也可增加。機軸相對于軸承箱的軸向位移量則由軸向滾道寬度決定。只需將機軸滾道機削加工到適當(dāng)?shù)某叽绾托螤罹?，即可獲得具有較高運轉(zhuǎn)精度的軸承配置。若機軸淬硬及研磨機軸不可能，或不經(jīng)濟時，可采用帶內(nèi)圈的滾針軸承。這時機軸對軸承箱的軸向位移限制在一定的限度內(nèi)。若需要較大的位移，可使用加長型內(nèi)圈取代標準內(nèi)圈。滿裝滾針軸承：分為RNAV（無內(nèi)圈）、NAV（帶內(nèi)圈）兩種。不帶保持架，裝滿滾針，適用于截面高度較低，而又需承受較大載荷的場合。 滾針軸承一般只承受徑向負荷，不能承受軸向負荷。當(dāng)有軸向負荷時，應(yīng)和其它軸承組合使用，它不限制軸或外殼的軸向位移。安裝軸承時，應(yīng)注意軸和殼孔中心的平行，軸承的外圈軸線和內(nèi)圈軸線不允許傾斜，否則會使?jié)L針和滾道面的線接觸破壞。滾針軸承的極限轉(zhuǎn)速較低，在主機轉(zhuǎn)速較高的情況下，應(yīng)盡量選用帶保持架的滾針軸承。和其它類型的滾動軸承相比，在徑向尺寸相同的情況下，滾針軸承的負荷容量最大，它的剛性較高，但摩擦力矩也較大。 ⑹國內(nèi)發(fā)展： 在國內(nèi)，南通市工農(nóng)路正揚商務(wù)有限公司已于 2005年9月23日在《商貿(mào)機會——中國軸承機械網(wǎng)》發(fā)布關(guān)于自動滾針軸承裝針機（專利）的技術(shù)轉(zhuǎn)讓信息，尋求商業(yè)合作。但是使用效果仍不能令人滿意，所以我們對此項課題進行深入的專門設(shè)計研究。 3工作內(nèi)容和要求： ⑴設(shè)計要求： ① 裝針機以CA141汽車傳動軸中的萬向節(jié)滾針軸承為裝配件。 ② 機械手每分鐘裝配成品十只。 ⑵工作量： ① 主機及零部件設(shè)計4張A0圖紙（草圖）。 ② 設(shè)計說明書一份（20000字）。 ③ 外文翻譯（5000字）。 ④ 上機繪制4張A0圖紙（CAD）。 4課題的重點和難點： ⑴課題重點： 本課題內(nèi)容為設(shè)計研究CA141汽車傳動軸中萬向節(jié)滾針軸承的自動裝針機。整個裝配系統(tǒng)采用臥式裝針方式，軸承外圈采用直立狀態(tài)，開口端正對滾針進給端，滾針采用橫向進給方式，即采用一特殊裝配裝置使?jié)L針安排成圓周均布狀態(tài)，其圓周直徑和軸承外圈內(nèi)壁直徑相同，然后采用一推套把滾針推入軸承內(nèi)，滾針只需排入軸承內(nèi)壁即可，從而完成整個裝配過程。 為了實現(xiàn)上述方案，本裝配系統(tǒng)采用五大機構(gòu)：滾針的自動上料機構(gòu)，軸承外圈的上料機構(gòu)，裝配和卸料裝置，凸輪機構(gòu)，傳動系統(tǒng)。滾針需要整齊的排序，此機構(gòu)采用槽隙定向的上料機構(gòu)原理，采用一特制齒形輪進行上料，軸承外圈上料機構(gòu)采用重心偏移法定向的料斗裝置原理，具體采用特制一斜邊推塊上料機構(gòu)，裝配和卸料機構(gòu)采用一汽實習(xí)時所見之裝配原理，并進一步改進而成。為了實現(xiàn)裝配和卸料機構(gòu)中推套的運動，采用一圓柱凸輪擺桿機構(gòu)。由于本系統(tǒng)所需轉(zhuǎn)速低，各軸之間傳動比要求低，所以采用鏈傳動，并且結(jié)構(gòu)簡單緊湊，對本系統(tǒng)特別適用。 因此，解決滾針的自動上料機構(gòu)，軸承外圈的上料機構(gòu)，裝配和卸料裝置，凸輪機構(gòu)，傳動系統(tǒng)五大機構(gòu)的典型機構(gòu)設(shè)計成為本研究課題的重點。 ⑵課題難點： 本課題設(shè)計的自動機床主要包含兩部分裝置：自動裝配工件裝置和自動裝卸工件裝置。如果設(shè)計過程中其中一部分功能不能實現(xiàn)，就只能稱之為半自動機床，從而也就不能完成連續(xù)的加工循環(huán)，因此兩部分相輔相成，缺一不可。所以如何選擇滾針的自動上料機構(gòu)，軸承外圈的上料機構(gòu)，裝配和卸料裝置，凸輪機構(gòu)，傳動系統(tǒng)五大機構(gòu)的設(shè)計方案，并實現(xiàn)整個系統(tǒng)的全自動化成為了本研究課題的難點。 5可能用到的主要知識和技能： 機械原理，機械設(shè)計，機械制造，C語言程序設(shè)計，工程圖學(xué)，機床自動化與自動線知識，自動裝配知識，機電一體化知識，AutoCAD 繪圖技能，CATIA建模能力，ANSIS有限元分析，ADAMS仿真分析等技能。 6需要自學(xué)的知識和技能： 機床自動化與自動線知識，自動裝配知識，機電一體化知識，AutoCAD 繪圖技能，CATIA建模能力，ANSIS有限元分析，ADAMS仿真分析等技能。 二 工作計劃： 調(diào)研，譯文，參考資料 總體布置，草圖，開題報告 總體設(shè)計，總裝圖 部件設(shè)計，相關(guān)計算 零部件設(shè)計，論文 修改，完善圖紙、論文，準備答辯 三 參考文獻： 1 [英] R.M.韋布， B.D.喬特.自動裝配圖集——料斗進給裝置與控制系統(tǒng).上海：上?？茖W(xué)技術(shù)出版社，1983.7 2 [英] R.M.韋布，B.D.喬特.自動裝配圖集——傳送機構(gòu).上海：上?？茖W(xué)技術(shù)出版社，1983.7 3 [英] R.M.韋布，B.D.喬特.自動裝配圖集——定向機構(gòu)與擒縱裝置. 上海：上海科學(xué)技術(shù)出版社，1983.7 4 [英] R.M.韋布，B.D.喬特.自動裝配圖集——工件移置機構(gòu).上海：上?？茖W(xué)技術(shù)出版社，1983.7 5 華中工學(xué)院機械制造教研室.機床自動化與自動線.北京：機械工業(yè)出版社，1981.3 6 工業(yè)機械手圖冊.北京：機械工業(yè)出版社，1978 7 Techniques in automated assembling— the state of the art，12th June 1985，The Bowater Conference Centre，Knightsbridge，London，UK 8 An automated assembly system for a microassembly station，Computers in Industry，Volume 38，Issue 2，March 1999，Pages 93-102 ，A. Mardanov，J. Seyfried and S. Fatikow 9 A microrobot-based automated micromanipulation station for assembly of microsystems，Computers in Industry，Volume 36，Issues 1-2，30 April 1998，Pages 155-162，Sergej Fatikow and Mirko Benz 8 本科生畢業(yè)設(shè)計（論文） 翻譯資料 中文題目： 以微型機器人為基礎(chǔ)的自動化 顯微操作裝置為微系統(tǒng)裝配 英文題目： A microbot-based automated micromanipulation station for assembly of microsystems A microrobot-based automated micromanipulation station for assembly of microsystems Sergej Fatikow Mirko Benz Abstract： The development of new types of miniaturized and microrobots with human-like capabilities play an important role in different application tasks. One of the main problem of present-day research is, for example, to assemble a whole microsystem from different microcomponents.This paper presents an automated micromanipulation desktop station including a piezoelectrically driven microrobot placed on the highly-precise x–y stage of a light microscope, a CCD-camera as a local sensor subsystem, a laser sensor unit as a global sensor subsystem, and a Pentium PC equipped additionally with an optical grabber. The microrobot has three piezoelectrically driven legs and two autonomous manipulators as endeffectors; it can perform highly-precise manipulations （with an accuracy of up to 10 nm） and a nondestructive transport （at a speed of several mm/s） of very small objects under a microscope. To perform manipulations automatically, a control system, including a task planning level and a real-time execution level, is being developed. （C）1998 Elsevier Science B.V. All rights reserved. Keywords： Microrobots; Microassembly; Automated desktop station; Assembly planning; Piezoactuators 1.Introduction： There is a growing need for miniaturized and microrobots worldwide. Due to the enormous breakthroughs in conventional robotics and in the microsystem technology (MST),everyone is convinced that the development of remote-controlled or autonomous microrobots will lead to improvements in many areas. Above all, positive results are expected in medicine (microsurgery),manufacturing (microassembly, inspection and maintenance), biology (manipulation of cells) and testing/measuring technique (VLSI) . Medicine is one of the application fields which would profit by the microrobotics the most. The attention lies on artificial organs (prosthetics) , laparoscopy, implantable drug delivery systems (diagnosis and therapy systems) , telemicrosurgery, etc. The minimal-invasive surgery developed into an important field of medicine during the last years.Smaller and more flexible active endoscopes are needed in order to replace human hands, respond to outer incidents, penetrate into a body or a vessel through natural bodily orifice or a small incision by remote control, where they perform complex in-situ measurements and manipulations. In order to meet these requirements, microprocessors, several sensors and actuators, a light source and possibly an image processing unit should be integrated into an intelligent endoscope. Biotechnology requires special microstructured active tools which are able to perform micromanipulations like the sorting or reunion of cells or the injection of a foreign body into a cell under a microscope. In the gene research and the environment technique （cells as indicators for harmful substances）, precise and gentle manipulation of single cells are also required. Industry and especially manufacturing and measuring techniques need highly sensitive testing methods in the μm-range. An important task represents, for example, the inspection of wafers, where several check points have to be contacted by a temperature or voltage probe. The same is valid for inspection robots which are used in inaccessible or dangerous terrain in order to detect leaks or flaws and make repairs （e.g., in pipelines）。The adoption of MST-related developments by the industry has already demonstrated which kind of problems occur with the mass production of microsystems. These systems usually consist of microcomponents of different materials which are produced with various microtechniques; this leads to one or several very precise assembly step （s）of the individual components. The assembly of microsystems, i.e., the non-destructible transport, precise manipulation or exact positioning of microcomponents is becoming one of the most important applications in microrobotics. 2. Manipulation of microobjects： The availability of highly precise assembly processes will make it easier to economically realize operable microsystems. In order to efficiently produce microsystems and components in lot sizes or by mass-production techniques, it is absolutely necessary to introduce flexible, automated, precise and fast microassembly stations. Different concepts are being followed to do micromanipulation for particular classes of application. Purely manual micromanipulation is the most often used method today. In medicine and biological research, it is used exclusively. Even in industry, microassembly tasks are very often carried out by specially trained technicians, who, for example, preposition assembly parts using screws and springs, then position the parts with tiny hammers and tweezers, and finally fasten them in the desired position. However, with increasing component miniaturization, the tolerances become smaller and smaller, and the capabilities of the human hand are no longer adequate. The application of partially automatic micromanipulation systems of conventional size, which are teleoperated; thereby, the hand motions of the human operator are translated into finer 3D motions for the manipulators of the manipulation system by means of a joystick or mouse. Here, the dexterity of the human hand is supported by sophisticated techniques. However, the fundamental problem of the resolution of the fine motion and of the speed remain, since the motion of the tool is a direct imitation of that of the operator’s hand. The use of automated multifunctional micromanipulation desktop stations’ supported by miniaturized flexible robots which employ MST-specific direct- drive principles. These robots could be mobile and are able perform manipulations in different work areas. The transport and micromanipulation units performing the assembly may be integrated onto one chip. As opposed to the aforementioned micromanipulation technique, there is no direct connection between the operator’s hands and the robot. The assembly steps may be carried out with the help of closedloop control algorithms. The human assigns all tasks to miniaturized assembly mechanisms and, by doing so, tries to compensate for his limited micromanipulation capabilities. Many microrobots can be active at the same time in a desktop station. The use of many flexible nanorobot systems which solve the manipulation tasks in close cooperation. Here, the robot size is comparable to that of the manipulated object. This concept could be based on the human behavior, but its realization lies in the distant future. In general, manipulations vary from an application to another. However, approximately the same operation sequences are used in every case. They are: grip, transport, position, release, adjust, fix in place and processing steps like cutting, soldering, gluing, removal of impurities, etc. In order to be able to carry out these operations, corresponding tools are needed, such as microknives, microneedles to affix objects, microdosing jets for gluing, microlaser devices for soldering, welding or cutting, different types of microgrippers, microscrapers, adjustment tools, etc. Microgrippers play a special role, since they considerably influence the manipulation capabilities of a robot. Microgrippers can clamp, make a frictional connection or adhere to the material, depending on the physical and geometrical properties of an object. Adapting a gripper to the shape of the object to be gripped is often the best solution in the microworld, even at the cost of flexibility. This allows handling of a workpiece having a complex shape, such as a gear. Thereby, the gripper securely attaches to the contour of the part. For small, smooth parts, a suction pipette might be a practical tool. If the upper surface of a workpiece must not be touched or gripped due to technological reasons, it can be protected by a corresponding form-fit of the pipette hole. For contour clamping and frictional connections in manipulations involving fragile parts, elastic grippers made of soft plastics are preferred over metal grippers. Due to the variety of task-specific gripping tools in automated micromanipulation systems, a suitable gripper exchanger system might be necessary. It should be mentioned that it is not always possible to adapt conventional manipulation methods to the demands of the microworld. A major problem is the effect of various forces which is completely different from the macroworld. Gravitation only plays a minor role in the microworld, but attractive forces,such as electrostatic forces or Van-der-Waals forces, are significant. Liquid surface tension can also act as an attractive force in micromanipulations if humidity is high or if a manipulator is wet. This unusual sensitivity to forces can be very irritating in a micromanipulation station. For example, it can be easier for the robot to grip and manipulate an object than to release it afterwards. On the other hand, such an adhesion force can be used to develop new gripping methods which can fundamentally differ from the familiar mechanical and pneumatic methods. In Ref. 【1】 , several interesting ideas were shown for adhesive gripping, such as electrically charging a manipulator or wetting a gripper surface by special micromachined orifices. The performance and degree of intelligence of a micromanipulation station is low for a manual operation; it improves by going to a teleoperation and further to an automation; this is similar with conventional robots. Most micromanipulation investigations today focus on the improvement obtained by going from a purely manual to a teleoperated system【2–4】. As previously mentioned, attempts are being made to make the transmission of effects from the microworld to the operator as realistic as possible. It is important that the operator has the entire scene in his field of view and that he can see the workspace from different angles. Besides visual information, the operator should also be able to receive acoustic and force signals if possible; this may increase the accuracy of his movements and avoid destroying the microobjects. For this, force sensors are needed which are implemented into the microtools (e.g., a microgripper) . Suitable solutions are now being sought after to realize such sensors【5】. 3. Development of a flexible micromanipulation Station： Typically, in a conventional automatic or semiautomatic assembly station, standardized mechanical parts are assembled in well-defined work positions. The robots performing the work are usually of multi-axis arm design or they are gantry systems,usually driven by DC motors. Today, it is being attempted to use these type of familiar systems for handling and assembling of miniaturized components with dimensions in the millimeter range. For example, a modular microassembly system with four degrees of freedom is currently being developed【6】.With increasing workpiece miniaturization, however,it becomes more and more difficult to use conventional manipulation robots for assembling microsystems.The manipulation accuracy is mechanically limited for conventional robots, since disturbing influences which can be neglected in the microworld, such as small fabrication defects, friction, thermal expansion or computational errors, play a large role in the microscale. Due to the mechanical drives for the actuator’s motions, these robot systems must undergo regular maintenance and are subject to mechanical wear, which makes them expensive. The assembly process in the microworld is influenced by the mass-related dynamics of the objects being handled. Different processing conditions exist for manipulating microscopically small components. The positioning accuracy and the tolerances of the micro-components lie in the nanometer range, a few orders of magnitude lower than in conventional assembly. These accuracy requirements can only be obtained with manipulators which have highly accurately drives utilizing the MST and advanced closed-loop control. Therefore, a microrobot-based flexible desktop station is of particular interest. A new concept for an automated micromanipulation desktop station is now being investigated 【7】. The main part of the station are the piezoelectric microrobots which were presented in Refs.【8,9】.Each robot has a micromanipulating unit integrated in a mobile platform, which makes it capable of moving and manipulating. Tools can be easily exchanged. These robot properties are good preconditions for the complete sensor supported automation of manipulation processes in the microassembly station. Owing to the flexibility of the microrobot, this multifunctional desktop station can also be used for other things, such as handling biological cells or actively testing microelectronic chips with temperatureor or voltage probes. This flexibility can also be used to accommodate several robots in the station, which can cooperate and carry out manipulations. The schematic design of the micromanipulation desktop station is shown in Fig. 1. The operations of the microassembly station may be described as follows. The parts are first separated and placed into magazines in order to have them correctly positioned for automated assembly. This is necessary, since microcomponents are often delivered as bulk material. This step can also be automated in a powerful microassembly station, to avoid the expensive manual handling. A microrobot removes a micromechanical component from the magazine and transfers it to a processing cell where the component can then be prepared for microassembly by other microrobots. In this step, adhesives or solder can be applied, adjustment marks taken, or other simple operations carried out. After the part has been processed, it is gripped by a robot and brought to a microassembly cell. If necessary, the same operations are repeated many times in order to fetch the other necessary components from a supply container and prepare them for assembly. All components are positioned correctly, affixed to each other and adjusted. Thereafter, they are joined together by various interconnection techniques, e.g., laser spot welding, gluing, insertion, wire bonding, etc. After assembly, a robot brings the finished component either to another work station or a microassembly cell for further processing or to an inspection cell, where all functions of the microsystem are being tested. Finally, the finished system is transported to a storage. The entire assembly process occurs in the desktop station under an automated light-optical microscope which is equipped with a RS232-standard interface. The sphere of operation includes a highly precise positioning table with two translational degrees of freedom （x–y plane ）and a glass plate fixed on top of it. By controlling the movements of the table, each individual working cell on the glass plate can be brought under the microscope. The station has a central computer （Pentium PC） which is used for task-specific assembly planning. The necessary operational steps are defined and carried out successively. The commands of the central computer are then further processed on a lower control level, using a parallel computer system with the C167 microcontrollers. This system was reported in Ref.【10】. The central computer is coupled with the parallel computer system over serial and parallel interfaces. These commands are resolved into command sequences for all active system components （robots, microscope and positioning table） by an execution planning system, and finally performed. Thanks to the parallel computer system, the generated commands can be executed in parallel, which makes the microassembly station capable of real-time behavior. The movements of the positioning table, different microscope functions （objective changing, focusing, lighting） and every piezoactuator are controlled. In order to automatically control the manipulation processes in the microassembly station, there must be sensor feedback. Therefore, the light-optical microscope is equipped with a CCD camera. The camera and the microscope form the local sensor system with the help of which the position of the microobjects and the robot tools must be determined. For this it supplies visual information on the robot tools and the microobjects to the central computer. The gross position of the robots on the glass plate is detected by a global sensor system which includes a laser measuring unit and another CCD camera. The visual sensor information from both the local and the global sensor systems is used by the control algorithms to generate new commands for the robots, microscope and the positioning table. Vision is supported by a frame grabber in connection with fast real-time image recognition and processing systems. The vision parameters are passed on to the parallel computer system. They are used as a set point for the control loop. 4. Planning of the microassembly： The above description of microassembly station activities is very general and perhaps makes the assembly process sound too simple, but many problems must first be solved. After a microsystem has been designed, all tools and techniques necessary for its automated assembly should be determined, so that the microassembly station can be set up for a taskspecific operation sequence. The specified techniques and tools must take the geometry of the components of the microsystem into consideration, as well as their physical properties, such as rigidity, texture and temperature stability. Therefore, the planning phase of an automated microassembly requires a high degree of competence. Pure top-down planning in a microassembly station seems to be impossible since the selected robots and their tools determine the flexibility and the degree of automation of the station, and therefore, also determine its performance limits. One possible planning strategy is the meet-in-the-middle strategy; thereby, this intermediate interface can be on the tool level. Indeed, the main functions of assembly planning are the determination of the task-specific sequences of the elementary operations and the selection of necessary tools for carrying out the work (top-down planning) . On the other hand, the tools and the elementary operations needed for the assembly of a microsystem also require that the microrobots have specific functional properties, which may influence the robot design (bottom-up planning) . As mentioned, for more complex assembly tasks several robots must be used together in the desktop station. Individual robots can, for example, be specialized to take care of one or more certain assembly operations. In this case, the robots carry out their manipulation tasks in a sequence which is defined during the planning phase. For more complex operations, robots can be pooled together to do simultaneous actions with the help of several different tools (e.g., transferring or gripping of objects) . In this case, the operator’s commands are no longer transmitted one-by-one to the manipulator arms, but are applied to the entire multirobot system, e.g., by means of the ‘one-by-multiple’ method 【11】. Here, one microrobot acts as the leader of the group, it gets micromanipulation assignments from the operator and then coordinates the other microrobots to complete the task using an automatic process for communicating with the robots and then giving them the corresponding commands. If the cooperating robots are equipped with sensors, new object manipulation methods can be developed, which are based on the distributed observation of the objects. The object could be observed from two view points, for example, which would supply