外文翻譯 專 業 機械設計制造及其自動化 學 生 姓 名 陳 曦 班 級 BD 機制 031 學 號 0320110129 指 導 教 師 劉 道 標 . 1 模具數控加工計算機輔助刀具選擇研究 耿鐵 段修濤 譯 引言 數 控加工中包括刀具軌跡的產生和刀具選擇兩個關鍵問題。前一問題在過去的20 年里得到了廣泛而深入地研究， 發展的許多算法已在商用 CAD/ CAM 系統中得到應用。目前大多數 CAM 系統能夠在用戶輸入相關參數后自動產生刀具軌跡。比較而言，對以質量、效率為優化目標的刀具選擇問題的研究還遠未成熟，當前還沒有商用CAM 系統能夠提供刀具優選的決策支持工具，因而難以實現 CAD/ CAM 的自動有機集成。刀具選擇通常包括刀具類型和刀具尺寸。一般來說，適合一個加工對象的刀具通常有多種，一種刀具又可完成不同的加工任務，所以僅考 慮滿足基本加工要求的刀具選擇是較容易的，尤其對孔、槽等典型幾何特征。但實際上，刀具選擇通常和一定的優化目標相聯系，如最大切削效率、最少加工時間、最低加工成本、最長使用壽命等，因此刀具選擇又是一個復雜的優化問題。比如模具型腔類零件，由于幾何形狀復雜 (通常包含自由曲面及島 ) ，影響刀具選擇的幾何約束在 CAD 模型中不能顯式表示，需要設計相應的算法進行提取，因而選擇合適的刀具規格及其刀具組合，以提高數控加工的效率與質量并非易事。 模具型腔一般用數控銑的加工方法，通常包括粗加工、半精加工、精加工等工序。粗加工的原 則就是盡最大可能高效率地去除多余的金屬，因而希望選擇大尺寸的刀具，但刀具尺寸過大，可能導致未加工體積的增多 ；半精加工的任務主要是去除粗加工遺留下來的臺階 ；精加工則主要保證零件的尺寸及表面質量。考慮到目前完全由計算機進行自動選刀還存在一定困難，因而在我們開發的計算機輔助刀具選擇(Computer Aided Tool Selection ， CATS)系統中，立足于給用戶提供一個輔助決策工具，即粗加工、半精加工、精加工等，真正的決策權仍留給用戶，以充分發揮計算機和人的優勢。 1 系統基本結構 CATS 系統的 輸入為 CAD 模型，輸出為刀具類型、刀具規格、銑削深度、進給量、主軸轉速 (切削速度 ) 和加工時間等六個參數 (如圖 1) ，包括刀具類型選擇輔助決策 2 工具、粗加工刀具選擇輔助決策工具、半精加工刀具選擇輔助決策工具及精加工刀具選擇輔助決策工具等。 圖 1 計算機輔助刀具選擇系統的輸入與輸出 鑒于粗加工在型腔加工中的重要地位 (通常為精加工時間的 5 10 倍 ) ，粗加工時系統具有刀具自動優化組合的功能，以提高整體加工的效率。除了上述決策工具外，系統還具有查看刀具詳細規范、根據刀具類型和尺寸推薦加工參數及評估加工時間等功能，最后生成總的刀具選擇結果報表 (如圖 2) 。系統所有的刀具數據及知識均由后臺數據庫做支持。 圖 2 計算機輔助刀具選擇 系統的基本功能與模塊 2 關鍵技術及算法 1) 刀具類型選擇 根據模具型腔數控加工實踐，型腔銑加工的刀具一般分為平頭銑刀、圓角銑刀及球頭銑刀三種。設刀具直徑為 D，圓角半徑為 r ，當 r=0 時為平頭銑刀， 0R 刀具又可分為整體式和鑲片式。對于鑲片式，關鍵是選取刀片的材質，刀片材質的選擇取決于三個要素：被加工工件的材料、機床夾具的穩定性以及刀具的懸臂狀態。系統將被 3 加工工件的材料分為鋼、不銹鋼、鑄鐵、有色金屬、難切削材料和硬材料等六組。機床夾具的穩定性分為很好、好、不足三個等級。刀具懸臂分為短懸臂和長懸臂兩種，系統 根據具體情況自動推理出刀片材質，決策知識來源于 WALTER 刀具手冊，系統由用戶首先交互選擇刀具類型。對鑲片式刀具，基于規則自動推理出合適的刀片材質。例如，如果被加工工件的材料為 “ 鋼 ” ，機床夾具的穩定性為很好，刀具懸臂為短懸臂，則刀片材質應為 WAP25 。 2) 粗加工刀具組合優化 型腔粗加工的目的就是最大化地去除多余的金屬，通常使用平頭銑刀，采取層切的方法。因此， 3D模具型腔的粗加工過程，實際上就是對一系列 2.5D 模具型腔的加工。刀具優化的目的就是要尋找一組刀具組合，使其能夠以最高的效率切除最多的金屬。刀具組 合優化的基本方法如下： a.以一定的步長做一組垂直于進刀方向的搜索平面與型腔實體相交，形成搜索層。 b. 求出截交輪廓。 c. 計算內外環之間或島與島之間的關鍵距離，即影響刀具選擇的幾何約束，算法流程如圖 3 所示。 圖 3 求關鍵距離算法流程 d. 根據合并原則 (相鄰關鍵距離相差小于給定閾值 ) 對搜索層進行合 并，確定加工平面和可行刀具集，形成加工層。 e. 確定每一加工層使用的刀具，即型腔加工的刀具組合。 f. 根據刀具推薦的加工參數 (切削速度、銑削深度和進給速度 ) ，計算材料去除 . g. 根據加工層實際切除的體積，計算每一加工層的加工時間。 h. 計算型腔總的加工時間和殘余體積。 i.對該組刀具組合的總體加工效率進行評估。 4 j. 重復 a i，直至求出最優的刀具組合。如以時間為目標，即要求以整個型腔的加工時間 t 最短來優化刀具組合。基于上述方法，可建立如下形式化的優化模型。 MRRi=(dicij)(Nfz)( 切割截面積乘進給率 ) 式中： n 型腔加工層數量 ； m 每一加工層刀具的銑削次數 ； l 每一加工層中的搜索層數量 ； q 每一加工層可行的刀具數量 ； h 型腔深度 ； cij i 加工層第 j 次銑削深度 ； aj 第 j 切割層底面積 ； vi i 加工層的銑削體積 ；MRRi i 加工層的材料去除率 ； di i 加工層的刀具直徑 ； dip i 加工層可行刀具集合 ； rik i 加工層 k 搜索層的關 鍵距離 ；e1 控制搜索層合并的常數 ；e2 控制殘余體積的常數 ；V 型腔體積 ；DV 殘余體積 ； N 主軸轉速 ； f 刀具每齒進給量 ； z 刀具齒數。考慮到不同的搜索平面步長會產生不同的加工層，從而導致不同的加工時間和殘余體積，因此有時盡管總的加工時間較短，但殘余體積可能較多。由此可見，單獨以加工時間為目標進行優化有時并不一定科學。為此，提出了效率系數的概念，綜合考慮了加工時間和殘余體積的因素，加工時間越短，殘余體積越少，則效率系數就越高。令： 上式中前一項反映了加工單位體積的時間系數，其中 k =DV/V 為殘余體積百分數。這樣，效率系數可定義為 q = 1/ Q 。 3) 半精加工刀具選擇 5 半精加工的主要目的是去除粗加工殘留下的臺階狀輪廓。為完全去除臺階，銑削深度必須大于每一臺階到零件表面的距離 x。其算法步驟如下： 步驟 1 由零件實體模型獲得兩個相鄰截面的表面積以及相應的輪廓長度 ； 步驟 2 計算平均輪廓 長度 ； 步驟 3 計算臺階寬度 ； 步驟 4 計算臺階拐角到零件表面的法向距離 x ； 步驟 5 重復步驟 1步驟 4 ，決定每一臺階的銑削深度 ； 步驟 6 計算刀具直徑 D， 按經驗 D=x/0.6 或根據刀具手冊推薦 ； 步驟 7 選擇銑削深度大于 x 的最小刀具。 4) 精加工刀具選擇 5) 精加工刀具選擇的基本原則是：刀具半徑尺寸 R 小于零件表面最小的曲率半徑 一般取 R=(0.8 0.9)r。其算法步驟如下： 步驟 1 從零件實體模型計算最小曲率半徑 ； 步驟 2 從刀具庫中檢索出刀具半徑小于計算所得的曲率半徑的所有刀具 ； 步驟 3 選出滿足上述要求的最大刀具 ； 步驟 4 如果所有刀具大于最小的曲率半徑，選擇最小的作為推薦刀具。 3 系統實施及算例 CATS 系統在 UG/OPEN API 環境下應用 C語言開發而成。后臺數據庫為 Oracle 8i ，利用 ODBC 編程實現 UG 與數據庫之間的通訊。所有的刀具數據及知識來自德國WALTER 公司的硬質合金刀具綜合樣本。圖 4為一包含島及雕塑曲面的模具型腔， 根據上文提出的粗加工刀具組合的優化方法，該模具型腔粗加工刀具的優化組合為 20，12， 8， 5。計算中，工件材料選定為中碳鋼，切削速度推薦值 為 100m/min ，銑削深度為刀具直徑的 1/ 2 ，進給量根據刀具推薦值由程序自動修正計算。同時，假定刀具庫中現有平頭銑刀刀具規格為 f3， f4， f5， f6， f8， f10， f12， f16， f20。同樣，根據半精加工和精加工的刀具選擇算法，得到的球頭銑刀的刀具直徑分別為 4和 3。 6 圖 4 包含島及雕塑曲面的模具型腔 4 小結與討論 模具型腔加工的工藝規劃通常需要很高的技術與經驗，準備 NC 數據的時間幾乎和加工時間一樣多。因此，自動產生型腔加工的工藝計劃及 NC 加工指令的需求就顯得愈加迫切 。 本文系統研究了模具型腔工藝規劃中的刀具選擇問題，提出了模具型腔粗加工、半精加工、精加工刀具選擇的原則和方法，構造了相應的實現算法，并在UG/OPEN API 環境下進行了初步編程實現，開發了 CATS 原型系統。在刀具類型和規格確定的基礎上，系統還可根據刀具手冊推薦加工參數 (切削速度、銑削深度、進給量等 ) ，對相應的加工時間進行評估。其最終目 的是真正實現 CAD/CAM 的集成，繼而通過后處理產生數控加工指令。目前 CATS系統的界面還是獨立于 UG的 CAM界面， CATS的策結果還需要用戶重新輸入到 CAM。 需要指出的是，要提高模具型腔的總體加工效率，需要從粗加工、半精加工、精加工的整體上考慮，進行多目標組合優化，這將是我們下一步要進行的工作 。 7 Mould&Die NC computer-aided Tool of Selection Geng Tie State Key Lab.of Mould&Die Technology,HuaZhong University of Science and Technology Wuhan 430074 ,China Introduction NC machining tool path generation and tool selection, including the two key issues. Before a problem in the past 20 years has been wide-ranging and in-depth study, Many algorithms for the development of CAD CAM system has been applied in the business. Most CAM systems to the user input parameters with automatic tool path. Comparatively speaking, the quality, efficiency and optimization tool of choice is far from mature. Currently no commercial CAM system optimization tool can provide decision support tools. it is difficult to achieve automatic CAD CAM integrated. Tools typically include tools and tool type size. In general, usually for a processing tool targeting a variety of different processing tasks as well as the completion of a tool. Therefore, considering only meet the basic requirements of the tool selection process is relatively easy, particularly for the holes, ducts, etc. typical geometric features. But in reality, and we usually choose the optimal tool goals, such as the most efficient cutting, processing time at least, the lowest manufacturing cost, and the longest life expectancy, and so on tool selection is a complex optimization problems. For example, die parts, complex geometric shapes (usually free surface and Island) Tool choices affect the geometric constraints in CAD model can not explicit that it is necessary to design appropriate extraction algorithm, choose the appropriate tools and tool specifications portfolio to improve the efficiency and quality of NC is not an easy task. NC general cavity with the processing methods, including extensive and usually, semi-finishing and finishing processes. Snag is the principle of maximum extent possible the efficient removal of excess metal and therefore 8 wish to opt for large size of the tool, But cutter size is too large, might not lead to an increase in processing volume； the main task is getting extensive and semi-finished stage left； finished the main components of the size and surface quality assurance. Taking into account the current election entirely by the computer automatically knife there are still certain difficulties, therefore, in our development of the computer-aided tool selection (Computer Aided Tool Selection, CATS) system, based on the users to provide a decision-support tool, roughing, semi-finishing and finishing. the real decision-making power is left to users, and to bring into full play the advantages of computer. 1.System structure. CATS system for CAD model input, output types of tools, tool specifications Milling depth, feed, Spindle Speed (Speed) and the processing time of six parameters (Figure 1). Tool types of options, including decision-support tool, tool selection decision-support tool for roughing. tool selection decision-support tool for semi-finishing and finishing tool selection decision-making tools. Figure 1 computer-aided tool selection of input and output Given extensive and important position in the cavity (usually 5-10 times finishing time). snag when the system is automatically optimized combination of functional tool to enhance overall processing efficiency. In addition to 9 the above decision-making tools, the system also has a detailed standardized tools to detect, According to the recommendation of processing parameters and the size and type of tool to assess the function of processing time. Tool choice of the final total returns generated (Figure 2). Tool System data and knowledge are all in support of the background database. Figure 2 CAD tool selection and the basic function modules 2 key technology and algorithms 1) Tools for choice According die NC practice, the tool generally consists of Milling Cutter peace. Fillet cutter and the cutter ball three. Based tool diameter D, the radius r, r = 0 when the crew cutter. 0R tools can be divided into the overall style and framed chip. Tipped for the ceremony, the key is to select material blades, blade materials are determined by three factors : the workpiece material, Machine tool fixture and the stability of the cantilever. Workpiece material will be processed into steel, stainless steel, cast iron, nonferrous metals, materials and other hard-to-cut materials and hardware six. Stability jig into the well, less than three grades. Cantilever Tool cantilevered into short and long cantilever two, the system automatically inducted blade material under specific circumstances. WALTER Tool knowledge comes from the decision-making manuals, system users to choose the first interactive tool types. Tipped Tools of the rule-based Automated Reasoning blades suitable material. For example, if the workpiece material as steel, the stability of the fixture good tool for short cantilever cantilever. Blade material will be WAP25. 2) Portfolio Optimization Tool snag The objective is to maximize the cavity snag in the removal of excess metal, used Ping Cutter and take all the layers. Therefore, the 3D die roughing process is actually a series of 2.5D die for processing. Optimization Tool 10 Tool Groups goal is to find a combination that will enable it to the highest removal efficiency of most metals. The basic approach is as follows : Portfolio Optimization Tool A. To do a certain step in the direction perpendicular to the feed cavity search plane and entities intersect, the formation search layer. B. Deadline for submission of outline obtained. C. Calculation link between or outside of critical distance between the islands, the choice of tools to influence the geometric constraint, the algorithm shown in figure 3. Figure 3 key demand from the algorithm. D. Under this principle (the distance between adjacent key difference is less than a given threshold) level of the search for a merger Plane processing and identify viable tool sets, forming layers. E. Each layer processing tool used to determine that the combination of the tool cavity. F. According to the recommended processing tool parameters (cutting speed, feed rate and depth of milling), material removal rate calculation. G. According to the actual removal of the layer processing volume, the processing time is calculated for each layer processing. H. Calculate the total processing time and residual volume cavity. I. The overall composition of this group processing efficiency assessment tool. J. Ai repeated until the optimal combination of the tool. If time goal that the processing time t required for the entire cavity tool to optimize the 11 combination of the shortest. Based on the above methodology, we can establish the following formal optimization model. MRRi=(dicij)(Nfz) (X cross -sectional area of cutting feed rate) Where : n-layer processing cavity volume； M-layer processing each of the milling cutter number； l-processing layer in the search each layer volume； q-layer processing every tool possible number； h-cavity depth； cij -i time processing layer j Milling depth； aj - j cutting area of the base layer； vi -i processing volume ；MRRi -i milling layer, layer of material removal； di -i layer processing tool diameter； -i processing layer dip viable tool set； rik -i processing layer k ；e1 search of the key distance-control layer with a constant search； e2 - residual volume control constants； V-cavity volume ；DV - residual volume； N-spindle speed； f-cutter feed per tooth； z-cutter teeth. Plane taking into account the different steps in the search process will produce different levels, resulting in the processing time and residual volume, So sometimes, even though total processing time shorter, but more likely to residual volume. This shows that the target of a separate optimization of processing time are not necessarily science. Therefore, the coefficient of efficiency of the concept, considering the processing time and the residual volume, the processing time is shorter. less residual volume, the higher the efficiency coefficient. Possession : 12 On a middle-processing unit volume reflects the time factor, k =DV/V percentage of residual volume. Thus, the efficiency coefficient defined as q = 1/ Q. 3) Semi-finished Tool Selection The main purpose is to remove semi-finished roughing the residue level contour shape. For the complete removal of height, depth milling the surface must be greater than the distance between each level of x components. Algorithm steps are as follows : Step 1 model will have two parts from the adjacent section and the corresponding surface contour length； Step 2 calculation of the average length profile； Step 3 calculating height width； Step 4 Calculation of the corner to level the surface of the parts to distance x law； Step 5 Repeat steps from 1 to 4 steps, each step of the decision Milling depth； Step 6 curve of the diameter D, D=x/0.6 Tools manual or recommended by experience； Step 7 x greater than the minimum depth of choice milling cutter. 4) finished Tool Selection The basic principle is : finishing tool selection tool radius R size smaller than the smallest curvature radius r surface parts. General admission R = (0.80.9) r. The following steps : Step 1 algorithm model from the smallest radius of curvature parts entities； Step 2 retrieved from the database tool cutter radius less than the radius of curvature calculation for all tool； Step 3 select the best tools to meet these requirements； Step 4 : If all the tools than the smallest radius of curvature. Minimum recommended as a tool of choice. 13 3 implementation of the system and examples CATS system in the C language environment UG/OPEN API was developed. Background for the Oracle 8i database using ODBC programming and database communications between UG. All the data and knowledge from Germany WALTER Tool Company Carbide Tool integrated samples.Figure 4 contains a sculptured surface of the mold cavity and the island, according to the snag in the portfolio optimization tool, The die-extensive portfolio optimization tool for 20,12,8,5. Calculation, the workpiece material selected for carbon steel, cutting speed 100m/min recommended value. Milling depth of two sixth tool diameter, feed rate according to the recommended values from the process tool that automatically calculated. Mean