Chapter 2

Template-Based Integrated Design: A Case Study

Ali K. Kamrani1 and Abhay Vijayan2

Abstract This chapter discusses the application of concurrent engineering principles in the product development phase to achieve maximum cost reduction by foreseeing the manufacturability constraints and solving perceived problems beforehand in the concept-generation phase itself. This would entail integrating the different phases, namely, design calculation, computer-aided design, and process planning. This design methodology is applied to the process of shaft design and bearing selection in the product development phase of a motor. The design calculations include a force analysis and also bearing life calculations. The output from these design calculations is used to develop drawings in integrated design and engineering analysis software (I-DEAS). These designs are then used to carry out a time and cost analysis.

2.1

Introduction

The increased globalization of the manufacturing industry opened up new markets with varying customer requirements and also increased competition with more number of companies by introducing low-cost products. This caused an acceleration in the rate of product change which forced engineers and managers to respond with products having lower cost, better quality, and shorter development times. As a result, companies started reinventing their product development methodologies. One of the areas that have gained importance due to this is concurrent engineering. Concurrent engineering techniques have gained wide acceptance in the high volume manufacturing companies such as automotive and consumer electronics markets. It has also gained importance in the low volume, high innovation industry of aerospace. Concurrent engineering is defined as an integrated and systematic approach to the design of products and their related processes, including manufacture and support.
1 Associate Professor, Department of Industrial Engineering, University of Houston, Houston, Texas 77004 2

Department of Industrial Engineering, University of Houston, Houston, Texas 77004

A.K. Kamrani, E.S. Abouel Nasr (eds.) Collaborative Engineering: Theory and Practice, doi: 10.2007/ 978-0-387-47321-5, Springer Science + Business Media, LLC 2008

19

20

A.K. Kamrani, A. Vijayan

This approach causes designers to consider the whole product life cycle from conception through disposal while selecting a design [22]. One of the main reasons for concurrent engineering to gain broad acceptance was its impact on time. Second, it got the manufacturing and marketing departments involved early in the product development process. This enabled them to influence the design and obtain a cost-effective and high performance product. It also made full use of latest advances in information technology to develop libraries with complex accumulations of detailed design. Because of this, knowledge gained during the development of one product could be passed on to subsequent product developments [20].

2.2

Problem Description

Any manufacturing process starts with the product development phase. In a sequential engineering approach the design process and the actual manufacturing are considered as mutually independent. The design department develops a product depending on the customer requirements and design constraints and provides engineering drawings to manufacturing. Here the design department does not take into consideration the manufacturing capabilities and the process limitations on the shop floor. The manufacturing department then carries out a manufacturability analysis of the design and returns it back to the design section with a list of manufacturing concerns. Here each department passes the design or drawings back and forth until they achieve functional success. In the concurrent engineering approach, this back and forth movement of drawings is completely eliminated since manufacturability concerns are addressed in the design phase itself. Once the design is finalized, resources have to be assigned for manufacturing the product. This is carried out in the process planning phase. In a manual process planning environment, each process plan reflects primarily the experience and skills of the individual process planner; hence it is difficult to maintain consistency and discipline in the planning process [5]. It also acts as a significant break in the automated flow of data from CAD to CAM. Hence it is necessary to automate process planning along with the design phase and allow smooth integration of the two. Advances in information technology and increase in the processing speed of computers help to develop libraries which contain all the data required for manufacturing. This also eliminates the need to develop a new process plan from scratch if a similar product component has already been manufactured.

2.3

Problem Solution

Modifications made to the product in the design phase have maximum impact on the cost and functionality of the product [15]. Hence, we would be concentrating on applying concurrent engineering principles in the product development stage

2 Template-Based Integrated Design: A Case Study

21

and thus achieve maximum cost reduction by foreseeing the manufacturability constraints and solving perceived problems beforehand in the concept-generation phase itself. In an integrated CAD/CAM environment, parts are detailed using a computer graphics system. The system stores the geometric information necessary to create process plans. These process plans determine the machining processes, tools, and machines required to convert a part from its initial form to the desired form. To apply this methodology, we have to interlink the different phases of the manufacturing process. The template-based system can be divided into three different parts: 1. Design calculations. 2. Drafting. 3. Process planning. The system schema is illustrated in Fig. 2.1. The user requirements are analyzed and converted into input parameters for the design calculation phase. The output from the design calculations is utilized as the control parameters for selecting feasible drawings from the CAD system. The process planning module uses the CAD data and then generates route sheets as well as a cost analysis of feasible designs. This enables the designer to study the impact on the design on the basis of cost and manufacturing capability. The template-based design methodology is used for the bearing selection and shaft design process of an electric motor. The input parameters are speed, power of motor, and also the minimum life of bearing required. This information is then used to obtain the parameters for CAD models. The CAD modeler then returns feasible part drawings which act as the input to the process planning phase. The process planning module prepares the route sheet and carries out a time and cost analysis for different feasible part drawings and returns it to the user.

Input parameters Part Drawing Process Plan (Route Sheet) User Interface Computer Aided Process Planning

CAD Control Parameters

Drawing Templates

Process Templates

Fig. 2.1 Basic system schema

22

A.K. Kamrani, A. Vijayan

2.4

Electric Motor

An electric motor converts electrical energy to mechanical energy. Alternating current motors are classified into polyphase synchronous motor, and single-phase or polyphase induction (asynchronous) motors. In polyphase synchronous motors, magnetic field associated with the rotor results from a rotor winding (electric field) excited by direct current via slip rings or from permanent magnets on the rotor structure. In single-phase or polyphase induction motors, the rotor magnetic field is created by electromagnetic induction effects [9]. There are two common features for all types of motors. The first feature is that, for an average torque to be produced, magnetic fields of the stator and the rotor must be stationary with respect to each other. The second feature is, for a specific rotor length, air gap diameter, and speed, there is a maximum average power output rating. In induction motors, power is transferred from the stationary part (stator) to the rotating part (rotor) [10]. The stator is made of a stack of steel laminations with insulated slots opening to the inside diameter and holding stator coils or windings. The rotor is also made up of slotted steel laminations. The rotor conductors and end rings are formed using a die cast process. The rotor stack length is approximately equal to the length of the stator stack and its outside diameter is smaller than the inside diameter of the stator packet by twice the air gap. This rotor packet is fitted on the shaft. The shaft is supported on the bearings. The end of the shaft where the power is transferred through a pulley or coupler is called the drive end. The other end is called the nondrive end. An external fan is mounted on the nondrive end of the shaft for external cooling. The proposed design methodology is applied to bearing selection and shaft design of electric motors. The inputs to design calculations are speed, power of the motor, type of drive, and minimum life of bearing required. This information is used to calculate and select the minimum diameter of shaft drive end and also to select the bearings required. Dimensions of the bearings and shaft extension dimensions are then used as the control parameters for CAD template retrieval. The process planning module then carries out a cost analysis for the modified product and returns the minimum cost machine to the user.

2.5

Design Calculations

Power (P) and speed (N) are inputs to the design calculation process. The torque (T) acting on the shaft is given by T(Nm) = [1000 60 (P(kW)/N(rpm))]/(2p) The tangential force (Ft) is calculated as Ft(N) = [T(Nm)/(D(m)/2)] (2.2) (2.1)

2 Template-Based Integrated Design: A Case Study

23

where D is the nominal diameter of the pulley or the coupler [16]. Based on the types of transmission, the radial force (Fr) and the axial force (Fa) acting on the shaft at the point of mounting is determined [21]. Forces generated at the elements mounted on the shaft produce bending moments and the power transmitted produces shearing stresses. A shaft force analysis is then carried out for the section at which transmission element is mounted. This analysis includes both the maximum shear stress theory and the distortion energy theory [14]. The stresses acting on the surface of a solid shaft are given as follows: sx = 32M/pd 3 + 4F/p d 2 txy = 16T/pd 3 (2.3)

(2.4)

Assuming that the axial force produces the same kind of stress as the bending moment, max = (2/pd3)[(8M + Fd)2 + (8T)2]0.5 s' = (4/pd3)[(8M + Fd)2 + 48T 2]0.5 (2.5)

(2.6)

From the maximum shear stress theory, the allowable shear stress is given by all = Sy/2n From the distortion energy theory, the normal stress is given by s'all = Sy/n (2.8) (2.7)

where d diameter of the shaft; F axial force; M bending moment; n factor of safety; Sy yield strength; T twisting moment; max maximum shear stress; s' normal stress; all allowable shear stress; and s'all allowable normal stress. The factor of safety for maximum shear stress is taken as 1.5 and for maximum normal stress is taken as 2. The material used for the shaft is alloy steel C1045 with density 7830 kg/m3, whose allowable shear stress is equal to 60 kpsi and allowable normal stress is 92.5 kpsi [7]. Equating max with all and s' with s'all, the minimum required diameter of shaft extension is calculated. The shaft is primarily supported at the bearings. Elements mounted on the shaft include the rotor and the transmission device. A free body analysis of this shaft is carried out and the radial and axial forces acting on the bearings determined. A search heuristic is developed to select all the feasible combinations of bearings from the bearing database that will provide the required life. The process retrieves the bearing dimensions for each bearing and its associated cost. The bearings considered are deep groove ball bearings and roller bearings. The length and diameter of the bearing seat would be the same as the inner diameter and width of bearings used. The power and speed required for the application determine the electrical

24

A.K. Kamrani, A. Vijayan

design of stator and rotor. A database of electrical specifications is created which contains the rotor packet dimensions and its corresponding weight for each combination of power and speed.

2.6

CAD Modeler

Integrated design and engineering analysis software (I-DEAS) is used to generate and store shaft design templates. The shaft consists of the drive end and nondrive end shaft extensions, the drive end and nondrive end bearing seats, and the rotor packet seat. The different dimensions of the shaft are obtained from the design calculations. They include the minimum shaft extension and bearing seat dimensions. The rotor seat dimension is obtained from the electrical design database. The drawings are classified on the basis of the shaft ends as shown in Fig. 2.2. It is either cylindrical or conical. Second classification depends on whether the shaft has extensions on both ends or only one shaft extension. Both end shaft extensions are used either to run a tachometer or to run an external blower at the nondriving end. Further classifications are based on the applications: low speed high torque and high speed low torque applications. Once the template is selected, necessary changes are made in the dimensions and also additional features like undercuts and threading are incorporated. All the changes made to the template are recorded. These modifications are passed on to the process planning stage.

Fig. 2.2 Shaft classifications

2 Template-Based Integrated Design: A Case Study

25

2.7

Process Planning

Process planning is the development of a set of work instructions to convert a part from its initial form to the required final form [6, 11, 12, 19]. In the proposed system, the process plan provides detailed description of the manufacturing processes and machine tools required to make the part. This detailed information includes the list of different machines that are used, specific cutting energy for the work material, cost per hour of operation, standard setup and tear down times, and the number of tools utilized for the process. The process plans are saved in the system database. A search heuristic is developed for selecting the process plan. If the process plan does not exist in the database then a new process plan is generated by selecting machines on the basis of the features, dimensions, and tolerance required for the finished product. The process plan information is used to obtain the time required for manufacturing and also for cost estimation.

2.8 Cost and Time Analysis

Each process plan is linked to a cost analysis chart which gives the costs incurred in the manufacturing process. If a new process plan is generated then a detailed analysis for cost estimation is carried out. The time required for manufacturing a part can be classified into productive and nonproductive time [2, 4]. The productive time consists of roughing and finishing operations. Roughing operations are carried out at maximum power and the time taken is given by: tmp = (60rv psW)/(raW b) = ((60rv ps)/(ra)) W(1-b) (2.9)

where rv proportion of initial volume of workpiece to be removed by machining, ps specific cutting energy, density of work material, W initial weight of work material in pounds, and a, b constants for machine tool type. The time taken for the facing operation at maximum power condition (tmpf) is given by tmpf = 60 Vm ps /Pm (2.10)

where Vm volume of material removed and Pm maximum machining power. The time taken for the finishing operation depends on the surface generation rate Rsg for the material. It is the product of speed and feed and is the rate at which machined surface is generated. Rsg for alloy steel work piece and high speed steel tool is found to be equal to 23.7 in.2/min [3]. The finished machining time is given by tmc = (60Am)/Rsg where Am surface area to be generated. (2.11)

26

A.K. Kamrani, A. Vijayan

The tool replacement time is accounted for by modifying the roughing and finishing time equations as t = tmp [1 + (n/(1n)) (tmc/tmp)(1/n)] mp when tmc/tmp < 1, and t = ((60Am)/Rsg) (1/1n)) mc (2.13) (2.12)

where n is the Taylor tool life index and depends on the material of the tool. Value of n is taken as 0.123 for high speed steel tool [4]. The nonproductive time is calculated as tnp = (tsa + nt tsb)/Bs = tln + notpt (2.14)

where tsa basic setup time, tsb setup time per tool, Bs batch size, tln loading and unloading time, tpt tool positioning time, and no number of operations. The loading and unloading time depends on the weight of the workpiece and is given by tln = c d W (2.15)

where c and d are constants and have values 38 and 1.1, respectively, for a lathe [8]. The total rate for the machine and the operator is given by Rt = (km e Wf)/(2 ny ns) = ko Ro (2.16)

where Ro direct labor rate, km factor for machine overhead, ko factor for operator overhead, e, f constants for a given tool, ny amortization period for the cost incurred, and ns number of shifts. The total cost of manufacturing can then be calculated by multiplying the cost rate with the total time required for the process. The manufacturing cost Cman is given by Cman = Ro (tmp + tmc + tnp) when tmc/tmp 1, and Cman = Ro (tmp + tmc + tnp when tmc/tmp < 1. Bearings are normally procured from a vendor. If the bearing required can be found in the bearing database then its corresponding cost is retrieved and used in calculating the total cost. If the bearing cost is not available in the database, then the relative cost of the bearing is determined using polynomials [17]: (2.17)

2 Template-Based Integrated Design: A Case Study

27

P3(X) = a3X3 + a2X2 + a1X1 +a0

(2.19)

where a3, a2, a1, a0, coefficients, X standardized dimension, and P3(X) thirddegree standardized cost function. The coefficients a3, a2, a1, and a0 are obtained from the following set of equations. The total cost for the design including the bearing procurement cost and the manufacturing cost is reported for all the machines that can be utilized to manufacture the shaft. This estimate helps the designer to select the minimum cost alternative along with the machine information on which it can be processed.

2.9

System Components

The structure of the system is divided into four major components: design calculations, bearing selection, design template retrieval, and machine selection with process cost estimation. These components are illustrated in Fig. 2.3. In this system, the product specifications are the input for carrying out design calculations. The external forces acting on the bearings and the shaft extension is calculated. The design calculations also retrieve the electrical design data for rotor packet from the database. A search heuristic is implemented for selecting a minimum cost bearing which can withstand the external forces acting on it. The features required and dimensions of the bearings, rotor packet, and shaft extension are used to retrieve the CAD templates. The template information is used for selecting the machine and carry out a process cost analysis. The output from each component is reported back to the user interface. The system is implemented on a personal computer equipped with a Pentium-4 processor and Windows XP operating system. The graphical user interface, search heuristics, and design calculations are developed in Microsoft Visual Basic (.NET platform). The relational database is developed in Microsoft Access 2003. The 3-D models are generated in I-DEAS and imported into Visual Basic which acts as the drawing templates.

28

A.K. Kamrani, A. Vijayan

Product Specifications

User Interface

Relational Manufacturing Database

Design Calculations

Electrical Database

Bearing Selection

Bearing Database

Design Template Retrieval

CAD Templates

Machine Selection

Machining Database

Cost Estimation

Cost Database

Fig. 2.3 System structure

2.10

Design Calculations

The power and speed entered in the user interface is used to search the electrical database to retrieve specifications of the rotor packet. These specifications are used to calculate the radial and axial magnetic forces acting on the shaft [1]. The radial force acting on the shaft is given by: Fmr = (0.1 + 0.005SID) [(0.25RL.SID)/ (5AG)] (2.21)