t e c h n o f o c u s

manufacturing

Rapid product development

survival

K P Karunakaran & Manoj K Tiwari

In todays competitive market, time to market and agility to respond to changing customer needs are critical parameters. They are not only the keys to an organisations success but also its survival. Thus, techniques such as rapid product development (RPD) that enhance a manufacturing organisations capabilities are being increasingly adopted to hone its competitive advantage. Read on

ith the advent of globalisation, a wide variety of highly sophisticated products are available at competitive prices. Due to the ever changing taste of the customers on one hand and the technological obsolescence on the other, product life cycle has been decreasing as illustrated in Figure 1a; this leaves less time for innovations as demonstrated in Figure 1b. In a buyers market today, customers dictate the price, and hence, the only way to earn profits is through cost reduction. Apart from the low prices, the customers have to be kept delighted through better quality, aftersale-service, and other benefits. Every organisation can be visualised to be on a tread mill (Figure 2); to retain their position in the market, they have to keep moving; to expand their share, they will have to run. This constant race applies tremendous pressure to launch new and innovative products. When a python captures an

animal, it tightens it a little every time the animal exhales, leaving less and less room for subsequent breathing, till the prey finally dies of suffocation. Cycles similar to breathing occur in organisations too; the inventory of raw materials, work in progress and finished products are not constant, cash flow fluctuates, and so on. Whenever each of these reaches a low level, the competitors lose no time in taking away the market share, just as the python does. Therefore, organisations can never remain complacent with their current performance. The present scenario is compounded by social commitments/ liabilities. Table 1 provides the comparison of this present industrial scene with the past. The survival, leave alone the success, of Indian industries against multinational corporations in such an industrial scenario depends on their ability to launch high quality, technologically advanced products in the global market with dramatically short lead times, and at prices that customers are able to pay. Being the best alone is not enough today; one also has to be the first to respond. To be the first is in fact more important than to be the best. In the buyers market, ability to compress time to market and agility to respond to changing customer needs, decide the survival of manufacturing organisations. Rapid product

development (RPD) enhances these capabilities of organisations, and hence their competitive advantage. RPD is the synergic integration of the various time compression technologies, which may be virtual or physical in nature. Time compression can be achieved by doing the various design, testing, and manufacturing activities in the virtual world as far as possible. When this no longer becomes possible, one can shift to the physical world where the aim is to do them rapidly. In order to achieve speed, a marginal sacrifice in performance is acceptable with the clear understanding of the implications of such a sacrifice. For instance, a mould that has less life is acceptable as long as it produces the components with the required quality during its known life. This is the philosophy behind RPD. The following are the three pillars of RPD (Figure 3): Virtual prototyping and testing technologies Physical prototyping and rapid manufacturing technologies Synergic integration technologies

Virtual prototyping & testing technologies

When an activity in the product development cycle is carried out in a virtual

Figure 1: Reducing product life leading to shorter development time

Figure 1a: Reducing product life

Figure 1b: Shorter time available for development

76

MODERN MACHINE TOOLS o c t o b e r - n o v e m b e r 2 0 0 5

t e c h n o f o c u s

Figure 2: Threat from competition

are for virtual testing or CAE tools. Geometric modelling: Geometric modelling (GM) helps in the fast creation and modification of virtual models of the products. Wire-frame, surface, and solid models are the three types of geometric models of increasing complexity, and are also rich in information. Solid models, which are an unambiguous and informatically complete representation of the products, are the most popular form of geometric models today. Solid models can be created elegantly and easily even by less skilled persons since 3D is the most natural way of designing. Constructive solid geometry (CSG) and feature-based modelling (FBM) are the two user-friendly solid modelling schemes, the latter being more popular. Boundary representation (B-Rep) is a system-friendly representation of the exact type and several inexact representations such as voxel, octree, etc, which are used in certain niche applications such as volumetric NC simulation and medical modelling. As one requires both userfriendliness and system-friendliness in modelling, any modeller will have one for each with a unidirectional converter to convert the interactions of the user into a system-friendly representation. AutoCAD, SolidEdge, SolidWorks, are some of the popular modelling packages and Unigraphics, Pro/Engineer, and CATIA are CAD/CAM/CAE packages as they have seamlessly integrated modules for doing all sorts of virtual prototyping & testing functions. Reverse engineering : Reverse engineering (RE) is used to create geometric models of physical objects. This is an important step in design iteration, especially of form design. RE also comes handy when it is required to reproduce existing objects, especially old or military equipment where the supplier no longer exists or is unwilling to help. RE in Present Open economy Global competition Buyers market Profit = Price - Cost Customers delight Shorter product life Stringent laws for liability and social commitments Competition

conjunction with RP&T can revolutionise the way in which prosthesis is practiced, although it is yet to be exploited in our country. Using these tools, it is possible to construct the broken bones and prepare the implant (Figure 4). Virtual reality & augmented reality : Virtual reality (VR) is a more sophisticated real time multi-modal simulation of existing or imaginary environments. It involves modelling of an environment and presenting it to the user and permitting him to interact with it intuitively & effortlessly such that he feels as if he is in the real environment. In other words, he feels immersed. To achieve the feeling of immersion or inclusion in the environment, a VR system will have the following five characteristics: Real time response Photo-realistic stereo rendering Intuitive, spontaneous and effortless interaction Special input and output devices Multi-modal interaction VR helps in better visualisation as well as evaluation of the design. Cave is a VR equipment that has stereo-displays on all sides of a cubical as shown in Figure 5a. Figure 5b shows two engineers studying the complicated assembly procedure of a car door lock in the virtual 3D environment using a cave. When one is able to mix real objects with the virtual ones, it is known as augmented reality (AR). In VR the objects exist only in the virtual world, whereas they exist both in virtual and physical form in AR. These are useful for assembly simulation, ergonomic analysis, simulation of surgical procedures, etc. AR has been used successfully in carrying out minute laproscopic operations (Figure 6a), and Figure 6b illustrates AR for the surgery of brain tumor. Finite element & boundary element analysis : Finite element analysis (FEA) and boundary element analysis (BEA) are tools for virtual testing. These can Past Protected economy Local competition Sellers market Price = Cost + Profit (often as a % of cost) Conformance to design Longer product life Less liabilities and social commitments Complacence

or soft environment, the primary benefits are savings in time and cost. There are also a few other important benefits as follows: Minimum risk to human life: VR-based flight simulators help the pilot prepare himself well before the fighter plane is ready for flying Upward or downward scalability of time in virtual world: The pace with which an activity takes place can be increased or decreased as required if it happens in a virtual world. It is possible to simulate creep of components in a short time to assess their fatigue life, which otherwise would take several hours when tested physically. On the other hand, one can analyse the trajectory of a missile leisurely at a chosen pace, which otherwise may get over in a few minutes Less fatigue to the personnel: Monotonous steps in various activities can be automated. For instance, drawings with various types of views and dimensioning can be automatically generated from the solid models with a few commands The technologies for virtual prototyping are known as computer-aided design (CAD) tools and the same for virtual testing are known as computer-aided engineering (CAE) tools. Geometric modelling, reverse engineering, and virtual & augmented reality are used for virtual prototyping or CAD tools, and tools such as finite element & boundary element analyses, motion analysis, tolerance analysis and synthesis, design for X, etc,

Table 1: Comparison of industrial scene - present and past

MODERN MACHINE TOOLS o c t o b e r - n o v e m b e r 2 0 0 5

77

t e c h n o f o c u s

Figure 3: Three pillars of rapid product development

be extended to various other activities such as disassembly, repair, recycling, mass customisation, etc. DFX, essentially is a set of CAE techniques that help in evaluating the design at a very early stage. DFX and concurrent engineering (CE) together help in early detection of design flaws. The later an error is detected, the more it costs to correct. Based on the result of these evaluations, design modifications are made.

Physical prototyping and rapid manufacturing technologies

be used to assess the stress levels of components under various types of static and dynamic loads. These loads may be purely mechanical or a combination of fluid pressure and thermal loads. These are also helpful in analysing fluid and thermal flow problems. Although FEA and BEA were originally developed for stress analysis, these can be applied to all situations in which the parameters exist in the form of a field or continuum. Problems in fluid flow, heat flow, vibrations, stress distribution, electric potential, magnetic potential, etc, can be studied using FEA and BEA. Both are divide and conquer strategies in which the object is divided into easy-to-analyse shapes, and the results of the analyses of all these constituent elemental shapes are combined to predict the performance of the whole object. Although BEA was developed earlier, its complex integral mathematics was responsible for its trailing behind FEA in popularity. Today, several commercial packages are available for FEA. With the advent of symbolic algebra, BEA is becoming popular but only in certain niche areas. ANSYS and MSC-NASTRAN are some of the popular FEA packages. Motion analysis : Several software tools are available that can conduct the motion analysis of assemblies to find out their working volume, accelerations at various positions, etc. Interference or collision detection is also one of the most important uses of motion analysis. Design for X : Design for X (DFX) is an important CAE tool. Design for manufacture & assembly (DFM&A) came first. Subsequently, these concepts could

In order to compress product development cycle time, as mentioned earlier, activities should be carried out in the virtual world as far as possible and when this is no longer possible, they should be done as rapidly as possible in the physical world. It may be noted that, in order to achieve the speed, following compromises on material, quality, and cost have been found acceptable: Material: Most RP machines make prototypes out of soft non-metallic materials, that too of proprietary nature. These prototypes are not isotropic. In spite of these limitations, these prototypes help in rapid assessment of form, fit and functional performances of the design, to varying extents. They are also useful as masters in indirect rapid tooling. Quality: The accuracy and surface finish of RP parts are nowhere near those of the machined parts. However, such prototypes enable several design iterations within the given time. Cost: The cost of a RP part may be much higher than the one made by

a pattern maker or by conventional or NC machining. However, RP can make it faster because the entire process right from design to physical prototype is totally automated. Therefore, although it may appear sub-optimal locally, money spent in RP parts will lead to net profit where timeto-market matters. The following are the technologies for physical prototyping and rapid manufacturing: Computer numerical control Robotics Computer-aided process planning Rapid prototyping & tooling These are essentially computer-aided manufacture (CAM) technologies. Computer numerical control and robotics: Computer numerical control (CNC) machines and robots are automatic machines working on similar principles of drives and controls. They primarily differ in their kinematic configurations and hence differ in terms of accuracy, rigidity and dexterity. In fact, a CNC machine can be loosely called as a Cartesian robot. The other subtle differences between CNC machines & robots are the extent of sensory perception and the method of programming. CNC machines and robots are known as machines of flexible automation type since change over from one product to the other is achieved simply by changing the control programmes. CNC machines are used primarily for component manufacturing whereas robots are mostly used for material handling, assembly, arc welding, deburring, spray painting etc.

Figure 4: Model of a skull used for operation planning & visualisation at an hospital in Mumbai in August 2004

Figure 4a: CAD model developed from CT scan

Figure 4b: Physical model built using FDM1650RP machine

78

MODERN MACHINE TOOLS o c t o b e r - n o v e m b e r 2 0 0 5

t e c h n o f o c u s

Figure 5: Applications of virtual reality

Figure 5a: A VR cave

Figure 5b: Study of the assembly of a car door lock in a cave

While the West placed strong hopes of robots as humanoids, it was the Japanese who treated them as industrial machines and exploited them first. A comparison between CNC machines and robots is given in Table 2. Computer-assisted process planning: Process Planning (PP) converts the design information available in the form of geometry, tolerances, and material specifications into manufacturing information in the form of motions on machine slides and other process parameters. Therefore, this is considered as a bridge between design and manufacture. When use of computers is employed for process planning, it is known as computer-assisted process planning (CAPP). CAPP is considered as a link between CAD and CAM. In a conventional manufacturing environment, CAPP will generate a process plan that indicates the sequence of operations to realise a component, sub-assembly, or assembly along with the details of the process, tooling requirements, set up and unit times, etc. When the acronym CAPP is expanded as computer-assisted part programming, the outcome is not a conventional process plan, but the NC programmes that carry out the operations on a CNC machine. There are primarily two types of CAPP systems: Variant process planning Generative process planning In variant process planning, a process plan existing in the company for a similar component is retrieved and suitably

modified so that the new component can be made using it. This is a semiautomatic method, which relies on group technology (GT) to classify and retrieve similar components. Generative process planning is fully automatic. It interacts with the various databases of machine tools, cutting tools, production tools, etc, and knowledge bases consisting of heuristics & thumb rules extracted from expert process planners, and generates an appropriate process plan. The preferred geometric input to automatic process planning is feature-based. If only a conventional B-Rep model is available as input, then a pre-processing step called feature extraction is used. Unfortunately, only limited success has been achieved in generative process planning that too for simple geometries such as axi-symmetric and prismatic features. Rapid prototyping & tooling: The
Figure 6: Applications of augmented reality

benefits of both CNC machines and robots can be fully realised only when their fool-proof tool path programmes for any operation are generated automatically. Due to the limited success in CAPP systems, this is still not possible, and hence, a skilled NC programmer has to generate the control programmes. An alternative to CNC is rapid prototyping & tooling (RP&T) technology wherein complete automation exists from design to manufacture. RP is an automatic process for manufacturing parts, prototypes, tools, and even assemblies directly from their CAD models without the use of any cutters, tools or fixtures, specific to the geometry of the object, and without any human intervention anywhere in the process. As success in CAPP has been elusive for subtractive manufacturing, all RP machines use additive manufacturing. In other words, todays RP machine is a CNC machine with an embedded CAPP system for additive manufacturing. RP is also a divide-and-conquer manufacturing strategy in which a complex 3D manufacturing problem is converted into several 2D manufacturing problems which can then be automated. RP&T can make prototypes only out of proprietary non-metallic materials, and the accuracy & surface finish are not adequate. However, direct as well as indirect methods are emerging to overcome these limitations. Hybrid technologies, which exploit the best features of CNC and RP, are also evolving. With these efforts, objects can be made automatically from their CAD models with the required material and quality. The huge investments on RP machines, as well as the high cost of these prototypes, can be amortised better if the same technology is able to produce tools through direct or indirect routes. It may

Figure 6a: Surgery of breast cancer

Figure 6b: Surgery of a brain tumour

MODERN MACHINE TOOLS o c t o b e r - n o v e m b e r 2 0 0 5

79

t e c h n o f o c u s
Characteristic Applications Robots Primarily for material handling and assembly. Online programming i. Teach pendant method method ii. Walk-through method iii. Tele-operation Offline programming Same for both, ie, (i) language-based; (ii) model-based interactive (iii) CAPP-based automatic methods. method Sensory capability Low High Kinematics Popularly Cartesian Many kinematic arrangements are 3 to 5 axes only possible, normally 6 axes. Accuracy High Low Rigidity High Low
Table 2: Comparison of CNC machines and robots

CNC Machines Primarily conventional and non-traditional machining. Written in G, M code format

Computer-integrated manufacturing

be noted that processes like selective laser sintering (SLS) and 3D printing (3DP) are capable of producing prototypes from a wide variety of materials right from wax or polystyrene to metals. When the same technology is extended for manufacturing parts in small batches, it will be called rapid manufacturing (RM).

Synergic integration technologies

parallel as possible. This is enabled by a technology called concurrent engineering (CE). The reduction in time happens not only by parallel processing, but also by an early detection and correction of error due to a feedback from related groups. Therefore, sometimes, concurrent engineering is also called collaborative engineering. Product data management : Since the customers vary in terms of taste, buying power,etc, the same organisation has to make several models of the product, each with its own variations and options. Furthermore, they have to release new products year after year while they are still committed to support earlier products for a certain period. All these make manufacturing very complex, requiring sophisticated systems called product data management (PDM) systems to keep track of various activities and entities. Metaphase and WindChill are some of the popular PDM packages. PDM, originally developed for managing complex and huge design databases, subsequently evolved into product lifecycle management (PLM) by expanding its scope through the addition of more databases & functions such as integration of dealer network, etc. Roughly speaking, the numeric databases of a company are managed by engineering resource planning (ERP) and its geometric databases are managed by PDM. This distinction is slowly vanishing with the advent of PLM. Interfaces between ERP and PLM exist so that data transfer between them can take place. Very soon, both ERP and PLM will merge into a single, seamlessly integrated system that will manage all kinds of databases of an organisation beyond geographic boundaries.

Computer-integrated manufacturing (CIM) is an embodiment of all the above integration technologies. Another popular term having the similar meaning is intelligent manufacturing system (IMS). There are three prerequisites for CIM/ IMS implementation in an organisation: All activities must happen on computers There must be reliable connectivity among these computers through LAN/ Internet/ intranet There must be well-established communication protocols among the computer systems Furthermore, there should also be a reliable security system to monitor and control the data that may be distributed across the network.

In conclusion

Product development is a network of several activities. The time taken by an activity may be constant or variable. For instance, an automatic operation may have a fairly constant time whereas the cycle time of a manual operation may have wide variations. Some activities can happen only if a few others are complete whereas there could be many sets of activities in the network, which can progress together. Therefore, RPD cannot be fully realised if these factors are not accounted for and managed properly. In other words, we need certain managerial tools that ensure synergy among the various activities so that: Parallelism among activities is exploited Mistakes are detected and corrected early in the process The benefits achieved at individual activities are carried forward without significant attenuation till the end The technologies used for the synergic integration of the activities of product development cycle as well as product life cycle are: Concurrent engineering Product data management Product lifecycle management Computer-integrated manufacturing Concurrent Engineering: An important technique to cut down product development time is to carry out as many activities in

Organisations can no longer remain complacent with their current performance in a buyers market with global competition. It is not enough to be the best any more; you also need to be the first to respond to customer needs. In the buyers market, ability to compress time to market and agility to respond to changing customer needs, decide the survival of manufacturing organisations. And RPD helps doing just that. Thus, adopting technologies for RPD today makes more sense than ever before. So we even find competitors working together to develop technologies of mutual interest, while aggressively protecting their domain expertise through patents and copyrights. F
Dr K P Karunakaran is an Associate Professor in the Mechanical Engineering Department of IIT Bombay. He is an Humboldt Fellow having research interests in RP&T, CNC, automation and computer graphics. E-mail: karuna@iitb.ac.in Manoj K Tiwari has been with National Institute of Foundry and Forge Technology (NIFFT), Ranchi since 1998 as an assistant professor and professor respectively. He is currently associated with the research promotion cell (RPC) of NIFFT and has contributed more than 100 articles in various leading international journals of

80

MODERN MACHINE TOOLS o c t o b e r - n o v e m b e r 2 0 0 5