AP05_3.vp 1 Introduction Thermal desorption is a highly-specialised technique for the capture, concentration and analysis of trace-level volatile organic compounds (VOCs) in real world samples. A typical application is the measurement of airborne VOC pollutants in an industrial setting, such as a chemical processing plant [1]. The thermal desorption technique, in conjunction with gas chromatography and mass spectrometry, allows minute quan- tities of these substances to be detected and measured [2]. However, the demand from customers for greater sensitivity over a wider range of applications is increasing. The recent se- curity concerns over chemical and biological weapons within the UK and USA provides a vivid example of the need for an accurate and sensitive technique to monitor air quality. A collaborative research programme was establishe in this area to identify optimal product design and development tech- niques for the novel thermal desorption process. The thermal desorption process combines analytical chemistry, materials science, transient heat transfer, mechanical drive systems and electronic control. Thermal desorption products are tradi- tionally labbased capital equipment, and the disparate areas of expertise have meant that product development has tended to be treated as somewhat of a ‘black art’. The devel- opment of these products has been driven by the application of analytical science, whereby thermal desorption is now a proven, reliable concept, but it has been reliant on ‘tradi- tional’ design and manufacturing techniques. New market opportunities, such as portable instruments and embedded on-line products, have highlighted the need to investigate a wider range of product design and development techniques in order to progress the technology. This paper reports the application of new design and development tools within a small company specialising in thermal desorption equipment. The context of this study is important because new product development activities within small and medium-sized enterprises provide higher financial returns than practically any other type of similar investment [3]. SMEs are often in a prime position to identify innovative new products as a consequence of their close working rela- tionships with customers and suppliers; however, the majority of the product development literature focuses on design tools and activities within large well-established companies. The literature on design and development within SMEs is more limited in scope [4]. This is surprising given the fact that SMEs play a key role in most European economies. It has been estimated, for example, that 95% of the three mil- lion businesses in the UK employ fewer than 20 people [5]. Thus their performance as product developers is a matter of no small concern. SMEs typically operate in a resource con- strained environment, and there can be a tendency for small companies to conduct product development in an ad hoc manner [6]. In order for SMEs to maintain their competi- tive advantage in an increasingly harsh international market, they need to adopt best practice design and development techniques. Advanced design tools, such as computer-aided design (CAD) software, represent an important route for improving competitiveness through enhanced quality and productivity. Integrated CAD-based systems have made a major impact across a range of industries. CAD speeds up design and engi- neering activities by rapidly capturing design intent and reducing the errors between development stages [7]. A num- ber of researchers have highlighted the benefits of integrating CAD systems within the overall product development process [8,9]; however it has also been noted that CAD systems can inhibit innovation if implemented inappropriately and used ineffectively [10]. Therefore the methods by which design and development techniques are implemented are an impor- tant consideration. This paper will assess the impact of a CADbased structured design process, which has been applied to the novel design challenges of thermal desorption prod- ucts. The specific requirements of thermal desorption are presented through a product design specification, and the core component of the thermal desorption instruments is highlighted for re-design and prototype development. The resulting design improvements are assessed through an ex- perimental test programme. The overall impact of the design © Czech Technical University Publishing House http://ctn.cvut.cz/ap/ 37 Czech Technical University in Prague Acta Polytechnica Vol. 45 No. 3/2005 The Design and Development of Enhanced Thermal Desorption Products R. Humble, H. Millward, A. Mumby, R. Price This research study is based on a knowledge-transfer collaboration between The National Centre for Product Design and Development Research (PDR) and Markes International Ltd. The aim of the two-year collaboration has been to implement design tools and techniques for the development of enhanced thermal desorption products. Thermal desorption is a highly-specialised technique for the analysis of trace-level volatile organic compounds. This technique allows minute quantities of these compounds to be measured; however, there is an increasing demand from customers for greater sensitivity over a wider range of applications, which means new design methodologies need to be evaluated. The thermal desorption process combines a number of disparate chemical, thermal and mechanical disciplines, and the major design constraints arise from the need to cycle the sample through extremes in temperature. Following the implementation of a comprehensive product design specification, detailed design solutions have been developed using the latest 3D CAD techniques. The impact of the advanced design techniques is assessed in terms of improved product performance and reduced development times, and the wider implications of new product development within small companies are highlighted. Keywords: thermal desorption, product design, SME. and development techniques is reported in terms of deve- lopment times and project costs, and the wider implications for new product development within small companies are discussed. 2 Case-study methodology Markes International Ltd was established in 1997 and now operates with a staff of approximately 20. They specialise in the design and development of thermal desorption capital equipment, and all the key development activities are con- ducted in-house. This encompasses early concept design, experimental testing and final assembly. The company has successfully developed a range of state-of-the-art laboratory- -based products, and their core products are based on the ‘Unity’ and ‘Ultra’ thermal desorption instruments, an exam- ple of which is shown in Fig. 1. A typical selling price for this type of unit is approximately 20k. The company’s client base includes the process industry, key regulatory agencies and the service laboratory sector, with applications ranging from envi- ronmental health to materials testing. The company has established an international reputation as a leading supplier of quality thermal desorption equipment, accounting for a 12 % market share across the world and 60 % of the market within the UK. As a result of emerging markets (e.g. building materi- als emissions and contamination in food and beverages), Markes International has identified new products that will drive significant business growth. In order to bring these new products to market, Markes International formed a partner- ship with PDR, and this identified the need to introduce a 3D CAD-based design capability to add value through enhanced product performance in combination with a more effective development process. A collaborative KTP (Knowledge Transfer Partnership) programme was established in March 2003, and this paper reports the preliminary results during the first 12 months of the programme. PDR have employed the KTP model as an effective mech- anism for partnership and collaboration with a wide range of SMEs, predominantly in Wales. The KTP strategy has been in operation for over 20 years, and is a government-backed knowledge transfer scheme. The aim of the scheme is to strengthen the competitiveness and wealth creation of the UK by stimulating innovation in industry through structured col- laborations with universities and research organisations. KTP is run for the government by Technology Transfer and Inno- vation Limited (tti). A typical KTP programme is a two-year partnership between one company, one university and tti. Each individual KTP programme is designed to address the key elements central to the successful development of the spe- cific company. The two-year project provides employment for a wellqualified graduate KTP Associate for the duration of the programme. It should be noted that KTP was formerly known as TCS (Teaching Company Scheme) prior to 2003. All the PDR-based KTP programmes are, or have been, fo- cused on product design, and the numbers reflect the UK trend in that the majority have been based in SMEs. PDR have successfully completed 12 KTP programmes since 1995, seven of which have been with small companies. A typical PDR-based KTP programme implements a new design capa- bility within a ‘traditional’ manufacturing company. In line with other researchers [11], PDR have found that the KTP model is an ideal vehicle through which to analyse the de- sign-to-manufacturing interface and the associated elements that impact upon new product development within SMEs. The well-defined management and structure of the KTP process promote a detailed analysis of the company from the university partner’s perspective. A close working relationship is developed with the company in the early stages during the drafting of the KTP grant proposal. This is written by the university in collaboration with the company, describing the company and the financial benefits, and quantifying the aims and objectives of the programme. A key feature of the pro- posal is a detailed 104-week Gantt chart, which defines a programme of work to address the strategic needs of the com- pany. Following the approval process, a grant is awarded and PDR employs an Associate to work full time at the company for two years to meet the project objectives. The Associate is assigned at least one PDR-based supervisor and at least one supervisor from the company. The regular contact with the company fosters a level of trust and co-operation that gen- erates an indepth understanding of the subtle issues and problems inherent in any small company. KTP programmes are characterised by a commitment to disciplined effective project management through manda- tory monthly and quarterly meetings. The monthly meetings between the supervisors and the Associate not only focus on the technical issues within the programme, but also address training and personal development requirements. The quar- terly meetings are designated Local Management Committee (LMC) meetings, and are underpinned by support from a tti consultant. LMCs act as the programme’s steering group to ensure that the longer-term objectives for the company and Associate are met. The documentation (technical reports, 38 © Czech Technical University Publishing House http://ctn.cvut.cz/ap/ Acta Polytechnica Vol. 45 No. 3/2005 Czech Technical University in Prague Fig. 1: Example of a ‘Unity’ thermal desorption laboratory-based instrument executive summaries and presentation material) arising from the structured KTP meetings, in parallel with the weekly informal contact with the company personnel, results in a comprehensive portfolio of case study material, and this pro- vides the foundation to this research study. 3 Product design specification The generic sampling and analysis process for the lab- -based thermal desorption instruments can be summarised as follows: a) The VOCs are ‘captured’ in sample tubes that contain an absorbent matrix. The sample tubes are located at key points in the area to be analysed (two sample tubes are shown in Fig. 1). b) The sample tubes (typical volume 100–200 ml) are sealed and transported to the lab-based equipment. c) The sample tubes are fed into the thermal desorption equipment and an inert gas stream (usually Helium) ex- tracts the VOCs from the sample tubes. d) The gas stream refocuses the VOCs into the cold trap module, which is electronically cooled and contains an ab- sorbent with a strong affinity for the VOC. e) The cold trap is then rapidly heated, which desorbs the VOCs and concentrates the sample into a smaller volume (typically 100–200 �l). f) The concentrated sample is then introduced into a gas chromatograph or mass spectrometer for analysis. It can been seen that the cold trap module acts as the con- centration engine at the heart of all the thermal desorption instruments. The cold trap technology has been developed for the Unity range, and this provides the basis for further design improvements that can be incorporated into replace- ment products and new products for emerging markets. The first stage of the re-design process was establishing a product design specification (PDS). The PDS is a central ele- ment within the various design documentation, and needs to be reviewed and updated in response to other design output (e.g. test reports and risk analysis). The implementation and maintenance of an appropriate PDS for assemblies and criti- cal components is the first step toward a systematic design process, although this is not always acknowledged by small companies [12]. The cold trap module PDS can be used as an exemplar, and key sections are highlighted in Table 1 in order to define the main design challenges. © Czech Technical University Publishing House http://ctn.cvut.cz/ap/ 39 Czech Technical University in Prague Acta Polytechnica Vol. 45 No. 3/2005 Performance • Constant cold trap cooling to give a minimum temperature of –15 °C. • Intermittent ‘firing’ to heat the trap and achieve desorption at a maximum temperature of 300 °C. • Materials compatible with temperature extremes and absorbent matrices. • Electrical supply to power both the ‘cooling’ and ‘heating’ elements. • Product will run for ten years, and service stock will be required for a further five years. • Product will be usable for up to 24 hours per day, operating a seven-day week. • Product must avoid ice formation. Interfaces • Upstream connection and input from sample tube. • Downstream connection to Heated Valve (constant operating temperature of 200 °C), and output to analyser. • The critical connection between the Cold Trap Module and the Heated Valve must avoid condensa- tion, and there should be no interconnecting tubing. • The cold trap module will need to operate as a key sub-assembly within a range of new products. Environment • Upstream ambient conditions assumed to be 30 °C. • Relative humidity in the range 5 %–95 % (non-condensing). • Module must withstand transit-vibrations and shock loading up to an acceleration of 10×9.8 ms�2. • Noise levels must adhere to laboratory regulations. Quality • No failure should cause any hazard for any operator. • The functional failure rate should be greater than 1 in 50 – based on primary testing. • Product must comply with all relevant EC legislation and BS regulations. Table 1: Key sections of the PDS for the cold trap module 4 Advanced design techniques The selection and implementation of a 3D CAD system was a key aspect of the first year of the KTP collaboration. Prior to this partnership, the company were reliant on a basic 2D CAD package that was used predominantly for final man- ufacturing drawings. With the focus on new product chal- lenges and inherently complex assemblies, there was a clear need to make the transition to a solids- and surface-model- ling package. Early 3D systems required costly UNIX-based hardware, but with the advent of cost-effective, mid-range packages, full 3D CAD functionality has become a viable option for SMEs. During the early stages of the KTP programme, the com- pany’s current and future design requirements were reviewed and documented. A number of mid-range CAD pack- ages were short-listed and tested against the company’s requirements. The system selected was the AutoDesk Inven- tor CAD package. Once the first phase of training had been completed, the 3D CAD software was employed on all new projects, and the old 2D system was kept running in the back- ground for comparison. The preliminary indications are that 3D solid modelling is a more efficient and productive ap- proach to the design of thermal desorption products. The main benefits of the new 3D CAD system can be summarised as follows: � The ability to clearly communicate the design intent of new product concept is a central feature within 3D CAD sys- tems. Different drawing views can be created instantly from a solid model, and this removes the ambiguity associated with 2D drawings. This allows non-technical personnel to provide meaningful feedback much earlier in the design cycle. Furthermore, at the end of the development process the 3D CAD images significantly enhance brochures and scientific presentations. � Thermal desorption products are essentially large, com- plex assemblies comprising hundreds of mechanical parts and precise interfaces; the 3D environment efficient- ly manages these assemblies. Furthermore, the 3D CAD can assess fit and tolerance problems early in the design process. � Design lead-times have reduced by approximately 15 % be- cause it is much simpler to make quick design changes, and error-checking time has reduced. This, in turn, has re- duced the number of engineering change notes. Changes to a 3D model automatically propagate through to all relevant drawings, and this is further enhanced through bi-directional associativity and parametric design. The immediate impact of the 3D CAD has been to speed up the design and development cycle; however the new de- sign software provides the foundation for strategic changes in the way thermal desorption products are designed and de- veloped. The 3D design data can integrate with downstream systems to facilitate analysis and verification, and expand manufacturing options. Analysis techniques, notably finite element analysis, have been applied to a wide range of indus- tries. The main design challenges within thermal desorption products are driven by the extremes in temperature, therefore thermal analysis software to model the transient heat transfer between critical components would link well with the 3D solid models. This form of thermal analysis should provide a route for evaluating and optimising cold trap performance prior to full experiments and verification testing within the labora- tory. It has been highlighted that low-volume manufacturing techniques are needed for the specialised thermal desorption market, therefore zero-tooling rapid manufacture directly from the 3D CAD data may be commercially viable – particu- larly for bespoke products. The design team have already started to consider stereolithography parts direct from CAD, and this could be extended to sintered metal parts for func- tional components. 5 Prototype development The development process established a new design con- cept for the cold trap module, based on the criteria set out in the PDS. Taking into consideration factors such as materi- als selection and design for manufacture, a detailed design evolved, and the CAD model is shown in Fig. 2. The key temperature-dependent functionality is driven by the absorp- tion/cooling and the desorption/heating components. The absorbent matrices are contained within the central quartz cold trap tube, and this is surrounded by a ceramic cylindri- cal channel. Continuous cooling is provided by two Peltier coolers located beneath this cylindrical trap. The trap box is positioned on top of a heat sink that helps to regulate heat conduction away from the trap channel. The desorption/heat- ing stage is driven by a metallic-strip heater located on the inner circumference of the ceramic channel. Both the Peltier coolers and the trap heater are powered through the electrical supply. The operation and performance of the new cold trap module can be characterised by a temperature traverse 40 © Czech Technical University Publishing House http://ctn.cvut.cz/ap/ Acta Polytechnica Vol. 45 No. 3/2005 Czech Technical University in Prague Materials & Manufacture • Materials and manufacturing techniques must conform to in-house assembly and approved sub-con- tracted manufacturing processes. • Products will be built individually to order, with approximately 40 cold trap modules assembled per year. • Manufacturing techniques must be appropriate for low-volume, batch fabrication. • The competitive driver is the delivery of required performance at a reduced cost. Maintenance • The cold trap module must be maintenance free. • Product must be designed to minimise the servicing requirements. • Absorbent traps must be changeable without the risk of breaking the traps. through the quartz trap during the key cooling and heating phases. This experimental testing is part of the rigorous in-house design verification process, and the data was gener- ated by positioning a thermocouple at known axial positions within the quartz trap. Fig. 3 shows the performance of the default condition, namely continuous cooling, whereby the Peltier coolers are always on and air recirculates through the trap box. It should be noted that the upstream junction of the ceramic trap (x � 10 mm) is at ambient temperature, and the downstream section connects directly to the heated valve assembly (x � 120 mm), which is maintained at a constant temperature of 200 °C. The critical absorbent matrices are lo- cated from x � 40 mm to x � 80 mm, and it can be seen that most of this section is maintained at sub-zero temperatures, close to the specified value of minus 15 °C. The design chal- lenge for this phase of thermal desorption is to generate a distinct step change in temperature at the upstream and downstream interfaces, and thereby create abrupt thermal junctions. The gradual ‘S curve’ temperature profile evident in Fig. 3 is not ideal, and derives from the thermal communi- cation between adjacent components. During the desorption phase, the trap heater is ‘fired’ for approximately five minutes, and the temperature profile for this operation is shown in Fig. 4. The objective for this phase is to maintain the core temperature above 200 °C from the matrices (x � 40mm to x � 80mm) through to the heated valve (x � 120mm). It can been seen that the peak temperature of 300 °C occurs within the quartz trap (x � 70mm), but the tem- perature drops just below 200 °C in the zone ahead of the heated valve. The reason for this is that the close proximity of the heated valve to the cold trap module creates conflicting design constraints. The step change in temperature indicated for the absorption phase acts against the need for uniform temperature during the desorption phase. The Peltier coolers remain on when the trap is ‘fired’, and this generates a cold pocket between the trap and the heated valve. However, the Peltier coolers must remain on in order to produce the abrupt temperature change when the absorption/cooling phase re- -commences. One option would be to increase the peak temperature of the trap heaters, but this is an inefficient use of power and may increase the price of the overall unit. These preliminary results provide a sound basis for further develop- ment. It is anticipated that innovative solutions will lie in the design of the cold trap module in terms of refined materials properties and configuring precise thermal communication between the key interfaces. 6 Conclusions & future work The implementation of structured design documentation and advanced design tools has been shown to add significant benefits to the design and development of a specific thermal desorption product. The first phase of a two-year knowl- edge-transfer collaboration has defined the design challenges through a detailed PDS, and has developed functional proto- types by employing a 3D CAD system. The preliminary results indicate that design-cycle lead times have reduced as a conse- quence of the new design tools, and product performance is in line with the requirements of the specification, which provides the basis for further design improvements. These findings support the view that, if properly implemented, design tools and techniques can have a significant operational impact within SMEs. © Czech Technical University Publishing House http://ctn.cvut.cz/ap/ 41 Czech Technical University in Prague Acta Polytechnica Vol. 45 No. 3/2005 Fig. 2: CAD model of the cold trap module Fig. 3: Absorption/cooling temperature profile within the cold trap module Future work in this area will expand on the research ques- tions highlighted in this study, and evaluate techniques for optimising the thermal desorption products. The thermal characteristics and constraints within the cold trap module have now been defined; therefore, the next phase of research will employ specific analysis tools to refine the thermal de- sorption process. For example, the 3D CAD data could link directly to thermal analysis software in order to model the transient heat transfer at precise locations within the cold trap module. In order to verify any theoretical design results, a thermal imaging system could also be employed to support the experimental test programme. CAD-based design opti- misation should provide a mechanism for reducing project costs, as well as lead times, and thereby deliver tangible commercial benefits directly linked to the new product devel- opment process. References [1] Woolfenden, E., Broadway, G.: “An Overview of Sam- pling Strategies for Organic Pollutants in Ambient Air.” Liquid and Gas Chromatography International, Vol. 5 (1986), p. 166–171. [2] Zha, Q., Oppenheimer, J., Weisel, C.: “The Automated Analysis of Ambient Air Samples by a Thermal Desorp- tion-GC/MS System.” Proceedings of the 38th ASMS Conference on Mass Spectrometry and Allied Topics, Tucson, Arizona, 1990, p. 617–625. [3] Berliner, C., Brimson, J.: Cost Management for Today’s Ad- vanced Manufacturing: The CAM-I Conceptual Design. Boston: Harvard Business School Press, 1988. [4] Brown, R., Lewis, A., Mumby, A.: Enhancing Design Effec- tiveness in Very Small Companies. Cardiff (UK): Design En- gineering Research Centre, 1996. [5] Daly, M., McCann, A.: “How Many Small Firms?” Em- ployment Gazette, Vol. 100 (1990), p. 14–19. [6] Millward, H., Lewis, A.: “Product Development Limita- tions within SMEs: The Drive to Manufacture in the Absence of an Understanding of Design.” Proceeding of the 10th International Product Development Manage- ment Conference, Brussels, Belgium, 2003, p. 713–723. [7] Robertson, D., Allen, T.: “CAD Systems Use and Engi- neering Performance.” IEEE Transactions on Engineering Management, Vol. 40 (1993), p. 274–282. [8] Droge, C., Jayaram, J., Vickery, S.: “The Ability to Mini- mise the Timing of New Product Development and In- troduction.” Journal of Product Innovation Management, Vol. 17 (2000), p. 24–40. [9] Sanchez, A., Perez M.: “Cooperation and the Ability to Minimise the Time and Cost of New Product Develop- ment within the Spanish Automotive Supplier Industry.” Journal of Product Innovation Management, Vol. 20 (2003), p. 57–69. [10] Kessler, E., Chakrabarti, A.: “Speeding up the Pace of New Product Development.” Journal of Product Innova- tion Management, Vol. 16 (1999), p. 231–247. [11] Lipscomb, M., McEwan, A.: The TCS Model: An Effec- tive Method of Technology Transfer at Kingston University. UK: Industry and Higher Education, December, 2001, p. 393-401. [12] Lewis, A., Walters, A.: “Implications of Inadequate Spe- cification Procedures on New Product Development in Small to Medium-Sized Enterprises.” Proceedings of the 9th International Product Development Manage- ment Conference, Sophia Antipolis, France, 2002, p. 541–550. Robert Humble Dr. Huw Millward, Ph.D. e-mail: hmillward-pdr@uwic.ac.uk Alan Mumby The National Centre for Product Design & Development Research (PDR) University of Wales Institute Cardiff (UWIC) Cardiff CF5 2YB Wales, UK Ryan Price Markes International Limited, Llantrisant Business Park Pontyclun CF72 8YW Wales, UK 42 © Czech Technical University Publishing House http://ctn.cvut.cz/ap/ Acta Polytechnica Vol. 45 No. 3/2005 Czech Technical University in Prague Fig. 4: Desorption/heating temperature profile within the cold trap module