Evaluation of Instrument Panel Designs for Cost of Manufacture and Environmental Impact

Dr. Peter Dewhurst
University of Rhode Island
Kingston, RI
02881

Abstract

This paper considers possibilities for redesign of a current model Truck Instrument Panel in order to reduce manufacturing costs and to increase the likelihood of end-of-life material recovery. A new design concept for the Instrument Panel is proposed which would have had the potential to achieve the following benefits:

• a reduction in the number of assembly steps for the main assembly and the principal mechanical subassemblies from 482 to 121.
• a reduction in assembly and manufacturing costs of $35.60 per Instrument Panel.
• a dramatic improvement in ease of disassembly at the end of useful life with the possibility of substantially reducing environmental impact from Instrument Panel manufacture.

The cost savings predicted in this work are based only on comparisons of assembly costs, and the manufacturing costs of the main structural parts of the instrument panel. For this reason the predicted savings are conservative ones, since they ignore the savings in parts cost which will result from the simplification of many of the sub-units and, to a lesser extent from the elimination of numerous spacers and separate fasteners.

INTRODUCTION

This paper is concerned with the design of a current model truck instrument panel (IP). The first stage of the work described here was an evaluation of the IP with the goal of generating design ideas which could improve both ease of initial manufacture and efficiency of end-of-life material recovery. A DFA analysis [1,2] of the IP design, which was the basis of this evaluation, included all of the mechanical assembly work in the IP. Electrical and electromechanical devices such as the instrument cluster, the switches, the cigarette lighter and so on, where treated as parts so that only their installation in the main assembly was considered.

Design ideas, produced during the course of the analysis, have formed the basis for an alternative structural IP design. This will be described briefly below. However, before proceeding with this description it is important to state that the intention here is not to design a new instrument panel. This must be carried out by skilled automotive designers who understand the performance requirements of the next generation of IP’s and who are also aware of the evolving tastes of automotive customers. The intention of the present work is rather to suggest how the current truck IP might have been designed for easier manufacture, assembly, and disassembly, but with the same product specifications. This necessarily means that the design is unlikely to be optimal with regard to any proposed new methods of manufacture. However, if the concepts are considered to be worthwhile then a future design, which evolves around these same concepts will likely be far superior. These statements are particularly true with regard to the proposed welding of two major injection moldings in the proposed design. Since the shape of the current IP was not chosen to facilitate welding, it will be seen that the perimeter of the weld lines is not likely to lend itself to efficient ultrasonic welding. However, the purpose is to show that the concept is flexible enough to allow an alternative structure for the current design.

PROPOSED DESIGN

The proposed design is based around two structural injection moldings. The Front Molding provides the main structure of the IP and is also a retainer for numerous sub-units which fit into and onto the IP. The molding consists of an essentially horizontal rib structure which is intended to provide the required lateral bending stiffness and strength. Cantilever snap elements to retain sub-units into the Front Molding are incorporated only into vertical ribs so that the horizontal rib structure is not weakened. It is anticipated that compliant foam or rubber washers/ gaskets will be assembled with each of the sub-units to be snap fitted into the Front Molding. This will serve two purposes. First, compression of these items during assembly will eliminate any looseness in the snap assemblies. Second, protruding material from the compliant washers/ gaskets will form a seat for the Front Trim Molding to eliminate squeaks between this and the Front Structural Molding. Also, the compliant gasket around the instrument cluster will extend down below the cluster to provide gap closure with the steering column and replace a separate sheet rubber part and five staking operations in the current design.

The Back Molding contains channels, which will form the required air distribution ducts when the molding is ultrasonically welded into the upper and front inner surfaces of the Front Molding. It can be seen that an inner straight channel supplies air to the slots in the top of the Front Molding. The channel, from the rear of the Back Molding, running around the perimeter, supplies air to four vent outlets across the face of the IP. It can also be seen that the welding together of the two moldings will produce a hollow box section across the top of the IP which, with its large second moment of area, will add substantial extra bending strength and stiffness to the IP. Also greatest strength will be provided around the glove compartment opening where otherwise the system would be at its weakest. The assembly of these two moldings in the proposed design would replace an assembly of four moldings (a main retainer molding plus three moldings which are ultrasonically welded to form the current duct system) plus a support structure of two assemblies, which are weldments of steel tubes and steel stampings. These weldments span the width of the instrument panel. The weldments consist of a short structure on the left of the steering column and a long structure on the right of the column which spans the passenger side. Together they consist of two steel tubes, 16 steel stampings, 11 rivets, 4 bolts, 8 screws, and 25 plug welds.

With respect to the main moldings, the rear of the Back Molding contains channel sections into which the wire harness can be push fitted for support. This will replace separate injection moldings on the wire harness itself which clip onto the current IP main molding and onto the steel structure at the rear of the molding. The proposed wire harness support comprises a simple support channel with gaps through which legs of the wire harness pass to make connections at appropriate positions on the rear of the Front Structural Molding.

The mating surfaces between the Back Structural Molding and the Upper HVAC Molding were changed in the proposed design so that the two upper dampers could be assembled vertically onto the corresponding bearing surfaces in the top of the HVAC molding. The upper dampers will then operate in the two passages of the Back Structural Molding. This design change also allows parts consolidation in the design of the HVAC dampers since they no longer need to be assembled through the wall of the HVAC molding.

The remainder of the proposed IP which is to be built onto the main moldings is intended to satisfy two requirements. First, it should be easy to manufacture and assemble and second it should involve the minimum of disassembly for material recovery. The second requirement is ideally satisfied if sub-units are manufactured from the same polymer as the main moldings so that disassembly for recycling is not necessary. With these goals in mind, new design concepts were produced for the cigarette lighter/ power outlet unit, the glove box, the cup holders, and the air flow outlet vents. Because of space limitations only the cigarette lighter/ power outlet unit will be discussed here. A full description of the proposed designs can be found in reference 3. Counting the lighter and power sockets as parts, the assembly involves 20 parts made from steel, rubber, thermoplastics and thermoset resins. After installation into the retainer molding it also requires the assembly of the electrical connectors at the rear of the IP.

The proposed design of the cigarette lighter/power compartment consists of 5 parts instead of 20 in the current design. The Ashtray Cover molding is a simpler molding than the cover in the existing design but it provides three addition functions.

First, molded pivot hooks on the cover engage with open bosses on the inside walls of the Front Molding to provide the hinge action. The Glove Compartment in the current design already uses this pivot design. Second, v-shaped projections in the center of the cover side walls engage with cantilever snaps in the side walls of the Front Structural Molding to provide closure. Third, cantilevered projections from the back edges of the cover molding engage stop projections from the Front Structural Molding side walls to hold the cover in the open position. The cantilevers are intended to cushion the contact with the stops and eliminate the need for the separate rubber stops in the current design.

Rib structures at the back of the Front Molding are to be provided to support electrical connectors from the wire harness in the positions required to make the electrical connections. Wherever possible, the electrical connectors then fulfill the dual function of both connecting and securing. The outer body of the electrical connector is secured to the rib structure, and the lighter socket is then inserted from the front of the IP to engage the electrical connector snap-fit elements. Note that a compliant rubber or foam washer is assembled to the lighter, as discussed earlier, to eliminate looseness in the snap fit assembly. The outer diameter of the washer is large enough to provide a seating surface for the region of the Front Trim molding which will surround the lighter. As discussed earlier, this seating flange will be a part of all compliant washers/ gaskets on the IP sub-units to be snap-fitted into the Front Molding.

DFMA COMPARISON OF ORIGINAL AND PROPOSED DESIGNS

The assembly structure of the current IP design includes the assembly of 28 subassemblies into the main assembly and involves a total of 314 assembled parts. In addition 137 separate assembly operations, such as attaching electrical connectors, applying adhesive, and so on, have to be carried out during the assembly process. This adds up to a total of 482 assembly tasks. The estimated total assembly time for these tasks is 3780 seconds which, for a burdened labor rate of $42/hour, gives an assembly cost of $44 per IP. A breakdown of this assembly cost per IP is shown in Fig. 1. The first bar labeled "necessary" in the figure represents a cost of $3.41 to assembly the items which were considered to be theoretically necessary during the DFA analysis. There are 41 of these items, some of which are sub-units such as the wire harness, the instrument cluster, and so on, which were not disassembled for further analysis. It can be seen that the cost, to assemble these main functional items, is completely outweighed by estimated costs of $11.89 to assemble the separate fasteners, $16.44 to assemble items which were considered to be candidates for elimination, and $11.26 to carry out separate assembly operations such as welding, staking, adhesive applications, and so on. All three of these categories of items and assembly tasks were targets for elimination in the proposals for the new design concepts.

Figure 2 shows a breakdown of the large number of assembly difficulties in the current IP design. The bars in this figure represent the number of occurrences of each type of difficulty in the assembly of one IP. From left to right in Fig. 2 the bars represent: (i) items which are not secured on insertion and need to be held down during assembly of subsequent items (50 occurrences per assembly); (ii) items which are not easy to align for insertion during assembly (148 occurrences per assembly); (iii) items for which the operator has obstructed access or restricted view of the mating location during assembly (95 occurrences per assembly); (iv) items which have resistance to insertion or which tend to hang up and require some manipulation for insertion (97 occurrences per assembly). It should be noted that the direct effect of this large number of assembly difficulties will be a higher average assembly time per operation with a resulting increase in assembly cost. However, the indirect and even more important consequence is likely to be an increase in the number of assembly defects and resulting negative effect on vehicle quality.

The proposed IP redesign could be assembled in 121 assembly steps compared to the 482 assembly steps for the current design. A detailed comparison of the occurrences of assembly steps in the current and the proposed redesign is given in Fig. 3. The number of theoretically necessary assembly operations, given by the first bars of Fig. 3, is the same for both designs. This is to be expected since both designs contain the same required functional items and sub-units. The number of separate fasteners in the redesign has been reduced to 23, compared to 123 in the present design; see the second bars in Fig. 3. Eighteen of the 23 fasteners in the proposed design are spring steel clips assembled to the trim molding for securing the trim to the front of the IP. These are candidates for elimination and could be removed from the design if appropriate snap elements could be molded into the trim molding. However, it is recognized that the trim may have to be removed for service tasks on the IP and that for this reason steel clips may be required. On the subject of service, it may be appropriate at this point to mention that the sub-units and parts which are snap fitted into the front structural molding will not be removable with standard assembly tools. However, is assumed that all of the snaps will conform to a standard which will allow a simple extractor tool to be used for removal of any IP snap-fitted item. The number of other candidate items for elimination from the redesign has been reduced to 39 compared to 173 in the current IP design, as shown by the fourth pair of bars in Fig. 3. Of these 39 items, seven are the compliant washers and gaskets which are used to provide snug and rattle-free snap-fit assembly of items into the front structural molding. The majority of the other candidate items for elimination in the redesign, are the separate molded parts in the four air deflector assemblies. In theory an entire deflector assembly could be a single spherical molded part, and so all the deflector parts except the case are considered to be redundant. Finally, as shown by the fifth pair of bars in Fig. 3, the number of separate assembly operations has been reduced to fifteen compared to the 130 separate operations required for the present IP design. The fifteen separate operations in the proposed design constitute an ultrasonic welding operation for securing the main moldings, a staking operation for the glove compartment, and thirteen operations to route the wire harness and snap fit connectors to the rear of the main moldings.

An assembly bonus occurs when the numbers of items and separate assembly tasks are drastically reduced as described above. This is because it is now possible to allow adequate access for assembly tasks to be carried out without difficulty, and so the assembly time reduces by even more than the proportionate decrease in the number of tasks. The end result of the reduction of both assembly tasks and assembly difficulties is a dramatic decrease in final instrument panel assembly cost from $44.12 per IP to only $8.52 per IP. This represents a potential assembly cost reduction per IP of $35.60.

This potential saving in assembly cost is based on the use of large complex front and back structural injection moldings, and a large trim molding as described above. Using ABS/PC engineering thermoplastic, with 0.14 in. wall thickness for the structural moldings and 0.08 in. wall thickness for the trim molding, would require an estimated combined weight of approximately 17 pounds of the thermoplastic at an estimated cost of $1.25 per pound. These are clearly likely to be more expensive individual manufactured parts than in the present design. For this reason, cost estimates were obtained for the injection moldings to be used in the proposed new IP design, and for the main items which they would replace in the current IP design. For the current IP design these items include the steel tube and sheet metal structural fabrications. The cost estimates were obtained by using the Boothroyd Dewhurst, Inc. DFMA software modules for Injection Molding and for Sheet Metalworking [2]. The part cost comparisons obtained in this way are summarized in Table 1 below.

It can be seen from Table 1 that the four injection moldings in the proposed IP design are estimated to cost almost exactly the same as the parts which they directly replace in the current IP design. The estimates include amortization of the mold costs but do not include finishing costs. For the latter, it is assumed that finishing costs are significant only for the IP molded outer surface, and that these costs will be similar in the two designs.

It should be noted that a full comparison of manufactured item costs would most likely show significant cost savings for the new design. This is because the comparison in Table 1 is only a partial one neglecting both the numerous small items which were simply eliminated from the current IP design, and also the reduction of item costs in the sub-units such as the Glove Compartment, Cigarette Lighter, and so on.

A possible concern with the proposed IP redesign is the investment in tooling required for the large complex molds. Estimates for all of the molds for both the current IP design and for the proposed IP redesign were obtained from the BDI Injection Molding DFMA software [2]. These are presented in Tables 2 and 3 below. It can be seen that the required investment in injection molds is estimated to be approximately $500,000 more for the proposed new design than for the current IP design. This is almost entirely due to the increased cycle time for the main structural moldings compared to the main moldings in the current design. This in turn gives rise to the need for an increased number of molds to satisfy production requirements. However, it should be noted that the cost comparisons given in Table 1 are based on amortization of the total tooling costs given in Table 2 and 3.

Combining the assembly costs with the manufacturing costs given in Table 1, initial estimates for the possible cost savings through implementation of the proposed new IP design are given in Table 4.

Design for the Environment (DFE): Recycling and Material Recovery

The final goal of the IP project was to quantify the difference between the current design and the proposed redesign for their suitability for end-of-life disassembly for material recycling. DFE analysis software developed by BDI [2], in collaboration with TNO in Europe [4], was used to make this comparison. The reader is referred to reference 5 for a description of the MET (Material use, Energy use, and Toxicity effects) environmental scoring procedure. The MET scores have been built into a database of materials and manufacturing processes, using life-cycle evaluation methods which are becoming widely accepted in Europe. MET points are based on a somewhat complex normalization of eight different quantifiable adverse environmental effects. The magnitudes of MET scores for mass produced appliances, with no special environmental problems, range from approximately one MET point for a small appliance such as a home coffee maker to about 25 MET points for a home washing machine.

These scores apply if the appliances are simply discarded when they are replaced by new ones. However, if products such as these are disassembled at the end of their useful life to recover materials for recycling or parts for reuse, then MET points are recovered and the MET score for the product is reduced. The logic of MET points recovery is based on the removal of the need in society for re-producing the same amounts of materials or manufacturing more of the same parts; with a repeat of the estimated negative environmental effects.

The DFE software uses the results of a DFA analysis to establish a disassembly procedure for the product. Responses can then be made concerning the materials and processes used and the intended end-of-life destination of parts. The program can then be used to optimize the disassembly sequence for most rapid profit rates or for quickest MET score recovery rates.

The DFE analysis of the current IP gave an estimated disassembly time of 1560 seconds to reduce the IP to parts and sub-assemblies destined for landfill, and to separate the entire collection of parts with potential for material recycling. However, with a burdened labor rate of only 15 $/ hour, the results indicated that further disassembly beyond 894 seconds would be uneconomic. After this time a maximum profit of $3.74 could be obtained, mainly from recovery of polycarbonate and copper. If complete disassembly took place then the net profit would reduce to approximately $2.00. The highest net profit rate would occur after 894 seconds and would correspond to 15.2 $/ hour. At this point the MET score of -33.7, corresponding to simply discarding the IP, would be reduced through material recycling to -6.7 MET points.

A DFE analysis of the proposed new IP design indicated that it would be dramatically easier to disassemble. The total disassembly time for the proposed design is estimated to be about 260 seconds. However, all of the material with recycle potential can be recovered in 195 seconds, at which point the bulk of the PC/ABS moldings are separated for recycling. In the proposed design, the main structural moldings, the parts of the cup holder assembly, the glove compartment, the ash tray cover, and the air deflector assemblies, are all molded from PC/ ABS blend resins. These can then be removed as a "clump" for recycling without disassembly, and at this time most of the MET score improvement has also been made. With the larger weight of engineering thermoplastic in the design, the potential net profit from material recovery is $6.39, almost twice that for the current design.. This profit assumes a burdened labor for disassembly of $15 per hour as before, and PC/ ABS blend resins are assumed to have a value for recycling of $0.18/ lb which is 15 percent of the current price for large volume purchase.

The profit rate from disassembly of the proposed redesign is approximately eight times higher than for the current design. Moreover, it is now feasible to consider complete disassembly since this would take only a further 50 seconds after separation of the main structural moldings. At this point the MET score would have decreased from -29.7 points if the IP was to be simply discarded, to only -2.4 MET point when the trim molding has been recovered.

It should be noted that in the discussion of environmental impact above, it is assumed that substantial numbers of the IP’s are to be collected for disassembly in large batches at the end of life. If, say, 85% of the approximately 5,000,000 current model truck IP’s were to end up in recycling centers, then we are discussing the possibility of a total environmental impact of -169 million MET points without recycling, being reduced by over 90 percent. And the activity being paid for by a net profit of almost $30,000,000 in recovered copper and PC/ABS.

Clearly the above scenario is unlikely to occur in isolation of other changes in automobile design for ease of recycling. However, if we are to bring about a meaningful change in material recovery, then it is imperative that we look beyond the present crude shredding and sorting methods to a more ordered and efficient system of end-of-life treatment of our products.

Conclusion and Discussion

The current truck instrument panel offers the potential for substantial structural simplification. The results of such simplification would be a significant reduction in total manufactured cost plus a dramatic improvement in the ease with which materials could be recovered from the IP at the end of useful life.

The cost savings predicted in this work are all attributed to the estimated reduction in assembly time. This is because the cost of the main structural items are estimated to be approximately the same for the current IP design and for the proposed redesign. However, this part cost comparison neglects the cost of the numerous small parts in the current design which have been eliminated in the new design. For this reason it is felt that the estimated cost saving of $35.60 per IP is probably a conservative one.

The concepts described in this paper were considered to be novel and initially a parallel concept, involving the use of a structural aluminum diecasting, was being evaluated. However, it was later discovered that a structural plastic IP had recently been developed for Chrysler Corporation for use on the Dodge Dakota truck [6]. Moreover a life-cycle analysis (LCA) of this IP has also been published in the literature [7]. The conclusion of this instrument panel LCA work supports the DFE conclusions presented here. To quote one of the researchers on the Dodge Dakota IP: "a key conclusion of the work is that thirty two pounds of PC/ ABS is a much more attractive recovery target than the diverse material mix of the previous design. [8]"

ACKNOWLEDGEMENTS

The author would like to thank General Electric Plastics Division for their support of this work through a grant to the University of Rhode Island.

He would also like to acknowledge the assistance of the following graduate students who helped with the various DFA, DFM and DFE stages of the work: Ted Andes, Peter Hardro, Sharath Ramamurthy, Frank Sienkiewicz, and Marcello Trolio. The students were assigned special design projects by the author while carrying out thesis research on other topics for their graduate degrees.

References

Boothroyd, G., Dewhurst, P, and Knight, W.A., Design for Manufacture and Assembly, Marcel Dekker, NY 1994..r Manufacture and Assembly Software, Boothroyd Dewhurst, Inc., Wakefield, Rhode Island

Dewhurst, P., Final Report on URI/ GE Plastics Instrument Panel Project, URI, Kingston, 1997.

TNO Institute of Industrial Technology, P.O. Box 5073, NL-2600 GB Delft, Netherlands.

Kalisvaart, S. and Remmerswaal, J., "The MET Points Method - a single figure environmental performance indicator", Proc. of Conference on Integrating Impact Assessment into LCA, SETAC, Brussels, October 1994.

"Dodge Dakota takes instrument panel to new level’, engineering news, Design News 12-16-96.

Lee, R.A., Prokopyshen, M.H., and Farrington, S.D., "Life Cycle Management Case Study of an Instrument Panel", SAE Paper No. 97-1158, 1997

Farrington, S.D., personal communication, October 1997.