Category Archives: Case Study

Case Study for Dfma Category

Redesign of Aerospace Project cuts costs up to 74%

aerospace-helicopter ah64d

Design for Manufacturing and Assembly Application on the design of the AH64D Helicopter

Alfredo Herrera
Dimensional Management Technical Lead
McDonnell Douglas Helicopter Systems
Mesa, Arizona
Presented at 12th International Forum on DFMA
June, 1997, Newport, RI.

Abstract

This study examines the effectiveness of Design for Manufacturing and Assembly (DFMA) methodology used by the design, manufacturing, quality, and supporting engineers for the development of the Longbow Apache Helicopter. Data were obtained through the Integrated Product Development (IPD) team for several redesigned areas of the Longbow prototype Helicopter Crew Station. Results of the study show that DFMA can be an effective approach, as indicated by a significant cost and weight savings.

Introduction

Design for Manufacturing and Assembly (DFMA) is a design philosophy used by designers when a reduction in part count, a reduction in assembly time, or a simplification of subassemblies is desired. It can be used in any environment regardless of how complex the part is or how technologically advanced this environment may be. It is gaining popularity where manufacturing costs are a concern. DFMA encourages concurrent engineering during product design so that the product qualities reside with both designers and the other members of the developing team.

DFMA is utilized by hundreds of domestic and international companies in an effort to cut down concurrent manufacturing and assembly time. Domestic companies like Allied-Signal, Motorola, Hughes Aircraft, and McDonnell Douglas Corporation have already implemented the DFMA philosophy throughout their product lines. The DFMA implementation process may be done at two different stages: when a new design requirement is established or when an existing design requires product optimization, such as the case of the Longbow Apache Helicopter. At the initial design stage, the designer develops a simplistic conceptual design by envisioning an assembly that requires a minimum of parts to perform the requirements previously established, and is easy to install. In the second stage the designer redesigns existing assemblies or designs new assemblies in order to implement design optimizations to ease manufacturing, and installation. This also meets reliability and maintainability requirements, moving the design towards cost reduction and customer satisfaction. In order to maximize the benefits of DFMA the designer must have a good working knowledge of the manufacturing processes available, and process capabilities to produce the parts. The design and manufacturing elements must work closely to determine the best manufacturing approach. A review of the State-of-the art manufacturing processes which increase the effectiveness of DFMA provides a means to understand this synergism, as well as the availability of Statistical Process Capabilities (SPC).

High Speed Machining (HSM)

HSM was the primary tool used by the DFMA process in the airframe structural design area. It can be defined as the act of machining at speeds higher than 10,000 RPM. A High Speed Machine is a manufacturing tool that, when used in a DFMA environment, allows part count reduction by providing the machining capability to rapidly create complex geometrical parts normally designed with many mechanically fastened sheetmetal parts.

HSM supports DFMA with the utilization of improved machine cutting spindle technology which has created spindle speeds of 24,000 to 40,000 RPM. It combines rough and finish machining of material into a single machine operation. It reduces heat buildup and thermal growth allowing stable machining operations and close tolerance features.

It has been so successful that it is currently used by McDonnell Douglas Aerospace on several on-going aircraft programs, including the F/A-18 E/F Avionics Shelf, and Aileron Closure Rib , the T-45 Nose Landing Gear Door, and C-17 Cargo Floor Ramp.

With its application, complex assemblies are converted into simple part assemblies reducing part count and simplifying the assembly process. Reducing parts count cascades into savings in other areas. For example, it reduces part cost, fabrication and assembly time, and reduces tool design and fabrication costs. The tooling manufacturing process can be eliminated since the designs are transferred directly from a computer aided design system model like Unigraphics II (UGII), to the high speed machine itself, representing a major saving. This is for the Numerical Control (NC) and the Coordinate Measuring Machine (CMM) programming. Additionally, it improves quality, provides design flexibility and reduces weight due to the absence of fasteners.

Composite Design

Composite design helps in the assembly process since it can help in minimizing the number of parts by combining several parts into one. The assembly time is reduced benefiting manufacturing in a similar manner to HSM. However, this manufacturing process is labor intensive due to the time it takes to set all the plies that build up the thickness of a part. Each ply must be cut to shape, stacked up, and then bonded together with adhesive at a later time as it is cured in an autoclave. It also creates the need for expensive tooling, since tools are used to form and shape complex parts. Tool wear and tear is another problem encountered in the composite manufacturing process, caused by the frequency of tool usage.

Composites are more susceptible to damage and ply dilamination due to humidity, creating costly future repair problems, depending on where and how they are used.

Superplastic Forming (SPF)

Superplastic forming (SPF) is the process in which a specific type of aluminum alloy, aluminum 2004 for example, is formed by blowing hot air against a sheet of metal, and over a tool inside an oven in order to shape the metal into the part’s configuration. An advantage is the low part rejection since there is virtually no tool wear. Complex parts are simplified by integrating auxiliary parts into one part. Compound curvature shapes are also made possible since metal is formed at its plastic point. However, SPF is limited to small parts due to oven size limitations.

DFMA has already proven to be a savings tool as used by McDonnell Douglas Corporation in the military and commercial divisions. In the military division, significant savings have been obtained in the F/A-18 C/D Bay 4R shelf; savings of 84% in the number of parts, 73% cost, 11% weight, and 89% assembly time. In the conversion of the F/A-18 C/D into F/A-18 E/F there was a part count reduction from 1,744 to 1,048. In the commercial division; the MD-11 cargo liner had cost savings per aircraft of $86,000, the MD-11 #2 Bulkhead had cost savings of $4,000. In general the part count reduction done by McDonnell Douglas has been in the range from 36% to 96% on several of its aircraft component assemblies.

In the development of the Avionics upgrade on the U.S. Army’s A Model Apache Helicopter, McDonnell Douglas Helicopter Systems (MDHS) utilized all design and manufacturing tools available. DFMA was one of these tools. DFMA was used on the AH-64D helicopter (Longbow) Program along with three auxiliary tools: High Speed Machine (HSM), composite design, and Superplastic Forming (SPF). HSM and SPF were used in the airframe structure design and composites in the Environmental Control System (ECS).

Statement of the Problem

During the engineering development of the Longbow Apache Helicopter, MDHS found that traditional design methods would not provide adequate support to the challenging budget and schedule. New design methods had to be obtained and implemented in order to support the program objectives.

Review of the Literature

For over 500 companies worldwide, DFMA has become a vital design tool in their effort to compete in domestic and world markets. The data collected from published literature on over 50 case studies conducted by McDonnell Douglas at its St. Louis plant, outlines the power of the DFMA methodology. Some of the results are: reduction in manufacturing cycle time, part count reduction, part cost reduction; time-to-market improvements; quality and reliability improvements; reduction in assembly time.

According to Geoffrey Boothroyd, Professor of Industrial and Manufacturing at the University of Rhode Island, the practices now known as Design for Assembly (DFA), and Design for Manufacture (DFM) had their start in the late 1970’s at the University of Massachusetts. Of all the issues to consider, industry was most interested in Design for Assembly.

When developing a product, the maximum potential cannot be achieved without considering all phases of the design and manufacturing cycle. DFMA meets this demand by addressing key assembly factors before the product goes on to the prototype stage. These key factors are the product appearance, type, the number of parts required in the product, and the required assembly motions and processes.

Statement of Hypothesis

Based on the success of DFMA to reduce cost and time on other programs, it follows that its application to the Longbow Apache Modernization Program should have similar savings. Utilizing DFMA for the design of the Longbow Apache crew station will reduce part count, manufacturing time, and assembly time, all of which help to reduce cost. The simplification of assemblies and reduction of fasteners will also reduce weight in the crew station area.

Methodology

During the years of 1994 and 1995, MDHS redesigned and optimized one of the six Longbow prototype helicopters. An Integrated Product Development (IPD) team was formed to conduct this redesign. The IPD team is a concurrent engineering team where representatives of several organizations such as engineering, manufacturing, procurement, suppliers, product support, quality, and others, work together to develop a product design. This design is then brought into production in a short period of time without the budget and lengthy schedule usually encountered by other organizations without a team concept in place.

Six helicopters were completed in the prototype phase and the experience obtained from this phase was applied to the Longbow Initiatives Project. During this project, Design and Producibility Engineering and Planning (PEP), which was developed and implemented with the purpose of improving the previous prototype aircraft configuration used DFMA as an aid to accomplish that established objective.

DFMA was applied to a limited number of parts within the crewstation, and the Improved Extended Avionics Bay (IEFAB) of the Longbow Apache Helicopter. Data were gathered and recorded by the IPD Team and compared to the baseline prototype helicopters which were designed without using DFMA.

Each DFMA case study was conducted by redesigning existing assemblies. The IPD team met and analyzed its requirements, including material, function, and location of parts. Once a preliminary design was done, the team studied it in order to reduce the part count, weight, and assembly time.

Data was obtained from each IPD team member that was involved in the DFMA process. Their estimates, tables and schedules were analyzed. All data that could be found relating to DFMA applications on the Longbow Apache Program including: producibility analyses, design concept descriptions and lists, weight data analysis, schedules based on the design and manufacturing plans, cost estimates, and detailed DFMA plans on at least four assemblies, were used to assess the impact of DFMA. Data were collected and summarized as they were made available by the IPD Team.

Collected data were loaded into the Boothroyd Dewhurst Inc.’s (BDI) DFA 7.1a software. This software analyzes the design, manufacturing, assembly process, and materials used. It then summarizes and provides recommendations on how to improve the design using DFMA philosophy.

Results

The first assembly examined is the Pilot’s Instrument Panel which is comprised of a combination of sheet metal angles and extruded stiffeners. The panel itself is attached to an existing airframe structure with rivets. It consists of 74 parts with a weight of 3.00 Kilograms. The fabrication time for this instrument panel is 305 hours. This panel also requires a final assembly tooling fixture in addition to tooling needed to form all brackets and angles. Utilizing DFMA in conjunction with the IPD Team concept and availability of HSDM, resulted in the redesign of the pilot’s instrument panel, into only 9 parts.

Table 1. Pilot’s Instrument Panel Estimate Summary

Present Instrument Panel DFMA Proposed Instrument Panel
Part Count 74 pieces 9 pieces
Fabrication Time 305 Hours 20 Hours
Assembly/Installation Time 149/153 Hours 8/153 Hours
Total Time 697 Hours 181 Hours
Weight 3.00 Kilograms 2.74 Kilograms
Cost 74% Reduction

Subsequent analysis yielded data indicating that the fabrication time could be reduced to 20 hours. The total manufacturing and assembly time would be reduced from 697 hours to 181 hours, weight reduction would be to 2.74 Kilograms, and the total cost was reduced by 74%. The pilot’s instrument panel DFMA concept is shown, and Table 1 provides a summary of the estimated comparison for the Pilot’s Instrument Panel.

Panel graphic

In addition, data were obtained for three other areas: The Co-Pilot Gunner (CPG) instrument panel was a good candidate for DFMA due to its assembly complexity, number of parts and rivets required to assemble it. It included the Up-Front Display (UFD) tray and the Multifunction Display (MFD) tray. These last two sub-assemblies of the CPG instrument panel made it very difficult to assemble and require extensive labor for the assembly activity and the final installation.

The total original part count was 87 parts. Presently it has been reduced to 12 parts, where 7 are machined parts and 5 are sheet metal/composite parts. The original instrument panel is a combination of sheet metal parts representing more than 90% of the total parts and a few machine parts being fastened mechanically. Bench tooling is required to perform the sub-assemblies of the UFD and MFD trays, making the task very difficult.

With the simplified DFMA instrument panel, sub-assembly is minimal, representing a considerable amount of time and cost savings, as well as weight savings. The DFA Summary of Results show part of the DFMA assessment done by utilizing the BDI software analysis on the Instrument Panel Top Assembly Drawing, 7-511171010-1. These tables provide data indicating what it takes for manufacturing to produce, and assemble the CPG instrument panel, substantiating the numbers of parts, hours, processes and cost to complete the task

The BDI software does a complete analysis of all the tasks required, providing a summary of results. A general overview of how long and how much it takes to build and assemble specific components or parts is done. Also, an analysis profile is provided with suggestions to improve the current design.

This software is used by the manufacturing team members to estimate and predict the savings that can be obtained. Data are entered and the system does its analysis in different areas. A complete listing of all the activities required to perform an assembly, including the count of tasks, figures the minimum items required for assembly, and the item(s) cost. This provides a complete overview of the task to be performed.

The BDI software does an assembly analysis profile on a set standard format where it theorizes the number of tasks to be performed, fasteners required, connectors to be installed, candidates for elimination, acquisition of items not in reach or on stock, acquisition of tools not on hand, standard operations, library operations, and reorientations. After all these activities are numbered and plotted, it automatically provides suggestions for improvement.

The system provides suggestions for design and for assembly, by giving instructions, and indicating every task with its time saving, and its percentage reduction. It indicates specific instruction to perform the related tasks in order to obtain the suggested savings.

It also lists, under what is called Design for Assembly Analysis Totals, all the parameters used for the analysis such as total assembly time, total assembly cost, total assembly weight, number of parts and sub-assemblies, theoretical minimum number of parts or unanalyzed sub-assemblies, and the hourly labor rate.

All the suggestions and comments included within the computer generated tables are automatically provided to aid the designers and manufacturing engineers to obtain a better view of the job.

Design for Assembly Suggestions for Redesign provides detail analysis of the UFD Tray Assembly (-13), MFD Tray (-57), MFD Tray (-49), Bracket (-119), Closeout Assemblies (-29 and -31), UFD Tray Assembly (-3), and Tee (-125) which are installed on the copilot’s instrument panel, and the pilot’s instrument panel. It gives suggestions for hardware reduction such as rivets by incorporating integral fastening elements into functional parts, or by doing a different securing method. Part reduction is recommended by combining parts with others. Hardware can also be reduced by the addition of chamfers, lips, leads. Assembly redesign is suggested to provide unrestricted vision for Bracket (-119), Closeout Assemblies (-29 and -31), UFD Tray Assembly (-3), MFD Tray (-57), and MFD Tray (-49). Detailed time saving and percentage reductions are provided.

There was a considerable cost, weight, and schedule savings of approximately 74%, 8%, and 74% respectively in the areas where applied.

Conclusions

There are many lessons from which the aircraft industry and, in this case, MDHS can fruitfully benefit. From the experience of other companies, it seems that various attitudes and practices must be nurtured for DFMA to be fully implemented. Many commercial companies attribute their world-class competitiveness to DFMA. John Deere and Company says, “the first companies to implement DFMA will be the leading world-class competitors, the last companies to implement DFMA won’t have to worry about it.” So, applying DFMA can help make MDHS a world-class competitor in the helicopter manufacturing industry, adopting a trend that has already been started by McDonnell Douglas in St. Louis and that has proven to be very successful.

The utilization of DFMA has not been extensive. Indications are that DFMA can successfully contribute when cost, weight, and schedule are the prime drivers for the development of a program.

It is suggested that DFMA has been successful in the Longbow Apache Program, even though its utilization was limited to a few components only. If it could have been used across all the design activities, it would have been more helpful in reducing the parameters indicated previously (cost, weight, and schedule). The part count was reduced by 87%, the fabrication time was reduced by 93%, assembly time was reduced by 94%, the weight was reduced by 10%, and the cost was reduced by 74%.

Training is a must. DFMA could be more successful if all the team members understand what it is about and what are the ultimate goals. The management commitment to DFMA can bring the success of the Program by making the decision on educating the employees. It is understood that every professional individual possesses a basic training and that it will give an easier transition towards the goal of the project, but, if these professionals are not provided with the adequate tools, the success of the program could be in jeopardy. That is why training is a key item for the success of any program.

It is clear that for DFMA to be effective, the design team must understand the capabilities of the production process that will be used to produce their parts and set requirements within those limits.

Recommendations

During the course of this study it was found that the Design for Manufacturing and Assembly utilization by the aircraft industry can present great advantages. Still, there are areas that could not be covered due to the study’s scope. Areas like designing for disassembly where designs also consider the future dismantling of assemblies for environmental purposes.

DFMA was applied on structural, and ECS designs only, but it can also be used in areas like flight controls, engines, transmissions, hydraulics systems, and electromechanical components used to house and support electrical and avionics components. These areas have not been studied, and may be topics for further future studies.

Implementation of DFMA is not an easy task. It takes the correct attitude in order to successfully use it and overcome all barriers created by people used to work under a different approach. Additional study in this area may be beneficial.

For a successful DFMA implementation, more management participation and concern are needed, as well as providing more empowerment to the different team members so that they consider themselves as participants in the development of the program as any others within the organization. Management involvement and commitment may be a good topic for study, given the results and successes of DFMA implementation in the general and aircraft industry.

A closer look at process business reengineering tools should be considered. Dimensional Management (DM) is one of the tools that can help this process. DM is an analytical and quantitative approach used to manage assemblies through disciplined techniques such as the proper identification of control datum, prediction of allowable variation, definition of key interface characteristics (KC), part count reduction and/or clearly defined product acceptance criteria established early within the product definition life-cycle. While listening to the voice of the customer, DM can provide a set of associative concepts and structural tools utilized within a disciplined Integrated Product Definition (IPD) process that establishes product characteristic requirements that ultimately yields ease at assembly, driving reduced operating costs and a reduction in non-conformance parts and tools.

Disclaimer

The author wishes to express that the views and findings are those of the writer and in no way intended to reflect the official opinion of McDonnell Douglas Helicopter Systems (MDHS).

New Generation of Medical Test Equipment benefits from DFMA input

new-genaration

DFMA software from Boothroyd Dewhurst lets designers produce a significantly improved case molding for Ciba Corning’s 800 Series blood-gas analyzer. The outer shells hinge together like a suitcase, making assembly easier and faster, with a closer fit than the previous 200 Series analyzer.

When Ciba Corning Diagnostics Corp. set out to upgrade its line of blood-gas analyzers, it established aggressive cost and high-quality targets. The new family of analyzers had to share a single base platform that could be upgraded at the customer’s site to change with clinical demands. The analyzers also had to be durable, easy to assemble, and simple to use.

Design for Manufacture and Assembly (DFMA) software, developed by Boothroyd Dewhurst Inc., Wakefield, R.I., was selected to help cut costs and part count. Says David Yeo, new product manufacturing engineer at Ciba Corning, “Each design engineer was teamed with a manufacturing engineer throughout the project to focus on manufacturing and assembly methods. We set up an area where those who had input on the design could work together.”

All assemblies were run through the DFMA analysis which provided estimates for assembly costs. But more importantly, the analysis prompted discussions about material selection and part fit. Some of the most significant changes made to the new model were in the hydraulic system in the upper caseworks. For example, the reagent manifold was made from a solid block of acrylic, eliminating need for numerous tubes. It reduced hydraulics parts content by 80%.

The upper caseworks presented a particular challenge. From the solid-model geometry, stereolithography was used to develop the first model, followed by rubber and cast urethane, and finally structural foam. “DFMA reduced total parts count, but the product still had to be highly functional with numerous features. Packing more functions into a single component presented a challenge to conventional prototyping methods. Three-dimensional solid modeling and stereolithography for rapid prototyping were able to meet that challenge,” said Yeo.

DFMA analysis helped reduce the overall number of parts by 48%, fluidic connections by 60%, and the cost by 22%. The designers also found trade-offs in molding snap fits and holes for screws or Tinnerman fasteners that helped reduce cost.

The difference between the old 200 Series and the new 800 Series is dramatic. The new model contains fewer components, most of which are self-aligning. “We paid special attention to self-locating features. We didn’t want to reorient any part more than we needed. Now, when assembling the 800 Series, the pins match up to specific holes and parts fit mating parts,” says Yeo.

One example of how the company might use DFMA software on upcoming projects relates to machining. According to Yeo, a machining software module would have helped even more in deciding if it was economical to produce separate parts. He will look closely at future DFM shape-forming modules for this reason.

Reprinted with permission from the 10/14/96 issue of Machine Design
Copyright 1996, Penton Publishing

Motorola Solutions uses DFA software to benchmark designs and measure improvement in its global product portfolio

moto(1)

Every product in Motorola Solutions’ extensive portfolio—which includes handheld RFID readers (top), mobile radios (middle), multimedia micro kiosks (bottom), and many others—is benchmarked and analyzed during design/redesign using Boothroyd Dewhurst’s Design for Assembly (DFA) software to cut part count and assembly time.

How good are our product designs? Are they better than past versions? How do they compare with those from other companies?

With competition in the global economy fierce, management at telecommunications giant Motorola Solutions wanted to know the answer to these questions. It was up to Rich Darrell, engineering manager at the company’s Holtsville, NY design center, to provide the answers. Says Darrell, “Senior leadership challenged us to come up with a methodology for measuring whether our designs were improving.”

For Darrell and his team, that would mean getting a handle on a product portfolio that was hundreds of items deep: It encompassed government and public-safety devices from the parent Motorola plus mobile-data-capture and handheld-computing units added after a 2007 merger with Symbol Technologies. To further complicate the task, the various product lines were distributed among separate design and manufacturing teams at six centers around the world—three in the U.S, and one each in Israel, Mexico, and Malaysia.

Choosing how to measure

In response to the challenge, Darrell turned to Design for Assembly (DFA) software, a tool developed by Rhode-Island based Boothroyd Dewhurst, Inc. and first adopted by Motorola more than 25 years previously. Part of a broader software strategy called Design for Manufacture and Assembly (DFMA™), DFA is used to simplify product designs by eliminating any unnecessary parts, making assembly easier and lowering labor costs. The other module, Design for Manufacture (DFM), provides concurrent-costing estimates for alternative material and fabrication choices.

Following its introduction, DFA had become a driving engineering principle at the company and was integrated alongside lean and six sigma initiatives in their internal training program Motorola University (now called Motorola Solutions Learning). Over time, though, as the global enterprise grew more complex and the product portfolio expanded, engineering procedures evolved to keep pace—but not always in sync. “While every design center was using the software, each site had a different way of applying it to development,” says Darrell. “So in 2012, we formed a global DFA team and worked to come up with a unified process.”

Darrell along with colleague Chris Foley, a process engineer for new product development also based in Holtsville, led the initiative. Together, they honed in on a calculation from the software—the DFA index—as a key measure of overall product design. In simple terms, the index is a number arrived at by dividing an “ideal” assembly time by the actual assembly time, where an “ideal” assembly is defined as one having a theoretical minimum number of parts (for a more detailed explanation see Sidebar “Demystifying DFA”).

“DFA clearly shows what parts are required by product function and what parts can be eliminated,” says Foley. “As the software steps through the process, it illustrates how you can achieve significant cost avoidance by identifying opportunities for simplification of product structure.”

The two experienced engineers felt strongly that the DFA index would work well as a metric to quantify product improvement. “It was perfectly suited for the purpose,” says Darrell.

Creating product families

While the DFA index provides a data-driven way of measuring new product designs and tracking improvement during redesign, Darrell and Foley thought it would be useful if comparisons could also be made between the designs of different products.

“We have hundreds of products and pretty diverse product groups,” says Foley. “It isn’t really meaningful to compare apples to oranges. It would be better if we could compare one red apple to another.” So they decided to divide the company’s extensive portfolio into a series of product families and then calculate a DFA index family range for each product grouping.

For starters, they separated legacy Motorola products from Symbol’s original product lines. Beyond this broad classification, however, things quickly got more complicated. Consider scanners: A first distinction was whether the device was corded (such as for retail checkout) or wireless (as used for inventory control). Hardwired units are much simpler and have a lot fewer parts than mobile ones, so it was determined that each classification warranted its own product family. Then there was further division into sub-families, depending on whether the product was laser-based or digital. Within groupings, they also included custom products whose only difference might be a minor change in the software, the color of a keypad, or an added logo for a specific customer.

When lumping and splitting in this way, all products within a family or sub-family, while not identical, were closely related by their technology. This ensured that items in a group were more likely to share design strategies and product features. “We got as close as we could to comparing the same kind of apples,” says Foley. The field of products, according to Darrell, included numerous families within both the fire/police radio and scanner/mobile computing portfolios—with each grouping containing between five and ten items.

Once families were created, all products were analyzed using the DFA methodology. This gave the engineering team a DFA index for every product and a range (and average) for each family or sub-family (see Figure 1). They also calculated a DFA index for competitive products. “Benchmarking is always key for us,” says Darrell.

Scanner Product Family Number of Products Minimum DFA Index Maximum DFA Index Average DFA Index
Scanner Laser-based 5 4.6 15.6 10.3
Scanner Imager-based 6 7.3 15.3 10.8
Scanner-on-a-stick Imager-based 1 14.9 14.9 14.9
Scanner-on-a-stick Laser-based 2 10.9 16.4 13.7
Scanner w/ radio
Imager-based
1 14.7 14.7 14.7
Scanner w/Radio
Laser-based
2 10.6 14.4 12.5
Mini-Kiosk 2 3.2 5 4.1
Imager Slot scanners 1 2 2 2.0
Mobile Computing Product Family Number of Products Minimum DFA Index Maximum DFA Index Average DFA Index
Retail 2 11.2 19.5 15.35
Light Industrial 3 6.9 10.4 8.7
Industrial 8 9.5 15.3 12.80
Retail w/WAN na na na na
Light Industrial w/WAN 2 12.3 19.6 15.95
Industrial w/WAN 4 13.4 18.5 14.9
Headsets 1 4.3 4.3 4.3
Smart Badges 1 12.8 12.8 12.8
Tablets 1 10.6 10.6 10.6

Figure 1. Using Design for Assembly software, DFA indices were calculated for products in Motorola Solution’s scanner (top) and mobile-computing (bottom) product families. Those values were then used to arrive at ranges and averages for each product family, which serve as goals for future new design and redesign programs.

These metrics helped the global team identify how close designs were to the company’s best-in-class goal. They also gave them a target to shoot at when looking to quantify whether their products were improving. “That’s how we rolled the program out,” says Darrell.

Redeploying DFA globally

The next step was to realign the company’s separate engineering groups in the use of DFA. To do so, Darrell embarked on a world tour of the design centers to retrain teams and get buy-in.

In order to sustain momentum following the tour, Darrell and Foley instituted a weekly global meeting (which is ongoing) with teams from all of the centers participating. Included are experts from design, manufacturing, and operations. “We have the mechanical, process, and quality engineers on the call,” says Foley. “We also have the electrical and software experts and a participant from purchasing as well. There are representatives from every discipline.”

Working with cross-disciplinary groups and using baseline DFA index values has improved the efficiency of the product development process enterprise-wide, says Foley. “In the beginning of a program, for example, we might be worried about cost,” he adds. “We might remember that this family of products had high-cost drivers in labor or maybe some unnecessary parts. Using DFA, we can look back and at the same time think ahead to avoid those sorts of problems. The tool really helps with these design discussions.”

DFA not only focuses engineers on concrete improvements—such as snap-fits, modularization, or fewer connectors—according to Darrell, it also helps the Motorola Solutions team keep up with rapidly developing communications and computer advances, which are integral to many of their units. This includes the design changes needed to adopt technologies such as digital imaging for data capture or Bluetooth. “We need to look at tomorrow’s technology today,” Darrell adds. “Most of our products have revolutionary designs.”

In addition, DFA is instrumental in all three stage-gated phases of the company’s design process—from “concept” to “critical” through the “final” phase, says Darrell. “As we initiate a redesign or new-design project, we’re always trying to improve,” he adds. “Our goal is to be within the index range—or better.”

Measuring for management

“We’re management driven,” Darrell continues. “Every month, I compile the DFA data and present it to senior leadership so they can see how our product development programs are tracking in real time.” If a product has an index lower than the target, it immediately triggers questions at all levels. This built-in early warning system generates targeted scrutiny and redesign activities early in the development cycle.

As part of reporting, Darrell also tabulates the financial benefits the software delivers. “If we come up with a savings of one dollar per unit, for example,” he says, “and we make one hundred thousand units a year, then we saved the company one-hundred-thousand dollars by using DFA upfront.” Quality is number one at Motorola Solutions, but cost reduction is an all-important second.

“We all know that using DFA early in design development is where you get the biggest bang for the buck,” Foley adds. As a result, in all of their design meetings, he and Darrell remind the global team of the hidden costs of adding on parts. It’s a well-documented fact—becoming engineering dogma—that the majority of expenses in a manufacturing organization originate from the designs of its products.

When the final accounting was done, Darrell notes that management’s continued support of the DFA initiative is ensured. Part reduction efforts have successfully cut assembly times on numerous products, and DFA indices almost universally have shown steady improvement—for every product family across the portfolio (see Figure 2). That bottom line allows the company to focus its precious resources on other improvement initiatives.

Product Family Type of Product Model/Name DFA Index Product Release
MCD Mobile Computer – Brick/Gun MC9060 9.0 2003
MCD Mobile Computer – Brick/Gun MC9090 13.6 2006
MCD Mobile Computer – Brick/Gun MC9190 14.5 2011
Product Family Type of Product Model/Name DFA Index Product Release
Industrial light Mobile Computer – Brick MC1000 6.9 2001
Industrial light Mobile Computer – Wearable Scanner Onyx 8.8 2012
Industrial light Mobile Computer – Brick MC2100 10.4 2012

Figure 2. In each of these charts, DFA indices were calculated for either redesigns or similar designs of a product as it evolved. The increasing value of the index over time indicates product improvement as defined by decreasing part count using the DFA methodology.

Demystifying DFA

In the complex world of product development, nothing could be more fundamental than product simplification as a means for improving designs. This strategy, when based on the part-reduction methodology of Design for Assembly (DFA) software, leads to reduced assembly labor and cost, improved quality and reliability, and shorter cycle times.

At its most basic, DFA involves the elegant incorporation of multiple parts into a single multifunctional component that shows improved economy and performance. Product engineers are increasingly savvy to the fact that 85 percent of all manufacturing costs are determined in the early stages of design—well before fabrication and assembly processes are locked in.

The DFA index has historically provided a quick, objective measure of assembly efficiency that can serve as a basis for quantitative comparisons of design alternatives, both internally and against competing products. The metric is automatically calculated by taking the assembly time for an “ideal” design (with the theoretical minimum number of parts needed for functionality) and dividing it by the total calculated assembly time of the design with its real-world part count. This result is then expressed as a percent. As the actual assembly time decreases through part reduction efforts, the DFA index increases. Higher indices indicate product improvement. The tool encourages redesign by pointing to specific changes and expresses the penalty associated with poor, or complex, design.

The index calculation, as conceived, was based on “small” products with “ideal” components defined as having the following characteristics: standard symmetry (alpha of 180 degrees and beta of zero degrees); being within “easy reach” of the assembler; secured by a simple snap-fit operation; and without handling or insertion difficulties. These same assumptions are obviously not always the case for larger products and, therefore, use of the assembly index in these instances was not encouraged. Engineering teams, however, can make highly detailed and accurate design comparisons for products of any size using the process time, process cost, and total cost results that have always been an integral part of the software.

Redesign of Forklift Hydraulic Cylinders

Redesign-of-Forklift-Hydraulic-Cylinders(1)

Matthew Miles
Mechanical Engineer
The Raymond Corporation
May 2008

Introduction

Customers demand reliable and high quality products at the lowest possible cost. Accordingly, it is incumbent upon manufacturers such as The Raymond Corporation (“Raymond”) to look continually for opportunities, when possible, to streamline the manufacturing of its products to meet customer needs. In this regard, Raymond was introduced to Boothroyd Dewhurst Design for Manufacturing and Assembly cost reduction software in 2006. Since that time, Raymond engineers have been able to incorporate the costing tool in the design and manufacture of Raymond’s full line of electric forklift trucks. The DFMA software was used to assist in the redesign of the hydraulic cylinders on the Model 7400 Reach-Fork® truck, in which part count was reduced significantly while maintaining product and component quality, functionality, reliability and reducing cost.

Raymond

The Raymond Corporation is the North American leader in the design and manufacture of the most reliable narrow-aisle electric forklifts. From orderpickers and Swing-Reach® trucks to stand-up counterbalanced and Reach-Fork® trucks Raymond provides material handling solutions that improve the efficiency for its customer’s warehousing and distribution operations.

Located in Greene, New York, Raymond started as a family-owned business in 1922. Over the last eighty-six years, Raymond has revolutionized the material handling industry by leading the way through its innovation and reliability of its products. In 1934, Raymond invented the first commercially successful hydraulic hand lift truck. The 1950s saw the introduction of the first narrow-aisle and first Reach-Fork® trucks. The Model 700 Spacemaker was the first narrow-aisle lift truck to work in aisles less than seven feet wide. Raymond also introduced the first computer-controlled drive system for material handling trucks in the 1980s and followed in 2001 with the first AC-powered Reach-Fork® truck.

A variety of industry leading Raymond lift trucks are produced at three plants in Greene, New York, Muscatine, Iowa, and Brantford, Ontario, Canada. Raymond is part of the Toyota Materials Handling Group, which is the largest lift truck manufacturer in the world.

Introduction to Boothroyd Dewhurst DFMA Software

Raymond first used Boothroyd Dewhurst DFMA software in 2006. Small teams of four to five design and manufacturing engineers were assigned to use the software on two small trial part count analysis projects. The two items to be analyzed were a lift truck overhead guard weldment and an operator’s display mounting weldment. The goal was to use the DFMA software to determine part count for multiple design iterations and alternate manufacturing opportunities while maintaining product and component functionality, quality and reliability. These two projects let Raymond engineers learn the software and essentially test the part count and costing tool in the Raymond development process. Both projects yielded opportunities for reducing part count and associated costs that have been implemented to offset material and transportation cost increases.

The next phase of using the DFMA software consisted of two members of Raymond’s Mechanical Design Engineering (MDE) group using the software more extensively on part count reduction activities. The timing of using the software was coincidentally advantageous as the MDE group was assigned to a product improvement project of considerable magnitude. Looking to offset material and transportation cost increases on Raymond’s Model 7400 Reach-Fork® truck, the MDE group performed a detailed part count analysis on the elevating section of this lift truck. The elevating section of a lift truck consists of three telescopic frames that are raised and lowered by hydraulic cylinders. As the cylinders extend, two sets of lift chains mounted to the telescopic frames lift the fork carriage that carries the pallet and load to the desired height for storage or retrieval in warehouse storage racks.

This investigation into the Model 7400 Reach-Fork® truck found many areas in which to apply the DFMA software. All areas of the truck seemed to have some possibilities where a manufacturing cost savings could be pursued. From the elevating section to the tractor and operator’s area, the MDE group found parts and assemblies that were analyzed using the DFMA software. The truck’s hydraulic cylinders yielded the most significant of these opportunities.

Case Study – Reach-Fork® Truck Hydraulic Cylinders

This study will detail how the Boothroyd Dewhurst DFMA software was used to assist in the part count reduction of the Model 7400 Reach-Fork® truck. There are two sets of hydraulic cylinders installed in the elevating section of this truck that provide the lifting mechanism to the forks. They are referred to as the “main lift” cylinders and the “free lift” cylinders. Part counts were reduced for both sets of cylinders but more focus was applied to the free lift cylinders due to the intricacy of the design.

The main lift cylinders elevate the outer telescopic mast, which in turn elevates the inner telescopic mast via lift chains connected to both inner and outer masts. The free lift cylinders provide elevation to the fork carriage. The fork carriage is an assembly that rides along the interior of the inner telescopic mast and has the lifting forks assembled on the carriage front. The purpose of the free lift cylinders is to provide lift to the fork carriage without extending the elevating section and maintain the overall collapsed height (OACH) of the truck. The OACH is the measured height of the truck when the elevating section is fully collapsed. This is beneficial when transporting loads through warehouses without the mast extended and clearing any transitional doorways at a fixed height. At the full extension of the free lift cylinder, the hydraulic system starts extending the main lift cylinders. The term used to describe this event is called “staging.” Essentially, it is the transition of the free lift cylinders extending followed by the start of the main lift cylinders extending. Raymond has developed a patent number US 6,557,456 B2, staging free lift cylinder that cushions the fork carriage and load during the staging event. This patented cylinder reduces noise and shock during staging impact.

The first step was to examine the current staging cylinder design and brainstorm ideas as to how to reduce part count of the patented cylinder without losing any cushioning or having an increase in noise during staging. A part count and resulting cost reduction was required to offset the rise of material and transportation costs. Each idea was investigated to determine the effect on performance and part count of the staging cylinders. This is where the DFMA software was effectively used to quickly estimate potential changes in cost with each new design idea. The value of the DFMA Software came to be realized as Raymond engineers could, in a timely fashion, project the cost of new designs and then through determining the feasibility of the new design, decide the best course of action. The new designs were efficiently reduced to the most beneficial idea.

The analysis of the staging cylinders started with determining a baseline cost using the DFMA software. From there, the new design was analyzed for cost to compare with the baseline cost. The promising results led to building prototype versions of the redesigned staging cylinders. After cycling the prototypes to acceptable levels without any adverse effects, Raymond decided to introduce the redesigned cylinders into production. The newly designed cylinders delivered virtually the same performance from the cushioning and noise standpoint.

This redesign was a collaborative effort that included Raymond’s Greene Manufacturing Operations. During the course of the product improvement project Greene Operations was concurrently updating their hydraulic cylinder assembly area with a goal of increasing the operating efficiency. In addition to the redesign of the staging cylinders and cost analysis with the DFMA software, Greene Operations contributed to the redesign with suggestions for improved manufacturing methods. Both main and free lift cylinders housings are made from drawn-over-mandrel steel tubing. This tubing is cold drawn electric resistance welded tube with all flash removed. Greene Operations suggested that eliminating machining operations currently performed on the cylinder housing would enhance manufacturability in the hydraulic cylinder assembly area reduce cost. The new design saw a traditional two-part design reduced to one part. A combination end cap/manifold eliminated welding the existing manifold to the tubing side. This removed machining from the tubing and reduced welding times considerably. The top end cap also changed and went from being retained by a snap ring to becoming a screw-on end cap. This also reduced machining to the tubing and made for an easier removal of the end cap for seal replacement. The free life chain anchors were also reduced from a right and a left hand part to a common one used on both right and left free lift cylinders. Lowering the chain anchor also allowed manufacturing to weld the anchor in a different order in the assembly process that reduced assembly time.

Analyses of the internal assembly also led to using more common parts on both the main and free lift cylinders. While the main lift cylinders also benefited by using the one-piece end cap/manifold, the staging cylinders were able to borrow the lower cushion assembly from the main lift cylinders. The lower portion of each cylinder, main and free lift, acts as a cushion as well when the cylinders retract and collapse the mast. Using common parts increased annual part quantities used and lowered part cost in some cases.

The hydraulic cylinder assemblies were analyzed with the Design for Assembly (DFA) software. Some existing and all new cylinder piece parts were analyzed using the Design for Manufacturing Concurrent Costing® (DFM) software. These analyses quickly showed that the new designs would lower assembly part count and allowed Raymond Purchasing to use the DFM results when procuring the new parts. The following table shows the comparison of the old and new free lift staging cylinders.

  Existing Free Lift Cylinder New Free Lift Cylinder
Number of Parts 37 28
Estimated Labor (sec.) 964 734
DFA Index 8.7 12.7

 

Summary

Using the DFMA software allowed Raymond engineers to identify the highest potential part count reduction early in the redesign process. Overall, the MDE group was able to reduce manufacturing costs to the Model 7400 Reach-Fork truck main and free lift cylinders to offset material and transportation cost increases. The free lift staging cylinders benefited the most by part count reduction. The existing design used 37 parts. The new design now uses 28 parts. Other benefits resulted throughout the course of this cost reduction exercise. The MDE group was able to reduce part number inventories for each main and free lift cylinders. The Model 7400 Reach-Fork truck went from having four different main lift cylinder parts numbers to now only one. The free lift cylinder part numbers went from eight to six. The hydraulic lines had to be altered to accommodate the new cylinders and benefited from a part count reduction and improved efficiency in the assembly process. This product improvement project was a tremendous success for the MDE group and The Raymond Corporation and resulted in components of the same quality, functionality and reliability.

Moving Forward

Because of the success from last year’s product improvement projects combined with the experience of using the Boothroyd Dewhurst DFMA software, Raymond is now in the process of integrating the use of the software into its culture. Members of the MDE group that were involved with the product improvement efforts have taken part in developing a recommended practices procedure for using the DFMA software in-house. The goal now is to use the DFMA software at the beginning of development projects as a cost attainment tool to determine more accurately part count and cost from the beginning to the end of a project. Also in process is developing a database of Raymond’s Greene Operations machinery in the DFM software. These steps are being taken to ensure further success with the DFMA software during the development of Raymond’s leading brand of lift trucks.

John Deere Harvests Savings with DFMA

John-Deere-Harvests-Savings-with-DFMA(1)

New EPA-certified Tier 3 engine prompts cost-effective redesign

Manufacturers of offroad vehicles have been meeting increasingly stringent emissions standards for the past several years, and Deere & Company is no exception. Starting in 2006, in particular, Tier 3 regulations from the Environmental Protection Agency require specific reductions of oxides of nitrogen in nonroad diesel engines. Deere has developed a family of new EPA-certified Tier 3 engines for use in its range of offroad equipment.

In the case of one combine harvester, fitting surrounding components to a new Tier 3 engine led Deere engineers and two of the company’s suppliers along a path of intensive cost analysis and redesign. The Tier 3 engine was a different size than the previous engine, so modifications to the basic combine platform were needed.

A major assembly slated for redesign was the swingout landing deck, which has an integrated ladder that provides access to the combine during maintenance. The entire device swivels from the side or rear of the combine at a height of about 6 feet from the ground. The service person can then pull out the ladder, climb up several steps to the secure footing of the landing deck, and attend to equipment maintenance.

Because of the change in engine size, the new landing deck assembly had to be larger. The engineers also wanted to increase the structural rigidity of the assembly at the least possible extra cost without adding weight. As work on the new design progressed, the Deere team identified some cost challenges. First of all, estimates showed that the redesigned assembly would come in at 8 percent above target cost. Second, and of more concern, the supplier quote for the redesigned assembly was 26 percent higher than target.

Deere tackled the cost issues using Design for Manufacture and Assembly (DFMA®) software from Boothroyd Dewhurst, Inc. (Wakefield, R.I.). The software helped them analyze the design for the landing deck and simplify the assembly to save cost. DFMA brainstorming sessions with design and manufacturing engineers, supply management, and suppliers responsible for fabricating and assembling the landing deck yielded 83 design improvement ideas.

“We selected the landing deck for DFMA analysis because we had to take cost out of the design without affecting its integrity, and the software helps us determine how to do that,” says Matt Saxton, cost management specialist at Deere. “We also used the DFMA process to elicit cost-reduction suggestions from suppliers and get them engaged with our design team. We have a system that rewards suppliers for being cost-conscious. A lot of times they’ll tell us how we can reduce their costs by changing something in our design.”

In their redesign work, Deere focused on three tactics:

Shorten the deck sheet. The landing deck sheet is a section of perforated metal flooring that personnel stand on while servicing the combine. A decision to shorten the deck sheet meant Deere could get lengths of steel flooring more economically from a standard-size sheet. “We went to a sheet size that was less expensive to begin with and saved dollars per pound,” says Saxton. “Plus we utilized the sheet much better. Better raw material cost and less scrap represent a two-for-one benefit.” Changing the size of the deck sheet reduced raw material cost by 60 percent.

Strengthen structural support. In the previous design, fabricated C channel provided the assembly with internal structural support. The design team decided to use rectangular steel tubes as support members instead. This saved cost by eliminating a metalforming step and achieved more structural rigidity at less weight. “Weight reduction was a good customer benefit,” says Saxton. “Less effort is required to pull the landing out from the side of the combine, and a lighter machine uses less energy during operation. Any reduction in weight helps manage the use of engine horsepower.”

Replace the ladder rails. In the concept design, the siderails for the pullout ladder were made from round steel tubes. After some investigation, the design team discovered that changing the geometry from round to rectangular tubes would gain them structural support at lighter weight, reduce the manufacturing operations required for mating parts, and reduce cost. With input from their supplier, they also pared 3mm from the tube length, which allowed three sections of tube to be cut from a standard 20-foot length.

The cost harvest for Deere paid off. The original 17-part ladder is now a streamlined 10-part assembly, and the engineering team beat their original target cost by 7 percent for the entire landing deck. Saxton is gearing up for other cost-reduction projects coming down the line. “I just saw a request for a cost review related to the 2008 production timeframe,” he says. “We’ll use DFMA to keep asking the right questions.”

Dell Builds a Framework for Success

dell

Design for Assembly and Design for Service drive the design of Dell Corporation’s Optiframe® Computer Chassis.

Dell Computer Corporation (Round Rock, Texas), the world’s leading direct computer systems company, has long been recognized as a provider of easily serviced, readily installed, customized computers. By 1998, Dell was associated as well with something else—an explosive growth 2.5 times the industry average.

Customized assembly of computer equipment requires operators, time, and space. Dell was glad that customers were enthusiastic about its products, but the company balked at the physical and personnel expansion that would be needed to meet the increasing orders. “Brick and mortar are expensive,” says Bradley Keup, a systems engineer at Dell. “So are added personnel.” Instead of adding facilities and people, the company took a less-expensive route: it redesigned its products to make them easier and faster to assemble and to service.

In fact, redesigning one product, the Optiframe® chassis for personal computers, saved Dell an estimated $15 million dollars in reduced direct labor costs. The company saved millions more by increasing throughput and thus postponing facility relocations that otherwise would have been required to boost manufacturing capability. The savings in materials cost from chassis integration and the related supply chain optimization program was $11.6 million for 1998 and is predicted to be $35 million in 1999. For a single design project, these are impressive savings.

The Optiframe design challenges

The goals that the Optiframe design team set for product and process design were no less impressive:

  • Create commonality throughout a product line. The Optiframe chassis is used as the corporate desktop platform for all seven models in the Optiplex line of personal computers. Each model is customizable. The new chassis design would have to accommodate all Optiplex models and all configurations of each model.
  • Reduce purchased part count at least 17 percent and mechanical assembly time at least 25 percent. This would increase factory throughput and capacity so that Dell could avoid having to relocate and build new manufacturing facilities.
  • Reduce screw-type count at least 67 percent and screw min/max count (a measure of the minimum and maximum number of screws used in the customized computer configurations for each model) at least 20 percent.
  • Make the product even more service and customer friendly by reducing average service time by at least 25 percent. The redesign had to support Dell’s well-known commitment to service and reliability, which earned an ‘A’ in June 1999 for the fourth straight year from the PC Magazine Reader’s Choice Service and Reliability survey.

Dell accomplished all of these goals through a methodology they call DFX, or design for X. Design for Assembly (DFA) and Design for Service (DFS) design analysis software from Boothroyd Dewhurst, Inc. (Wakefield, R.I.) are fundamental to this methodology.

Tools of the DFX design team

At Dell, DFX is a multi-disciplined development path that encompasses design for velocity, assembly, quality, manufacturing, service, total cost, logistics, safety and ergonomics, integration, environment and modularity. The goal of DFX is to drive optimization of product design, material flow, logistics, product warranty, and other areas vital to manufacturing through concurrent engineering, metrics-based improvement, impact analysis and strategic planning. The cross-functional, international design team shares data on the company intranet to evaluate and improve product designs.

The DFX design team includes representatives from procurement, manufacturing engineering, manufacturing quality, customer service, process engineering, new product engineering, supplier quality engineering and logistics. Membership from a wide range of expertise promotes early design definition and ensures manufacturability up front in the product development cycle.

To the DFX team, metrics are all-important. They drive operational and product goals, provide benchmarks, and allow the team to evaluate results impartially. “Our slogan,” says Doug Dewey, a DFX engineer, “is ‘If you can’t measure it, you can’t control it.’ Once metrics define product characteristics, we can set goals and measure our design progress against them.”

DFX metrics at Dell measure throughput, time and cost, and they are defined and implemented through several tools. Among these tools, DFA and DFS software are critical for establishing metrics and for promoting concurrent design evaluation and innovation.

DFA software enables engineering design teams to evaluate the functional purposes of each assembly component in a conceptual design. Data accumulates as the engineers answer three questions about each part in the product and its assembly sequence:

  1. Does the part move relative to other parts already assembled?
  2. Must the part be of a different material or be isolated from the other parts already assembled?
  3. Must the part be separate from other parts for purposes of assembly or disassembly?

Once the designers answer these questions, the software produces a summary report of a theoretically attainable optimum design. The DFA software then guides the engineer in rating each component on its ease of orientation and assembly. Finally, the software generates estimates of total assembly time and costs. Dell uses these estimates to help build its design metrics.

DFS software focuses on the disassembly sequence required to service components in the design assembly. Data transfers directly from the DFA structure chart to the DFS disassembly worksheet. The resultant data provides a time estimate for disassembly and reassembly, from which the software arrives at cost estimates for service.

“DFA and DFS software were key tools in setting first-pass initial goals for our Optiframe project,” Keup says. The team used the software to evaluate the designs of existing products—both Dell’s and their competitors’—and to establish benchmarks. Keup says, “Benchmark and competitive analysis is a huge part of our job. We have design data bases on nearly every product that our major competitors have released in the last six months.”

In addition to setting achievable design goals, the team established metrics for product throughput. Based in part on the estimated assembly and disassembly times and on cost data derived from DFA and DFS analysis, the team was able to establish metrics for units assembled per hour per operator, direct labor costs per unit, mean time to repair, and even units assembled per hour per square foot of shop floor. This last metric was fundamental for meeting throughput goals that would enable Dell to avoid costly plant expansion and relocation.

Design review process

During basic layout review of the Optiframe design, data from the Boothroyd-Dewhurst software analysis was linked to an animated model on Dell’s intranet so that worldwide design and process teams could shape the product and the process concurrently. “Our electronic product models provide motion so that the team can see the characteristics of part assembly and disassembly, clearances and orientation,” Keup says. “In design analysis, if a picture is worth a thousand words, an animated model is worth a thousand pictures.”

The design team incorporated suggested changes in the model, creating multiple iterations of virtual prototypes throughout the DFX stages of concurrent design and process development: CAD and Process Modeling, Concept Layout and Process Tool reviews, and Final Design and Manufacturing Process reviews. At each stage, the designers conducted DFA and DFS analyses of the evolving model. Keup notes, “By using DFA throughout the design process, we found ways to streamline the product still more.”

For example, during virtual prototyping, process engineers identified assembly points that would cause manufacturing bottlenecks, and the design engineers focused on redesigning those areas. In this way, redesign for reduced part count and assembly time also resulted in the greatest time savings on the production line.

Innovations in the new Optiframe design included a hard drive bracket that rotates out of the chassis for drive assembly and a patented hinge-lock mechanism on the stand for the power supply. The hinge-lock allowed a supplier to install the power supply ahead of time, which cut Dell’s assembly time nearly 40 percent.

Another time-saver for assembly is the patented hook-and-lock retention system for the mother board. The system requires only one fastener, a vast improvement over most competitors’ designs.

But it’s the tool-less cover, a product of DFS analysis, that provided the greatest reductions in disassembly and reassembly times. “The cover rotates up and allows a customer a top-down look into the chassis within ten seconds, without his needing to use a screw driver or any other tools,” Brad points out. “That dramatically reduces the MTTR, or Mean Time To Repair, which is the number that our service contractors use to set charges for Dell.”

Reducing assembly steps resulted in increased quality. By keying connectors to parts so only one fit into the assembly is possible, the number of “touches,” or human assembly motions, are reduced, and quality rises. “The fewer the touches, the lower are Defects Per Hundred Units,” Keup says.

The Results

The final design for the Optiframe S exceeded the original design goals:

  • Mechanical assembly time was reduced an average of 32 percent, exceeding the goal of 25 percent.
  • Purchased part count was reduced 50 percent, exceeding the goal of 17 percent.
  • Screw type count met the overall 67 percent reduction goal. Screw min/max count was reduced 55 percent, exceeding the goal of 20 percent.
  • Average service time was reduced 44 percent, exceeding the goal of 25 percent.

These reductions resulted in substantial gains in productivity for Dell. Throughput per hour per square foot in the factories increased from 0.009 units/hr/ft2 to 0.016 units/hr/ft2, a 78 percent improvement. Throughput per hour per direct labor operator increased from 1.67 units per operator to 3.07 units per operator, an 84 percent improvement.

The Future

Dell is already looking at the next generation of Optiframe. The 1999 Optiframe S will be the benchmark for the next generation Optiframe, which will demonstrate significant improvements in assembly and serviceability over the present design.

A previously optimized assembly can make a tough benchmark for additional optimization. “Screw count was important to the previous design, but this one already has no screws,” Kemp points out. “You can’t reduce zero any further.”

Even so, the design engineers are looking forward to the challenges of the next design “With DFA and DFS, we’ve been able to set aggressive goals for next year,” Keup says. Dewey concurs: “We’re extremely proud of this year’s Optiframe design, but I’m already getting a warm feeling when I think about what the product will be like next year.”

LEAN CUISINE Whirlpool Sweden puts DFA to work to cut parts by 29 percent and assembly time by 26 percent.

Whirlpool

Training is a valuable investment, but companies don’t expect immediate profits from the training sessions themselves. At Whirlpool Sweden in Norrköping, though, this was exactly the case as they trained cross-functional teams of in-plant personnel to perform design for assembly (DFA) analysis.


Training Objectives

Whirlpool Sweden in Norrköping is a Whirlpool Global Technology Center for microwave ovens. It manufactures over one million ovens a year. Whirlpool Sweden produces several platforms of microwave ovens, with varying options on some of the base models. The company’s primary market is Europe, though there are some overseas sales as well.

The plant produces 5000 microwave ovens a day. Assembly is almost exclusively manual. Because the European market favors simple, low-cost appliances, oven designs must be as lean as possible, and assembly of the final product rapid and inexpensive.

A DFA training program at Norrköping was initiated with two goals. The first was to teach two teams of eight people each how to perform DFA analysis using specialized software. The cross-functional teams included mechanical and electrical engineers, microwave technicians, designers for air ventilation systems, and personnel directly involved with production and assembly.

The second goal was to reduce in-plant costs and generate a lean oven design by using a proposed production model as part of the DFA teaching process. Cost reduction was expected to come from reducing the number of parts in the oven assembly and from reducing overall assembly time as a result of integrating parts and simplifying or eliminating assembly processes.

Whirlpool Europe’s tool of choice for DFA analysis is DFMA®, or Design for Manufacture and Assembly, cost and design analysis software from Boothroyd Dewhurst, Inc. (BDI), Wakefield, R.I. The software guides design teams through a systematic analysis that enables them to benchmark existing designs and propose new designs that consolidate parts and eliminate assembly difficulties. In so doing, the teams reduce assembly time, increase assembly efficiency and quality and produce lean, functionally integrated designs.

Stefan Wohnhas, a DFA champion at Whirlpool, says, “It was important for us to develop a DFA implementation strategy with a target, a running project, in order to train people in using DFMA as a standard tool in the product development process.” His object was to focus the design teams on existing products and ongoing design projects so that participants would see measurable benefits from learning and applying the DFA methodology and the DFMA software.

Training Focus

Fortunately, there was an ideal project in-house that presented both a new design and a benchmark. Whirlpool Sweden was planning the introduction of a new oven. Since the new oven will replace Whirlpool’s current microwave oven, the VIP 20, the decision was made to benchmark the VIP 20 and use DFA analysis to create the design for the new oven.

The Whirlpool VIP 20 is a 900-watt microwave oven with the cavity-ceiling quartz grill option. Controls are mechanical (pushbuttons and dials), which are preferred in the European market over touchpads and digital displays. They include seven presets, a twenty-four-hour clock, and a ninety-minute timer. The quartz grill allows for the grilling of meat in the oven. In 1997 the VIP 20 represented nearly one-third of Whirlpool Sweden’s total oven production.

The overall assembly of the VIP 20 shares characteristics common to most microwave ovens. The outer chassis is bent and clinched sheet metal. The sheet metal cooking cavity, fitted snugly to the chassis, reflects the microwaves during cooking. A hinged door, either side-opening or drop-down, includes a perforated steel wall that reflects microwaves but allows users to view the food or beverage being cooked. The cooking cavity includes a lamp for observation.

Most of the oven components are mounted in the area behind the control panel, in the pocket between the cooking cavity sidewall and the chassis sidewall. The microwave source is a magnetron, which in turn receives power from a transformer. Screws and other fasteners hold components in place. Cabling connects the components, nearly all of which are electrical, and fasteners guide the cabling and hold it in place.

Redesign of a microwave oven presents a number of design and assembly challenges. To save kitchen counter space, the footprint of a microwave oven is as small as possible, but a small footprint reduces the available component space. Inside the oven, the spatial relationship of some components must be maintained. Because the magnetron can reach 200°C during operation, air from a fan must flow across the magnetron and the transformer that powers it. Finally, the fastening and wiring of each component can require inserting tools and hands into this small component space. As more components are installed during assembly, there is less room for tools and hands to maneuver.

Analysis Tools

DFMA software was not the only tool available to the DFA teams. To benchmark the present assembly, they reviewed videotape of the VIP 20 assembly line, noting any awkward operations and assembly reorientation. To review proposed design revisions, they used a 3D product modeling program, Pro/ENGINEER. They employed stereo lithography for physical prototype examination. They also created laser-cut steel sheets, which were then folded to the exact shape and size of components. This allowed the designers to physically assess component arrangement and fit.

But it was the DFMA software that focused discussion and analysis. Wohnhas says, “DFA software facilitates communication within cross-functional teams, providing fact-based data that is easy for everyone to understand and verify.”

This communication resulted in discussions that were eye-opening for the design engineers. “The early involvement of the production engineers was new and very valuable,” says Johan Dahm, a mechanical engineer in the development department at Whirlpool Sweden. “We discovered that, before their involvement, we ran the risk of designing assembly problems into the oven even in the concept phase. Now, the production engineers help us to eliminate assembly problems before the design is finalized.”

For each assembly workstation, the DFA teams prepared a performance matrix that evaluated each assembly operation for ease or awkwardness. Is the assembler taking advantage of gravity or fighting it? Is assembly conducted in the open, or in a constricted area? Does an operation require reorienting the assembly? How much assembler time is spent holding components and fastening them in place?

From the evaluations came simple changes to save assembly time. In the original design, for example, the torsion spring for the drop-down door was fastened to the bottom of the door, underneath the chassis. In order to fasten the spring in place, the worker had to turn the chassis over, fasten the spring, then turn the chassis back over for the next operation. This slowed the assembly line. Moving the spring fastening point so that it was accessible from the front of the oven eliminated the need to reorient the chassis.

The rest of the door assembly was a target for parts reduction. Dahm says, “Formerly, fixing the door hinge to the oven chassis required fasteners. Now we use the line where the cavity and chassis fit together as a fastener that automatically holds the hinge in place.” The new design also reduced the tolerance build-up between the fasteners, the chassis, the cavity and the hinge. In the long run this will reduce wear and improve product quality.

In this and in other instances, the data from the DFA enabled the team to make decisions which simplified assembly. Gradually a new oven design took shape. Dahm says, “The main advantage of the DFMA software is that you can structure a bill of material and easily see the number and type of parts, how they are assembled and in what order. You get a good picture of the product you are building before it is even a prototype.”

Open Assembly, Lean Design

Based on its DFMA analysis of the VIP 20, the DFA teams developed an assembly strategy of “assembly in the open.” For instance, the fan and air guide for cooling the magnetron and transformer are assembled separately. They are then mounted along with all the other components, either on the outer oven chassis or on the oven cavity. These two large sub-assemblies are docked later in the assembly line. Dahm points out, “This gives technicians more workspace and accessibility, and we receive savings both in time and in product quality.”

The design teams also identified unnecessary fasteners for the cabling, and reduced the number of parts still further by standardizing the cabling itself. Lastly, they set a goal for future product models of having as few unique components as possible, thereby reducing in-plant inventory and MRP costs.

Results

The proposed design for the new oven was leaner than that of the VIP 20 despite adding a forced convection feature. Whereas the VIP 20 had 150 parts in its original design, the new oven has only 106, a 29 percent part reduction. That alone would reduce assembly time, but because many of the eliminated parts were fasteners, assembly operations were further reduced. In addition to the lean design, the open-assembly strategy and the elimination of reorientations will lower assembly time an estimated 26 per cent when the product line comes out in late 1999.

The Norrköping plant won’t have to wait to see cost benefits. It has already implemented the standardization of cables on present product lines. A secondary benefit of the DFA sessions was the standardization of chassis on the different product lines. Improved assembly methods and parts reduction have thus far produced savings greatly exceeding expectations. At that rate, payback time for the entire training project was six months.

Whirlpool Sweden also ended up with two trained DFA teams, ready to undertake new design programs. Dahm says, “I think we may take advantage of the cost-reducing feature of Boothroyd-Dewhurst’s Design for Manufacture software on the next project, now that we are familiar with the DFMA software. We also hope to focus on weight reduction, which will reduce shipping and transportation costs.”

Betting on design analysis

Appropriately headquartered in Reno, Nevada, International Game Technology specializes in the design, development, manufacturing, distribution and sales of computerized gaming machines and systems products. IGT’s video gaming machines are instantly familiar all over the world.The company’s main product families are video and mechanical-reel slots. Cabinets for the machines have either upright or slanted fronts, depending on the angle of the LCD screen.

Inside the slot machine cabinet is a large array of electronic and mechanical components that perform the behind-the-scenes work for the games: generating random numbers for virtual reels, controlling the sights and sounds that make play entertaining, and managing and tracking payment in and out. “It’s very busy inside our machines,” Mikhail explains. “As a result, it’s also very crowded.” Numerous fasteners and limited access points can make assembly and service challenging and time-consuming.

For instance, the electrical box mentioned earlier originally contained two PCBs and a number of hard (plug) connectors wired to it and distributed around the box. The sheet metal box itself had multiple mounting points for parts and several electrical ground points where studs had to be attached by hand and ground wires hooked up. Because there were so many parts in a small space, access was tight and assembly operations difficult. “That box was an obvious place to start a redesign,” Mikhail recalls. “In terms of its complexity and our ability to rapidly put a new design in production, it was low-hanging fruit.”

Before the redesign, Mikhail and other IGT personnel began training on their newly acquired DFMA software, which combines two complementary, closely integrated analysis tools: Design for Assembly (DFA) and Design for Manufacture (DFM).

DFA software enables engineers to reduce a product’s complexity by consolidating parts into elegant and multifunctional designs that provide significant cost savings. DFM guides designers through the selection of materials and processes. Early in product development, at the concept stage, the software helps engineers starting with basic shapes or CAD models to cost out alternative materials and processes. The extensive process library and cost models in DFM Concurrent Costing help to identify major cost drivers such as machining time, part handling (set-up) time, materials outlay, or secondary manufacturing processes such as finishing surfaces.

The software was a hit at IGT, right from the earliest implementation and training of a team. An interesting by-product of the design discussions during training was a new level of communication and spirit. “Our design teams have always interacted, with members giving their best input from their specialties,” Mikhail notes. “What was different with DFMA was the level of creativity it drew out. It turned the team training exercise into an environment that was fun to work in.”

Once training was over, it was time to begin a practical, hands-on design exercise.

Hitting the Jackpot with DFMA

dfma-casestudy

International Game Technologies is a worldwide supplier of computerized gaming machines. Using Boothroyd Dewhurst Design for Manufacturing and Assembly software engineers score 40 percent total savings on slot machine part and assembly costs.

Design analysis software helps engineers score 40 percent total savings on slot machine part and assembly costs.

Long before the lights start flickering in the video slot machine corner of a casino, a light bulb has to go on over a design engineer’s head.

And over the heads of an electrical engineer, a mechanical engineer, and a manufacturing engineer as well. “The simple part of a gaming machine is putting in your money and pushing a button,” says Sam Mikhail, Engineering Manager at International Game Technology (IGT). “The difficult part is keeping every other task relating to the machine nearly as simple. Our customers aren’t only the brand-name operators—it’s everyone who touches the machine.” That includes installation personnel, casino employees reloading the cash system, service technicians, players, and others.

Specifications for the gaming system are quite rigorous inside and out. Because it is an electromechanical product, there are EMF and ESD requirements as well as safety standards to meet. Beyond these, there are security requirements imposed for gaming equipment. “Safety, security, quality, functionality and cost reduction are our main design goals,” Mikhail says. In addition, the slot machines must be customizable for a wide variety of games, with new ones coming out every few months.

But the more complex the internal assembly, the more time-consuming assembly and service can be. Unless the design team strives for simplicity, their product can build in costs up front. That can raise IGT’s manufacturing costs—and the customer’s service costs as well. “Bear in mind,” Mikhail says, “Every moment that these machines are shut down they make no money for the casino. Ease of service is extremely important so, for designers here, that’s a constant challenge.”

To meet that challenge, and coupled with his experience in implementing DFMA programs in a number of companies, Sam Mikhail led the initiative to implement DFMA in IGT, turning to Design for Manufacture and Assembly (DFMA) analysis software from Boothroyd Dewhurst, Inc. (Wakefield, RI), as an integral requirement that complemented the workshops that emphasized DFMA principles and techniques.

Through the use of the software, one of the design teams in the pilot workshop managed to pare 30 percent off part cost and an impressive 50 percent off assembly times and costs for a critical electrical box used in many of their machines.