|
Figure 1: The SamarEye competition version monocopter has a single foamed wing, a thermo-plastic
fuselage, a foam landing gear or foot, a carbon-fiber spar to give the wing rigidity, and a
plastic fan. Figures: Christopher Hockley, Embry-Riddle Aeronautical Univ.
|
Software analysis of design concepts can reduce manufacturing and parts costs.Since the time of DaVinci, inventors and engineers have wrestled with aviation design
challenges by taking inspiration from the natural world. Even today, in the time of jet
aircraft, this way of thinking can solve long-standing puzzles in aeronautics. These two
aeronautics challenges, for example, have defied easy answers:
Aeronautics challenge #1: Build a flying machine that imitates the aerodynamics
of a maple seed—a nature-mimicry design problem that has stumped engineers for the past 60
years.
Aeronautics challenge #2: Fly an unmanned vehicle inside a closed structure—one
of the last remaining frontiers in the growing world of unmanned aerial systems (UAS) and
vehicles (UAV). Now, both of these challenges have been successfully solved with one design—a
biologically-inspired, robotic monocopter.
The aircraft that accomplished these aviation feats was designed and built by a team from
Embry- Riddle Aeronautical Univ. of Daytona Beach, Fla., and was entered in the
19th International Aerial Robotics Competition (IARC) sponsored by the Association for
Unmanned Vehicle Systems International (AUVSI). The craft with the unusual flight-style
placed third in the competition and won an award for “Most Innovative Air Vehicle”. With
thoughts of commercialization for the toy and hobby market, the student-led team brought their
UAV design into a fall-term course on Design for Manufacturing and Assembly (DFMA) software to
analyze the costs of materials and processes in the design stage.
According to Sathya Gangadharan, a professor of mechanical engineering at Embry-Riddle, the
DFMA class is a reality-based course that is geared to bridge the gap between academics and
industry.
“A lot of times when students and practicing engineers do a design, they don’t look at the
practical aspects or cost implications of manufacturing the product,” says Gangadharan.
Students have to select a product that has between 15 and 30 components and then use DFMA to
come up with modified designs. These new designs explore alternative materials and manufacturing
processes which, in the end, allow the teams to preserve or improve features and functionality
while reducing part count and cost.
Graduate student Christopher Hockley brought the competition-tested monocopter to the DFMA
course looking to make improvements. Prior to the course, his team had considered several
designs, ultimately choosing a version inspired by the shape of the maple seed, with a single
foamed wing, a thermo-plastic fuselage, a foam landing gear or foot, a carbon-fiber spar to give
the wing rigidity, a plastic fan, and only two moving parts. (Figure 1).
The maple seed is a type of fruit called samara, meaning winged, one-seeded fruit. This and
the sensor requirements of the design inspired the team to name the aircraft the SamarEye.
Samara-derivative aircraft, as well as monocopters in general, have not gained much attention or
acceptance, so very little existing design information existed. At the time, there were fewer
than ten papers on powered monocopters. As a result, the team was on its own when it came to
design and materials, as well as cost, assembly, and manufacturing considerations.
|
Figure 2. A CATIA model “cartoon view” of the competition SamarEye monocopter before DFMA
redesign.
|
Manufacturing time took approximately 40 man-hours, and hot glue was the fastener of choice.
After the competition, some critical thinking was needed to improve assembly and manufacturing,
and to see the aircraft not as a collection of separate parts, but as an integrated, holistic
system. The DFMA course provided the perfect vehicle for rethinking the design.
Asking the right questions
DFMA is an integrated software suite from Boothroyd Dewhurst, Inc., Wakefield,
R.I., that helps engineers ask critical questions about their product designs early in the
development process. Such an evaluation can have major impacts downstream on the manufacturing,
assembly, and cost of those products.
Design for Assembly (DFA) software guides engineers to simplify a design using queries—such
as whether parts move with respect to one another, or whether they can be made of the same
materials—the answers to which lead to reduced part count and cost. Functional efficiency, fewer
parts, and ease of assembly are the goals.
Design for Manufacturing (DFM) software provides engineers with a structured way to examine
process technology and material choices in order to anticipate manufacturing costs early in the
product development life cycle. Manufacturing knowledge and reduced costs are the payoffs. By
asking the right questions up front, all of the cost-implications of designs can be taken into
account, rather than popping up later after the design has been locked-in.
“With DFMA we’re trying to solve a number of engineering problems,” says Hockley. “By
reducing part count, we’re hopefully going to have a flying machine that not only works better,
but is more weight-efficient, more weight-economical, and structurally stronger. We’re also
looking to make it more repeatable for manufacture. That’s a big thing.”
Students in the DFMA class follow a realistic design process, incorporating computer-aided
design (CAD), finite element analysis (FEA), and tolerancing.
“First, they model the part or product using CATIA (Dassault Systèmes) to
produce 3D views,” says Gangadharan (Figure 2). “Next, they perform FEA and calculate data on
stress, deformation, and frequency constraints. Then they must tolerance the parts.”
Following this baseline design work, students then use the DFMA software modules to evaluate
and refine their designs, using FEA in a feedback loop to prove out the functionality of their
DFMA design changes.
Analyzing the baseline design
The student team’s SamarEye monocopter design was 71 cm long, 184 grams, and had 18 parts, eight
of which were provided by outside sources. Of the remaining ten parts to be manufactured, six
were extruded and four were thermoformed. All were of similar size except for the wing. This
kept manufacturing costs low, as common injection molding and thermoforming equipment could be
purchased, and the only variance was in the dies. Despite the aircraft’s light weight and
comparatively weaker materials, the baseline design was structurally strong and had an ample
factor of safety under typical operating conditions.
Using a projected product life volume of 100,000 and a batch size of 12,500, Hockley ran a
DFMA analysis of the original design and determined that the cost of tooling was $2.55, the
piece part cost was $4.25, and the assembly cost was $7.08 for a total cost-per-product of
$13.88. Despite the fact that the baseline design had relatively few parts and that the
manufacturing methods were already relatively simple, the analysis demonstrated that there was
still room for improvement. Of the ten manufactured parts, three of them—the fuselage top,
fuselage bottom, and the wing—were the most expensive. As for assembly, more than 50% of the
total product cost resulted from this activity, representing the greatest room for improvement
of any design-to-cost variable.
Refining the baseline design
Following the DFMA analysis, Hockley evaluated the design to see where improvements could be
made. In the baseline design, the wing—the single largest piece—was made out of an expanded
polystyrene thermoplastic (used as floor insulation), while the fuselage was made out of a PETG
thermoplastic (used for clamshell packaging).
|
Figure 3. CATIA model of the DFMA redesigned SamarEye monocopter illustrates the simplification
of the design, achieved by combining the wing, fuselage, and main gear into a single
injection-molded foam piece.
|
In the modified design, the team decided to combine the wing, fuselage, and main gear into
one injection-molded polystyrene foam piece, with the main spar and fan housing molded in place
(Figure 3).
“These changes not only removed a number of components,” says Hockley, “but reduced the
number of operations required in assembly.”
With the switch in materials from the stronger PETG to the weaker polystyrene, additional
iterative FEA simulations were required to ensure that the aircraft could withstand all loading
scenarios.
“In the redesign, I cut the parts down from 18 to 13, eliminating five manufactured parts,”
says Hockley.
This consolidation was the result of combining parts that shared materials or did not have
motion relative to one another. Reducing parts and streamlining the assembly process can cut
manufacturing costs in a number of ways, according to Hockley: “It can reduce the number of
molds required for part production, decrease the type and quantity of machinery required, and
simplify the storage of parts on the shop floor.”
Another benefit of parts consolidation is a reduction in part interfaces, which improves
quality by helping eliminate stress at joints and fasteners.
“Consolidation of parts,” Gangadharan adds, “improves FEA performance—an outcome that often
gets overlooked by designers and analysts.”
With commercialization of the monocopter in mind, Hockley is excited by the final DFMA
results for his class project: piece part cost reduction of 25%; overall product cost reduction
of 51%; assembly labor time and cost reduction of 74%; and a grand total savings of 625 days and
$717,000 for a production run of 12,500. Such savings are huge—when you need to keep an eye on
what rings up at the register.
“In the aerospace industry, and more specifically in the UAV market, it is becoming
increasingly important to maximize functionality while minimizing cost,” Gangadharan says. “DFMA
is the perfect tool for accomplishing this.”
“The engineering students’ success in industry depends on how close to reality they are able
to think,” he adds. With the AUVSI competition having demonstrated the SamarEye monocopter’s
navigational skills, DFMA should help the craft’s designers begin navigating commercial markets
as well.
Published in R & D magazine: Vol. 52, No. 3, June, 2010, pp. 34-35.
