User:Medelen8/ENES100/Example Design Report

Problem Statement
The project is to redesign and/or replace the existing AFSM100 mail-injection shock-absorber assembly, which is not preventing bucket failure.

Requirements for each element or component derived from system level goals and requirements
The system level goals and requirements are documented in the CDIO Conceive report. These requirements drive the individual requirements for each component of the shock absorber assembly. Some of the requirements apply to all components, while others are applicable only to certain components. System level requirements (copied from Conceive report):
 * 1) reduce shock load on buckets
 * 2) limit mail bounce
 * 3) must not damage mail (damage rate < 0.1%)
 * 4) required maintenance with max frequency of once per year
 * 5) durability to last 30 million injections (impacts) per year for 10 years (300 million total impacts)
 * 6) retrofit installation time less than 2 hours
 * 7) assembly cost and complexity similar to existing assembly
 * 8) all of the above must be satisfied for all types of mail (TV Guide, newspapers, thin envelopes, heavy magazines, plastic-wrapped, etc)

Most of the above requirements apply to the assembly as a whole, or to all components of the assembly. Requirements 1, 2, 3, and 4 tend to apply more specifically to the system components that interact directly with the mail. In particular, any moving parts (dampers, springs, shafts, etc) must be designed with the durability requirement in mind. Urethane components (pads, rollers) are sensitive to the shock, mail bounce, and mail damage requirements.

Alternatives in design
Several design concepts were originally considered for this project. These original concepts are documented in the Conceive report. After additional brainstorming and concept selection, the following three design concepts emerged as the primary alternatives. These three concepts will be tested prior to the selection of one final design.

Concept 1: Urethane Wear Strips
The first concept is a minor modification to the existing design. In this concept, the existing impact pads (stainless steel on rubber) will be replaced by a custom polyurethane elastomer. Using a local urethane manufacturer, the polyurethane material can be customized to obtain the desired properties (durability, energy absorption, low elasticity).

Concept 2: Urethane Roller
The second concept features polyurethane "compliant rollers" that are free to rotate. Injected mail pieces impact the compliant roller, causing it to compress and rotate. As the wheel rotates, successive mail pieces strike the roller at different positions around the perimeter of the wheel. This behavior should increase durability of the impact surface, which extends around the circumference of the roller.

Concept 3: Pivot Arm
The third concept utilizes L-shaped metal arms that rotate about a pivot point. The rotation is constrained by a spring-loaded shock absorber. The combination of the spring and the weight of the arm return the arm to its original position after each impact. On the impact surface of the arm, it might be necessary to include a polyurethane impact strip to prevent mail damage.

Notably, all three concepts use polyurethane materials with customized properties. The tailoring of urethane properties is a key step in this design process.

The initial design
As described above, there are three initial designs. Each of these design concepts has sufficient merit to prototype and conduct preliminary testing. Due to the simplicity of the designs, it is feasible to build prototypes and conduct testing while staying on schedule and within budget.

Experimental prototypes and testing conducted during design
For each of the three design concepts, a single prototype unit was fabricated. Concept 1 was a simple prototype, since most parts could be used "as is" from an existing assembly. The old impact pads are removable, so these pads were removed and replaced with custom-manufactured polyurethane strips of the same dimensions. Concepts 2 and 3 are new designs, but the base plate and mounting bracket are almost identical to those used on the existing design. Prototype drawings were developed for all new parts. These drawings were sent to a local machine shop for fabrication, which took approximately three weeks. Parts were received, inspected for quality, and assembled into each of the three assemblies. The preliminary tests were derived from the "System performance metrics" defined in the Conceive report for this project. The test setup aims to measure two key metrics: peak impact force on the bucket and max height of mail bounce. The third performance metric, mail damage rate, is typically a very small percentage and therefore difficult to measure without large volumes of mail. For this reason, mail damage will only be observed anecdotally (none is expected); measuring the mail damage rate will be deferred until the field test of the final design.

Measuring Impact Force

In order to measure the impact force (shock load) on the bucket, a compression load cell was used in the configuration shown. The load cell was mounted in a fixed position, just below the bottom edge of the stationary bucket. By using a stand-alone infeed line, which is already owned by the engineering company, mail could be injected into a stationary bucket mounted just below the injector. When a mail piece is injected, the shock absorber assembly absorbs some of the impact, but the mail piece always exerts a downward force on the bucket. The bucket immediately impacts the load cell, which measures the compressive force as a function of time. The load cell signal was recorded using an oscilloscope, and later analyzed via Excel.

Measuring Mail Bounce

When a mail piece is injected into a bucket, it generally bounces back upward. Sometimes, this rebound is high enough that the mail piece escapes the bucket and lands horizontally across the top of the bucket. Such a scenario leads to subsequent mail jams and lost mail. Using high frame rate video of mail injection and rebound, the bounce height was measured for each prototype. The testing revealed that stiff mail pieces (e.g. envelopes) exhibit the highest bounce.

Test Results

The test results are summarized the following graphs. Note that all three concepts were tested, along with the existing design as a baseline (denoted by "default"). Legend Key: default = existing design, C1 = urethane wear strips, C2 = urethane roller, C3 = pivot arm

As shown in the above graph, all three design concepts reduced the shock loading on the bucket, as compared to the existing design. The improvement was most dramatic (40-50% reduction in force) for heavy magazines, which generate the largest shock loads and are likely responsible for bucket fracture. Concept 3 (pivot arm) provided the best shock absorption. Legend Key: default = existing design, C1 = urethane wear strips, C2 = urethane roller, C3 = pivot arm

Overall, all three of the concept prototypes performed better than the baseline, in terms of bounce height. For light mail, the bounce height was comparable between the existing design and Concept 1. Depending on the type of mail, Concepts 2 (urethane roller) and 3 (pivot arm) resulted in the lowest bounce heights. These tests confirm that the existing design is not an effective shock absorber, which exposes the buckets to significant repeated shock loads and produces high rebounds of injected mail.

Appropriate optimization in the presence of constraints
There were two key design features that required optimization, while abiding by the constraints of the application.
 * 1) For the pivot arm concept (Concept 3), the geometry of the pivot arm needed to be optimized for proper balance.  If the pivot arm is not correctly balanced, then the arm will not rotate back into position quickly enough to receive the next mail piece.  Based on the injection timing, and some rough calculation of the pivot arm's rotation dynamics, the geometry was optimized to ensure that the arm returns to its original position within the time constraints.  Simultaneously, it was important to keep the arm from being too heavy, as this could create too much inertial damping.
 * 2) For the urethane roller concept (Concept 2), the size of the wheel needed to determined.  The wheel diameter needed to be large enough that all mail pieces would impact the surface of the wheel.  There is a range of approximately 2.5 inches in which mail pieces land. If the wheel diameter was too small, then mail pieces near the ends of the landing zone would not strike the wheel.  However, the wheel diameter was limited by the geometry of the guide fins and base plate.  The optimal wheel size was large enough for a sufficient landing zone, but small enough to have clearance from other assembly components.

Iteration until convergence
In addition to the testing described above, there were two areas in which iterative design was required.
 * 1) Attachment method for shock pads - In Concepts 1 and 3, there are urethane pads attached to metal backing.  In order to simplify manufacturing of these parts, the concepts were originally designed with adhesives holding the pads in place.  After preliminary tests, it became clear that the chosen adhesives could not withstand repetitive loading.  In the second iteration, the urethane pads were attached via screws.  This method is more secure, but provides the risk of reduced shock absorption near the fastener heads.
 * 2) Polyurethane formula - In all three of the prototype concepts, polyurethane components are featured.  This material is used in many industrial settings for a variety of applications, but the material is often customized for each use.  The material properties of polyurethane can be tailored to the requirements of each component.  In this case, the urethane needed to be durable, compressible, low-rebound, abrasion resistant, and impervious to dust/dirt/debris over many years.  In order to find an appropriate combination of properties, several urethane formulas were tried.  Multiple different varieties of urethane were used to produce multiple urethane rollers and pads.  A series of tests were used to eliminate certain urethane formulas; some were too hard or too soft, too elastic (high-rebound) or susceptible to cutting/gouging from mail pieces.  Using a trial-and-error approach, the urethane design converged to a material with the desired properties.

The final design
Based on the results of preliminary testing, the "pivot arm" concept was selected as the final design. This concept, while the most complex of the three, is still a fairly simple design with few parts. This concept provides the best shock absorption, but there are concerns regarding the long-term reliability of the shocks themselves. The parts list (excluding fasteners) for the final design is shown below. The concept includes industrial shock absorbers and accessory components from Enidine. The shock is an oil-filled hydraulic damper with 10 mm stroke length. It is non-adjustable, as the customer does not desire regular inspection and adjustment. A side-load adapter will be attached to the top of the damper, since the load from the pivot arm will not be exactly horizontal through the entire length of the stroke. In order to prevent the pivot arms from rotating backward/upward into the moving buckets, a "Stopper Bracket" is included. Small plungers are screwed into this bracket, acting as hard stops against the face of the pivot arms. As an example, the drawing of one key component is shown below:

Technical and scientific knowledge
Damper specification
 * In order to determine the specifications for the shock damper, it was necessary to calculate the maximum impact energy (for a single impact) and the rate of energy absorption (energy per hour). Before calculating these quantities, which depend on the mass and speed of injected mail, tests were conducted to determine the injection speed.  Using high frame-rate video, it was determined that mail is injected at a speed of v = 3.9 m/s.  Given a maximum mail piece mass of 0.68 kg, the kinetic energy is calculated:
 * $$KE=\frac{1}{2}mv^2=5.17 J$$
 * The kinetic energy that must be absorbed in each hour of continuous running is then calculated. During normal operation, approximately two mail pieces are injected every second.
 * $$E_{TC}=KE*\text{mail pieces per hour}=5.17J*2*3600=37200\frac{J}{hr}$$
 * This information was used to determine the specific model (PRO 25) of oil-filled damper, such that the damper is sized for the appropriate loading.

Load cell specification
 * A load cell was purchased for the preliminary testing to measure impact force on the buckets. Each type of load cell is rated for a particular loading magnitude, so the maximum expected shock load had to be estimated.  This force was calculated using the principle of impulse and momentum applied during the short-duration impact:
 * $$F\Delta t = m\Delta v$$
 * Using an estimated impact duration of 10 milliseconds, the average impact force can be calculated:
 * $$F=\frac{m\Delta v}{\Delta t} = \frac{0.68 * 3.9}{0.01} = 265N = 59.6\mbox{lbf}$$

Based on the above calculation, the LC204-100 Omega load cell was selected. The 100 lbf limit significantly exceeds the max expected impact force.

Polyurethane material specification
 * Selecting the appropriate polyurethane formula required an understanding of various material properties (as described above), including hardness, stiffness, ductility, etc. In this application, it was critically important to find a urethane material that was compressible and energy absorbing.  Some materials, like rubber, are highly compressible, but their restitution is elastic.  These materials tend to generate elastic collisions.  In contrast, some urethane compounds absorb energy, leading to inelastic (plastic) collisions.  To minimize mail rebound, an inelastic collision was desired.

Creativity, problem solving, and group decision-making
Creativity was particularly important in the early stages of the design (and in the conceive phase), when design concepts were generated and refined. One specific example of creative design is the urethane roller component. The roller design could be varied in terms of size, material, and internal "spoke" geometry, allowing for an infinite number of possibilities. Perhaps the most interesting design feature was the internal geometry of the wheel, which plays a role in the overall compliance of the roller. The compressibility of the wheel, in response to radial loading, is a function of both the material and the spoke geometry. In this design, several different spoke geometries were considered; a few examples can be seen here.

One example of problem-solving was the question of how to test the shock loads on a bucket. This type of test would be very difficult and expensive to set up under normal operating conditions, since the buckets are moving while the mail is injected. Using a stationary bucket, but with the actual mail-injection assembly, was the first part of the solution. Second, the shock absorber assemblies were mounted under the stationary bucket, and a load cell and oscilloscope were used to capture the impulsive force. This approach allowed for the collection and analysis of quantitative data, even for very short duration (5-20 millisecond) impacts.

The most important decision in this phase of product development was the selection of one final design from among the three protoyped concepts. A weighted decision matrix (Pugh method) was used to compare the three concepts to the existing design, which served a datum. The evaluation criteria are derived directly from the original design requirements. The test results presented above were used to assign scores for the relevant criteria. Weights were assigned based on input from the customer and engineering judgment. All three prototype concepts were rated as improvements over the existing design, with the pivot arm concept getting the highest overall score. This concept was presented to the customer as the final design for implementation.

Prior work in the field, standardization and reuse of designs (including reverse engineering and redesign)
Because this design was intended to replace an existing assembly, some of the existing assembly components could be re-used with little or no modification. For each concept, a handful of newly design parts were design, drafted, and manufactured. However, many of the components were COTS (commercial off-the-shelf), including the industrial shock absorber in final design. Using these standardized parts provides higher reliability and ensures that system performance will meet expectations.

Engineers relied upon the expertise of our vendor for custom urethane components. The vendor has decades of experience in urethane production for a variety of automated material-handling applications, including mail-sorting. This prior work in the field allowed the vendor to provide valuable insight during the design and testing of the prototypes. This video shows one example of a compliant urethane roller used in automated mail-handling.

Modeling and/or Simulation
Not applicable. The design was not modeled (except for 3D CAD models) or simulated. Instead, testing was performed.

Aesthetics and human factors
Aesthetics are not important, as this assembly is not visible and is found in an industrial setting. Human factors are not important on this project, since the assembly does not require human interaction, except during the initial installation and infrequent maintenance.

Implementation, verification, test and environmental sustainability
The plan for implementation is to first verify performance with an extended field test. This test is meant to be a final design verification, identifying any reliability or durability issues, and confirming that mail damage is not a problem. After the field test, any necessary design changes will be made before the submitting the final design for customer sign-off. It is expected that the customer will contract the same engineering company that conducted the design effort for the production of the retrofit assemblies. The engineering company will produce a drawing package and send the drawings to one or more manufacturers. Components will be assembled by the engineering company, inspected, and delivered to mail distribution centers across the country. The customer will coordinate the retrofit installation at over 100 distribution centers across the country.

This design did not account for environmental sustainability, but this is not critical for such a design that has no consumable component parts.

Maintainability, reliability, and safety
In accordance with stated requirements, the design does not require any scheduled maintenance. Operators will inspect the assembly every 6 months for signs of wear, or for shock failure. The reliability depends largely on the the shock absorber component, which are designed for high-cycle industrial use. There are no safety concerns with this design. It is located in a guarded area that cannot be seen or easily accessed. Sufficient safety measures (e.g. emergency stop buttons) are present for the parent mail-sorting system.

Robustness, evolution, product improvement and retirement
There are no plans for product evolution or continuous improvement. The customer desires that no adjustments/improvements take place after the initial installation. The assembly is meant to last for 10 years, per requirements. It will likely be retired in less than 10 years, when the AFSM machines become obsolete.