Best Paper Award, ASME CIE
Design of a Competitive Self Contained Trash Compactor
Document Actions
Team Members
- Patrick Chang
- Benjamin Lee
- Aditya Shah
- Ryan Stewart
Project Overview
Figure 1. Self Contained Trash Compactor
(Ref. http://www.fac.unc.edu/OWRRGuidelines/images/McCollCompactor2.jpg)
The design decision
will be to determine the characteristics of the hydraulic components (the pump,
cylinders and electric motor). Along
with the hydraulic components, cost is a factor that will be considered and the
objective will be to minimize the cost of the hydraulic system (both component
cost and the cost associated with the power consumption). Below are the design objectives and associated attributes that will be modeled in this project to design a competitive trash compactor. The objectives that are expected to be modeled in this project are represented by an '*'.
|
Objectives |
Attribute
(units) |
Can it be
Modeled in Modelica? |
|
*Maximize Trash Capacity |
Compaction Ratio |
Yes |
|
*Minimize Cycle Time |
Cycle Time (s) |
Yes |
|
*Minimize Operating Cost |
Input Power/Cycle (kW-hr) |
Yes |
|
*Minimize Purchase Cost |
Total Component Cost ($) |
No |
|
Maximize System Efficiency |
η= (Power at Cylinder / Electrical Input Power) |
Yes |
Modelica Model of the Compactor System
Figure 2. Compactor System Model
Modelica Model of the Motor Power Unit
Figure 3. Motor Power Unit Model
Modelica Model of the Hydraulics Unit
Figure 4. Hydraulics Unit Model
Modelica Model of the Trash
Figure 5. Trash Model
Simulation Results
Figure 5. Simulation Results
Uncertainty Analysis
Uncertainty Analysis was done for both the inherently uncertain variables as well as for the uncertain variables in the conceptual design phase. They are Bore Diameter, Rod-Bore Ratio, Pump Displacement, Max System Pressure, Trash coefficient, and Market factor.
A Central Composite DOE was run to obtain the main effects of the variables on the outputs. A Monte Carlo and Latin Hypercube Sample were also performed to find the expected output and standard deviation. Finally, a Sensitivity Analysis was carried out using the Method of Morris. Following are a portion of the results obtained.Central Composite Results for Max Trash Density
Monte Carlo Results for Total Cost (1000 Samples)
Latin Hypercube Sampling Results for Total Cost (500 Samples)
Sensitivity Analysis using Method of Morris
Design Space Exploration
In order to get a better understanding of the overall shape of the design space, the system model was combined with the elicited preferences and simulated using a full factorial DOE. For the DoE, three design parameters were considered: Max System Pressure, Bore Diameter, and Pump Displacement.
DoE Results for the Design Space with Max System Pressure = 2e7 Pa
Solving the Design Problem Deterministically
ModelCenter Setup for the Deterministic Optimization
Optimization of the Utility for the Deterministic Case
Solving the Design Problem Under Uncertainty
In solving the design problem under uncertainty, the expected utility can be estimated by taking
a Latin Hypercube Sampling (LHS) for the uncertain variables.
Optimization of Utility for the Uncertain Case (LHS = 25)
Sensitivity Analysis of the Trash Coefficient
A sensitivity analysis of the trash coefficient, k, on the design optimization was performed. To accomplish this task, k was varied to its high and low extreme values. Then, the design was optimized with these new values of k.
PDF of the Trash Coefficient, k, at its High Extremity (Average Centered Near 350 Instead of 260)
Optimization Results for k = 350
Homework Assignments
All reports are in portable document format (.pdf). You will need Adobe Reader to view them.- Homework Assignment 2 - Planning your Simulation Based Design Study
- Homework Assignment 3 - Energy-Based Systems Modeling with Modelica
- Homework 3 Files (for Fluid Power Library contact Dr. Chris Paredis)
- Homework Assignment 4 - Uncertainty Analysis
- Homework 4 Files
- Homework Assignment 5 - Preferences Modeling and Optimization
- Homework 5 Files
Forgot your password?
