Vibration DoE
This project is based on reducing the vibrational amplitude that a certain vehicle component experiences in a new application, as it contains a vulnerable element which often breaks after withstanding high acceleration loads. In a unique approach, a Taguchi experimental design was conducted to determine in what manner this component can be mounted to minimize this vibrational load.
Team Members
Maya Kota, Ryan Lin
Affiliation
Bosch, MIT
Motivation
An existing automobile component was tested for a new application (new position) in the vehicle system. It was found that in this new position the vibration loads experienced by the component exceed the component limits. Hence, a study needed to be done to understand the correlation between 3 different variables and the output i.e. vibrational loads on the component
My Role
Design of Experiments, Data Collection and ANOVA
.png)
.jpg)
Defining Parameters
Three parameters of the component placement namely mounting boss thickness, mounting angle, mounting distance from the engine were varied and the effects of each on the vibration loads of the component are measured.
.png)
.png)
Taguchi L9 DoE
To efficiently determine which factors affect vibration with minimal experimentation, a 3x3 L9 Taguchi array was used. This design, requiring 9 experiments testing each factor at 3 levels, balanced cost-effectiveness and data utility.
Vibration Analysis and Guideline Applicability
The analysis assessed whether factors A, B, and C affected vibration loads (2000–3000 Hz) in vehicles with engine A. ANOVA, regression, and residuals analysis were conducted. A t-test compared maximum accelerations across engine types, evaluating mounting guideline applicability.
.png)
Process
Defining Parameters
Three parameters of the component placement namely mounting boss thickness, mounting angle, mounting distance from the engine were varied and the effects of each on the vibration loads of the component are measured.
Vibration Analysis and
Guideline Applicability
The analysis assessed whether factors A, B, and C affected vibration loads (2000–3000 Hz) in vehicles with engine A. ANOVA, regression, and residuals analysis were conducted. A t-test compared maximum accelerations across engine types, evaluating mounting guideline applicability.
.png)
Taguchi L9 DoE
To efficiently determine which factors affect vibration with minimal experimentation, a 3x3 L9 Taguchi array was used.
Results
01
Factor C (mounting distance) showed the most influence, with further distances reducing vibrations.
02
Mounting Guideline Applicability across engine families
A t-test comparing 5 engines with the original engine was performed. One of the engines showed comparable results and the guidelines can be used for that engine.
Mounting Guidelines
The component should be mounted as further away from the engine as possible.
The angle of the mounting should be the component lies at an angle <x° to the direction of the reciprocation of the engine piston.
The mounting boss thickness should be maintained at y mm above the mounting location.
.png)
Future Work
A key lesson learned is the need for more data to improve the regression model. A full factorial DOE, focusing on component distance and orientation angle with at least 5 replicates, is recommended for better accuracy. The Taguchi approach, suited for sparse data, only indicates relative significance and lacks insights into factor interactions. Additionally, the analysis relied on simulated data (ANSYS); validating findings with real-world measurements is crucial for reliability.
New Design
Existing Design
New Design
Results
Cost reduction of 95% and weight reduction of 50% was achieved. The prototypes of the design were built and tested extensively. The process plan has been defined and documented. The supply chain for the materials has been set up. The assembly line has been set up in MIT Building 35. 25 FrEDs have been successfully manufactured and are ready to be deployed to the end users.
Old Design | New Design | |
|---|---|---|
Minimum Extruded Fiber Diameter | ~0.2 mm | 0.15 mm |
Weight | ~10 lbs | 5 lbs |
Unit Costs | $5428 | $269 |
