ROBUST DESIGN

1
ROBUST DESIGN MANUFACTURING APPLICATIONS AT
FORD MOTOR COMPANY
Dr. Yavuz Goktas Reliability Technical
Specialist Ford North American Family Vehicles
Quality
1st Industrial Engineering Spring
Conferences Izmir Efes Hotel, Izmir, Turkey May
11, 2001
2
ROBUST ENGINEERING PROCESS-DFSS
Vehicle
System
Sub-system
Component
3
DEFINE
Inputs
Outputs

• HISTORICAL DATA
• - Campaign Actions
• - Quality History
• - Satisfaction Surveys
• - Lessons Learned
• WANTS DATA
• - Customer
• - QFD
• - Kano Analysis
• Regulatory
• Requirements
• GENERIC DATA
• System Design Specification (SDS)
• Worldwide Customer Requirements (WCR)
• - Benchmarking

Capture the Voice of the Customer

• List of Critical to
• Satisfaction
• Characteristics (CTS)-Ys
• Campaign Prevention Plan
• Program Specific SDS
• Design Assumptions
• High Priority Systems
• Targets

4
CHARACTERIZE
Outputs
Inputs

• Functional Mapping
• VDS/SDS Interfaces
• Brainstorming
• P-Diagram
• D.O.E.
• CAE Models
• FEA
• Real World Usage Environmental Profile
• Generic Design FMEA
• Supplier Quality History

System Design Functional Mapping

• Relate CTSs (ys) to
• CTQs(xs)
• System DFMEA
• Component DFMEA
• Functional Targets
• Updated P-Diagram
• Noise Factor Management
• Strategy
• High Impact Supplier List

5
OPTIMIZE-MANUFACTURING
Inputs
Outputs
Design for Producibility

• Model Process
• (process flowchart)
• Process Data
• Gage RR
• Process FMEA Generic
• Critical to Quality
• Characteristics(CTQs-Xs)
• Characteristic and
• Correlation Matrix
• (Ys Xs)
• Process Capability Model

• APQP Assessment
• - Program PFMEA
• - Flow Diagram
• - Control Plan
• Process Capability for Xs

Design for Producibility
6
OPTIMIZE-DESIGN
Inputs
Outputs
Design for Robust Performance

• PARAMETER DESIGN
• - P-Diagram
• D.O.E.
• Optimization
• TOLERANCE DESIGN
• Customer Usage
• Environmental Profiles

• Design Verification
• Plan
• Quantitative
• Assessment
• Engineering
• Specifications

7
VERIFY
Inputs
Outputs

• Engineering Specifications
• Design FMEA
• Customer Duty Cycle
• Environmental Profiles
• Noise Factor Management
• Strategy
• Design Verification Plan

Test and Verify

• Design Verification
• Plan Report

8
CRITICAL DESIGN PARAMETERS

• COST
• QUALITY
• TIMING
• WEIGHT
• PACKAGING

9
COST REDUCTION EFFORTS IN ROBUST ENGINEERING
PROCESS
CASE STUDY 1 COST REDUCTION ROBUSTNESS
STUDY IN THE DESIGN OF A NEW COMPOSITE NYLON
INTAKE MANIFOLD
CASE STUDY 2 A 6 SIGMA APPLICATION ON A
EUROPEAN VEHICLE LINE FOR NOISE REDUCTION IN THE
PASSANGER CABIN
10
COST REDUCTION ROBUSTNESS STUDY IN THE DESIGN
OF A NEW COMPOSITE NYLON INTAKE MANIFOLD
11
OPPORTUNITY DESCRIPTION
The conversion of cast aluminum intake manifold
to glass-reinforced nylon for COST and WEIGHT
improvements has uncovered high frequency
radiated noise sources in the air intake system.
The objectionable noise was described as hiss
noise that can be easily mistaken for engine
vacuum leak. The team believes that hiss noise
can easily mis-lead dealers for mis-binnings
in Warranty which in turn can increase total
WARRANTY COST. When no vacuum leak was discovered
and the noise was traced to the intake manifold,
a cross-functional team was formed to address to
resolve this problem.
The goal of the team was to identify the causal
factors contributing to the hiss noise and
concentrate on implementing a robust,
financially, technically and timely feasible
solution
12
TEAMWORK
A cross-functional team was set-up to resolve the
intake manifold hiss noise phenomenon.
13
PROCESS IMPROVEMENT METHODS USED
14
PROBLEM RESOLUTION PROCESS FLOW CHART
SET-UP TEAM
DEVELOP CAUSE EFFECT DIAGRAM
PLAN DOE
PERFORM DOE
ANALYZE DOE
CONFIRM DOE
IMPLEMENT DESIGN
PERFORM BENCHMARKING
15
CAUSE AND EFFECT DIAGRAM
INTAKE MANIFOLD
THROTTLE BODY
T-Body Hole
No Tape
Rib Inserts
Yes
Taped
Plenum Surface Texture
Nominal
Sharp
None
Sharp
T-Body Plenum
Thickness
Rods Rounded
Blended
50 thicker
HISS NOISE
IACV Location
Smooth
Direct
Remote
IACV Channel Surface
yes
Diffuser
Tapered
no
Yes
IACV
No
Scoop
16
SCREENING DOE
The goal of the screening DOE was to identify
significant design parameters contributing to
intake manifold hiss noise and carry out further
robustness studies to recommend design actions
to minimize/eliminate hiss noise.
17
FACTORS LEVELS OF DOE
18
INTAKE MANIFOLD
19
THROTTLE BODY PLATE
20
HONEYCOMB DIFFUSER
21
IACV MOUNT LOCATION
22
T-BODY/SCOOP INSERT/IACV
23
DOE TEST MATRIX
24
QUALITY CHARACTERISTICS
1. Subjective Evaluation A
jury of 10 engineers aged between 20 to 50 years
old and came from different fields were asked to
listen recordings of noise and rate their
preferences based on the following table
25
2. Objective Evaluation
Overall sound pressure level with a pre-set high
frequency bandwith is determined to be an
appropriate index to represent hiss noise. Two
high frequency bandwiths as 6kHz-16kHz
and 8KHz-16KHz were chosen as the noise indices
for hiss noise due to strong correlation between
subjective and objective measurement of the hiss
noise.
26
EXPERIMENT AND DATA ACQUISITION
27
DATA ANALYSIS
The significance of the nine main factors on hiss
noise was determined statistically by using
General Linear Model(GLM) procedure of the
statistical package MINITAB.
Analysis of Variance and Main Effect plots were
utilized to draw conclusions regarding
significance of the nine main factors on hiss
noise
28
DATA ANALYSIS (continued)
ANALYSIS OF VARIANCE P-Values
29
DATA ANALYSIS (continued)
MAIN EFFECTS PLOT SUBJECTIVE MEASURE
Decision Criterion The bigger-the better
30
DATA ANALYSIS (continued)
MAIN EFFECTS PLOT OBJECTIVE MEASURE (6K-16K Hz)
Decision Criterion The smaller-the better
31
DATA ANALYSIS (continued)
MAIN EFFECTS PLOT OBJECTIVE MEASURE (8K-16K Hz)
Decision Criterion The smaller-the better
32
FOLLOW-UP ROBUSTNESS STUDIES
Two full factorial DOEs were conducted on IACV
location and T-Body to further study their
contribution to the hiss noise.
DOE Matrix for Throttle Body Hole
DOE Matrix for IACV Pattern
33
CONCLUSION

• Statistical analysis using GLM on both subjective
  and objective
• hiss noise measurements concluded with high
  confidence that
• the following factors are significant for hiss
  noise
• Intake manifold thickness
• Throttle Body Hole
• IACV Location
• The team recommended that the following
  factor/level settings be
• Used to minimize the hiss noise
• 50 thicker Intake manifold- Implemented for
  2000 MY
• Holes in the Throttle Body - Implemented for
  2000 MY
• Remote IACV - To be implemented for future
  program
• Honeycomb diffuser - Implemented for 2001 MY

34
RESULTS
AN 11 dB(A) IMPROVEMENT IN THE INTAKE MANIFOLD
HISS NOISE!!!!!!
35
MAINTAINING THE QUALITY IMPROVEMENT
A. Lessons Learned/Awareness

• Two presentations given to EDQR during the hiss
  noise resolution process
• A presentation given at PT NVH PAT
• The full report of the project is included in
  the Ford Web at
• http//www.poee.ford.com/VEE/doc/Components/B/In
  take/news/intakdoe.html
• Published at the 99 International SAE
  Conference (Ref. 1999-01-1228)
• The team is in the process of submitting the
  full report of the project to the
• Ford technical Journal
• Shared the findings of the project with Puma
  Diesel Engineering

36
B. Benchmarking