Lever 3000 (Load class 2000 Nm) Missing Teeth Segment – Adient

SEOCONTENT-START

Adient –
RESTRICTED
Adient –
RESTRICTED
Improving the experience of a world in motion
Lever 3000 (Load class 2000 Nm)
Missing Teeth Segment
20 March 2019
Adient Proprietary and Confidential
2
Adient – Improving the experience of a world in motion
Adient –
RESTRICTED
Adient –
Lever 3000 RESTRICTED
Recliner – Process Flow & Controls – Lever II
Place Driver assembly
into Guide
Place Teeth Segments
– B into Guide
Greasing of
Teeth Segments
Feed and load
Control Disc
Place Teeth Segment – A
into Guide
Teeth segment
adjustment
Presence check Camera check
Greasing
Process flow & Controls
Area of interest
Assembly Steps
Presence check
– Control Disc
Potential for fall-off
Adient Proprietary and Confidential
3
Adient – Improving the experience of a world in motion
Adient –
RESTRICTED
Adient –
RESTRICTED
Why? Missing front Toothed Segment B
> Why? Toothed Segment B fell out of guide before
assembly of control disc due to collision between
Work piece carriers
> Why? Toothed Segment B was not properly placed
into guide slot during assembly process (Placed in
tilted condition)
> Why? Gripper of pick and place Toothed Segments B
not properly adjusted
Lever 3000
Recliner –Why made?
• Optimization of grippers for pick and place of toothed segments B to reduce chance of
misplacement. implemented on 8 Feb 2019
PCA
Previous system
Updated system
Locator to Guide on Base Plate
Fixed unit without any adjustments to
compensate variation
Additional Units in updated system for
accurate positioning with additional DoF
1) AFAG Module
Additional rotary freedom for proper lead in
2) Positioning Plate
Additional vertical flexibilty for Locator to
touch Guide before Gripper
3) Push-out Fixture
To slide Teeth Segments directly into the
slot and push down by compressing Grease
1
2
3
Adient Proprietary and Confidential
4
Adient – Improving the experience of a world in motion
Adient –
RESTRICTED
Adient –
RESTRICTED
Why? Missing front Toothed Segment B
> Why? 100% online detection of front Toothed Segment
B failed
> Why? Sensor for detection of presence of front Toothed
Segment B not adjusted to detect sloped Toothed
Segment B (sloped Toothed Segments B can fall out of
sub-assembly before next process step / assembly of
control disc)
> Why? Failure of sloped Toothed Segments B not
considered in PFMEA
Segment A Sensor Placement – Difference in height
is significant to be detected
Segment B Sensor Placement– Difference in height is
insignificant to be detected in case of tilted teeth
segment
Lever 3000
Recliner –Why shipped?
• Optimization of sensing device for detection of presence
and position of toothed segments B
a) Diameter of pins enlarged
b) Position of pins optimized to be at the middle
• implemented on 21 Jan 2019
PCA
Adient Proprietary and Confidential
5
Adient – Improving the experience of a world in motion
Adient –
RESTRICTED
Adient –
RESTRICTED
Algorithm:
If W2 > Moving Median of 100 data + 2,43°
=> Suspect
=> Otherwise, OK
Principle : Use increase in W2 angle to rest of the population to identify Outliers as Suspect
All NOK Parts
identified
Safety margin
“Moving Median”
of 100 parts
2,43°
(Based on
statistical analysis)
Limit line
Lever 3000
Recliner – Detection method DW2
1. All parts identified for NOK W2 angle increase were opened and found to have missing segment.
2. Mean shift in W2 angle was able to identify recliners with 3 segments vs 4 segments. W2 is 100% monitored and recorded
Upper Limit Standard (31 deg) (Control limit is set at 30.5deg)
Lower Limit Standard (23 deg) (Control limit is set at 23.5 deg)
Adient Proprietary and Confidential
6
Adient – Improving the experience of a world in motion
Adient –
RESTRICTED
Adient –
RESTRICTED
Detection method – Explanation
Principle : Use increase in W2 angle to rest of the population to identify Outliers as Suspect
W2 – 4 Teeth Segments – OK Recliner
W2 – 3 Teeth Segments
NOK Recliner
4,855° (Mean shift
based on Layout)
+3S (3*0,5405°) = 1,621°
(Max sigma of 10 Campaigns
-3S (3*0,5405°) = 1,621°
(Max 1,6123° sigma of 10 Lots)
Sorting Procedure:
For each lot,
• Sigma = 0,5405°
• Calculate „Moving Median“ for each data point based on
next 100 data
• Median is chosen over Average to improve stability
against Outliers
• Moving Median curve is the Base line
• Offset by 2,43° to define the Limit curve (Offset Factor =
4,5 * Sigma)
• Any data point lying above the limit curve is declared as
Suspect
Explanation – Mean Vs Median
“Moving Median”
of 100 parts
Limit Curve defining
boundary between OK
and suspect parts
2,43° Used as limit curve to
determine suspect parts
= =
Lever 3000
Recliner – Detection method DW2
Mean Shift
NOK variation
Adient Proprietary and Confidential
7
Adient – Improving the experience of a world in motion
Adient –
RESTRICTED
Adient –
Lever 3000 RESTRICTED
Recliner – Detection method DW2
Derivation of Mean Shift in W2
Left and Right Toothed Segment
(shown = Left; Right is 180° rotated)
Upper and Lower Toothed Segment
(shown = Upper; Lower is 180° rotated)
Gap
Gap
W2 angle depends on,
• Variation in functional
geometry of all components
in recliner
• Variations are relatively
constant for particular batch
of production; but, changes
over time with mixup of
batches
Functional principle
4x Contacts between
Excenter to Teeth segment
Excenter is allowed to float to
support all 4 Teeth segments at
the same time
With missing Teeth segment – B,
Excenter looses support at one of four
contact points and has to turn more to
find the next possible support
Contact points
When all other conditions remain the same,
Increase ΔW2: +4.855° (Based on CAD study on Nominal geometry)
Adient Proprietary and Confidential
8
Adient – Improving the experience of a world in motion
Adient –
RESTRICTED
Adient –
Lever 3000 RESTRICTED
Recliner – Detection method DW2
Max Standard deviation 0,5405° is chosen to set the sorting criteria
Derivation of standard deviation in W2
Max 0,54045
W2 – Recliner with 4 Segments – OK
23,1 24,2 25,3 26,4 27,5 28,6 29,7 30,8
LSL 23
Target *
USL 31
Sample Mean 24,7078
Sample N 225
StDev(Overall) 0,404899
StDev(Within) 0,386348
Process Data
Pp 3,29
PPL 1,41
PPU 5,18
Ppk 1,41
Cpm *
Cp 3,45
CPL 1,47
CPU 5,43
Cpk 1,47
Potential (Within) Capability
Overall Capability
PPM < LSL 0,00 12,33 4,93
PPM > USL 0,00 0,00 0,00
PPM Total 0,00 12,33 4,93
Observed Expected Overall Expected Within
Performance
LSL USL
Overall
Within
Process Capability Report for OK
24,0 25,2 26,4 27,6 28,8 30,0
LSL 23
Target *
USL 31
Sample Mean 29,4421
Sample N 225
StDev(Overall) 0,424322
StDev(Within) 0,427133
Process Data
Pp 3,14
PPL 5,06
PPU 1,22
Ppk 1,22
Cpm *
Cp 3,12
CPL 5,03
CPU 1,22
Cpk 1,22
Potential (Within) Capability
Overall Capability
PPM < LSL 0,00 0,00 0,00
PPM > USL 0,00 120,60 132,52
PPM Total 0,00 120,60 132,52
Observed Expected Overall Expected Within
Performance
LSL USL
Overall
Within
Process Capability Report for NOK
Mean : 24,71°
Sigma : 0,405°
N : 225
Mean : 29,442°
Sigma : 0,427°
N : 225
W2 – Recliner with 3 Segments – NOK
CONTROL BUILD for verification
Adient Proprietary and Confidential