Slide 1: Organizing a Successful Problem Solving Project
m Organizing a Successful Problem LSL T USL Solving Project Michael R. Amann CQE Metallurgical Consultant September 10, 2019 M R Amann CQE 1
Slide 2: Background
Background ❖ 55 years of experience in metallurgy of steel – 40 in tinplate ❖ Fourth generation steelmaker ❖ Retired Sparrows Point in 2008 – Metallurgical Consultant for Quaker Chemical ❖ Member of the ASTM A01.20 committee on tinplate ❖ Former and current Chair of the AIST Committee on the Metallu
Slide 3: The basic Six Sigma premise
m The basic Six Sigma premise . . . LSL T USL ❖ All processes have variability ❖ All variability has causes (independent variables) ❖ Typically only a few causes are significant ❖ To the degree that those causes can be analyzed and understood, they can be controlled, and therefore, defects reduced .
Slide 4: Understanding the defect
Understanding the defect... ❖ If we can’t measure it, we don’t know much about it. ❖ If we don’t know much about it, we can’t control it. ❖ If we can’t control it, we are at the mercy of chance. M R Amann CQE 4
Slide 5: The key is data-driven decision making
The Six Sigma problem solving process promotes an organized methodology to solving problems This protocol applies statistical tools to solving practical problems, avoiding the shotgun approach The key is “data-driven” decision making M R Amann CQE 5
Slide 6: Getting Started
Getting Started ❖ Problem Solving Project Phases: Define - Measure - Analyze - Improve – Control ❖ Let’s focus on the “Define” Phase M R Amann CQE 6 Used With Permission from GE
Slide 7: Problem Definition Examples
Problem Definition Examples 1. Define the problem: a. What is the problem? b. What does it look like? (Photographic Evidence) c. When did the problem begin? d. Where does the problem manifest? i. If you have multiple lines is the defect on one or all lines? e. How often does the problem occur? f. Do
Slide 8: Other Problem Definition Examples
Other Problem Definition Examples 3. Product data requirements: a. Substrate source/s involved: i. Is one substrate better or worse than others? b. Single reduced or double reduced: 1 CR or 2 CR? c. Continuous Anneal or Batch anneal: CA or BA? d. If BA was the material processed thru a cleaner line
Slide 9: So, the customer comes to you with a problem
So, the customer comes to you with a problem…. Problem = Response Variable M R Amann CQE 9 Used With Permission from GE
Slide 10: Defining the Problem as Y, the Response Variable
In each project, we defined the “Problem” as “Y” … the “Response Variable” ❖ Response (Dependent) Variable “Y” = f (x1, x2, x3, x4, …. xn) ❖ The response variable “Y” is a function of a number of independent variables x1 through xn. ❖ The response variable “Y” can be: ❖ An attribute variable – not q
Slide 11: Problem = Response Variable Y
Problem = Response Variable “Y” “Response Variables” solved using Six Sigma methodology: ❖ Tinplate Surface “Zits” from Eastern Europe (SEM) ❖ “Grainy Surface” black spots (surface carbon contamination) on 2CR tinplate ❖ “Yellow Spots” light intermittent rust (FeO) on surface - very light tin coatin
Slide 12: Summary of the Zit Study
Summary of the Zit Study ❖ Y = Response variable = Presence / Absence of Zits ❖ X1 = Analysis of metal condition inside the open Zit ❖ X2 = Visual appearance of the Zit – normal and 200X ❖ X3 = Density of Zits on plate surface (good vs. bad) ❖ X4 = Presence of adjacent “popped” and “un-popped” Zits
Slide 13: Tinplate Surface Zits from Eastern Europe (SEM)
Tinplate Surface “Zits” from Eastern Europe (SEM) ❖ Inspectors noted grey spots on finished tinplate surface • The spot density is estimated by counting the number of eruption in a given ❖ Cosmetically objectionable image normalized to the measured area. The “Bad” sample indicates about 2X the spot
Slide 14: Popped vs. Un-popped Zits
Popped vs. Un-popped “Zits” ❖ The main factor in definition is the data we were able to acquire on the defect ❖ SEM investigation down into the “popped eruptions” ❖ No significant findings down inside the “volcano” ❖ Digital profilometry found “un-popped artifacts” next to the “popped eruptions” ❖ S
Slide 15: Summary of the Grainy Surface
Summary of the Grainy Surface ❖ Y = Presence / absence of dark linear marks on tinplate noted by inspectors ❖ X1 = Tin coating weight on the tinplate - light tin coating ❖ X2 = SEM analysis of the locus of the dark marks – Fe noted ❖ X3 = Study of marks with tin removed – amorphous carbon noted ❖ X4
Slide 16: Grainy Surface on Double Reduced Tinplate
“Grainy Surface” on Double Reduced Tinplate ❖ “Grainy Surface” was a defect encountered at Sparrows Point on double reduced tinplate ❖ Certain can interior enamel coatings used in the can making process highlighted the condition. ❖ An exchange of steel substrate between Tin Mills confirmed the criti
Slide 17: Iron exposure in the linear marks
First item of data analysis: Iron exposure in the linear marks M R Amann CQE 17
Slide 18: What was making the dark areas?
What was making the dark areas? We needed more data Defining the defect further - lab team members stripped the tin with HCl - the dark areas were analyzed by micro Raman spectroscopy to be amorphous carbon. M R Amann CQE 18
Slide 19: The grainy surface final fix
The “grainy surface” final fix ❖ Where is amorphous carbon coming from? ❖ Rolling oil exposed to elevated temperatures in annealing ❖ The mill rollers on the team suggested we look at “where the condition existed on the strip surface” ❖ The amorphous carbon was on the “mountain tops” of the Data and
Slide 20: Summary of the Yellow Spots
Summary of the Yellow Spots ❖ Y = Presence / absence of yellow spots by inspector ❖ X1 = Where did spots appear – Only on low coating weights (1.1/1.1) after coiling ❖ X2 = When - after some storage time wrapped (48 hours) – then unwrapped ❖ X3 = What do they look like under magnification? ❖ Rusty s
Slide 21: Yellow Spots on finished tinplate
“Yellow Spots” on finished tinplate ❖ Inspectors reported seeing “yellow spots” on the tinplate after 48 hours in coil form CLEAN AREA – NO SPOTS ❖ The team defined the location of the spots (where), and the state of the material (when) they were first noticed. ❖ The yellow spots exhibited mirror im
Slide 22: The yellow spot defined
The yellow spot “defined” In every “spot” location we examined, the spot presented itself as a rust contaminant lying in the “valley” of the strip roughness profile The atmospheric humidity during coiling was being “wrapped into” the laps of low coatings tinplate. This problem only occurs during the
Slide 23: Maximizing Joint Problem Solving Efforts
Maximizing Joint Problem Solving Efforts ❖ Utilize Quaker’s “Basics of Problem Solving” to define the problem – What, when, where, why, how often, etc. ❖ Response Variable “Y” definition is the starting point ❖ Provide a precise description that is easy to understand ❖ Present photographs or actual
Slide 24: Quaker is here to help!
Quaker is here to help! ❖ Using Quaker Houghton’s Six Sigma based problem definition process can save you time and money ❖ We have a team of hands-on experts whose tinplate careers span decades, so it is likely we’ve had a similar experience ❖ Quaker Houghton’s laboratory is capable of the sophistic
Slide 25: Contact Information
Contact Information ❖Michael R. Amann CQE ❖ Metallurgical Consultant ❖ Quaker Chemical Corporation ❖mramann24@verizon.net ❖Cell – +1 (410) 458-8751 M R Amann CQE 25