Data processing: measuring – calibrating – or testing – Measurement system – Dimensional determination
Reexamination Certificate
2002-01-14
2004-08-03
Shah, Kamini (Department: 2863)
Data processing: measuring, calibrating, or testing
Measurement system
Dimensional determination
C702S097000, C073S597000, C073S598000, C073S801000, C700S090000, C324S230000
Reexamination Certificate
active
06772092
ABSTRACT:
FIELD OF THE INVENTION
Statistical process control charts track coating processes by plotting data over time. C
pk
is a statistical index that indicates whether a process will consistently produce coated units with a film having a thickness that is centered within tolerance limits. This invention relates to a novel computer-implemented process that anticipates the effect of C
pk
for optimizing coatings usage and costs by reducing variability in film build thickness. The process minimizes the environmental impact of volatile organic compounds from the overspray of finishes being applied to painted parts. Improved quality is achieved by analyzing obvious non-random patterns of film variability. The system uses software that refers to the Industry and/or World Class Standards for C
pk
′s. Optimum film build range and average values selected from the process and/or the coating applicators' manufacturer guidelines are used for calculating the impact of changing coating thickness on all finished surfaces in terms of paint usage and costs.
BACKGROUND AND SUMMARY OF THE INVENTION
Manufacturers of painted products, such as automotive bodies and/or furniture companies, assess paint film build thickness by various measurement tools. These tools include the Elcometer, the Pelt Gage or a Wet Gage. These devices measure film build thickness at specific points on the painted unit. Data from these measurements are then downloaded into a commercially available software database. Numerous statistical process control (SPC) and trend charts are generated from this data. These control charts include {overscore (X)} (average thickness vs. time); R charts (range of thickness vs. time) [APPENDIX A], and Individual Moving Range Charts [APPENDIX B]. Appendix A charts the film thickness average readings of a paint coating, taken Oct. 10, 2000 to Mar. 21, 2001, in millimeters. Appendix B charts the film build average values on automotive bodies taken Oct. 10, 2000 to Dec. 21, 2000.
Control limits are defined as a line (or lines) on a chart used for evaluating the stability of a process.
Typical control limits are plus or minus three standard deviation limits using at least 20 data points. When a point falls outside these limits, the process is said to be out of control.
Additionally, schematics of the painted unit [APPENDIX C] can be prepared which highlight or animate the painted surface areas' compliance with either material and/or engineering coating specifications.
Process engineers review these charts and make corrective changes to the automation equipment, and any manual application equipment applying the coating. Considerable expenditure is spent on paint automation equipment and manual spraying techniques to ensure that the highest quality finish is produced at the lowest possible cost. This cycle repeats itself daily in many coating industries. Engineering reviews SPC charts and/or trend charts, and then adjusts automation based on historical data.
A statistically controlled condition exists when all special causes of film thickness variations have been eliminated with only common causes remaining. A “common” cause is a source of variation that affects all the individual thickness values of the process. An SPC chart that is described as “in statistical control” possesses data values that neither surpass the control limits of the charts, nor possess non-random patterns or trends within the control limits.
Engineers refer to Process Capability indices (C
p
and C
pk
) generated by the SPC charts to evaluate the total range of a process's inherent variation. C
p
is defined as a measurement of the allowable tolerance spread divided by the actual 6&sgr; spread data. C
pk
has a similar ratio to C
p
but considers the shift of the mean film thickness relative to the central specification target.
Industry groups have set different control targets for what they consider capable processes. Raw data in software statistical control charts generate C
p
and C
pk
data. However, one manufacturer may consider the process in control when a C
pk
of 1.33 is obtained, while another may seek a C
pk
of 1.5. The Coating Applications Industry does not have a way to quickly identify the optimum statistical data from process data that will produce the given industry standard for a given C
pk
value. Control Charts currently are used only for tracking purposes rather than control purposes.
Previous attempts at controlling processes within the Coatings Application Industry with only SPC (Statistical Process Control) charts suffer from a number of disadvantages:
(a) Engineers cannot calculate the minimal material usage for each painted surface area by increasing the process C
pk
′s using commercially available SPC charts.
(b) Engineers cannot calculate the cost savings that can be realized for each painted surface area by increasing the process C
pk
′s using commercially available SPC charts.
(c) Engineers only use SPC Control Charts for tracking rather than controlling purposes.
They must wait for a preset number of entries within the database before generating C
pk
, rather than selecting a lesser number of current values and quickly determining its impact on C
pk
.
(d) Most plant personnel do not realize that they can increase quality to industry standards, yet realize no savings in raw material usage and costs.
(e) Manufacturing facilities typically employ numerous types of coating applications as well as different colors among coatings. Substantial engineering and labor time is allocated for each manufacturing facility to analyze SPC charts that originate from the following categories:
1. each surface area measured;
2. each painting booth within the plant;
3. each style of manufactured product that is produced;
4. each color group or individual color that is used on the painted product; and
5. specific time frames where process improvements or evaluations are being conducted.
(f) substantial coating waste occurs because non-random trends are only identified a substantial period after the coating is applied, thus requiring excessive coating reworks of the painted products.
Typically, a coatings engineer will review an SPC chart, but fail to note the impact that continuing what is noted only as a temporary optimum trend, will have on the process in terms of savings in material usage and costs.
Appendix B highlights this point. This is an actual example from a coating process. Factors causing optimum Range value numbers: 7 (10/31), 8 (10/31), 12 (11/20), 13 (11/23), 14(11/27) and 20 (12/21) are observed but not scrutinized. The respective range values are: 0.01, 0.02, 0.03, 0.04, 0.01 and 0.03. Current commercially available statistical software packages do not correlate the effect that continuing an optimum range will have on reducing coating material usage and costs. At this one surface area, over a two month period (October 10 through December 21), the process was able to produce six optimum ranges which average 0.023 mil. versus an overall range average of 0.21 mil. for the same time period. Range values of coating thickness differ from unit to unit close to a magnitude of ten.
Standard prior art operating procedures continue tracking the data until a shift in the process data warrants the calculation of new control limits. Calculating new control limits usually require justifying to management why new control limits are necessary. Factors that warrant the calculation of new control limits, that are normally considered using the prior art method, include:
1. Waiting for a trend of seven consecutive points moving in the same direction, either upward or downward which indicates a gradual change in the process.
2. Waiting for seven points above the central line which indicates that the center of the normal distribution has started upward.
3. Waiting for seven points below the central line which indicates that the center of the normal distribution has started downward. Source: AIAG (Automotive Industry Action Group) Statistical Process Control (SPC) Re
Chandler Charles W.
Shah Kamini
LandOfFree
Method for generating coating film build usage and cost... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for generating coating film build usage and cost..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for generating coating film build usage and cost... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3301333