Evaluating Process Monitoring Charts

To be useful, you must be accurate with a control chart and you must analyze it at least daily.

PLOT PATTERNS.—Plots, or points, on a control chart should be thought of as patterns or groups and not as individual points. It is not enough to know where a plot is; you must also know how it got there. In process monitoring, you must be able to recognize patterns that indicate when a process is moving toward an out-of-control condition. After a value is plotted on a chart, the point is connected to the previous point with a straight line. This provides a graphic representation of variations in the process and whether the desired processing conditions are being maintained.

TRENDS.—A drift of plotted points either upward or downward, away from the established mean, with no sudden change in direction is called a trend pattern. A trend usually consists of at least five plotted points. An upward or downward trend usually indicates over- or underdevelopment, respectively. The processed images will be either increasing or decreasing in both density and contrast. A trend that is gradual may be an indication of too much or too little replenishment.

JUMPS.—A plot point that jumps or suddenly moves away from previously plotted points may be caused by contamination of the chemistry, improperly mixed chemicals, or mechanical breakdowns, such as replenisher systems or thermostats. A jump pattern is likely to occur after a process has been shut down, especially if the process has not reached the proper operating temperature.

RANDOM PATTERN.—Whereas trends and jumps must be analyzed to determine their probable cause and corrective action taken when necessary, a random pattern within control limits indicates that the process is in control and is not moving toward an out-of-control condition. When an in-control random pattern is maintained, solution strength is probably normal and no correction is necessary; however, when

Figure 2-15.—Control chart interpretation.

a random pattern includes a run, the process may need corrective action.

Correct Diagnosis

Sometimes two or more things may go wrong with a process at the same time and it is not clear which of several probable causes is resulting in trend, jump, or run patterns. In this case, each of the probable causes must be investigated until the cause(s) of the undesirable condition is/are found and eliminated. When there is more than one probable cause, you should start with the easiest one to correct, or the one that is most likely to occur.

When there is a sudden jump in the plot, look for factors that occur rapidly, such as changes in chemicals, temperatures machine speed, and so on.

When a plot falls outside of the upper- or lower-control limits, follow the steps given here, in the order of precedence, until you find the cause and correct it:

1. Check the appropriate instrument, such as the sensitometer, densitometer, pH meter, and hydrometer, to be sure it is calibrated and being used correctly.

2. Reread the sensi-strip or control strip densities or chemical sample to eliminate reading error as a cause of out-of-control plots.

3. Recheck the calculations and the plotting of values.

4. Ensure the processing temperature, agitation, and times are correct.

5. Review the chemical mixing and/or replenishment records to determine whether there is a discrepancy.

6. Prove or disprove the value of the questionable out-of-control plot by immediately running another control strip or taking another pH or specific gravity reading, as appropriate. This should eliminate the possibility of one improperly exposed, processed, or handled strip from indicating a problem in an otherwise normal process. When the values of this second reading are within limits, you can assume that the first or out-of-control value was the result of irregular or random conditions, and resume processing. On the other hand, when the second reading also gives an out-of-control value, stop processing production work until the problem has been found and corrected.

Monitoring black-and-white film processing is not difficult. A good-quality assurance program that relies heavily upon process monitoring can enable you to do the following:

Detect problems before they become serious.

Test processing solutions and evaluate their usefulness to determine when they need to be replaced.

# Maintain a continuous record of the process.

• Establish correct replenishment rates.

0 Eliminate processing variables.

0 0

Post a comment