DYNAMIC LOCAL-E EVALUATION

DYNAMIC LOCAL-E (GRADE INDICATOR) EVALUATION

It should be a key aim of any MSG control system to ensure that the dynamic local-E measured by the grading machine and the static local-E measured at the same positon correspond with one another within a certain maximum error band.

Tadpole provides a number of techniques to enable these dynamic/static local-E relationships to be monitored. Three of these are illustrated below.

(1) Calibration stick runs - comparing dynamic with static values

Calibration sticks are sticks that have been measured for static local-E at various positions and the data entered into the database. The static data is retreived and the stick is then run through the grading machine. The dynamic values are then compared to the static values.

The chart in the next Figure shows how well the dynamic local-E correlates with the static local-E for a single calibration stick. The continuous line showing 27 points displays the dynamic readings taken by the grading machine. The 5 static local-E measurements recorded for the stick are shown as a range, in this case 5% above and below the actual value. The chart shows that the maximum difference between the static values and the dynamic values for the five points is 3%.

As mentioned before a calibration stick is more than often a specially selected production stick that has been tested in a static testing machine to determine the local-E at one or more points. For production use, only certain sections on the sticks with the lowest local-E are normally measured, resulting in about 5 points being measured in total on each stick.

The operator also records a 'standard dynamic profile' for each calibration stick at a time the grading machine is known to be measuring accurately and this is also displayed on the chart shown above. The 'accurate' dynamic profile can then be compared to the current dynamic profile and any significant difference indicates a calibration shift which can be immediately investigated.

(2) Production samples - comparing dynamic with static values

Production samples are selected randomly in order to evaluate the compliance of specific batches of timber to the grade specification. During the processing of these samples additional checks are carried out to evaluate the relationship between the dynamic local-E and static local-E at specific locations on the samples. This provides a useful estimate of grading machine accuracy when processing production stock as opposed to grading machine accuracy when processing calibration sticks.

Tadpole provides a real time analysis of the relationship between static and dynamic local-E for these samples from production. The analysis provides a chart of individual static/dynamic values and a moving average as a group to indicate the true ratio of the two measurements. Also shown on the chart are the times/dates when the grading machine was re-calibrated.

(3) Production samples - comparing static values with grade thresholds

The analysis chart shown in the next Figure is provided for each individual stress grade of timber and is used to display two variables. Firstly, it shows how well the dynamic local-E correlates with the static local-E of production material (Stat/Dyn in the chart title) and secondly, it shows how well the static local-E measurements fit within the local-E threshold boundaries of the grade by plotting the individual local-E values, moving averages and simple averages of groups of data.

The individual static local-E values are shown as small circles and the averages of groups of data are shown as large circles. The moving average is not shown to retain some clarity in the chart. The horizontal lines show the preset grade boundaries (inner lines) and arbitrary control guide lines (outer lines). Data is plotted against the date of sampling on this particular chart.

For the data set used to generate this Figure the ratio of the static values to the dynamic values is shown as 1.04, indicating that the mean of the static values is 4% greater than the mean of the dynamic values. This information can be used to make adjustments to the machine calibration to bring the ratio closer to 1.

The individual points and the averages displayed on the chart give an indication of how well the production is being measured by the grading machine and how much scatter there is in the data over time. Arbitrary limits are placed on the maximum allowable scatter for individual operations and the results (pass/fail) are reflected in the results palette (discussed later).

Home