About this website

Granulometrics is a web tool provided by Nicola Liberty Holdings Inc. of Logan Lake, BC, Canada.

The website provides a plotting facility for particle size distributions generated by laboratory sieving tests. Please browse the Wikipedia sieve analysis page for background on what this actually means.

Using the website

You, the User, will upload information to the website that defines a particle size distribution from a laboratory sieve set. Once the information is uploaded, the website will present you with a preview table to verify that the data was uploaded properly, then (if you accept the data preview is correct) the website will display a particle size distribution plot of your data. You can manipulate the plot by changing the units of the axes, zooming in and out, and applying certain regression models to the data.

Uploading data

Various settings allow you to customise the chart to your preferences.
Click the button on the left side of the page "Create New Particle Size Distribution Plot".
Enter the sieve openings that were used in the laboratory test. This can be done by one of three ways: manually choose sieve opening using the checkboxes provided, paste a list of screen opening sizes (one per line), or upload a CSV or TSV file containing the sieve openings (one per line).
Click the 'Next' button at the bottom of the page when the data has been entered or the data file has been uploaded.
(If data is pasted or uploaded) - Specify what kind of data has been provided. For example, if the sieve sizes are in mm, then select 'Sieve sizes (mm)' as the data type for the appropriate column of data shown on the preview.
(If multiple columns of data is pasted or uploaded) - Also specify the column containing the data to plot and what kind of values are represented in that column. For example, if the second column of the pasted data or CSV file contains the "cumulative percent passing" each sieve, then select "Cumulative %Passing" from the column's drop-down header. Click the green 'Next' button at the bottom of the page.
(If sieves were specified by check-boxes or only one column of data was pasted) - Specify at the top of the page what form the screen data will be in. For example, if the data is raw screen weights, then select the green "Raw Screen Weight(g)" option. Then enter or paste the data from the test into the appropriate spots. The %retained and Raw Screen Weight options must also specify a value in the "pan" fraction; the pan fraction may be omitted for the two Cumulative options. Click the green 'Next' button at the bottom of the page.
Review the preview data of data and confirm that the website has correctly parsed the data that you provided. If it is correct, click the green 'Plot' button at the bottom of the page.

Chart manipulation

Various settings allow you to customise the chart to your preferences.
The main Y-axis (Y1) can display any of cumulative %passing, cumulative %retained, or a histogram of %retained. Choose the desired style from the right-hand side of the page.
A secondary Y-axis (Y2) may be added that can display any of cumulative %passing, cumulative %retained, or a histogram of %retained. Choose the desired style from the right-hand side of the page.
The colour and line style of the data associated with each axis may be modified by clicking the setting button below the axis selection button.
The behaviour of the tooltip (that follows your mouse around the screen) can be modified or removed using the "Cursor snap-to-series" or "Axis tooltips" toggles on the left side of the page.

Chart manipulation

Regression models are used to fit equations to the data and make predictions based on the fitted data. A regression will, in most cases, plot a regression line on the chart and provide the best-fit equation.
Nearest Neighbours - interpolates the two nearest points to provide a prediction. Cannot extrapolate, only works if the interpolation point is between two existing data points.
Bond \u221A2 model - shows the typical size distribution expected for Bond Work Index type calculations based on a fixed exponent of 2 -½. Uses a Gaudin-Schuhmann model to predict the 50% point of the data points.
Gaudin-Schuhmann model - commonly used granulometry model suitable for most mineral processing and earth science uses.
Rosin-Rammler model - commonly used granulometry model suitable for most mineral processing and earth science uses.
Swebrec model- commonly used granulometry model for characterising blasted rock.

All of the regression models can provide a prediction of the %passing a specific size, or the specific size that passes a specified amount. The bottom of the regression area provides a field where you can enter the desired value to compute the regression for. Most of the regression models also plot a curve on the chart; you can judge how well a particular regression model has fitted the data by looking at the curve.

Return to the main page