Example of Sieve Analysis Data Calculations, Graphs, and
Data Presentation.
Sample Description and Location : Upper Sand over shelly zone, GA 88, Eocene Sandstone
Weight of Sand and Packet 36.90 grams Weight of Packet 3.38 grams
Weight of Dry Sample before Sieving 33.52 grams (this value is calculated by subtracting the packet
weight from the Weight of Sand and Packet)
Screen Weight of Weight of Weight of Cumulative Weight Cumulative
Opening Beaker with beaker Sand Weight Percent Weight
(phi) sand (grams) empty (grams) (grams) Percent
(grams)
-1 (10) 5.27 2.32 2.95 2.95 8.83 8.83
0 (18) 7.27 2.32 4.95 7.90 14.81 23.64
1 (35) 15.66 2.30 13.36 21.26 39.98 63.61
2 (60) 11.86 2.31 9.55 30.81 28.58 92.19
3 (120) 4.44 2.31 2.13 32.94 6.37 98.56
4 (230) 2.78 2.32 0.46 33.40 1.38 99.94
5 (pan) 2.31 2.29 0.02 33.42 0.06 100.00
Total Weight of Sieve Fractions 33.42 grams Percent of Dry Weight 99.70%
The weight of sand is calculated by subtracting the weight of the beaker empty from the weight of the
beaker with sand. So for -1 phi data 5.27 - 2.32 = 2.95
Cumulative weight is the sum of the weights thus -1 phi value is the starting point and for the 0 phi line the
cumulative weight is the weight of the sand from -1 added to the 0 phi weight of sand. So 2.95 + 4.95 =
7.90, and the 1 phi cumulative weight is 7.90 + 13.36 = 21.26, etc.
Weight percent is the total weight of Sieve Fraction Divided into each weight of the sand fractions. Thus
for -1 phi 2.95 is divided by 33.42 then times 100 = 8.83 %. ( times 100 will convert decimal to percent).
Cumulative weight percent is calculated by dividing the total weight of sieve fraction into each of the
cumulative weight values. Thus for -1 phi 2.95 is divide by 33.42 times 100 = 8.83 %, 0 phi 7.90 is divided
by 33.42 times 100 = 23.64 %
The Percent of Dry Weight is calculated by dividing the Total weight of Sieve Fractions by the Weight of
Dry Sample before Sieving, times 100. So for our example 33.42/33.52 times 100 = 99.7 %. This gives
what percent of the sand you began with you recorded after sieving. A percent lower than 97% means very
poor laboratory technique. This number also shows the amount of error in your experimentation process.
To Graph this data you will use the Weight Percent for the histogram and the Frequency Curve. The
Frequency curve for this data looks like this.
The Cumulative percent data is used on the Cumulative Arithmetic Curve and the Probability Curve (which
is graphed on probability graph paper). The Cumulative Arithmetic Curve is shown below for this data.
The Probability Curve is where you will interpret various values you will need to plug into the Folk and
Ward Formulae. The graph you produce will look something like the graph below (image5a.gif). Your
graph will be more precise than this example. The formulae require the phi value at 5%, 16%, 25%, 50%,
75%, 84%, and 95%. To interpret these values you will find where your line for the data crosses these
percentages. You will project a line down to the phi scale and record the phi value (as the line on the graph
shows the phi value at 50%) Your graph you will be able to determine the phi value to the one-hundredth of
a phi (0.01 phi). Thus on this graph the phi value at 50% would be 1.43 phi (on your graph you will have
0.1 phi lines and you will be able to interpolate to the one-hundredth of a phi (0.01).
Thus the values for the data are:
Percent Phi Value
5 -1.13
16 -0.42
25 0.05
50 0.67
75 1.29
84 1.60
95 2.30
So using Folk and Wards Formulae, the various calculations would be:
From the charts in the lab exercise you can translate these numbers to words. Thus:
For the Graphic Mean the number represents the Mean grain size is Coarse Sand (0.62)
For the Inclusive Graphic Standard Deviation the number represents the sample to be Poorly Sorted (1.02)
For the Inclusive Graphic Skewness the number represents the sample is Near-symmetrical in its
distribution (-0.06)
For the Graphic Kurtosis the number represents the sample is Leptokurtic (1.13)
These numbers can be used to relate known samples to unknown samples numerically. What you will find
is the numbers will not match exactly, but they will be close. Thus if we had a known Depositional
environment which had a mean grain size of 0.72, sorting of 1.10, skewness of 0.06, and kurtosis of 1.23.
The sample we have just calculated since its values are so close to these we could conclude that most likely
these samples are from the same depositional environment.
Return to Sieve Analysis Lab Exercise
Return to Table of Contents for the Lab manual
Return to Dr. Anderson's page for the Lab manual
