Sand Grain Size Analysis
Materials Needed
Equipment: (per table)
1. 6 sets of sieves = 10, 18, 35, 60, 120, 230, pan (6 sieves and the pan) 2. Electronic Balances to measure mass of samples 3. Handlenses or stereo microscopes 4. Computers with a spreadsheet program 5. Dilute HCl
Materials:
1. 3 Sand samples 2. Large sheets of paper (butcher paper or flip chart) 3. Smaller sheet of paper - notebook paper will do 4. 7 containers to place sieved samples in; these can be weighing trays or beakers
Introduction
While sediments may seem removed from biology, they are in fact important. Sediments structure benthic communities because of grain size preference by various organisms. For example, grain size makes a difference in the ability of flatfish to bury themselves in the sediment. Sediments may have biological origins in the skeletal material of corals, macroalgae, phytoplankton, foraminifera, radiolarians, mollusks, etc. Suspended sediments have been shown to cause stress and gill damage in fish, smother coral reefs, and decrease benthic primary production.
Apart from the biology, sediment characteristics can provide information about source materials, the depositional environment (how much energy there is in waves and currents), and other physical and chemical factors.
When rocks are broken down into fragments, either through the mechanical means of weathering, or through chemical reactions, the fragments are called sediment. When that sediment is compacted or cemented together, it forms a sedimentary rock. Sediments are either clastic or chemical. That is, rocks are broken down through either mechanical or chemical means.
Clastic sediment Clastic sediment is what one usually thinks of when speaking of sediment. From the Greek word klastos (broken), it refers to the broken remains of rocks of all types, broken and altered by weathering processes such as wind, water and ice. Clastic sediment is also known as detrital sediment. Chemical sediment Chemical sedimentary rocks may contain fossils and other sedimentary characteristics, but their components were not broken up mechanically. Rather, rocks were dissolved in solution (as salt can dissolve in water) and transported, then precipitated chemically (as salt can precipitate out of a saturated solution).
This lab will investigate unconsolidated sediments. You will learn how to characterize the particles or grains that are present, the size and size distribution of those grains, and then make some interpretations from these observations. In addition you will learn some fundamentals of statistics. During the lab you will measure grain size in two different ways: a) using a settling tube and b) using sieves.
Texture refers to properties of a sediment such as particle size, shape, roundness, and sorting. A well sorted sediment is one in which the grains are all about the same size. In contrast, a poorly sorted sediment contains a chaotic mixture and large, intermediate and small grains. Shape is a measure of the sphericity of a grain. Some grains are almost spherical, whereas others may be elongate or flattened. Particle roundness refers to the smoothness of a grain, regardless of its shape. Grains may be rounded (i.e., no sharp corners), subangular or angular. The concepts of roundness, shape, and sorting are illustrated in Figures 1–3.
Figure 1. Roundedness is often a function of distance transported since the corners wear down from abrasion with other particles.
Figure 2. Sphericity is independent of roundedness, and measures how close a grain comes to being spherical or elongate. Figure 3. Sorting is a measure of how even the particle size distribution. Sample A is poorly sorted while sample B is sorted.
Phi is the negative log base 2 of the diameter in mm. Table 6.1 in the "Rapid Method" lab has conversions from mm or micometers to phi size.
There are several documents to the lab. I will bring copies of them so you don't have to print them out, but they are here for you to look at.
Grain Types
Coastal sediment is made up of weathered terrigenous rock (terrigenous detritus) for the most part, plus organic detritus, plants, worms, sea shells if marine, and pore spaces. There may also be small amounts of calcite cement. The type of terrigeneous detritus (lets call it TD to keep it simple) found in sediment is dependent upon the types of rocks in the source area of the sediment. If the sediment is down stream from weathering granite then certainly expect to see detrital quartz grains in the sediment. If on the other hand your sediment is sunning itself on the beaches of Hawaii do not expect to see any quartz, (why not?). Most students when they think of sand or sediment they immediately think quartz grains. This is a good guess but not a sure bet. Quartz grains in most cases are the dominant grain type in a sediment but there will also be rock fragments. These are chunks of rock and yes technically a detrital quartz grain is a chunk of rock but for descriptive purposes we keep it separate. Sometimes we can recognize what type of rock those rock fragments are from. We can expect to encounter sedimentary rock fragments (SRF), metamorphic rock fragments (MRF), and igneous rock fragments (IRF).
Grain Size
One way to characterize a sediment is to determine the sizes of grains in that sedeiment. So one could measure all the grains on a particular beach, for example. Rather than measure each grain, scientists rely on subsampling. In order to characterize the sediment, one would take a representative sample of the sediment and run it through a set of sieves to break the sample subset in to size classes and using statistics reconstruct what the population's size characteristic are (much easier and quicker). Statistics are a way to describe populations of things, like fish or trees, and grain size.
Definitions from the American Geological Institute Glossary, 6th ed., 1980
mean: an arithmetic average of a series of values. median: the value of the middle item in a set of data arranged in rank order. If the set of data has an even number of items, the median is the arithmetic mean of the middle two ranked items. mode: the value or group of values that occurs with the greatest frequency in a set of data; the most typical observations. standard deviation: the square root of the average of the squares of deviations about the mean of a set of data. skewness: the quality, state, or condition of being distorted or lacking symmetry. kurtosis: the quality, state, of condition of peakedness or flatness of the graphic representation of a statistical distribution.
Notes: The phi value is the negative logarithm to the base 2 of the particle diameter. To calculate phi size you can use the Excel function "-log(number, base)". Where number is the diameter in mm, and base is 2.Round the result to 1 decimal place.
Sieve Analysis Laboratory Procedure
(1) Take approximately a 100 gram split of a sample. Examine it briefly with a hand lens or microscope and make appropriate notes about its character. Put this into Table 1 and include what you perceive the size of the average grain to be (sand box sand in a playground is medium grained, if larger sized grains dominate then it is coarse grained, and if smaller then fine grained). How well sorted is the sample (all or most grains are the same size then well sorted, some range in grain size then sorted, and if there is quite a bit of variation in grain size then poorly sorted). Are the grains for the most part angular, sub-angular to sub-rounded, rounded, or well rounded? What is the sphericity of the grains; compact or spherical, bladed, elongate? What types of grains are present; quartz, feldspar, rock fragments, mica, shell material? Test a small amount of each sample with a drop or two of dilute hydrochloric acid (HCl). If it fizzes there is carbonate material (shell, coral, etc.) present. Next pick through the sample and remove all large chunks of vegetation and bugs.
(2) Weigh the sample on the balance and record the mass of the sample in Table 2.
(3) Take a set of sieves and make sure that they are stacked such that the screen with the smallest opening is at the base and the largest is at the top. Note that the screens have different numbers on them. These are referring to different types of size scales. The most common are the US Standard Sieve Mesh #, opening in millimeters (micrometers), opening in inches, and Phi Scale; see table below. Place the pan at the very base of the stack. Dump your sample onto the top screen and put the cover on the top screen.
(4) With a circular motion shake the sieves and occasionally rap gently it on the bench top. Do this for 5 minutes, no more and no less.
(5) Gently pry off the top cover of the screen set. You may need to use a dime to aid in this. In the same manor remove the first screen from the stack; being very careful not to launch any grains off across the lab (don't force it be gentle). Lay a clean sheet of paper that is larger than the area of the screen on the bench top. Turn the screen over and dump its contents on the paper. Transfer the sand on the paper to the weighing paper or pan. Then take the screen and turn it over and rap its rim once on the surface of the paper. Transfer the grains to the weighing pan. Rap it again but a little harder this time and then dump the grains. Then slam the sieve down on the paper such that the entire rim contacts the paper at once; dump the grains onto the weighing pan and set the screen aside. DO NOT ATTEMPT TO POKE LOOSE THE GRAINS THAT REMAIN ON THE SIEVE; NEVER TOUCH THE WIRES OF THE SIEVE WITH ANYTHING. You will probably leave behind some grains, big deal, you also will probably gain a few grains from the previous user of the screen. This contributes to measurement error and everyone understands that.
(6) Weigh out what you have dumped from the sieve and record the results on a sieve analysis form. Set aside the sample split (what you just weighed) for future observations. See the end of this for an example of how one might set up a data sheet for doing this.
(7) Repeat (5) and (6) for each screen and the pan.
(8) Add up all the weights from each screen and the pan. Does it compare to you initial sample weight? Take the difference between the two and divide by the total weight of the size fractions. This is you measurement error. What are the sources of this error? (9) Construct a histogram of your results. I know that you all know how to make a histogram, but I want you to think about what you are doing before you construct this. The columns of the histogram will have a width proportional to the size range of grains (expressed using the phi scale) for each sample split. For example if you used a -1,-2, and -3 phi screens then the sample which was on the -2 screen represents all the grains which were greater in diameter than -2 but less in diameter than -3. Your histogram must reflect this and if you were using odd steps in sieve size then the widths of the various columns will vary. The height of the column will be proportional to the percent retained on the respective screen and therefore it will be in units of percent, not units of mass or weight.
The tallest column indicates the mode of the grain size distribution. What is the mode of this sample? The correct answer is not the size of the corresponding sieve but the range from that sieve to the phi of the next larger sieve (the one above it in the stack). (10) Construct a plot of grain size (x-axis) versus cumulative percent (y-axis). The Scale of the x-axis will be in phi values (not meters) and the spacing between a phi of 1 and 2 will be the same as that between 2 and 3. The y-axis will have a scale of percent (0 to 100%) using a linear scale (uniform spacing). Plot each data point from your form on the graph. Since we are working with cumulative percent, we can plot as a point the corresponding sieve size phi value. Next draw a smooth line through each point using a French Curve if you have one, otherwise eyeball it. This is not a scatter plot so do not draw a line through the field of points. This plot will be know as a cumulative curve with an arithmetic ordinate.
(11) Construct a similar plot of grain size versus cumulative percent using the probability paper. Do it the same as in (9) except the cumulative percent axis is already set up for you. Note that there are two similar scales along the long sides of the paper. This will be the Y-axis. It will be a little odd, but use the right side scale (.01 at the bottom and 99.99 at the top) for the cumulative percent scale. The size scale expressed in phi units will run along the short side of the paper, with the lowest phi value (or largest grain size) at the left of the graph. If your data has a 'normal' distribution (bell shaped curve) then this plot will turn out to be a straight line running from the lower left to the upper right. This plot is known as a cumulative curve with a probability ordinate (Y-axis). (12) Using either of the cumulative curves determine the phi size for each of the following phi values: phi at 5% (φ5), phi at 16% (φ16), phi at 25% (φ25), 50%, 75%, 84%, and 95% (where the % refers to the cumulative percent). Record these in the Table 3 below.
(13) Use the values from Table 3 to calculate the other statistics listed below using the equations shown. Record these values in Table 4.
Mode – the most frequent size category
Is the largest column of the histogram.
Graphic Median – 50% above and 50% below this category
The phi value at 50% is the Median of the sample or grain population. Half the population of grains (mass wise) was smaller than this and half was larger).
Graphic Mean – the average size category 16 50 84 M 3 Inclusive Graphic Standard Deviation – a measure of sorting or variation in sizes
84 16 D 4
Inclusive Graphic Skewness – shows if the distribution is bell shaped or shifted to side
84 16 2(50) 955 2(50) S 2(84 16) 2(95 5)
Kurtosis – shows if the distribution is bell shaped, very flat, or very peaked
95 5 K 2.44(7525)
Verbal Descriptions
Descriptive Terms Derived from Grain Size Analysis – Use the above numbers and match them to the descriptive terms from the tables below and write those descriptions in Table 5.
Grain Size: (from graphic mean) boulder -12 to -8 phi cobble -8 to -6 phi pebble -6 to -2 phi granular -2 to -1 phi very coarse grained -1 to 0.0 phi coarse grained 0.0 to 1.0 phi medium grained 1.0 to 2.0 phi fine grained 2.0 to 3.0 phi very fine grained 3.0 to 4.0 phi coarse silt 4.0 to 5.0 phi medium silt 5.0 to 6.0 phi fine silt 6.0 to 7.0 phi very fine silt 7.0 to 8.0 phi clay 8.0 phi and smaller Poorly sorted Well sorted
Sorting: (from inclusive graphic standard deviation) very well sorted under 0.35 phi well sorted 0.35 to 0.50 phi moderately well sorted 0.50 to 0.71 phi moderately sorted 0.71 to 1.0 phi poorly sorted 1.0 to 2.0 phi very poorly sorted 2.0 to 4.0 phi extremely poorly sorted over 4.0 phi
Sorting skewness: (from inclusive graphic skewness) strongly fine- skewed +1.00 to +0.30 fine- skewed +0.30 to +0.10 near symmetrical +0.10 to 0.10 coarse-skewed 0.10 to 0.30 strongly coarse-skewed 0.30 to 1.00 Sorting kurtosis: (from graphic kurtosis) very platykurtic < 0.67 platykurtic 0.67 to 0.90 mesokurtic 0.90 to 1.11 leptokurtic 1.11 to 1.50 very leptokurtic 1.50 to 3.00 extremely leptokurtic > 3.00
Grain Size Scales US Standard Opening in Phi Sieve millimeters Scale Mesh 4096 -12 1024 -10 256 -8 64 -6 5 4 -2 10 2.00 -1 18 1 0 35 0.5 1 60 0.25 2 120 0.125 3 230 0.0625 4.0 325 0.044 4.5 0.031 5.0 0.0039 8.0 0.00006 14.0 Additional Materials:
Part 2 - Rapid Method of Sand Analysis - Lab Example of Sieve Analysis blank arithmetic probability paper spreadsheet example for sieve analysis example cumulative probability curve for sieve data
Read this background material - good figures and explanation of interpretations! The classic text "Petrology of Sedimentary Rocks" by Robert Folk.
What to turn in: A short discussion of the 3 samples, and possible reasons for any differences among them. Include the 3 graphs (histogram, cumulative percent, and cumulative probability) and 5 tables. Plot all three data sets on a single graph to make comparisons easier. You may turn this in as an Excel spreadsheet and/or drawn by hand. Name ______
Table 1. Visual Observations of Samples
Sample Sorting Avg. Size Roundedness Sphericity Type Material
Table 2. Sieve Analysis Data
Sample: ______
mm Phi Wt retained % retained Cum Wt. Cum % -1 0 1 2 3 4.0
xxxxx TOTAL xxxxxxxxxxx xxxxxxxxxxx xxxxxxxxxxxx
Sample: ______
mm Phi Wt retained % retained Cum Wt. Cum % -1 0 1 2 3 4.0
xxxxx TOTAL xxxxxxxxxxx xxxxxxxxxxx xxxxxxxxxxxx Sample: ______
mm Phi Wt retained % retained Cum Wt. Cum % -1 0 1 2 3 4.0
xxxxx TOTAL xxxxxxxxxxx xxxxxxxxxxx xxxxxxxxxxxx
Table 3. Measures for Statistical Calculations
% Example Sample Sample Sample 1 2 3 95 3.7
84 3
75 2.5
50 1.7
25 0.9
16 0.5
5 -0.5 Table 4. Statistical Calculations
Sample Grain Size Mode
Median Grain Size
Mean Grain Size
Standard Deviation
Skewness
Kurtosis
Describe your samples below in verbal terms (see Descriptive Terms above) from the numeric data gathered
Table 5. Sample Description
Sample
Grain Size
Mode Median
Grain Size Mean
Grain Size Sorting
Skewness
Kurtosis