|Important note for all data sets: Some particles may not exist at all times, but the tracking program has identified them as likely the same particle and thus with a continuous identity. However, it is known that occasionally the tracking program makes mistakes with this.|
The data are all available below. They used to be at eComploids, but I think that website is down.
These data sets were studied in:
Information on each individual data set is below. Each file contains one half-period of data, where the sample is being sheared with a triangle wave. The data are presented in the reference frame co-moving with the center of mass of the data. This is not the "de-sheared" reference frame that we discussed in the paper.
The data are in the form (x,y,z,t,id). t is an integer, to convert to seconds, multiply by the time step listed in the table. (x,y,z) are in microns and the mean particle radius is 1.05 microns plus or minus 0.04 microns; the error bar is the uncertainty of the mean radius, see our paper for details. Each particle has a unique ID number assigned to it. Some particles may not exist at all times, but the tracking program has identified them as likely the same particle and thus with a continuous identity. However, it is known that occasionally the tracking program makes mistakes with this.
Volume fraction #1 is the one reported in the paper, which we believe is more accurate. #1 is based on pretrack data. Volume fraction #2 is the apparent volume fraction if you look at the data set as presented below (that is, try to figure out its volume V, count the number of particles N, and compute N V_p/V using the volume of a particle V_P). For various reasons we believe the apparent N/V is likely over-counting.
(Meso strain rate) = (total strain_meso) / duration. The macro strain rate is what we applied to the plates, calculated as follows. The macro strain applied by the plates, as described in our paper, was always 1.4; due to shear-banding the strain_meso is always different from strain_macro=1.4. The macro strain rate can thus be calculated more accurately from strain_macro/(period*0.5) = 2.8/period. This is because we were applying a triangle wave, so a strain of strain_macro was applied over a duration of half a period in one direction, and then reversed. Thus the macro strain rate was a constant for each half-period.
The "figure" column indicates correspondence with figures in the paper. The data set "dchen07.txt" was used for many figures, and the time period used was t in the range between 11 and 41 (these are integers which match the t values in the file, multiply by 1.5 to convert to seconds for this data set.) For Figure 15, we've listed the delta-gamma value for the data set so that it's apparent which symbol it corresponds to. As described in the paper, the clusters for Figure 15 were found from a subset of the sheared data so that the particles weren't sheared out of the field of view, thus the dg values for Fig. 15 are less than "total strain_meso".
|dchen01.txt||0.51||0.536||1.5 s||54.0 s||1.224||0.023 1/s||0.018 1/s||150 s||Fig. 6: red-dashed; Fig. 15: dg=0.58|
|dchen02.txt||0.51||0.550||1.5 s||42.0 s||0.740||0.018 1/s||0.018 1/s||150 s||Fig. 6: blue-dotted; Fig. 15: dg=0.39|
|dchen03.txt||0.51||0.538||1.5 s||70.5 s||0.695||0.099 1/s||0.018 1/s||150 s||Fig. 6: green-dashed; Fig. 15: dg = 0.41|
|dchen07.txt||0.51||0.545||1.5 s||64.5 s||0.569||0.0088 1/s||0.018 1/s||150 s||Fig. 6: solid-black; Fig. 15: dg = 0.43; Fig 5 and others|
|dchen04.txt||0.51||0.507||1.5 s||120.0 s||0.746||0.0062 1/s||0.006 1/s||450 s||Fig. 6: purple dash-dotted; Fig. 15: dg = 0.32|
|dchen05.txt||0.57||0.613||1.7 s||78.2 s||1.515||0.019 1/s||0.016 1/s||170 s||Fig. 15: red square (dg=0.72)|
|dchen06.txt||0.57||0.599||1.7 s||62.9 s||1.226||0.019 1/s||0.016 1/s||170 s||Fig. 15: orange square (dg=0.49)|
|dchen08.txt||0.56||0.573||2.0 s||96.0 s||1.099||0.011 1/s||0.013 1/s||200 s||Fig. 15: black triangle (dg=0.89)|
|dchen09.txt||0.56||0.574||2.0 s||90.0 s||0.718||0.0080 1/s||0.013 1/s||200 s||Fig. 15: red triangle (dg=0.66)|