Wrapping Up

Having trouble with file size/type restrictions so I went with a dropbox folder for uploads. Contents include:

Final presentation

BESL2013 work
The 1-D analyses of step data and freecools for Cu55-xZr45Alx, x = {7, 8, 9, 10, 12} generated over the summer are in their respective /BESL 2013/STL13XYZ folders on the hard drive. Here is a .zip of all the S(q) files used in the final analysis, performed using the following

S(q) analysis Matlab code
Contents of the .zip:

  • S_peakfind.m is a function to determine peak information from S(q) .csv files
  • S_q1_evol.m is a function to perform peak fits across temperatures and a linear fit on S(q1) for a single composition
  • S_q1.m is the big-picture function to analyze across compositions and temperatures

Could run as follows: >> S_q1('/home/sussmanl/Data/', 3, 800, true, 1)
Parameter interpretation:

  • n_sigma = 3 is the error multiplier for each S(q1) value, used to determine uncertainty in linear fits across temperatures
  • T_min = 800 is the lower temperature cutoff for fits
  • do_plot = true will plot the results, and
  • file_out = 1 prints results to the console (otherwise prints to the file path specified).

Output will look like this:

x		m		err_m_u		err_m_l		
0.000000	8.884653	0.551413	0.576458	
2.000000	8.478724	0.323882	0.361227	
4.000000	8.145996	0.283907	0.114391	
6.000000	8.026923	0.237747	0.161743	
7.000000	7.972041	0.370155	0.321799	
8.000000	7.636627	0.221637	0.299425	
9.000000	7.833311	0.442020	0.313923	
10.000000	8.406356	0.441755	0.415206	
12.000000	8.127207	0.286038	0.179636

Thanks again for a fantastic summer!

Week the Fourth

Continued x-ray characterization of ribbons. Doing 1hr runs in both parallel and perpendicular orientations (generally for two sets of 6-10 ribbons) as well as one 16hr run for each composition. So far no samples that look decidedly crystalline, but a few questions. Still something of a mystery why the parallel and perpendicular have such distinctly shaped curves. By the end of next week intend to compile data for each composition.

In the data analysis world, focusing on liquids. Finished a first pass through Cu55-xZr45Alx for x = {12, 10, 9}, tracking S(q1). Still need to wrap up x = 8 and work on 7.

Week the First

Looking at temp step data for liquid Cu43Al12Zr45 system [STL13039], produced the following plots:

cu43al12zr45 g(r) temp steps [title says S(q), should be g(r)]

cu43al12zr45 S(q) temp steps

Just starting to delve into glassy data for same composition from STL13080. Trouble adjusting params to get smooth results, but general shape of S(q) comes through and follows pattern seen among temps in liquid despite being glassy.

cu43al12zr45 S(q) glassy 200

Question: is it okay to use darks from the beginning of a file of “empty” data if there are none at end?

Site doesn’t allow upload of .txt files (really, security issues?) so here are logs so far…

STL13038 – not processed

WashU (Kelton)    008    26    Cu43Al12Zr45    3    50.56    50.5    6/14/2013    23:46    Processed    STL13039    Yes    High temp and undercooled step data, free cool x-ray and density data

Used Empty at 1fps associated with STL13039
Frames 10-30

First exploration:

Look at hold, 900
Darks 28,29

exploration of first peak heights S(q1) at different temps:
900: 2.7
950: 2.65
1000: 2.61
1050: 2.57
1250: 2.43
1300: 2.41


Looking at step data for temps 850-1300 C
Goal: I(q), S(q), G(r)/g(r)
Using 1fps empty from inc4
Empty frames: 10-29
Empty darks: 1-8
For now only combining 1 frame (#10)
Using I(q) Settings -> Dark = 1, Bg = 1.00306, Empty Dark = 1 as loaded from hist file

dark: 28-29

dark: 28-29

dark: 28-29

dark: 28-29

dark: 28-29

dark: 28-29

dark: 28-29

dark: 28-29

dark: 28-29

dark: 22-29 (different b/c came from longer hold at 1300)


Made plots of S(q) and g(r) for Cu43Al12Zr45 temp steps 850-1300 C.

started looking at glass data

empty (inc3_700empty): 8-70
empty dark: 1-6

dark: 67-69
played with adjusting weights, couldn’t find anything too great (forgot to record ones used, bad me!).
still very choppy but general shape is clear for S(q) (gets worse in transform).
Made plot comparing with 850 and 1300 C liquid data, seems to fit into that trend despite phase change,
i.e. q1 and S(q1) increase with temperature.