Ice704: 3 x P-drop f.g. ice I + 10% by vol 0.3-micron Al2O3

T = 210 K, P = 50 MPa, s = 4, then 5 MPa step (1)
T = 210 K, P = 50 MPa, s = 7 MPa step (2a), (2b), (2c), [depressurize], (2d, (2e), [depressurize], (2f), (2g), (2h)
   

 

Legend: Upper graph of each pair is motor speed V (proportional to edot); lower graph is load (in volts). Translate motor V to edot thus: V1.00 ≈ 6e-6/s ≈ 2%/hr. Load to stress translation for ice704 is roughly 0.3 MPa/mV, with a 3 mV zero offset. Horizontal scale is time. 5000 ≈ 14 hrs.

     
Run log
704 working plot (with 687, 691)

After step (1), views at 0, 90 deg

After step (2c), views at 45, 135 deg

   
After step (2e), views at 45, 135 deg
After step (2h), views at 45, 135 deg
 
e=0.1297 Tues 5/28 0900 hrs, jump ahead 4 days to end of step (2h). Amazing stability of rig over Mem. Day weekend. End of run, presumably.
e=0.1094 Fri 2300 hrs, start step (2h).
e=0.1064 Fri 5/24 1100 hrs, skip to end of step (2g).
e=0.0978 Weds evening, step (2g) continues. These occasional slow-motion "blips" up (and sometimes down?) in s seem to be a fact of life. Stick-slip of piston, maybe?
e=0.0943 Weds 5/22 a.m., step (2g) continues. Bounciness relates to operator-induced confusion. Details in log.
e=0.0909 Tues 1715 hrs, note for the record: Data recorder has been off since the start of step (2g), and is just now turned on. This screen shot is the only record of that part of the run.
e=0.0905 Tues 1600 hrs, start of step (2g). Reload after zero check. Short transient as expected; got to steady levels in ~40 mins. [Note: data recorder was accidentially left off from the start of the step until 1715 hrs
e=0.0897 Tues 5/21 1300 hrs, end of step (2f).
e=0.0849 Mon evening 5/20, start of step (2f). Rig fired up again after weekend shutdown. Transient here shows the same kind sponginess as step (2d). Operator was better prepared this time, so compared with (2d) s got to target more quickly in time. The amount of strain to get through the transient might have been a bit less than (2d) but was definitely more than for step that didn't involve removal of the sample.
e=0.0818 Thurs 5/16 1010 hrs, end of step (2e). Pretty smooth. Sample out, photo (see above) shut down rig, restart Monday.
e=0.0781 Weds 2130 hrs, step (2e) continues.
e=0.0749 Weds 5/15/13 1130 hrs, step (2e) continues. The higher edot near V0.11 endures, see soliloquy immediately below. No sign yet of mechanical instability.
e=0.0706 Tues 0930 hrs, still first part of step (2e). Interestingly, the current edot level is near V0.11, 25% higher than in step (2d), yet all that was done between (2d) and (2e) was unload, wait for a few minutes, and reload. This "noisy" behavior of the sample is not unusual in creep experiments on icy materials (and of rocks); I attribute it to (1) high sensitivity of strength to small perturbations of known variables like T and stress boundary conditions from sample canting or buckling, and (2) stochasticity, the chaotic nature of deformation of polycrystalline solids.
e=0.0677 Tues 1300 hrs, start of step (2e), after unloading for gage zeroing. Transient on reloading is much more like the start of (2b) and (2c) and in stark contrast to the spongy behavior at the start of (2d) (where screen shot matches scale here). Spongyness must have been related to depressurization and handling of the sample for photography. Did this make it full of microporosity?
e=0.0657 Tues 1130 hrs, end step (2d).
e=0.0647 Tues 5/14 0615 hrs, first part of step (2d) still. Things have stabilized, and edot is now down below(!) level of previous steps, to <V0.08.
e=0.0621 Mon 5/13 1845 hrs, start of step (2d). Very odd behavior! Takes a very long time at unusually high amount of strain to climb to target stress. Operator hesitation at running at high edot (and thinking about possibility of mechanical instability) drew the transient out over several hours. Here, nearly two hours into the step, edot is still more than 2x greater than in previous steps. But sample is strain hardening; it is not unstable, more like "spongy".
e=0.0585 Sun 2130 hrs, end step (2c). Whatever is going on, edot just can't keep up. With no problems to machinery or sample jacket apparent, suspect sample itself. Funny jumps of s gage (not shown) suggest shearing issues (e.g., sample canting). Remove sample. See photos above. Sample has indeed started to cant. Can't get anywhere near 20% strain any more, but it is worth trying for a few percent more. Sample now back in, step (2d) about to begin.
  Close up of event at 1900 hrs. Was not sudden, but spread over 10-15 mins, and edot now seems to want to be higher by 25%. No signs of trouble with T, P, etc. etc.
e=0.0577 Sun 1900 hrs, step (2c) continues. Very odd jump down and up in s at approx 1500 hrs.
e=0.0546 Sun 5/12 0730 hrs, step (2c) continues. Edot over the past 24 hrs is ~8e-8/s. Too soon to tell if this is continued strain hardening or if is an artifact of the gage zero adjustment between steps (2b) and (2c).
e=0.0524

Sat late, start of step (2c), first 7 hrs. Edot is ~20% lower than in step (2b). Much of this can be explained by applying n = 4 to the ~1-mV zero-load adjustment just made to the force gage (see run log); i.e., (23 mV/24 mV)^4 = ~0.85.

back to top

e=0.0496 Sat 1422 hrs, end step (2b).
e=0.0472 Sat 5/11 0645 hrs, step (2b) continues. Avg edot over past 24 hrs is ~1.2e-7/s, close to 691 strength. This solidly confirms the result that WBD and DLG material are indeed different. (Working plot finally updated to something useful).
e=0.0427 Fri 1820 hrs, step (2b) continues
e=0.0376

Fri 5/10 early a.m., start step (2b) after force gage zero check. edot now pretty close to 1e-7/s.

back to top

e=0.0323 Thurs ~1600 hrs. Last 12+ hrs of step (2a). Note that about an hour before the end, I swapped out the 10:1 gear reducer for the 100:1 in order to get better speed control. Same edot now requires 10x faster motor speed.
e=0.0269 Thurs a.m. 5/9/13, work hardening continued overnight as expected, current edot a little above 1e-7/s, approaching strength of 691.
e=0.0172 Weds 1700 hrs, start of step (2); s = 7 MPa. Current edot is roughly 3e-7/s, and I expect this to strain harden like all the other samples. At this point, it has roughly the same strength as 687, but if it hardens as I expect, it will look more like the WBD half of run 691.
e=0.0079 Weds a.m. 5/8/13. Most of step (1). Conservatively chose 5 MPa as the One Step; turned out to be too hard. Before moving to the new One Step, take the sample out and confirm that it compacted ok, get true dimensions, etc. See photos above--absolutely gorgeous.