2015.02.05

Torqued servo gearhead to mounting block.
Torqued lower half of upper servo coupling to 70 Nm.
Applied dielectric grease to cabinet end of servo power and feedback cabled, and attached these to the servo and cabinet connectors.
Confirmed AKD parameters
- Slider tuning at 5
- Regen set to external
- Emulated encoder ouput at 5,000 lines per rev, or 1e5 including the gearbox.
Jogged turbine axis at 5 RPM and things looked good. Recorded some video.
Wired up torque transducer to 9205 module and 12 VDC power supply.
Attached Vectrino sync leads; applied dielectric grease to terminals.
Homed z-axis.
Checked Vectrino comms; good.
Checked Vectrino velocity measurement; very noisy, but probably due to no seeding.
z-axis has some trouble sticking when reversing direction. Will need to do some tuning.
Looked at torque, drag and RPM NI signals while rotating turbine at 10 RPM; look good.
Homed y-axis; looks good.
Jogged turbine at 15 RPM.
Torque transducer and arm measurements appear to be tracking eachother, which is good.

Vectrino trigger was not working initially. Moved positive lead to "Synch -" and it worked fine.

Have done a couple shakedown runs at 0.5 m/s:

$\lambda$	$C_P$
2.6	0.13
2.7	0.17
2.8	0.21
2.9	0.24
3.0	0.27
3.1	0.29
3.2	0.31
3.3	0.32
3.4	0.32

Greased z-axis ball screw to aleviate binding.

To-do

~~Re-tape Vectrino cable~~
~~Zero turbine angle~~
~~Take more pictures~~
~~Measure YZ traverse geometry~~
Measure salinity
Add seeding

Confirming YZ traverse geometry

The distance shown below was measured by tape measure to be 70 3/16" when the controller thinks $y=0$.



In [2]:

    
from IPython.display import Image
Image(url="https://unh.box.com/shared/static/qldih5xxwvp2mie9ph87xsl6zemkvusl.png")









    Out[2]:

The measurement corresponding to the image below was 74 1/8" via tape measure, so things look pretty good.



In [3]:

    
Image(url="https://unh.box.com/shared/static/h2lxd3k0ptr6poz0f7a5ecljbvqil1i3.png")









    Out[3]:

Zeroed turbine angle +/- a couple degrees. It's not that easy to tell even with the pointer bracket.

Tranferred Nikon camera images and video to Box.

Did another shakedown run at $\lambda=4.0$. Power coefficient was 0.25. At a tip speed ratio of 5, $C_P$ went down to 0.02. At 5.5, -0.10. Judging from the RVAT $Re$-dependence data, this means we'd be pretty close to zero at the higher Reynolds number.

$C_P$ went up to 0.34 for $U_\infty = 0.6$ and $\lambda = 3.3$. Hit 0.35 at $U_\infty = 0.7$ and 0.37 at $U_\infty = 0.8$.

Calling it quits for tonight. Will resume tomorrow.