Attached are best fits for cooldown runs on 01/14 and 01/31. The setup for both cooldowns can be found in the previous ELOGs. We noticed that the outer shield did not cool as significantly on 01/31 than on 01/14, hinting that there might have been more thermal contact between the outer shield and cold plate / copper bar.
The model considers the resistances of the following conductive elements (and uses these resistances as fit parameters):
1.) Joint between cold head and copper bar
2.) Bulk of copper bar
3.) Joint between copper bar and flexible strap
4.) Bulk of flexible strap
5.) Joint between flexible strap and cold plate
These additions helped the model more closely resemble our recorded data, with a few exceptions:
- At early cooldown times, the model seems to be underestimating the heat load on the inner shield and outer shield.
- The best fit was performed on the inner shield and outer shield data (to fine tune elements of the cold linkage), so the test mass fit is not optimized. (This will be performed next to refine emissivity predictions.)
To identify bottlenecks in the cold linkage, I used the 01/31 model and tweaked the resistances to see which would provide the largest gains in cooldown. The results from such tweaks are below:
- Halfing the resistance of the greased joints (1, 3, 5) made negligible change to the cooldown.
- Halfing the resistance of the bulk of the copper bar made significant improvement to cooldown (Attachment 3).
- Halfing the resistance of the bulk of the flexible strap made some improvement to cooldown (Attachment 4).
From these observations, it seems like the greased joints are thermally efficient, and the bulk area of the copper bar appears to be the largest bottleneck.
|