The current extraction method

Possible improvements

Tests

Test 1 - Re-freezing the slices

Conclusion: Fail, too ductile

We are going to mill samples out of turkey slices and then re-freeze them before trying to pop them out. This will be done with the old prototype without the added cooling method, only the vacuum gripping.

We will test on 5 samples on each of the 5 slices, totaling 25 samples. The first sample was milled front and back to see if double-milling could be done on the current prototype. The image overlay was not precise enough to cut the same place twice. In addition when flipping the slice, one would have to be able to flip it without any rotation. If one rotates the slice slightly, one would not be able to match the shapes.

Another observation was that the cut quality went down as the slices thawed. Keeping the slices frozen was of no importance during this test as they were to be frozen again before extraction.

The slices all stayed secured to the bed during milling. The slices were milled at 0.6mm/s and the mill bit head was between 0,55 to 0,05 mm above the base plate. This is due to the crookedness of the base plate.

The next step was to extract the samples from the slices using popping. The initial evaluation of the 5 slices was the the cuts made by the CNC were really difficult to locate.

The samples were pressed out using metal screws, and was later switched out for zip-ties, as plastic was thought to thaw the slice less. They were pressed out over the hole seen on one of the images below.

None of the samples were extracted successfully. Some samples i were able to almost push out, but it was always stuck at one place or another. Most samples, I were not able to extract at all. The slice caved in to the fore from the pushing rods. 

These were not as crispy and brittle as the ones stored in -80 degrees at MRCL (MR Core Labs), so i don't know if that extra brittleness will made the method successful or not.

Every slice was still frozen after extraction was performed, but they were not cold enough for the samples to pop out. 

Test 2 - Re-freezing the Slices in dry ice

Conclusion: Fail, too brittle

For the next test we will be re-freezing the slices in liquid nitrogen in order to test popping on a -80 degree slice. We will then mill a slice at TrollLABS and cool it further in dry ice before popping it out. 

We used cooled down metal prongs as well as a metal "stamp" to push the samples out.

The force did separate the sample from the slice with a quite high quality cut, but both the sample and the slice cracked into several pieces. Every sample we tried to extract ended the same way.

Test 3 - Re-freezing the Slices in the CNC-Freezer.

Conclusion: Successful, but risky.

For the next test we will be re-freezing the slices in the CNC freezer in order to test popping on a -20 degree slice. The difference from test 1 is that the slice will experience much less temperature fluctuations before extraction.
We will also try to assist the extraction by cutting/scraping  a scalpel inside the milled "track".

The test was split three ways.

The first two small samples and the first big sample was extracted using popping. A circular tool was used to apply force to the sample. The samples all popped out and left a nice clean cut to the slice. we pressed out the samples above a shallow hole. When the samples were pushed out they probably hit the bottom and was then hit by the circular pressing tool. 2 of the 3 samples came out cracked. This would probably not have happened had the pressing hole have been larger, but that would have to be tested before anything can be confirmed.

The second part of the test was done using the mill bit as the pressing tool. This did not loosen the sample from the slice and instead punctured the sample and cracked it.

The third part was done using a scalpel to cut around where the mill had milled. This worked quite well and both samples were extracted fully intact.

The technique used with the scalpel was applying force on some spots before moving to a new place. This weakened the link between sample and slice and eventually led to the link breaking, freeing the sample.

One takeaway from this was the re-frozen spooning providing extra binding between sample and slice. The effect of the spooning will be considered in test 4.

Test 4 - Milling several passes to remove spooning before re-freezing.

Conclusion: Easier removal

For this test 2 slices were tested. 

On the first slice, the program was run twice. Once to cut and once to force spooning out of the cut. The vacuum was used to suck away as much spooning as possible after each run.

The slice got sucked halfway into the vacuum, being ripped off the plate, leaving all the samples on the place(they had frozen to the plate).

Although this was not ideal, it did show a weakened bond between the samples and the slice.

The second test was altered slightly. Instead of one extra run to clear out spooning, 2 runs were made. In addition, in order to keep the slice from being sucked off the plate, i decided to blow away the spooning. 

The results from this was very clear. Running several times and removing the spooning each time both loosened the bond between sample and slice, and made the samples better. The cut looks way more clean on these samples than the ones from slice 1, and especially the ones from test 3.