A very good point, yes a crack will close when a loading is...

A very good point, yes a crack will close when a loading is removed in one direction, but once a crack occurs that bonding is lost and will not connect again unless thermally fused. Even under smallest microscopic gaps a vacuum will still pull air molecules through it but at a much slower rate. Remember CVM sensors detect the surface cracks, it does not have to wait for a crack to continue the thickness of the material under test which it will eventually do.
If it’s true that’s what they want, the cracks there let’s close it up and see if the sensor can see it, then they could do that testing in days not many months. A crack is a crack the sensor sensitivity has already been proven by Sandia labs. I was under the understanding they would cycle a sample under their low load requirements and when a crack appeared will the sensor detect it - long time frame. Just does not add up, still a little suspicious of their motives.