Twenty years ago, I was working on a product improvement project of a product called a mammography accreditation phantom (e.g.
https://www.supertechx-ray.com/BreastImagingandMammography/QCC/Gammex-156.php ). The phantom was required by federal law to be used qualify every mammography facility in the country. It consisted of a patterns of test objects which represented the types of cancerous objects that might be encountered in a breast exam. To qualify, a facility had to x-ray image the phantom and be able to see all of the test objects each day before taking patients.
One group of test objects consisted of star patterns of aluminum oxide particles in decreasing sizes, the smallest of which had and average diameter of .16mm or .006". The problem was, that we were sieving corundum grit to obtain the particles but the particles had widely varying shapes with some shapes appearing brighter. My intent was to use a sapphire ball instead of the corundum grit. The material would be the same chemical composition but the size and shape would be more uniform.
I ordered samples of appropriately sized sapphire balls from a supplier. When they came in, I thought that the plastic bag for the smallest balls was empty but on closer inspection, the were a couple of dozen balls at the bottom crease of of the bag.
The construction of the phantom consists of laying the test objects in a thin sheet of dental wax and pouring a top layer of wax to fix them in place. The test objects were placed using an alignment fixture, positioned in place with a "twee" (one jalf of a tweezers). The assembler, who had made literally thousands of these phantoms, expressed his extreme frustration with the sapphire balls. Unlike the corundum particles which stayed in place when positioned, the round balls would roll around. Furthermore, at .006" diameter, the smallest balls were nearly invisible and if one escaped, you could forget about ever finding it again.
As it turned out, the balls imaged extremely consistently but the images were easier to see because of the uniform shape. The solution would have been to change the standard to use a smaller ball but this was far easier said than done. Although our company had developed the original phantom, the accreditation program was governed by federal legislation and administered by the American College of Radiology and any changes to the phantom would need to be approved by unanimous agreement. Unfortunately, I left the company before all this happened and to the best of my knowledge, the product revision never occurred.
One thing about the sapphire balls of interest was they appeared to be perfect spheres, even under a microscope. It seems incredible that the manufacturing process was able to achieve that level of form and finish on an object .006" in diameter and as hard as sapphire.