These videos were captured at the SpaceX McGregor, Texas engine test facility from off-site by McGregor Live. Note that in the failure of Raptor 2 SN 144 on 2022-08-24, the engine plume turns green just before shut-down of the engine is commanded. This is indicative of “engine rich combustion”, where copper in the combustion chamber and/or nozzle throat is burning due to a coolant failure and/or burn-through. This is visible in some of the earlier failures as well.
It is not known whether these were intentional tests to destruction (for example, to establish absolute limits on operating parameters such as chamber pressure, or to verify that a failed engine can be shut down in time to prevent an explosion that would cause fratricide in other engines in a cluster) or were mishaps in tests intended to complete normally.
I recall a particularly stirring speech by a VP at a place I once worked. To paraphrase, he said that if none of our projects failed, then we were being too risk averse and would be missing out on potential opportunities.
Of course, the organization still preferred successes to failures!
There used to be a similar reasoning in medicine. In the old days, before abdominal CT scans made the diagnosis of appendicitis all but idiot-proof, it was generally believed that if a surgeon did not have about a 10% incidence of operating and finding a normal appendix, that he/she was not being sufficiently aggressive, i.e. erring on the side of operating too frequently was preferred to operating too infrequently. At that time, there was no definitive test, like CT, for appendicitis - one had to rely on a constellation of ‘soft’ signs & symptoms with inherent unreliability. The prospect of missing a true case of appendicitis - with the possibility of rupture and life-threatening peritonitis - was too dire to allow erring on the side of missing one. This seems the same reasoning in areas of engineering.
This leads me, at least, to some philosophical speculation. There is surely much overlap in reasoning between physicians and engineers - with good reason. It just may be - once one sets aside a very strong intuition of élan vitale, that biological creatures can be considered to be extremely complex carbon-based machines, just immersed in blood and other bodily fluids. Who knows, it may turn out that the biological cell (from bacteria to the mammalian cell or somewhere in between) represents the shortest route to development of micro-machines.
There is a whole genre in statistics dealing with Type I errors & Type II errors, i.e. False Positives and False Negatives. Hence it is easy to show that most of the people who tested positive for Covid were not infected, while a significant number of those who were infected were given a clean bill of health. Unfortunately, this is statistics, which most of us mortals can barely understand.
Not so sure about that. One has to make allowance for the reality that most people are not infected with Covid. Even a small percentage of False Positives applied to the great majority of the tested population yields a significant share of the reported Positives being False.
The study you cite uses PCR tests to verify the accuracy of rapid antigen tests. But a positive PCR test does not necessarily indicate an infection. The qualitative result of PCR, which was never intended as a diagnostic tool, is highly dependent on the Ct value that is used. Hence, the official number of COVID cases can be easily manipulated by changing the Ct value. Early PCR tests were too sensitive, often picking up nothing more than “dead nucleotides”. Nevertheless, the inflated case numbers (and by extension, “COVID-related” deaths) were used to justify public policy response including the terrible lockdowns.
