Arrow Mon Sep 23, 2024 11:36 am
https://t.me/kramnikcat/4935

Kramnik's Cats and Cat
0:17
Regarding the Sarmat tests. You can't skimp on technical projects, you can't skimp on tests. All design features of any product in accordance with GOST RV 15.203-2001, regardless of the modification of this GOST, are laid down at the technical project stage. At the technical project stage, the main technical decisions are made, after which the technical project is approved by the customer. The customer is periodically tempted to skimp on the technical project, which should, in theory, be stopped at the military-industrial complex level, but this did not happen here.
Second, you need to increase funding for testing components. Autonomous preliminary tests, factory tests, all these tests. The more stands, the more tests, the more likely it is that by the time you enter the LKI (flight design tests) you will come up with a normal product.
And there is only one recipe for success at the LKI: shoot more. This is not the case when modeling decides anything, it can cut off some things in advance, but a lot still depends on physical testing.
The R-36M (15A14) was tested in the Soyuz from 1973 to 1976, launching 43 missiles during the tests. 7 launches were unsuccessful. The R-36M UTTKh (15A18) got by with 19 launches (two unsuccessful) from 1977 to 1979, while its 1st and 2nd stages were identical to the 15A14.
The R-36M2 (15A18M) was tested in 1986-88, 26 products, 20 successful launches.
The Yars was tested very briefly, getting by with three launches without failures, but given the high continuity with the Topol-M, it was possible to afford it.
If the Sarmat had gone without a hitch, it would have been fit into 6-7 launches over 3-4 years and that would have closed the issue, but given that the missile had to be made from scratch - its predecessors, let me remind you, were made by the Yuzhnoye Design Bureau and Yuzhmash, it did not go without a hitch.
And in these conditions, only more firing, testing each significant addition/change with a launch, and firing more often, 4-5 products per year, better 7-8, and not once every six months, so that problems with cooperation and staff turnover do not creep in, those who are not on a chain, and then find a better job, then grow in their careers, then retire, then new ones come.
Here is a picture from a citizen who knows that you need to burn products during tests as much as possible.
266.8Kviewsedited Sep 22 at 13:43