Further to the discussion above on calculating reliability from destructive
firing trials, which took place several weeks ago, I stumbled across a Test
Plan at work the other day. It is a plan to test the reliability of two
elements of an explosive train (detonator to lead charge in this case, but
of course is just as applicable to other elements of a firing train, or to
other applications entirely).
I haven't been through the maths in detail, but essentially the
introduction to the test plan says that by increasing the detonator to lead
charge distance by 1.5 times the nominal value, carrying out 20 tests
(assuiming they all fire) gives you a theoretical reliability of 99,9999%.
This makes a few assumptions of course, and only takes into account one
factor (det->lead separation), but I found it useful, hopefully someone
else will too...
I've attached a scan of the two relevant pages (minus some reference
numbers and some other info that could possibly identify where the document
came from, although of course in this age of social media a quick search of
LinkedIn will give you a good idea ;-)
Hope someone finds this interesting or useful,
James.
On 1 November 2015 at 11:00, Uwe Klein <uwe@xxxxxxxxxxxxxxxxxxx> wrote:
Am 29.10.2015 um 03:44 schrieb David Weinshenker:
On 10/28/2015 09:27 AM, James Padfield wrote:
I suppose I should have given a bit more background...
I am testing the reliability of an explosive train, i.e. from booster to
main charge -
Ah - so the desired outcome is that the barrier -does- rupture under
pressure, allowing the explosion to propagate. If the object of the
exercise is to confirm practical (vs. statistical) reliability, then
your proposal for building test assemblies with slightly thicker
barriers does make some sense.
You can always do larger scale changes for one parameter and
see where you leave the comfort zone.
here forex reduce the size of the primer charge until you
get into nonworking territory ( you'll pass a point of marginality. Repeat
for some other
parameter changed slightly. Prove that charge size scales well with
the changed parameter.
uwe
Attachment:
Test plan - initiation train reliability.pdf
Description: Adobe PDF document