Troy:
That’s an ad hoc genius idea.
My concern would be that we have two different burst failure problems: the
low pressure side had to fail toward low pressure, the high pressure side
has to fight the pressure on the other side of that disc. The charge
required for the latter is obviously greater than that required for the
former. And in both cases we are venting hot gas into our system....
Bill
On Sun, Oct 25, 2020 at 5:31 PM Troy Prideaux <troy@xxxxxxxxxxxxxxxxxxxxx>
wrote:
I helped a guy with PSANCP research for his master’s thesis years ago. He
built a pretty snazzy static testing rig c/w burst discs. He pretty much
standardised on cutting all the discs from Al-Alloy soda cans whose
material specs are apparently all within a pretty narrow tolerance. Being
experimental PSANCPs and being tested a across a broad spectrum of pressure
ranges, the burst discs got to experience multiple situations of
“relieving”. He tested/characterised them with various seat designs from
sharp edge (sharp shearing of the BD) to large radius that provided more of
a typical burst. The sharp edge seats would obviously fail at much lower
pressures but provide a full open conduit when they did fail.
Maybe 1 possibility might be use 2 burst discs with the pyro between
them. Arrange the BD seating so there’s a sharp shearing failure in the
direction of the pyro pressure (against the pressure source) so the failure
point is both lower in that direction, but also provides a full flow when
open. This is one of my ad-hoc/on the fly ideas without much thought put
into it, so add appropriate portions of salt.
Troy
*From:* arocket-bounce@xxxxxxxxxxxxx [mailto:arocket-bounce@xxxxxxxxxxxxx]
*On Behalf Of *William Claybaugh
*Sent:* Sunday, 25 October 2020 8:10 AM
*To:* arocket@xxxxxxxxxxxxx
*Subject:* [AR] Re: Burst disc thickness
Steven:
Thanks, that conforms with my thinking.
I’m testing the initiators this next week: two forms of nitrocellulose
(cotton and cord) to get a working unit with an initial cut on the amount
of gas generated, I am assuming the cord will prove easier as it can be
cut to a desired length vs. the guncotton which will need to be weighed.
I’m having disks made in three thicknesses in accord with the circular
flat plate model and will test each to burst in the valve assembly once
those are finished.
Final testing will then use the initiators to punch through the disk under
pressure...the trick there will be to find an amount of actuating gas that
does the job with one charge working but does not over pressurize with two
charges igniting. (I’m using dual redundant initiation In accord with
standard practice for pyro actuated devices.)
Bill
On Sat, Oct 24, 2020 at 8:36 AM Steven Berg <skyshredder9488@xxxxxxxxx>
wrote:
From my experience in using the flat plate model to design burst disks in
a few actuated systems, we used a circular flat plate to go into testing.
The overall thickness was typically all that needed to be varied. Adhesives
and clamping in the design were also the other first points of design
review.
On Fri, Oct 23, 2020, 2:27 PM William Claybaugh <wclaybaugh2@xxxxxxxxx>
wrote:
Having inhaled deeply of the arcana of burst disc sizing; I find that the
entirety of the available math is about sizing the plumbing, not about
design of the burst disc.
For my current actuated burst disc design I have a 1.5" diameter disk that
is clamped at 1.25" diameter. Operating conditions are air at (probably
something better than) -50 degree F (cold soak on the way to 200 Km) and at
an original 125 psia. Using the standard flat plate model I get this to be
0.050" thick disc of 6061-T6 providing a safety factor of 1.43.
If I am going to manufacture my own disk (an assumption worthy of review),
then is it correct to just assume a circular flat plate for initial design,
pending testing?
I will be testing prototype initiators next week; for now I am using an
e-match with a very small bit of nitrocellulose to generate the pressuring
gas. The nitrocellulose is in both "guncotton" and cord form. I expect
the former will be faster but the latter possibly easier to control since
cord length can be directly measured.
All testing will be at a local energetic materials facility...
Bill