Thanks for all the help so far guys, lots of useful info. To expand on the
initial post I'm imagining LOX/Hydrocarbon and like-on-like injectors are
probably the most common selection an amateur would make, primarily due to
the vast amount of information out there but also because of relative ease
of manufacture (compared to coaxial) and propellant supply chain reasons.
I did just dredge up one piece of information I think is interesting from a
patent regarding Russian LOX/HC coaxial injector designs:
http://www.google.com/patents/EP1022455B1?cl=en
In particular when you look at the picture of the injector distribution,
other than some longer posts which form baffles (also used in SSME) you'll
notice that the remaining injector elements are all members of three
different groups. This paragraph explains the groups:
- [0041]
The injectors, except those located near the internal fire wall of the
chamber casing, should differ in mutual location and definite step in the
fuel flow rate, and therefore they are divided into groups according to
fuel flow rate growth (three groups are given consideration as an example).
In this case, the injectors of different groups are made in such a way that
the mass flow rates of fuel for adjacent groups at the nominal operation
mode of the engine differ by no less than 3% and no more than 10%. The
introduction of injectors with different flow rates is necessary in order
to reduce the effects of high-frequency oscillations at operating modes of
the engine. During the mounting of the injectors 13 on the mixing head 1,
the injectors are arranged so that injectors of different groups are
adjacent each other, the groups being positioned in the chamber according
to a definite law (for example, by cyclic sequential helical repetition of
the injector arrangement from the first to the last group). A schematic
representation of the arrangement of injectors 13 on the mixing head is
shown in Fig. 2 as a section A-A. Wherein, each injector of a corresponding
group is designated by a circle with a wholly colored sector of
corresponding size:
- An injector of the first group is designated by a colored sector
which is 1/4 the size of the circle;
- An injector of the second group is designated by a colored sector
which is 1/2 the size of the circle;
- An injector of the third group is designated by two diametrically
opposite colored sectors which are 1/4 the size of the circle.
So they deliberately vary the O/F ratio of each adjacent element by no LESS
than 3% specifically to mitigate the effects of high frequency combustion
instability. They also make mention of a 'definite law' with regards to how
the elements are varied (mentioning a spiral pattern as an example,
possibly not in reality). This would seem to confirm my thoughts of using
element pattern variations to introduce 'soft spots' into the combustion
region to absorb acoustic wave energy, although I was thinking of much more
organized/large scale discontinuities while this patent would appear to
show more of a 'white noise' type O/F ratio variation across the injector.
Andrew
On Sat, Jan 2, 2016 at 4:42 PM, Henry Spencer <hspencer@xxxxxxxxxxxxx>
wrote:
On Sat, 2 Jan 2016, Andrew Burns wrote:
...I'm not sure that there's really enough information available
in public literature to properly estimate the tendency of a given injector
design to instability. From what I've seen the traditional process has
been
testing and semi-trial and error redesign based on hot-fire results. That
said there are plenty of examples of calculated 'stability maps' so
somebody
must have either a theoretical or empirical method for estimating
stability.
Lots of them. The relation of any of those methods to reality, however,
is a different question. Combustion stability is still a black art, with
no convincing theory and at best rough empirical correlations, so it really
is ultimately based on testing and rule-of-thumb redesign.
(To be more precise: that's the situation for high-frequency instability,
which is typically driven by interaction with combustion; low-frequency
instability, typically driven by interaction with propellant flow and
injection, is better understood and typically less problematic.)
The bible on the subject comes in two parts. The Old Testament :-) is
Harrje & Reardon, "Liquid propellant rocket combustion instability", NASA
SP-194, 1972, a massive phonebook-sized slab that's long out of print but
can sometimes be had secondhand, and is probably available online as a PDF
now (but since I have the slab I haven't looked). The New Testament :-) is
Yang & Anderson, "Liquid rocket engine combustion instability", AIAA
Progress in Astronautics and Aeronautics vol. 169, 1995, which I think can
still be had from AIAA at an excessive price, might be findable used, and
almost certainly *isn't* available free (AIAA doesn't do free). Y&A was
meant as an update to SP-194, so for completeness one wants both.
It's getting to be about time for there to be a Koran :-) to follow the
New Testament, but if anybody's done one, I haven't heard about it yet.
Both books have considerable useful advice, but the most interesting bit I
found in either one is in the Muss paper in Y&A: plots (page 84), for two
different LOX/hydrocarbon injector types, of strong correlations between
D/V (orifice diameter over injection velocity) and a critical frequency,
with stability problems to be expected if the chamber is big enough to let
that frequency (or a lower one) resonate.
- The chance of instability is increased as injector performance
increases... Most tactics to increase performance also decrease stability
and vice-versa.
This is pretty definitely the case for the like-on-like injectors used in
most big non-LH2 engines. Whether it is universally true is less clear. In
particular, other injector types (triplet and coaxial, in particular)
probably don't entirely escape from this, but might shift the tradeoff
enough that it's no longer a big problem for a particular design.
For this topic in particular, beware of assuming that the design space is
well explored. Much of the conventional wisdom is empirical correlations
only, often derived from fairly small regions of the design space and quite
possibly not applicable elsewhere. Discovery of major improvements remains
possible; indeed, there are hints that some of the newspace companies may
have made real advances here.
Clearly there are also traditional approaches to controlling instability.
For example using baffles to compartmentalize the injector face...
Acoustic cavities can also be added around the injector face...
Another possibility: as you may have noticed in another discussion,
although transpiration cooling is generally in disrepute, there is
extensive use of it for cooling LOX/LH2 injector faceplates in particular,
and some of those engines appear to be unusually stable. There has been
speculation that the porous surface may function as an acoustic absorber to
some extent.
- Is there anything out there to read or that people know with regards to
controlling instability purely by injector pattern or element design?
The two books noted above have quite a bit of this, although it's not all
universally applicable or entirely consistent -- more of a collection of
hints than a coherent design methodology.
example it's my understanding that deliberately delaying mixing and
combustion such that the heat release zone is away from the injector face
increases stability but also decreases performance.
Delay certainly tends to decrease performance. How much it really affects
stability, except in the well-studied case of like-on-like injection (where
it does improve it markedly), is less clear. See note above about limited
exploration of the design space.
- I've seen references to linear and non-linear instability however I'm
still not entirely sure about what the difference is, anyone able to
explain?
As in many other things, it's basically the distinction between things we
can mathematically analyze (linear systems make that *so* much easier) and
things we have only vague clues about. There is unfortunately a lot of
nonlinearity in combustion instability, which is why the mathematical
analysis is of very limited help in practice.
- Most thrust chambers are simple cylinders with flat injector faces
however is it possible to modify the geometry of the combustion chamber to
increase damping of acoustic modes or to re-direct their energy?
In principle, yes. It's poorly explored. (Note that many injectors are
at least somewhat dished for other reasons.)
Henry
