.03 is a sensible number for a shell design. The reason people mill
channels in is partly to make them more manageable sizes. The other
option is to do something like the LR101, where it's basically a shell
but it has a wire spiral up the engine that forces the flow to be one
larger spiral channel, rather than one very thin annular channel of a
shell.
Aluminum film cooling is going to be frustrating. Film cooled engines
optimize to be as hot as possible while not failing structurally.
Haynes, inco, or other nickel alloy would likely be most robust.
Something refractory is plausible, though they tend to be fragile.
Stainless is most practical for a hobby or school project; machinable
and relatively cheap. Even mild steel will allow higher temperatures
and would be really cheap to make.
Armadillo's long line of film cooled engines were stainless.
Fabricated a lot of different ways, including casting, machining from
solid, spinning, and mixes between them. The 4000lbf LOX/methane
engine I built there had a beautiful machined nozzle that was welded
to a rolled cylinder, with stiffening rings so that the ID could be
lathed round to match the machined bolted flange.
Ben
On Tue, Jun 27, 2017 at 4:10 PM, Zachary Martinez <znm3m8@xxxxxxx> wrote:
I have considered a hand calculation. However I really have no idea how I
would calculate the convective heat transfer coefficient. Should I use the
gas side equation because it is on the gas side or the liquid side equation
because the surface will be wetted by the liquid? Should I consider the
convective heat transfer to the film from the combustion products to be the
same as it would be to the wall if there was no film there? I also don't
really have a good idea of how to model the radiation heat transfer from the
reaction products. RPA is great for this kind of calculation as it looks
fairly involved.
Even the Bartz equation is pretty involved in the number of things you need
to get right. I have gone through a regenerative design using Bartz a few
times on my own both by hand and using Matlab and at the end while I did get
an answer, I definitely did not feel very confident in my results. I kept
getting impossibly small required fluid flow gaps around .03" when using a
coaxial shell design even if copper was used. I have probably spent a month
on this problem on and off and was never able to get a different/better
number.
Back to the whole point of this. We plan to use 25% film cooling for the
engine (I am assuming this means of fuel flow and not of total mass flow
someone please let me know.) I am fairly sure the design I am using will be
fine thermally based on what I have seen others get away with. However, I
have no way to definitively predict the wall temperatures which is
especially important since aluminum is being used.
The most reliable hand calculations I have done have been from an energy
balance point of view.
Assumptions made in these caclulations:
Heat flux to the film is equivalent as the gas side heat flux to a cold wall
Gas side heat rate was determined by graphically integrating the heat flux
curves from RPA when you put in a wall temp of 300K
The film absorbs all the energy required for it to evaporate before it
leaves the engine
At a given time the entire engine is at the same temperature.
I got 4.5 seconds of burning as "safe" i.e. not heating the aluminum as a
whole past its weakening temperature. The bad part about these numbers is I
never get a throat temperature so I don't know if our copper throat insert
is going to be damaged or not and there definitely could be other hot spots
in the engine where the aluminum exceeds the safe temperature.
Zachary Martinez
Aerospace & Mechanical Engineering
Missouri S&T
On Tue, Jun 27, 2017 at 5:11 PM, William Claybaugh <wclaybaugh2@xxxxxxxxx>
wrote:
Zac:
Have you considered a hand calculation?
Sutton's liquid rocket engine example shows how--although I expect you
understand that part. So doing, or even setting up a simple spreadsheet,
will let you know what you should expect from your simulation. For a test
engine, checking heat transfer at the chamber wall, the throat, and the exit
is probably more than sufficient; no simulation required.
In any case, 20% BLC should be plenty: I once saw a 5000 lbf. stainless
engine do two 2.5 minute runs on the same day--pausing only to refill the
tanks--using nothing but BLC. It was pristine after the second run.
I have often thought were I ever to build another liquid, I'd just spin
the TCA from minimum gage stainless tube and carry 20% more fuel: it's net
way cheaper than even an ablative TCA. I'd also look hard at hydrogen before
falling back to methane.
Bill
Sent from my iPhone
On Jun 27, 2017, at 12:42 PM, Zachary Martinez <znm3m8@xxxxxxx> wrote:
Yes I understand that part. I was always assuming that the film was
transparent to the radiation. What I am trying to do is different from that.
Say radiation heat flux is 500kw/m^2 on the inner wall of the chamber.
Nothing I do can change that number. This effect alone makes the inner
surface in steady state conditions unbearably hot. The only way to reduce
inner wall temperature is a) reduce the wall thickness so and get rid of the
heat using radiation and natural convection on the outer surface or b)
actually transfer heat from the inner wall to the cooler film. As our engine
has a couple SAE ORB ports on the side for pressure measurement reducing the
wall thickness makes manufacturing more complex so we were trying to go with
option b and simulate this somehow.
Zachary Martinez
Aerospace & Mechanical Engineering
Missouri S&T
On Tue, Jun 27, 2017 at 1:34 PM, Alexander Ponomarenko <contact@xxxxxxx>
wrote:
Hi,
Indeed, within thermal analysis in RPA it is supposed that the liquid
film is fully transparent for radiation (that is, the it is using the
temperature and the properties of the wall to calculate the radiation heat
transfer).
Still, you may affect the value of radiation heat transfer by adjusting
the value of "Emissivity of gas-side wall surface" (default is 0.8) on the
tab "Heat Transfer Parameters". By reducing this value, you may simulate the
lower transparency of the film. Of course, such approach won't consider
heating of the film from absorption of some part of radiation energy. And
you have to find the reasonable value surface emissivity on your own.
Regards,
Alexander
On 06/27/2017 08:17 PM, Zachary Martinez wrote:
Good Afternoon,
Does anyone have experience in analyzing the the thermals of an engine
that only uses film cooling (ie no regenerative or ablative). I was trying
to simulate this using RPA but was coming up with strange effects.
It seems that in RPA if you set up a film cooling section it calculates
the convective heat transfer from the film to the wall independently from
the radiation heat transfer from the hot core. When I adjust the film
cooling parameters I can get the convective heat transfer to the wall to be
0 W/m^w in sections (ie. fully insulated). However because the radiation
heat transfer is calculated separately I am getting a steady state inner
wall temperatures in excess of 1000K which is far beyond the melting point
of aluminum. This means even if I simulate the engine with over 50% film
cooling I can't get a steady state solution that is feasible.
Are there any tools that I could use to simulate this heat transfer
problem where the convective heat transfer is negative (ie the film is
actually cooling the chamber). I know some on this list have developed
engines that can run steady state using only film cooling, how did you do
it?
Thank you,
Zachary Martinez
Aerospace & Mechanical Engineering
Missouri S&T
