Gallery, Projects and General > Project Logs

Liquid Fuelled Rocket Engine

<< < (2/4) > >>

British Reaction Research:
Hi Bill,

Thanks for replying and thanks for wading through what I had written. If you were to re-read it with a view to commenting on my approach to the problem then that would be most gracious of you. I would be very grateful indeed.

The boiler economiser you have designed sounds very interesting indeed. I seem to have found the right person to look at my results!

I'm planning on using a gas pressurised feed for the engine, at least to begin with. Some other calcs that I've done to determine the Reynolds number indicate that even with an annulus type parallel flow coolant gap in the jacket of 5mm - easily achievable - the Reynolds number would be in excess of 127000. Anything above 10000 in liquids and you have turbulent flow.

As I mentioned one plan I have is to use cross flow - the coolant flowing perpendicular to the axis of the nozzle. This would induce even greater turbulence and vortex generation, all of which are good for heat transfer. Knurling or milling grooves or slots in the nozzle outside wall would also induce turbulence and eddies, as well as becoming nucleation points for a boiling heat transfer regime.

I think that one thing I didn't mention in my above post is that the numbers would be for steady state operation. As the rocket engine is likely to be run for short durations it will probably never attain steady state conditions.

Another point I forgot to make is that the engine ought to be tested with water as a coolant to begin with. As is well known, water is a formidable thermal fluid and will easily cool the nozzle and chamber in the way described in my post above. The inlet and outlet temperatures can then be measured and from that the performance of the cooling system can be checked. Only when it is proven as safe will I try to regeneratively cool the system with the fuel.

Thanks again for taking the time out to read my post and to comment on it.

Carl.

PK:
Did I miss-read or are those calculations just for the throat section of the motor?
The throat is the hardest part to cool, but it's not the biggest source of heat. Particularly in a development motor where you tend to run conservative (read big) L* chambers.

Re strength. If you make the cooling jacket and saddle (I'm assuming you're doing it that way) such that it can take some load, then you can run thinner chamber walls and get better heat flow.

One important point, and I assume you are doing this. I know of no biprop that doesn't rely on film/boundary layer cooling. If you push your rich mixture to the edges of the chamber you can pull 300deg out of the combustion temp near the walls and still get reasonable ISP because the rest is nearer to stoichiometric.

British Reaction Research:
Hello Pk

Thanks for taking the time out to read my post and to comment. You did not misread, these calculations are for the throat section. I wanted to base them on the hardest section to cool to try to get a good feel for the problem.

I take your point that the biggest heat load is not the throat, despite it being the hardest part to cool.

I'm not currently planning on using film cooling as such. That said, the injection system will be putting a lot of fuel against the walls. What I haven't so far mentioned is that I also want to mix some silicon oil in with the fuel. A small percentage only. This will burn with the fuel and the theory is that a glass like compound is formed that deposits on the chamber wall. One issue with heat resistant coatings is that they can Spall and flake off. Using this method the coating is being continuously reapplied throughout the duration of the burn.

This acts as an insulating layer that will reduce heat transfer to the walls. I haven't factored this into my heat transfer calculations as yet.

The L* I'm using is about 0.8 metres, which from what I have read is on the low side for a nitrous/alcohol motor.

Thanks again for your wisdom. I'll definitely be looking at incorporating film cooling.

Carl.

billmac:
I wonder whether the silicon oil might preferentially coat the inside of your cooling ducts. If so that could have bad consequences for the rate of heat transfer that you can acheive. As you probably know, the presence of oil, even in very small quantities is a real no-no for steam boilers for that very reason.

PK:
NOX biprops are (to my understanding) poorly characterised. My own efforts don't really count because the motors were tiny, but I have made a LOT of NOX hybrids over the years.
Based on that I'd suggest a longer combustion chamber. The reason being that it takes time for the N2O to heat up to the 500degC or so that it needs to reach in order for it to let go of it's oxygen.  This (and poor injector design) is what gives that classic bell shaped regression in a hybrid fuel grain.
Maybe build a water cooled, boiler plate, motor first, just to remove some variables and arrive at a correct L*.

For the non rocket nerds: It takes time for a given fuel and oxidiser combination to atomise, mix, and burn and, when you are flowing kilograms of propellant per second, time==distance.

If you make the combustion chamber too short then some of that combustion might not happen inside the motor.  It will look great because there will be an enormous flame emitted from your creation, but the sad reality is that all that heat will contribute nothing to generated thrust. 

You can't (normally) just make the motor a lot longer because the amount of heat you need to remove from the combustion chamber walls is a reasonably linear function of the surface area of those walls and you never have an excess of cooling capacity. 

As such, much work goes into characterising the combustion process of a given propellant combination over a range of pressures, flow rates, and o/f ratios. The end result of this is a set of L* constants which you can poke into some simplified maths to get the length of combustion chamber needed for your motor.

Navigation

[0] Message Index

[#] Next page

[*] Previous page