[How do I?] Lathe/mill A Converging, Internal Angle Deep Inside Part

[knoq]

Attached is a schematic I was given by a friend to lathe up for him.

What I need is essentially a cone on top of a cylinder, but cutting a 30(deg) angle deep inside the part without a cnc lathe is.... challenging (I think).

I'm familiar with the design and have made things similar before (it's a rocket, yes) but he has some interesting angles/dimensions compared to my designs which usually allow me enough room to enter the nozzle-side with a boring bar and get the angle that way.

I've considered maybe a drill bit or something of that kin may exist, but no luck in my searches.

I also thought about a custom-ground form tool, but it's too large for pretty much any tool holder I could access.

I have no experience with a mill, but I do have access to one and someone who can help me.
Because of my lack of knowledge, there may be a solution there I do not see.

Note: The schematic is a cross-section, and I know it's a mess! I combined multiple views of it to compress it for the forum



Help?
Zach

 

[kd4gij]

Even on CNC that would be a challenge. Would need a 3/8" boring bar hanging out 4" Maybe in 2 pieces Screwed or welded together.

 

[Bob Korves]

Welcome to the forum, Zach.

First, help us! What is the material you are planning to make it from? Are the four significant digits to the right of the decimal really necessary to the design? That is a tough to build part, regardless. Could you tell us what it has to do, so there might be a easier to build solution?

 

[Steve Shannon]

That's a converging-diverging nozzle, also referred to as a de Laval nozzle. High pressure gas on the left passes through it and becomes high velocity gas on the right (in the illustration posted above).
They are frequently made of graphite, sometimes phenolic, occasionally stainless steel or othe very high temperature metals (steel is not allowed in most hobby or sport high power rocketry).
Most solid fuel designs do not include the long straight walled section on the left. That makes it much easier when to machine. Instead, the nozzle is sealed into a cylinder. Most designs have a shallower (more nearly perpendicular to the combustion chamber wall) angle on the left or converging side. This one looks more like one used for liquid fuel. Very interesting!
Here's an article and simulator:
http://www.engapplets.vt.edu/fluids/CDnozzle/cdinfo.html



Steve Shannon

 

[Chipper5783]

What tolerance are you looking for? How close to concentric does it really need to be? If you could work it from both ends, life would get much easier (concentricity will be pretty easy down to within 0.002"). Leave your work piece a bit long on each end for gripping. It would help to have a chuck with a bore large enough to tuck most of the work in side.

Obviously plan the sequence of operations. You will end up cutting away your work surfaces.

 

[kd4gij]

The only way I can see the way it is drawn Would be to step out the ankle then go back and smooth it out. Would be a long drawn out process.

 

[Steve Shannon]

Sometimes they are spun. Here's how Copenhagen Suborbital spins theirs, which is also a two wall design.

 

[knoq]

Sorry for the delay in getting back.

So, the material is 6061 aluminum (quite familiar with it, and used in this application often).

Steve is correct, converging-diverging.

As for tolerances, the most important tolerances are the throat (smallest radius you see) and the exit radius (far right).
They do not need to be 4-decimal-places accurate, no more than +/- 0.003 from some basic CFD on it.
Again, some radii need to be more accurate than others....

Hope that helps some!
Thank you all for the speedy response

 

[knoq]

(sorry for double post, but better than edit)

So, I was also considering a long-thread mating between the flat wall section and the conv-div section, but not sure how the pressure would hold up or if I could seal it sufficiently (pressure).

That's not something I expect input on, as I have to go model the forces myself to even prove it useful, but throwing it out there.

 

[Steve Shannon]

If you're using aluminum, your fuel must not be erosive at all. Also, that may make it really easy to just make a couple of tapered reamers, one converging and one diverging, so you can turn these out quickly. Nozzles are expendable items. Even the reusable graphite nozzles usually only get 20 burns or so.
We usually have these fit into a liner of some kind and then an o-ring sealed against the nozzle by the aft motor closure.



Steve Shannon