A decision that needs to be made on the overall structural architecture of OHKLA | Rocket body discussion | Forum

This thread is supposed to give a separate, logical home to an off-topic conversation rpulkrabek and I have been having in another thread.  It's about a relative major decision which needs to be made regarding the overall structural architecture of the OHKLA rocket.  The sooner we make this decision the better because it will to some extent dictate other design decisions.  I guess we should make it relatively soon.  I don't think we should necessarily use a poll like we did with the propellant thing because there are fewer choices and probably less subjectivity.

The decision is whether to adopt one of two obvious ways of constructing the body of the rocket:

Option A: we build the combustion chamber and oxidiser tank to have exactly the same outer diameter, join them together with some kind of connecting ring and allow this configuration to be the lower part of the rocket, i.e. the outer surface of the chamber and tank will be "wetted" by airflow during flight.  See the current OHKLA diagrams in the wiki (I wont' post them here because they're huge), or this diagram from CS:

Option B: we use a single tube, called an aerostructure, for the lower part of the rocket's structure and insert the chamber and tank inside of that tube, holding them in place with bolts through the tube, like this diagram:

As near as I can tell, Option B is by far the most common in the sounding rocket world.  The only example of option A I have found is Copenhagen Suborbital's rockets (which have never been flown).

These are some of the relevant issues:

• Mass: Option A is less massive because there is no aerostructure.  Depending on what the aerostructure is made of, this could be a considerable mass saving.  On the other hand, thin aluminium or fibreglass might make an aerostructure so light that this is not a huge deciding factor.
• Structural strength: Option B feels to me (and admittedly this is a gut-feeling, people with better structural knowledge please correct me if I'm wrong) considerably stronger than Option A.  I feel like with a rough landing the connection between the chamber and tank in an Option A rocket could break.  This could possibly even happen in flight due to turbulence?
• Aerodynamics: Option B results in a lower rocket body which has perfectly uniform diameter and can easily be painted and polished smooth, which are both good for reducing drag.  It may be possible to achieve this with Option A if we are clever but it will be more work and even the best results will probably not be as good as possible with Option A.
• Greater freedom of chamber dimensions: Option B allows greater control over the L:D ratio of the combustion chamber (and hence fuel grain) because we can size the chamber independently of the tank.
• Ease of fin mounting: Option A requires either bolting fins directly into the combustion chamber, which is a pressure chamber making this somewhat dangerous, or welding fins directly onto the combustion chamber, making swapping out damaged fins from a rough landing very difficult.

It seems to me like Option B is superior in most respects with the obvious exception of mass, but this could be the result of personal bias.  If anybody can think of other arguments in support of Option A, please share them here.  The same goes, of course, for additional arguments in favour of Option B.

Not as anything final but just to get a judge of what people's guts say, could everyone, in their first reply to this thread, say which option they'd pick if they had to make a gun-to-their-head choice right now?

Glad you agree.  I was trying to find some details on mass and costof appropriate aluminium tubing to act as the aerostructure, but was driven mad by how hard this is!  It seems like 90% of people who sell aluminium tubing sell it in forms where wall thickness is correlated with outer diameter, i.e. by the time the OD is up to around 25 cm (or a corresponding number of inches – finding metric suppliers in the US is tough :() the wall thickness is around 0.5 inches, which I feel is a lot thicker than we actually need for this application.  I haven't yet been able to find some aluminium tubing that's around 25 cm OD and around 2mm thick.  Still, I've just been blindly googling, people with more experience sourcing metal may be able to do better.

As an important point for those searching for metals, we definitely want tube and not pipe.  From onlinemetals.com:

A brief discussion of the difference between pipe and tube:

• Pipe is generally more rigid than tube, and is usually produced in heavier wall thicknesses.
• Pipe
  is specified by a nominal dimension which bears little or no
  resemblance to the actual dimensions of the pipe. 1" Schedule 40 pipe,
  for instance, has an actual OD of 1.32", a wall of 0.133", and an inner
  diameter of 1.049". Tube dimensions are actual dimensions.
• Pipe fittings are sized to meet pipe sizes, but not tube
  sizes. A 1" schedule 40 nipple will fit correctly on a 1" schedule 40
  pipe, but not on a 1" OD tube.

Translation: pipe is heavy and the people who manufacture it are braindamaged.

As an alternative to thin-walled aluminium tubing for the aerostructure, could we consider fibreglass?  I don't have much experience with fibreglass, I don't know how strong it is, or really what sort of stresses the aerostructure may be subject to.  I figure it would be lighter than Al, though.

rpulkrabek said:

I think that laying down a first model as a basis is the best way forward. I have tended to think that all along, which is the reason I drew up the first OHKLA concept. Thus far, I think it has worked quite well. We are now all in a common understanding where our thought process is going and what it looks like. It has now been evolving to the point that the aerostructure is better. I still think we should all decide on every aspect though. We can update the rocket model, make a note that certain things have been decided and that we are still investigating other decisions. So, I think we should go towards an evolution route, rather than an iterative route. We have a base design, but then decide the details as we go and then up date the design. We don't have to have a vote for everything, because some decisions are common sense. We should have a vote on the controversial decisions.

---

Thanks for replying rpulkrabek, I was looking forward to what you had to say on this.  Could you clarify exactly what you mean by an "evolutionary route"?  I'm not sure how what you describe differs from what we basically already had.  How does your evolutionary route plan deal with the problem of making optimal choices for individual components "in a vacuum", with other components not specified?

rpulkrabek said:

As for you FEA/CFD question. I am not sure if I can analyze a temperature profile based on velocity and atmospheric density (are you thinking the temperatures involved with atmospheric reentry?). What I can do is give a temperature and see how heat transfers. I should also be able to determine optimum aerodynamics, in a similar way as I have determined the best geometry for the nozzle before. I can do an analysis for the shear force based on the wind speeds, but I am not so sure how to go about the rotational rates. I, at least, haven't tried these things before. I'm sure some how it can be done, though.

---

That all sounds like it should be pretty helpful.  I do think we should try to strike more of a balance between numerical and theoretical techniques, however.  I was always kind of nervous with the nozzle being designed largely by CFD-trial and error when I know there is properly theory out there about ideal expansion ratios, etc.

joe.haydu said:

Didn't a mechanical engineering senior just join up? He might be the ideal person to address CFD questions. I haven't done any CFD related stuff in years, but if needed, I can brush up and do the calculations. In either case, this http://dspace.mit.edu/bitstrea…..367392.pdf pdf has some great information, and section 3.4 deals specificly with re-entry heat transfer.

---

 

Indeed, one did, although he didn't reply to my welcoming post and hasn't commented anywhere else yet.  I'll comment on his introductory thread again pointing him here and asking about, and we'll see what happens.  One thing you'll get used to if you hang around CSTART long enough, though, is people turning up very enthusiastic, making 1 or 2 posts and then disappearing forever.

Thanks for the link!

rpulkrabek said:

What I mean by the evolutionary route is that we take the design we have now (with the aerostructure) and we then optimize things from this point.

---

 

By "the design we have now" do you just mean the use of an aerostructure, or are you including all or part of the design proposed in this other thread?  Sorry for the confusion, I actually originally thought that your reply was on that thread and not this one.

By iterative I basically meant specify everything, shapes, materials, etc. based on what seems best now and then go back and better investigate individual choices and make changes where necessary – each change results in a new iteration of the design.  This sounds kind of like what you had in mind, I think?