[AR] Re: Intertank coupling design
- From: Robert Watzlavick <rocket@xxxxxxxxxxxxxx>
- To: "arocket@xxxxxxxxxxxxx" <arocket@xxxxxxxxxxxxx>
- Date: Fri, 11 Oct 2013 19:17:37 -0500
Thanks for the suggestions. The tank walls only need to be around 0.05 or so
to take the 500 psi load with a 1.5 FS. I went with 1/8 wall instead of 1/16
because I figured it would have better tolerance to handling damage and also be
easier to true up round. I can always turn down the OD on the lathe to save
some weight. I didn't mention it below but the load from the tank bulkheads
drove the design from 12 to 16 radial fasteners to stay within the bearing
allowable on the skin (after truing up the end, there was only 0.105 wall
left). The actual arrangement is a bit different than below with the tank
bulkheads included.
The threaded ring is one piece to aid in assembly.
As far as shear stress on the threads, I originally was going that route but I
was worried it might be a source of galling and/or fatigue cracks. There wont
be a zillion cycles on the tanks but if its only one extra piece to eliminate
that issue its worth it.
Your point about the joint strength in general is well taken. I need an
estimate of the flight loads ideally. Since its not actively controlled, they
should be small except for weather vaning right after launch.
-Bob
On Oct 11, 2013, at 17:46, Ben Brockert <wikkit@xxxxxxxxx> wrote:
> That seems like overdoing it. The tanks on a pressure fed rocket are
> really strong, much stronger than they need to be for flight load
> purposes. If the rocket crashes in a semi-survivable manner I'd much
> rather destroy the coupler than the tank. By making them the same
> strength you've lost control over what breaks.
>
> Regarding the threaded ring: wouldn't it need to be two rings, or the
> web between them to be really thin or slotted, for it to carry none of
> load? You show it thinner but not completely strengthless.
>
> Why so much effort to not put shear stress on the thread? Sure the
> thread is weaker than a shoulder of the same major diameter but that's
> only a concern if the total fastener count has to be minimized as the
> very first design driver. I don't understand why it would be.
>
> Ben
>
> On Fri, Oct 11, 2013 at 3:23 PM, Robert Watzlavick
> <rocket@xxxxxxxxxxxxxx> wrote:
>> I'm designing the couplings that will connect each of the propellant tanks
>> to the rest of the structure for my rocket. The primary structure for the
>> tanks is 5 inch OD, 0.125 wall 6061-T6 tubing. The question is: what wall
>> thickness and length does the coupling tube need to be to connect two of the
>> 5 inch OD tubes together? To simplify the discussion, ignore the tank
>> bulkheads or any other structure for now. I don't have an estimate yet of
>> the expected flight bending loads but for a first pass, one philosophy would
>> be to make the joint at least as strong as the rest of the structure in
>> bending so it won't fail at the joint. There would actually be 3 concentric
>> elements: 1) the two primary structure elements butted up end-to-end, 2)
>> the coupling sleeve inside, and 3) another sleeve inside the first one. NAS
>> 623 fasteners, installed radially from the outside in are used to hold it
>> all together, with the primary structure and coupling sleeve drilled to
>> match the "grip" diameter of the fastener. The only purpose of the
>> inner-most sleeve is to have something for the fastener to thread into. By
>> using a second sleeve, there are no shear loads on the fastener threads,
>> only tension. Even though the inner sleeve (with fastener threads in it)
>> will provide additional stiffness, I wasn't going to count its contribution
>> since it will be cut down as thin as possible. The material for the two
>> sleeves would also be 6061-T6.
>>
>> If the goal is to match the bending capability of the primary structure,
>> then I would think the coupling sleeve only needs to be thick enough to
>> match the moment of inertia of the primary structure. Then, the length of
>> the coupling should only be a function of the shear tearout allowable for
>> the joint. Am I on the right track here? Of all the allowables for the
>> joint (fastener shear stress, bearing stress, shear tearout, net area
>> tension), shear tearout and bearing stress appear to be the most critical
>> for this design. One thing I'm not sure about is how to convert an
>> arbitrary bending moment into the shear load per fastener.
>>
>> See attached sketch for details. Any advice would be appreciated.
>>
>> -Bob
>
>
Other related posts:
