Is there any way to wrap a spiral around a spline defined revolved surface?

robert_lilly

I'm trying to define a spiral cooling channel in a rocket nozzle.  This is strictly a CAD exercise - I'm trying to sketch up the thrust chamber and nozzle of the LR-101 Vernier rocket engine.  I think I understand how to do this using a spiral defined path followed by a sweep of a disc.  Unfortunately the spiral can only be defined on cylindrical and conical surfaces.  Is there any way to get a spiral to wrap around spline defined revolved surface.  This area is the throat of a supersonic nozzle, so it's converging then diverging.  

I was thinking about cutting this section in half, then trying to project a flat spiral (and there's a great discussion somewhere on the forum here on how to generate a flat spiral in the first place) onto the surface...but I don't see how to make the project fro the flat sketch onto the spline defined surface.  

Do folks have any suggestions how I might proceed?  Thanks, Bob

andrew_troup

If you generate a helix (based on a cylinder) and they sweep a radial line up it to produce a helical surface (like an auger) and intersect that surface with a revolved surface sharing the same central axis .... would that give you what you want? (As a path for sweeping your "disk"?)
If so, I can go into more detail on the procedure.

nav

Hi @robert_lilly  share your document and post the link here.

for the video gif I use http://www.cockos.com/licecap/
for screen grab I use Press Command (⌘)-Shift-4. https://support.apple.com/en-us/HT201361

nav

Modify the helix as is not entirely contained in the  thrust chamber you get an error when using the split command.

nav

is something like this what you´re trying to model ?

andrew_troup

If you generate a helix (based on a cylinder) and they sweep a radial line up it to produce a helical surface (like an auger) and intersect that surface with a revolved surface sharing the same central axis .... would that give you what you want? (As a path for sweeping your "disk"?)
If so, I can go into more detail on the procedure.

robert_lilly

Andrew,

...I'm speechless...that's brilliant!  At least, if it works.  I had not thought of creating the auger then INTERSECTING it with the desired surface.  Rueful smile.  You are crafty.  I'll try it on my own, but when you get this I suspect all I need is how to do the surface intersection.

Bob

robert_lilly

Andrew,

A bit of a head bending experience, but I did create the auger.  So cool.  Reminds me of my complex calculus class way back in the 90's (talking about Riemann surfaces and such, which is pretty close to an Auger - as I remember).  Anyway, if you get a moment to share how to do the intersection bit - because I have no idea.

B

nav

@andrew_troup brilliant solution as @robert_lilly I´m also having trouble with the intersect part

andrew_troup

@robert_lilly 
The easiest way is probably to use whichever surface overlaps the other surface to split that surface (using "Split" feature).
If you like, you can use "Delete Part" to get rid of the excess surface (in @nav's case, it would be the outer part of the auger) to get easier visibility of the split edge at the intersection (again, in the case shown in the graphic) with the revolved surface (which you can hide), and that edge can be selected to act as your path. 

robert_lilly

Andrew,

So (everybody), please forgive the messy screen grab.  I've gotten myself into a pickle on this.  The items in question are in red and blue.  I recreated the auger (had it stuck inside the defining cylinder first).  Reworked it so the spiral surface extends beyond the defining cylinder.  Lovely.  

It won't let me split the spline defined revolution with the spiral surface (says the geometry is incorrectly generated), and in the other direction it appears to be an invalid selection.  Trying to cut back on the spiral surface using delete wipes out the auger in its entirety.

Suggestions?  Again, what I am ultimately after is the trace that spiral leaves on the spline defined revolution (in blue).

B

andrew_troup

@robert_lilly 
The surface doing the splitting of the other surface needs to either overlap it at BOTH ends (or be strictly flush)
It's not possible to see from your graphic if that's the case, also it's possible your blue surface is actually a solid (again, can't tell)
It should preferably be a surface, (it could possibly be a single face of a solid, but I haven't got a good handle on that yet)

andrew_troup

@nav: did it work for you?

nav

andrew_troup said:

@nav: did it work for you?

It did but see below it was weird as you tend to select the auger as the splitting surface ,it worked the other way arround 

andrew_troup

@nav
Thanks for that, good to know it works in practice as well as theory!
If the auger was at least as long as the revolved surface, you should be able to use it as the splitting surface.
If they both ended exactly flush, either surface could be used to split the other.

robert_lilly

Andrew,

Unfortunately, I am still stuck.  I reworked the "surface to be divided", i.e. spline defined revolved surface .. it had been a solid, as you speculated.  The auger is larger than the spline defined revolved surface.  It doesn't work.  The diagnostic says the surface is unable to split the part.

I don't know how to post my models cleanly like nav does, but I will share the document with you.  Perhaps you can enlighten?  (It is a bit dark in here, still).

Thanks,

Bob

nav

Hi @robert_lilly  share your document and post the link here.

for the video gif I use http://www.cockos.com/licecap/
for screen grab I use Press Command (⌘)-Shift-4. https://support.apple.com/en-us/HT201361

robert_lilly

Here's the model link (that I've made public).  

https://cad.onshape.com/documents/9def4ef0771b40d398859822/w/b27a2dc56aca4e4ab456f910/e/8c9c6a56099f4497a2685bff

nav

Do the  spiral cooling channel goes here ?

nav

Modify the helix as is not entirely contained in the  thrust chamber you get an error when using the split command.

robert_lilly

Nav,

Well, to be precise, a "wire" will follow the spiral on the blue surface.   A second layer will follow on top of it.  I"ve grabbed an image off the internet, just to show you what I'm hope to draw:


So you see that the whole point of the spiral is to guide the wire (i.e. I'll be sweeping a disc to represent the wire).  The actual cooling channel was defined by the space between the loops of wire.  

Bob

robert_lilly

Nav,

So you are telling me that the helix should NOT extend beyond the spline defined revolved surface?  Err... I had thought Andrew told me that it needed to extended beyond the surface to be cut.  I'll try (I see you've already figured it out)....

B

nav

is something like this what you´re trying to model ?

robert_lilly

Nicolas,

Yea, pretty close.  Except that, as you can tell from the photo of the LR-101, that cooling channel has go all the way through throat and the nozzle.  Harrumph.  Well, I'll get this beasty.  I'm an obvious newbie with CAD, but even at my advanced age, I still have my dreams.... (Thanks Nic, and Andrew too!) - B

andrew_troup

robert_lilly said:

Nav,

So you are telling me that the helix should NOT extend beyond the spline defined revolved surface?  Err... I had thought Andrew told me that it needed to extended beyond the surface to be cut.  I'll try (I see you've already figured it out)....

B

@robert_lilly

Well no, I didn't say (or mean) that. Apologies if I was unclear, I guess I was trying to be concise.

Here's a dotting i's and crossing t's version of what I actually said : 

Regardless of which surface (call it A) you are using to split the other surface (call it B ) into two surfaces :

Surface A needs to be at least flush, or longer, than Surface B, at both ends. 

It is an open question whether surface A should be the helix or the revolved surface.

But in the first graphic @nav posted, which I referred to, in illustration of my point, because it was the only graphic available at the time showing a clear overlap at either end : the natural choice to use as the splitting surface A (by my rule) would be the revolved spline, and in fact his subsequent graphic :
 
https://us.v-cdn.net/5022071/uploads/editor/c2/q28afm3i0wtd.gif

makes it clear that it only worked when he did it that way round, in accordance with my proposed rule.
(In other words, it only worked when he split the auger, which he called "weird", suggesting it was contrary to what his intuition suggested)

HTH