loft not smooth, how to adjust surface?

[mossnd]

I am modeling in 'solid' a fuselage for airplane.

im using sketches for cross sections/ profiles, and also a couple of guide rails.

between most of the profiles, im getting areas of 'sink' on the surface (concave type appearance)

can anyone suggest what im doing wrong, or another approach. or can change the surface from here using the 'surface' functions??

[piotr.kornafel]

Hello @mossnd 

Could you share your design?

If it's ok for you, here you will find an article How to share files in Fusion 360.

I can look at your design and I will try to give you some suggestions about modeling.

[mossnd]

attached

[davebYYPCU]

Short critique - Not enough rails, and profiles in areas that could be better.  You have too many spline points - everywhere.  

Tangency across the mirror line will be much better when using extruded dummy surface bodies, but not with these combs, the fine red line should be a smooth as possible.  Will come from way less spline points and handle adjustment.

The area in front of the windscreen would likely be a crease that is later filleted, but would need more detail from the fullsize to be sure.

Might help....

[TrippyLighting]

In addition to what @davebYYPCU has already written, lofting this in one go in not going to produce desirable results.

You probably need different lofts even for the main fuselage. I made this vid a while ago to explain what I mean with that. 
I would usually embed one of the Tutorials I've made, but unfortunately he forum software now is even more broken than yesterday. So here is a plain link. Hopefully that works.

https://youtu.be/3GjraefaAuo

[piotr.kornafel]

I agree with @davebYYPCU  and @TrippyLighting 

I suggest you follow this tutorial shared by @TrippyLighting.

Here I found one more inconsistency in your design:

What's more, I'm a little bit worried about your wings model, you create there a lot of surfaces instead of one smooth shape.

Below you will find a short screencast of how I would prepare geometry like yours.

Screencast

If you have any more things I can help you with or additional questions, please feel free to ask.

[davebYYPCU]

I agree the airfoil should be traced, you too, are tracing with too many spline points.

For top outline I would use 4 points, with handles constrained to vertical at front, and horizontal at top.  Adjust with handles.

In the other pictures, what is your arrow indicating, profile or rail problem?  I did not check top rail, as side rails are not nice.

Adding curve comb almost crash, hanging on 48 percent compute but wait and it did resolve.

So I predict very unstable with so many spline points that are unnecessary.

@mossnd Are you designing for a plug, or formers and skin?

Might help.....

[mossnd]

Thanks everyone for the response. 

Fwiw this my first attempt at a fuselage, in any CAD software. 

I will take on board whats been said, i will look at the tutorials over the weekend, then redo the fuselage first.  
I have troubles trying to get spline points to move when i go back into a sketch  and want to adjust things or even try make Them coincide with purple dots. 

The wing  Does have lots of surfaces. The loft on this originates from dxf files brought into the model, and i loft straight off these. 
I will not trace if it doesnt produce an identical  airfoil!  Airfoil fidelity is important. Please let me know if there’s a way to “copy” the dxf file, turning it into a usable sketch for lofting, that has less points in it. 

What im  designing will be 3d printed, and then i will pull female moulds off the prints. Then i can hollow mould composite planes. These will be RC models. 

cheers

[mossnd]

The plan view sketch was started but not implemented yet in the loft of the fuselage.

I now know i need more profiles and rails, so i will need get this sketch up to speed to use it for the loft in future. 

additionally, all sketchs and profiles in the this file are a guesstimate as i dont have enough info from the original aircraft. 

[TrippyLighting]

More literature for the weekend. It will make all the difference!

Lofting creates NURBS surfaces. The blue link is to a Wikipedia article, but this one is to Rhino's documentation which also offers a good explanation.

Here is a link to the Autodesk Alias Theory builders. The tools in Alias are very advanced and many are not available in Fusion 360, but the underlying concepts are the same.

[davebYYPCU]

You can’t adjust green lines.  Lines and arcs should be black, when done make the splines green.

Fidelity to the dxf airfoil - just went out the window with mention of a 3d printer, bog and sandpaper.

so mirror the four point spline, will be the efficient way.

Jump on the Internet, download photos of the aircraft walk arounds, when examining them, you will pick up lots of clues.

Much better results with Lofting off body edges, as taught to me in here.

Might help....

[mossnd]

Im looking at cnc machining moulds for wings in future.

So id be great full if you would share any methods other than tracing for airfoils. 

generally fuselage moulds start with printed plugs because they arent so dimensionally critical 

[mossnd]

I don’t understand when you say 

“You can’t adjust green lines.  Lines and arcs should be black, when done make the splines green.” 

is fusions expectation that the sketch  is correct or perfect in the first instance, and the ability to Go back into a skecth to change or edit not allowed? 

[davebYYPCU]

Sketches are editable, any time you can Edit sketch.  When you do so, like a time machine in the history of the design, edit Sketch 2, and all the later sketches will become unavailable, until you finish the edit.  Changes are then computed right through the design.

Fully defined sketches are recommended, they will contain, black, green and purple articles.

When not fully defined you have a combination of those colours with blue, orange articles and white dots.  Purple dots are your friend, snap to them for successful Lofting.

In your case I would be tempted to use longeron lines as profiles in the loft, with formers as Rails.  Loft has geometry rules to follow, rails must have tangent connections, but profiles do not, so deciding on which way to run either depends on the shape now.

Will demo the airfoil Trace, I did not check the wings, as the fuse combs stopped me dead in my tracks.

Metal moulds, will keep that in mind, will let you decide on the fidelity standard, when done.

Might help....

[mossnd]

Thanks i would like to learn best practice from the beginning for the wing side of things. 
There will be future projects and they also require attention to detail and accuracy, so practice from the beginning makes perfect in the end. 
cnc moulds wont be an issue, im a toolmaker buy trade. 

I will post back here in a couple days after ive reworked  the model. 

thanks

[mossnd]

thanks for the screen cast , yes that looks like a good option for symmetrical airfoils. 

i can only try this for the main wing airfoils and panels and see what happens

[davebYYPCU]

Have traced the root rib, and created the body for it.  This was then copied and scaled down, for the other 3 ribs.

You have no washout, and the dihedral angle is not the same as the canvass.  Therefore have not done the tip, (Tip rib will be included in the top skin,) in case there is a dihedral change or washout influence to include....

Lofted Trailing edge body to the Leading Edge sketch, with rib edges as rails.

This photo shows the front of windscreen highlighting, helps diagnose the former lines.

You will note (pic and canvass) that the mainspar is parallel out to the aileron (common in fullsize for fuel tanks) so rib 2 & 3 are currently too thin from the scaling.

Might help....

[davebYYPCU]

Splines were fixed, and yep, good outlines, means this one has no rails.  Won't know if this is accurate until the windows are cut out.

Might help.....

[mossnd]

I was lofting as a solid , do you suggest i loft surfaces then close it up later?

[davebYYPCU]

Yes, solid loft is no better and surfaces are more versatile with less work.  Solid loft this fuse is just plain hard work, due to the compound curved side datum line.

Might help....