As a newcomer to MOI, I am struggling to create an example of an aircraft fuselage.

Any guidance will be greatfully recieved !


The attached file shows how far I've got to date. Can't seem to create a network surface.

maybe this can help you ;)
With the Network function http://moi3d.com/2.0/docs/moi_command_reference7.htm#network
trimed if you want separate facetts as real airplane

"It is not necessary for the curves to exactly intersect" yes but ...
if all curves are well "connected", must be no problem for have predictable result :)

Hello Pilou - Many thanks for your prompt reply !

Hi John , were you able to get your network created with Pilou's advice?

Just in case I've attached here a modified version of your file which should now work with Network.

Basically that involved trimming the long outside curve so that it was in 2 pieces instead of all one long closed curve.

The way Network works, it needs to have a kind of grid layout of the curves that can be unwrapped down to a 2d pattern. It's pretty similar to how latitude and longitude lines on a globe unwrap to a flat map. When you have things coming to a point you need to have all the curves that come to that point end there, you don't want to have a couple of curves end there and other ones be longer.

Please see some of these previous discussions which have some more information and some illustrations about how the curve layout needs to be structured in order for it to work with the Network command:

http://moi3d.com/forum/index.php?webtag=MOI&msg=4384.3
http://moi3d.com/forum/index.php?webtag=MOI&msg=1905.1
http://moi3d.com/forum/index.php?webtag=MOI&msg=1778.6
http://moi3d.com/forum/index.php?webtag=MOI&msg=1484.1
http://moi3d.com/forum/index.php?webtag=MOI&msg=2315.6

- Michael

Hello Michael, At first I thought I understood Pilou's reply - unfortunately tests then proved that I didn't !

However your response made everything clear - so hopefully I will have no more problems of this sort.

I plan to follow up on the discussions you suggested - to cement my understanding.


As an aside - I might mention that in the mid 1950's the UK aircraft design industry employed quite a few "loftsmen" who laid out aircraft sections full size on large grey anodised alloy plates using black ink pens with the use of thin wooden splines and lead weights.


Times have certainly changed !


I must commend you on a superb piece of software. I look forward to learning more of it.


Thanks


John

But the next year when you re-read it that will be also clear :)

Yes Pilou I am sure you are right !!!

John

Hello Pilou ! Now I understand - sorry I didn't at first.

I am in the process of modelling a Dehavilland TK4 Racer from 1937. Perhaps not the best model to start with - but in at the deep end.

However - I am struggling with creating and blending the air intake into the front of the engine cowl.

Again any suggestions from forum members most welcome.


John

No problem : use of nurbs is not trivial as polygones ;)
A little training is necessary!
Hopefully the ergonomy of Moi compensates this no triviality :)

I modeled a lot of airplanes with the old Rhino beta back in the '90s. I always used the 2-Rail Sweep instead of Network. The guide rail option in MoI is a big help because it makes it a 3-Rail Sweep. For the type of plane you're modeling, I'd break it up into front half and back half. I made a copy of the hilited curve (in yellow) and used the copy to model the taller X-section curve at teh same location. You can cut out the cockpit area by going into side view, drawing a curve, and projecting it to the surface.






Steve S

Hi Steve S, many thanks for that suggestion.

I will give it a try.

I'm still a novice user just finding my way around.


John

With semi-organic shapes like airplanes (or ship hulls, my main nemesis), NURBS can be quite a challenge as continuity and fluidity of shapes is hard to maintain. If possible, I'd try to stick to one continuous shape nose to tail and then edit out the cockpit area, rebuilding the nose and engine cowl to its proper shape thereafter, as per the example below.



Jean-Paul Binot

Hello Jean-Paul !

Many thanks for your comments - I agree totally with your logic.

The problem that I am finding (when trying to model older aircraft) is that it is not always obvious what the overall shape of the fuselage should look like before modification.


John

Hi Steve,

How would you go about smoothing the front and rear sections of your design at the intersecton and the mirrored surface that would make the opposite side of the fuselage?

Cheers

Ramon.

Hi, Ramon. Since the cross-section curves where the front and back meet are exact copies of each other, they should match perfectly. Just be sure that when you cut out the cockpit area, the cutout extends below where the two surfaces start to diverge from each other (the hilighted yellow point in the pic below.)

As for the mirrored sides, make sure that the control points at the top and bottom stick straight out from the ends of the curve. This will ensure that the curve ends at the proper angle to form a smooth transition.





Steve S

JPBWEB wrote: "If possible, I'd try to stick to one continuous shape nose to tail..."

I'd have to disagree. With my method, both the cross-sections and the side profiles of the nose are accurately drawn. With your method, neither set of curves is drawn accurately. Everything is an approximation. Just my $0.02.

Steve S