I have "hit a brick wall" trying to model the complex wing root fillet on my TK4 model aircraft.

This old photograph is the only real evidence that shows what it should look like..............



The following screen shots show my - unsuccessful - efforts to date ..............









ALL SUGGESTIONS & HELP would be very welcome !

John

Hi John, could you please also post your 3DM file (preferable with just the pieces adjacent to the wing juncture that you need help with) so that people can look at your geometry and give some suggestions based on that?

It looks like you have some curve structure in place that could be used to build a surface with Network or Sweep.

Another possibility is to work off the edges between the 2 surfaces and use the Blend command to put in a smooth blend surface between them, see here:
http://moi3d.com/2.0/docs/moi_command_reference7.htm#blend

- Michael

Hey... Theres a break in that planes surface....It needs some mojo from modo!

Hello Michael, Sorry I forgot to include the most important file of all !!!

The photograph shows a "bump" just underneath the pilots head - I am ignoring that on my model.

John

PS: BurrMan,

Well spotted !
The original aircraft does show discontinuity between the prop. spinner profile and the engine cowling - my guess is that apprentices who designed and built the machine had to make do with a standard spinner off an existing DeHavilland production model.

Hi John, so one thing that will make things more complicated is that your leading edge of your wing has little thin slivery surface fragment in it.

You can see it if you zoom in a bit on this area here:



There's a little slivery surface in there:




You probably want to get that wing constructed more cleanly without that little slivery piece in there - possibly you have something like some generator curves used to create that surface were made up of some small sub segments instead of just one smooth single segment curve - when you have multiple joined segments in your curves that will usually produce segmented face output as well. You want to do things like draw one long curve instead of multiple sub segments, or use the Rebuild command to remove segmentation.

Anyway, along with that little sliver means that the big edge pieces of that wing don't quite come to a single touching point as well, they've got a little tiny edge in between them.

- Michael

Hi Michael,
So I fooled around with his area, and come to that "n-sided patch" command... (I think)

Here is the model as a solid, with a pre-fillet model in it and a couple different results from it... Mostly, the fillet makes a sharp transition in the rear of the fillet. (Style Blue-Object "problem area") I've left a couple different types of open areas in the different models.

Looking at filling these in, Whether starting from the front or back as the crux, could you elaborate on the tool for me.. I looked at patch in Rhino and didnt make sense of what i needed to do.

Also, perhaps a variable fillet will perform here? as a smaller fillet will work, but to get the large front part, that transition goes out.. Using blend, I always end up with the last patch that needs filled while maintaining more than 2 tangencies...

Thanks for the time.

I had a similar problem earlier with a propeller. I just wanted you to know it can be done with MoI, and once you figure out how to do it, it can be done very easily and quickly. The thing is you have to experiment a lot to figure out how to get it done. I can't say specifically for you but the biggest things I found was that switching to the mm unit system helped. Then switching back to the meters system that the model was acutally in. Also making sure that in your case the wing is a solid and then the body is a solid and unioning them. Then make the fillet. If MoI can't make the wing a solid or the body a solid then there is a problem. Like Michael said you have to clean up the surfaces. That was something I had to work on too. Its not intuitive at first, you think it should just work no matter what. But in reality it will only work under certain circumstances. If you are willing to spend the time on it and not give up you can reach your goal. Michael and BurrMan helped me out with my problem, which is very similar to your problem. So definitely try out what they are saying and more. I tried a lot of stuff after I understood what they were saying. After awhile I got it done to a very simple process.

Hi Burr - re: n-sided patch - that hole you've got in the blue one I don't think is feasible to fill with n-sided patch since it's got that pointy part that's kind of sticking up 90 degrees away from the side loop part. That's kind of too much variation within a single patch to fit very well.

N-Sided patch basically starts with a plane and then squishes points on that plane up or down until it is deformed to fit the outline. But with that much variation the outline there will be difficult to fit to.

I don't think you can put a smooth single surface in that spot - I think it needs to have a ridge down the middle of it to match with the sharp point on the other side.

Something more like put a line down the middle of it and then patch each side separately, see attachment.

- Michael

Hi Burr, so I guess what you would really want there would be for the fillets to just continue and meet each other in a sharp crease instead of it trying to put in a corner patch in that spot. I don't think there is any way to get the fillter to do that, but actually if you look at each fillet piece, they're actually trimmed back parts of a larger fillet like this:




Unfortunately they don't quite extend far enough to totally finish crossing over each other, so it doesn't quite work but you can see in the attached 3DM file how that could work if they just extended a bit more.

I think what you would want here would be a way to get a maximally extended fillet - that could be possible for me to add as an option on surface/surface filleting - that's the variation of filleting where you work between just 2 individual surfaces instead of picking edges.

- Michael

""""""""I don't think you can put a smooth single surface in that spot - I think it needs to have a ridge down the middle of it to match with the sharp point on the other side.

Something more like put a line down the middle of it and then patch each side separately, see attachment"""""""""

Yes, i was working that, but couldnt get the surface to maintain tangency/curvature against the other ones (The existing blend and the body surfaces).. I was trying various sweeps and different ways to draw that "spine", but couldnt get a smooth result... I was wondering if ther was a Rhino command that would do this in this situation.

For instance, running a blend of 10 on these 2 curves here, gives a great result, but you can see it deviate away from the body, so thats not a good command for it.. Maybe when MoI gets a multi edge blend then these would be possible?

Hi Burr,

> I was wondering if ther was a Rhino command that would do this in this situation.

Rhino's Patch command should do that, on the version that I posted that has a dividing line running down the middle and the long edge split into 2 pieces. You would run it in 2 separate passes though, select 3 inputs at a time - 2 edges and the centerline curve.

- Michael

Hi Burr,

> Maybe when MoI gets a multi edge blend then these would be possible?

That wouldn't really apply for this situation - it looks like you want to pin down the side parts of the blend. Multi-edge blend means being able to pick a longer chain of edges for each side of the blend, not really about doing something for the other ends of the generated blend surface.

- Michael

OIC regarding the patch.. :o

Thanks for the help! Patch worked great!

Hello prop-design !

Thank you for your encouraging words. Wing root fillets do tend to be among the most complex surfaces on any aircraft. So I new from the start that was not likely to be easy.

But this sort of thing is the reason I invested in MOI - I certainly couldn't do it in Alibre' - which is very good in other ways.

I'll keep trying.

John

Hello Michael,

It looks to me as though you've cracked it !

I am afraid that I am going to expose my ignorance here - did you intersect the wing and fuselage surfaces and then apply various fillets ?

I confess that I am struggling to understand how you achieved your result.

The "sliver" you pointed out was an attempt by me to fill the the original gap I discovered, between the upper and lower wing surfaces. I am still puzzled as to how I created the gap in the first place.

Is there any sort of "history" file created within your software that would allow me to review previous actions ?

John

Hi John,

> I am afraid that I am going to expose my ignorance here - did
> you intersect the wing and fuselage surfaces and then apply
> various fillets ?

That's actually Burr's model that he attached above here:
http://moi3d.com/forum/index.php?webtag=MOI&msg=4488.6

It looks like he cleaned up that little slivery part and got the wing and body joined together as one connected piece and then used Fillet to produce the rounding - here it looks like he used the "Shape: G2 blend" option for generating the fillet, which puts a kind of more semi-organic shaped blend surface for the generated fillet rather than one that has an exact circular cross section like a regular fillet. That option can be good when you want to have a somewhat more seamless looking connection between 2 pieces.

The fillet seems to have worked well for the leading edge area, but the small area on the trailing edge side needs some rework since the filleter tries to build a corner patch there that is kind of squished up so that part needs to be deleted and something else filled in there.

But it's pretty important to get the base surfaces cleaned up with that little slivery piece removed - having a little slivery bit like that will tend to get in the way of a lot of kinds of techniques of trying to make a good connection.

- Michael

Hi John,

> Is there any sort of "history" file created within your software
> that would allow me to review previous actions ?

No, there's nothing yet for reviewing a log of all actions that were used to build a particular model. That is something that I want to work on in the future.

If you are still in the same modeling session you can try to use Undo to rewind things though.

- Michael

This short video shows what I did to your model...I also added a slightly different method to the leading edge of the wing, to get what "I" thought was a better surface.. But not needed.. Basically just showing the making of 2 solids that will union properly for a fillet.. An n-sided patch is need to finish off the trailing edge.. I cant tell you how many times I need that command!!! :o (still havnt purchased Rhino)

Sound is very good this time! ;)

Cool tutorial!

Ps What is the code for put the Youtube video inside the forum?

Hello BurrMan !

I was just about to post a message asking why I could not unzip your original fillet model - when I noticed that you had posted your video.

To say that I am blown-away by your video response is an understatement !

I shall watch it repeatedly until I fully understand all of the contents.

Incidentally, I had come to the conclusion that I had chosen the wrong method of creating the wing in the first place. But hey-ho, one learns by mistakes !

Many thanks again.


John