Fusion 360 Forms Not Allowing Edges to Properly Combine.

[Jason502]

 

Hi everybody,

I am working in Fusion 360 Forms Environment to create the bodywork for a car and have encountered a problem with two rings of edges that I am trying to combine to create a smooth tunnel. The rings have the same number of edges and hypothetically should allow for each of the individual vertices to align, but after welding all vertices Fusion refuses to convert the model back to smooth display and remains in box mode. What am I doing wrong?

I have attached the model. Currently, to access the problematic body, you must go back in the timeline because until the two rings are connected Fusion believes there is a point overflow error and will not display the body in the main workspace. It is best to use either of the two sectional analyses as the area is within the model and otherwise mostly inaccessible.

If you could provide information on how to fix the problem so I can do it myself and why things went wrong so I don't make the same mistake in the future, that would be fabulous.

Thanks in advance!

[jeff_strater]

I'm not sure about the edge merging, but when I edited the Form feature and just exited, it converted OK for me.  Which was weird, because the first time I tried it, it did not, so I did a Repair Body, and exited, which showed some intersecting faces.  Which I then deleted and re-created, pretty much the same, and then it converted OK.

[Jason502]

That's great and all, and I appreciate the effort, thank you, but I really need these two edges to merge otherwise the model is effectively useless to me.

I have tried using the Repair Body tool numerous times. It's really confusing to me because before I try to merge the two rings, it doesn't have a single problem and understands the line flow exactly as I desire. The second I merge the two edges the Repair Body tool seems to throw errors all over the model at nearly every T-point and while I would usually have no problem solving the errors as the computer wishes for the sake of having a workable model, the Repair Body tool distorts the car beyond usefulness.

I know that in Box Display both of the rings line up almost perfectly, have the same number of edges, and that, even after merging, all faces have four sides and all of the basic fusion form errors shouldn't be a problem. Does anybody else have any idea as to how I should best combine the two edges?

Thanks in advance!

[TrippyLighting]

Your model is so full of undesirable N-Gons that Id fix that first.

[jeff_strater]

Can you indicate a bit more clearly which edges in which body are not combining?  I did not see any structure like in your screen shot, nor did I find any unmerged edges.  Can you provide some more information here?

[Jason502]

Ok, thank you I'll try that. I was under the impression that as long as, in box display, a T point occurs along a straight line it was ok but I can see how with conversions to smooth display and all that it might cause issues.

[Jason502]

Hi, I really appreciate all the help I have been getting and I'll do my best to explain because the concept is a little difficult to describe using non-car-specific lingo, but as I said, I'll try. Basically, racecars have something called diffusers. This is the long expanding in-height tunnel that is located between the rear wheels. If you look just after the point where the diffuser begins to expand, there are two identical holes in the bodywork, mirrored, one on the left and one on the right. These in turn lead up through the internals of the bodywork, join in a "y" like fashion, and terminate in the rounded rectangular hole at the back of the car. **I'm not sure if fusion allows sectional analyses to transfer with the model so if they don't forgive me.** If you unhide either of the sectional analysis you should be able to easily view the internal structure that makes up these two tunnels (obviously, if you use the sectional analysis that cuts the car in half you'll only be seeing one tunnel). Approximately 3 to 5 rows of faces up from where the hole starts in the underside of the bodywork, you should see two dark rings, it may at first look like the area is simply creased, and the two rings will probably only be separated by about a quarter of an inch. This is probably much easier to detect in Smooth Display as Fusion misaligns the two edges; but it will be easier to modify in Box Display as Fusion warps the edges into entirely different positions otherwise. These are the two rings I am looking to combine, they have the same number of edges and to the best of what I can see line up well. I apologize for the proximity of the two rings to each other and the resulting difficulty of detection, I attempted to get them relatively close together in my efforts to try to make Fusion understand what I was trying to do.

The start of the two internal tunnels.

Here the two dark lines denoting the break in each tunnel (not creases) are barely visible.

I know it's a confusing and extremely complex model, something that is probably only made worse by my misconceptions, but I would appreciate all the help I can get. If there is anything that I can try to further explain or do so in a different way, please let me know, I would be more than happy to try to help.

Thanks again,

Jason

 

[TrippyLighting]

---

@Jason502 wrote:

Ok, thank you I'll try that. I was under the impression that as long as, in box display, a T point occurs along a straight line it was ok but I can see how with conversions to smooth display and all that it might cause issues.

---

It is usually OK if 5-sided N-gons are created very sparingly, but in you model they are all over the place. Not only that, but you also have a good number of N-gons with more than 5 sides. The key here that I believe most people don't understand is that under the hood, Fusion 360 has to create untrimmed NURBS surfaces from these T-Spline polygons in order to create either a surface model or a solid body from the model.

Untrimmed NURBS surfaces for mathematical reasons have to have 4 sides, so Fusion 360 has to find ways to insert the necessary edges itself. In cases where there are too many n-gons it fails.

As such my usual recommendation is to work with quad topology, avoid triangles at all cost and only use 5-sided n-gons sparingly. 

You can use weld vertices to merge these rings. 

[jeff_strater]

Thank you, @Jason502.  That helps.  However, this is very strange.  Your understanding is correct - when you see that dark line in your model, it usually indicates that there are double edges.  However, I do not see that dark line, nor can I find, using Select Other, any double edges in that area.  I do not know why that is, because I believe that you ARE seeing those...

[Jason502]

I don't know if your computer allows you to render the fusion model in higher definition, maybe resulting in the dark line shrinking, but you are actually looking right at it. I think it would be easier to see if you used a sectional analysis to cut the car in half down its length (so you are left with only the driver or passenger side) and then view the internals. However, I will try to explain using your screenshot. 

The rectangular area that I circled in red is actually part of the quarter-inch gap I am trying to close up and is allowing you to see geometry other than the tunnel. The blue line is a portion of the ring that is closest to the bottom of the car and you can't see the rest of the ring because the entry to the tunnel blocks it. The green line is a portion of the second ring (the one that is farther away from the floor and leads back toward the hole in the rear of the car).

Again, let me know if I need to clarify further.

Thanks,

Jason

[Jason502]

Ok, I understand the idea of Fusion trying to extend lines that dead-end in T points in order to create four-sided faces. The reason I have all of these T-points is that I needed a large number of faces on top of the car due to the geometric complexity and precision I desired in the upper surface, but I don't need as many faces on the bottom of the car. Basically, the bottom need not be as complex as the top. So, I guess the question I'm asking is, is there a proper or any way that T-points are allowed where a necessary edge becomes unnecessary, so as to reduce unnecessary complexity?

[TrippyLighting · Accepted solution]

Apologies for the delay. It's been a busy week 😉

After removing all creases and "Make uniform" I was able to convert it.