Can you teach me how to constrain these lines?

[maula006]

So let's say we start with this:

As you can see, there are two lines connecting the inner profile and the outer profile.

The thing that defines them is their distance from the top most horizontal line.

Which is the problem... because the following is practically defined the same way as the shape above:

I need to find a way to keep the lines in their own "halves" of the shape.

Because... well, I've used the above technique in the following project and changing a parameter caused the loft to go crazy:

I used it to define the front and rear supports of the shape... and you can see how the loft turned out for the rear end when I stretch out the model (via a parameter change)

I am incredibly close to finishing this model so I'll really appreciate if someone can help me iron out this bug!

[hamid.sh.]

Can you check if this works? I broke each of your spines into 2; this way each of your tilted lines can have only one possible solution (because they are constrained to be coincident with half splines). 

[maula006]

Hm... I'm sure that'll work but it seems that I'll need to make many adjustments to make it work.
In case, do you have any alternatives that doesn't involve splitting the spline into two parts?

[hamid.sh.]

Maybe this one works. I made an offset of your spline and split that instead of the original one. I did it for only one of the splines but you should repeat it for the other one as well.

By the way, neither this one nor previous one is a 100% fix. Because of infinite extension of entities, geometries often have multiple solutions. Here I tried to take advantage of the fact that extension of splines is more complicated than that of say lines, so perhaps there is less chance of another geometric solution popping up first in Fusion's engine. 

[laughingcreek · Accepted solution]

In your attached example the inner profile appears to be a trimmed ellipse.  I strongly advise against trimmed ellipses in fusion, there have been lots of bugs related to the behavior of those and I'm pretty sure they haven't all been worked out.  could be contributing to your woes here. 

projecting the splines into a second sketch, and adding the lines there will simplify how to solve each sketch for fusion and will probably also result in  better performance also. 

if you can get it to work with your design intent, removing one of the distance dims and replacing with an angled dim to the horizontal line may also behave better. 

[maula006]

Regarding the trimmed ellipse, don't worry about it. I didn't use a trimmed ellipse but the projection of some lines put together. Frankly, I probably used some practices that'd make a senior engineer want to off 'emself, 

(the various dimensions are there so the fit-point-spline can squash and stretch while retaining its shape).

But I digress...

The first option you offered will definitely work and will definitely be consistent, but requires a bit of time to implement.

The second one is significantly simpler but has the side effect of causing the front/rear supports to vary greatly with the changing length... But still, I think I can live with it. I will be using your second option!

The only issue is only but it's a minor aesthetic thing that affects only the most extreme examples and is barely noticeable. Only finicky thing the length of the distance dim has to be 25mm or else the line may brake again.

273mm, 600mm and 150mm (Realistically only 219-335mm)

Regardless, I thank you for your advice! I never considered that limiting the length of the line itself could be so effective!

[maula006]

Thanks for your advice!
I've seen the solution and can tell that this will definitely work and will likely be consistent among all possible parameters too.
But I'm afraid I decided to use laughingcreek's advice.
The solution is a lot less robust than your solution, but it was a very simple change. But if things go wrong, I will definitely consult your solution.

Thank you for your time and help!