Sweep with guide rail - anoying problems

[lnonnato]

Hello,

I am trying to build a helical body using sweep, but I am getting the most anoying problems.

I know this is a subject that have already been discussed in this forum, but I couldn't find a conclusive solution.

As a simple way to present my problem, I tried to reproduce a single turn square-shaped coil. My results are shown below. Left body was created using COIL (which was the expected figure), while the right one by SWEEP with rail. The complete drawing process is presented in the attached screencast.

 

I used the COIL body as a reference for the SWEEP one. The coil cross-section was used as profile and its centerline as the sweeping path, while one of the 4 edges was employed as rail. A first trial, using the lower outer edge produced a distorted figure (see the screencast). The upper outer edge produced a better result. But the result is not ideal:

- except at the sweeping origin, the body's axial cross-section is not attached to the rail

- body's axial cross-sections does not have the same size and orientation: the original section rotates and tilts around the sweeping path

- as a result, the inicial and final faces are not parallel, with is a nasty problem in my case

Considering that the oiriginal COIL body is geometrically sound (what I have no reason to doubt) and that the sweeping path was exactly at the center of the coil (and the rail was one of the coil's original edges), I cannot see any reason for the sweeping operation not producing

a body identical to the original one.

A side question - Why the choice of the rail causes such an impact on the final body (the lower and upper coil edges are parallel) ?

So, I would be very glad for any hint I could get about this issue.

Cheers,

Luiz

[Anonymous]

Luiz - I don't see a link to the screencast,

[lnonnato]

Hello all,

Due to some strange reason, the screencast disapeared from my message.

Here it goes.

Luiz

[I_Forge_KC]

I was tinkering with this in Inventor just a bit ago and I think I see what the issue is and I have a thought as to its cause (though no work around).

In Inventor you have a sweep rotation control that is not an option in Fusion (IdeaStation!!!!). Fusion seems to draw its sweep twist from the curvature of the rail in relation to the sketch plane. For example, the center (your path) is perpendicular to the sketch plane. The other four rails are not... and so Fusion is using whatever that instantaneous angle is to calculate the twist. This is why the top of the coil is not perpendicular either. It also explains why each of those four rails creates a different result.

I don't know of a good workaround right off the top of my head - though if you have Inventor, that might be the place to go in the short-term.

[lnonnato]

Hello I_Forge_KC,

Thanks for your message. I agree Fusion should use some awkward algorithm for generating the swept body.

I can accept (but I am not happy with it) that the sweeping profile rotates around the path, but I don't understand why it tilts. Both the path and the rail slope are constant for the whole body, so there is no reason for this gradual tilt.

By the way, contrary to what you mention in your message, I think my path is not perpendicular to the original profile - I will check it.

Cheers,

Luiz

[innovatenate]

I was able to find two solutions for this. The location of the guide rail makes a big difference. Here's one solution:

Solution 2: 

If you use a profile that is truly normal to the Path (see the Plane Along Path command in the construct menu for base plane), then you can use a vertical line/centerline of coil as the guide rail.

Below is a screencast of making a helical thread that is tapered using as sweep with a guide rail.

I hope that helps! Let me know if you have any questions!

Thanks,

[innovatenate · Accepted solution]

Attached is a sample file with the two solutions highlighted above.

[lnonnato]

Hello Nathan,

Your help was much appreciated. I already suspected that the path should be normal to the profile and now you confirm this.

Regards,

Luiz

[cekuhnen]

Maybe elevate to the team that we really need improved adv modeling options in Fusion. It just pales compared to what is commonly a default toolset and requires you to spend a lot of time to figure out how to build a shape with Fusion limitations in mind instead of using a tool just to build your shapes.