Loft with Helix Rail failing for Irregular Shapes

[karl.zammit.18]

Hi all

Currently stuck using the loft tool with a helix as the rail. The sketches (Sketch15 and Sketch16) seem to be 'Loftable' as shown in Fig. 1. However, I am unable to use the helix (3D Sketch4) as a rail or centreline for the loft, as in Fig. 2. 

Using the helix as a centreline for a Loft operation using simple geometry works, as shown in Fig. 3. I have also tried adding a circular end to the inner side of the flights, but as shown in Fig. 4, this did not work either. The latter attempt also begs the question on whether the helix has to pass through the exact centre of the sketch to use it as a centreline (instead of a rail). If so, any ideas on how to find the centre of such shapes?

Fig. 1:

Fig. 2:

Fig. 3:

Fig. 4:

Any ideas on what could be done? Sweep or Coil features are off the table (I think) as the helix has a changing pitch and diameter in later portions of the auger.

Attaching my file for reference and in case anyone wants to have a go at this.

TIA

[blandb]

I don't have 2025 on this machine, but if it is just a cylinder with a constant helix, can you just use the coil command?

[kacper.suchomski]

Sketches must be perpendicular to rails.

[karl.zammit.18]

Sweep or Coil features are off the table (I think) as the helix has a changing pitch and diameter in later portions of the auger, as mentioned in the latter part of the question (beneath the figures)

[karl.zammit.18]

Thanks for the heads up. 

Its progress I feel, but now, I'm greeted with a new error message about self-intersecting surfaces.

Any ideas?

[karl.zammit.18]

File added again, updated with sketches to be swept now perpendicular to the helix rail.

[johnsonshiue]

Hi Karl,

I would not use the Loft command for this particular model. I would try to use the Sweep command. However, it is better to have smooth path. If possible round off the sharp edges on the shaft. Make sure the path is at least tangentially continuous. Then sweep from the top (select the top planar face as the guide surface).

Trim off the blade using an offset surface at the bottom.

Many thanks!

[karl.zammit.18]

Thanks for this, but I encountered the same problem with Sweep anyway (Self-intersecting error).

I feel that sweep would not work either, as the crosssection also varies from one end of the attempted loft to the other.

[johnsonshiue]

Hi Karl,

Please share the ipt file that you tried sweeping with me johnson.shiue@autodesk.com. I would like to take a closer look.

Many thanks!

[karl.zammit.18]

Thanks for this.
Sent as requested.
Note that Sketches 19 to 22 correspond to Sketches 15 to 18 but are oriented perpendicular to the helix curve.

[johnsonshiue]

Hi Karl,

Many thanks for sharing the file! I thought Guide Surface Sweep would help. But it seems that the profile has to vary along the path. So Guide Surface Sweep isn't a good option. Another workflow to try is to use Guide Rail Sweep. The path and the rail can help scale the profile along the path.

Thanks again!

[karl.zammit.18]

Thanks for this.
Does that mean that I would have to construct a helix for each of the four corners and then sweep the lines individually? Note from the images below that both the radius of curvature of the crosssection and the angle of the flight change from one end to another.

Would a sweep still work in this instance? My concern is the change in curvature.

Crosssection at the start of helix:

Crosssection at the end of helix:

[johnsonshiue]

Hi Karl,

Unfortunately no, Guide Rail Sweep will not work in this case. You do have to use Loft.

Many thanks!

[karl.zammit.18]

Hi all,

I started from scratch and tried tackling the problem from another angle. Now, I feel as if I'm almost there. Having done some calcs to add guide rails for the loft, I was struggling to find a perpendicular plane to loft the entire flight at once. Therefore, my current solution is to loft the individual surfaces and then join them. 

I am currently stuck trying to loft the curves in yellow (see below) as the existing rails do not seem to suffice (possibly because they are not perpendicular to the helices as is, but I am struggling to see how I might orient them as such.

Would creating two helices from the points circled in red (Fig. below) be enough to use as a rail? This, I feel, would involve some tricky arithmetic to get the taper angle right. Any suggestions or alternative methods are welcome (and much appreciated). File also linked for reference.

@johnsonshiue @kacper.suchomski @blandb - Any comments on the topic further to those above on considering the new way of going about it by any chance?

TIA!  

[karl.zammit.18]

Hi,

I've finally managed to Loft the curved faces, but the edges don't seem to follow the rest. The figure below explains it better than words. 

Would you suggest anything to fix this? I've attached the file once more for reference.

[blandb · Accepted solution]

You can boundary patch the top and bottom, then patch the ends of the flight and stitch with what you have.

Please see attached.

[karl.zammit.18]

THIS IS IT!

Thanks a lot.