Sweep failure

[Vagulus]

Every time I try this Sweep it fails.

Its siblings work okay, and I have lost track of the times I have redrawn the track.

What have I done wrong?

Thanks

[IgorMir]

Hi Vagulus,

The problem is with a messy modeling practice. Unconstrained sketches, overlapping lines do tell the story.

Once it is all cleaned up - the model should work. Just as in the attached IV2018 format file.

Cheers,

Igor.

[Vagulus]

Hi Igor

Please note that the positions of the Boss and the Foot Plane (and hence the Rail Profile at Foot) are derived from SR-01-001.  Are they not, therefore, constrained?  Only the interpolation spline in the RH Track sketch is unconstrained and its curvature is defined using Activate Handle.  How does one make an interpolation spline constrained?

It is a puzzle that both the other Sweeps in the part worked okay despite the same technique being applied to the construction of their Tracks.

I assume you used a Control Vertex spline which makes me ask how you defined your waypoints.

[IgorMir]

Hi Vagulus;

Here is updated part of yours. I am not quite clear on the final design of yours, but as I have said before - the modeling technique of yours does require some improvements.

Cheers,

Igor.

[Vagulus]

Hi Igor

Please clarify:

1.  Did you use a Control Vertex spline?

2.  How did you define the waypoint?  Did you just play round with the dimensions until you had a curve you liked, or was there something more scientific involved?

Thanks
 

[Xun.Zhang]

Hello Vagulus,

Here is the model without change any geometry for your reference.

Hope it helps!

[Vagulus]

Okay - I got as far as creating Control Vertex Splines.  Now the LH Rail will not Sweep.  I even created LH Profile at Foot as a separate entity thinking that maybe AI didn't like re-using profiles.

I note the neither of the profiles is constrained.  Please, it does not help to be told that something is unconstrained without being told/shown how it can be constrained.  I am not trying to avoid this working

Updated ZIP attached.

[Xun.Zhang]

Hello @Vagulus ,

Have you taken a look the part from my side, there is nothing change from your original part.

Hope it helps!

[Vagulus]

Hi Zhang

I did look at your part.  Direct Edit and Combine are changes.  Furthermore, they are changes I do not understand.  It will take some research to find out what you did.  Some descriptive video would be a great help, but please do not use Screencast because it is usually too hard for the viewer to see what you are doing.  I upload my videos to YouTube and link them in from there.  That way you can get a copy of it if you need it.

[IgorMir]

I have attached a file to my previous post. There is very little I can add to it. Except, that on the Left Hand profile Sketch I forgot to add a Coincident constraint on the construction line at the top. But it was not a show stopper. All other two sketches were fully defined. That what Inventor tells me anyway.

Cheers,

Igor.

---

@Vagulus wrote:

Okay - I got as far as creating Control Vertex Splines.  Now the LH Rail will not Sweep.  I even created LH Profile at Foot as a separate entity thinking that maybe AI didn't like re-using profiles.

I note the neither of the profiles is constrained.  Please, it does not help to be told that something is unconstrained without being told/shown how it can be constrained.  I am not trying to avoid this working

Updated ZIP attached.

---

[Xun.Zhang · Accepted solution]

Hello, 

OK, here is a simple one for your reference. The reason why the sweep failed which is caused by self-intersection.

If you move the start point 8mm higher, the problem is gone.

Please check out the enclosed part again and review the video below.

https://autode.sk/2XaFHCU

Hope it helps!

[Vagulus]

I don't understand why AI took exception to self-intersection here while allowing it at the two Bosses.

However, I got a hint from your Screencast to put in an extra profile above the bottom profile and to track to that,

That did the trick.

Thanks

[IgorMir]

Here is an updated part of yours. As I have said to you - your sketching technique is in due for a review. 

Cheers,

Igor.

---

@Vagulus wrote:

Okay - I got as far as creating Control Vertex Splines.  Now the LH Rail will not Sweep.  I even created LH Profile at Foot as a separate entity thinking that maybe AI didn't like re-using profiles.

I note the neither of the profiles is constrained.  Please, it does not help to be told that something is unconstrained without being told/shown how it can be constrained.  I am not trying to avoid this working

---

[IgorMir]

Hi Vagulus;

Here is your part. If you want to change some dimensions here and there - feel free to do so. I was trying to get as close to the dimensions of yours as I could. But when I see the angular dimensions to the second decimal - that's just not on. I have changed that to more round numbers. Apart of that - the part is pretty close to what you have asked for as far as I can tell. To view the differences - open up both parts, yours and mine. Move EOP all the way up on both of them and than go down one notch at a time.  Compare both parts to each other. That's all.

Cheers,

Igor.

---

@Vagulus wrote:

Okay - I got as far as creating Control Vertex Splines.  Now the LH Rail will not Sweep.  I even created LH Profile at Foot as a separate entity thinking that maybe AI didn't like re-using profiles.

I note the neither of the profiles is constrained.  Please, it does not help to be told that something is unconstrained without being told/shown how it can be constrained.  I am not trying to avoid this working 

[IgorMir]

Hi Xun;

It is interesting to see that when on the part I have posted the latest - making dimensions d241 and d253 equal to each other it results in the sweep failure. Granted - in the OP model these dimensions are 8mm and 10mm accordingly, so in my model it works OK. But there is still something to look into.

Best Regards,

Igor.

---

@Xun.Zhang wrote:

Hello, 

OK, here is a simple one for your reference. The reason why the sweep failed which is caused by self-intersection.

If you move the start point 8mm higher, the problem is gone.

Please check out the enclosed part again and review the video below.

https://autode.sk/2XaFHCU

Hope it helps!

---

[johnsonshiue]

Hi Guys,

This is a bug. The sweep should work. This is a Boolean failure. The sweep does work by itself (create a new solid). But, if you join it to other geometry, it will fail.

I use a different modeling technique to approach the design. Instead of Sweep, I used Thicken. The design is not completely the same as the original intent but it is similar. Please take a look.

Many thanks!

[IgorMir]

Hi Johnson;

Well, the Sweep works too but there is a condition when it fails. I have described what I have found in my earlier posts.

All of a sudden that exercise got interesting! 🙂

Cheers,

Igor.

---

@johnsonshiue wrote:

Hi Guys,

This is a bug. The sweep should work. This is a Boolean failure. The sweep does work by itself (create a new solid). But, if you join it to other geometry, it will fail.

I use a different modeling technique to approach the design. Instead of Sweep, I used Thicken. The design is not completely the same as the original intent but it is similar. Please take a look.

Many thanks!

---

[Xun.Zhang]

Hello @IgorMir,

Thank you for the feedback!

As Johnson indicated, it is a bug, please stay tuned for a while, try to reach out to project team for more details.

Thanks again! 

[IgorMir]

Hi Johnson and Xun;

Here is a part which updates and behaves properly. IV2018 format. It is a simplified version of the OP part, of course. But maybe I will try to recreate the original part from scratch and see if it works better.

Cheers,

Igor.

[IgorMir]

Hi Johnson and Xun;

Playing with the files a bit more - here is what I have discovered.

The rails can be created either by using Sweep or Loft.

Yet, when using Sweep (_C file), making dimension d273 equal to d241 one delivers an error message and the Sweep LH Rail fails.

In (_D file) the loft is used to produce both rails. In here the dimensions d273 and d241 can be made equal without a problem.

In Part3 - Sweep is used to create both rails. Dimensions d3 is equal to d19 one. No problems are here.

All of the above leads me to believe - the failure of the Sweep in (_C file) is file specific. Yet I would be interested to hear your opinion on this one.

Cheers,

Igor.