Helical boss failure

[jhiker]

Helical boss failure

I get this effect fairly frequently and it's becoming an annoyance. I'm still not sure whether it's an Alibre problem or something to do with the graphics (ATI Radeon X600) drivers. Perhaps someone could help.
It happens when I'm doing a helical boss operation with a perfectly legitimate sketch and input parameters (pitch, height, etc). Alibre goes quiet for a while then displays the misshapen blob you see at the LHS of the image. However, if I alter the display to wireframe the swept path looks OK.
I've checked out the display tab on the 'Preferences' (I think) menu and can happen whether the 'smoothness' is set to 'automatic' or 'manual'. It IS usually possible to obtain a proper 'solid' sweep by reducing the smoothness (and hence making the curves more coarse) but this all goes to pot when if you insert the coil shape into another assembly that has different settings for the curve smoothness. In any case, I would expect the 'automatic' setting to handle this.
Incidentally, I don't think the above applies in every case - I'm sure I've had results where the wireframe sweep has been screwed too.
On the other hand, many helical boss sweeps work OK but I can't seem to determine any pattern and the outcome is always somewhat unpredictable!

 

[MilesH]

Can you post the file of a Helical Boss that doesn't mesh/render properly?

 

[jhiker]

Sure - what do I do - just post the Alibre model file?

 

[MilesH]

As it's just a Part file, yes. Zip it and then attach the same way as you do the image files

 

[W. Wagelaar]

Go to properties/display/minimal circular facets. Lower down the current resolution and see if this helps. It worked for me.

 

[jhiker]

Sorry - I'm struggling to upload here. Keep getting this message..

Upload Error: Could not upload Attachment to ./files/coilexample.zip.

 

[jhiker]

Can't get it uploaded just yet.
If it helps I was trying to sweep a 'hollow spring' - two concentric circles.

OD 3.175mm ID=2.5mm on a radius of 26.98mm with a ref line on the vertical axis.
Had it set to height & pitch. Height=200mm - pitch = 5mm

Incidentally I was just messing with that same model - set the helical operation to 'revolution & pitch'
Rev=5
Pitch=5mm
Pass!

Increased the no of Rev to 50 without changing anything else - Fail!

There's something odd here...

Does it matter how high (long) the reference line is - I mean, does it have to be 'longer' than the height you're trying to 'helical'?
Jeff

 

[MilesH]

No it doesn't matter how long the Reference Line is.

I've sent you a PM

 

[MilesH]

I opened this file on my machine and it looked pretty much like the illustration above. The tessellation was set to 30. Amazingly, reducing this to 29 fixed all the mesh/render problems.....

This seems to be a weakness in whatever meshing engine Alibre uses.

 

[indesign]

Re:

I opened this file on my machine and it looked pretty much like the illustration above. The tessellation was set to 30. Amazingly, reducing this to 29 fixed all the mesh/render problems.....

This seems to be a weakness in whatever meshing engine Alibre uses.

Had to set my smoothness to 27 :? . I wonder if this could be graphic related or a bug or both. Anything above 27 would only get worse.

 

[indesign]

Re:

I opened this file on my machine and it looked pretty much like the illustration above. The tessellation was set to 30. Amazingly, reducing this to 29 fixed all the mesh/render problems.....

This seems to be a weakness in whatever meshing engine Alibre uses.

Had to set my smoothness to 27 :? . I wonder if this could be graphic related or a bug or both. Anything above 27 would only get worse.

 

[MilesH]

Re:

Had to set my smoothness to 27 :? . I wonder if this could be graphic related or a bug or both.

Well, if it happens more or less the same on ATI and Nvidia cards.....

 

[jhiker]

FWIW I've logged this as a Support Incident. Will advise if I get something meaningful back from the guys at Alibre.

 

[patmul]

I logged this problem as an Incident several months ago (June).

It is independent of the graphics card...it is clearly a renderer problem.

I was told to reduce the number of display segments...which worked!

I haven't tested 9.1 yet

 

[jhiker]

Yeah - it works but that's clearly a 'workaround' rather than a solution and is less than satisfactory for a number of reasons.
It can look awful!
If you subsequently use the model in an assembly where the number of facets is set higher than the 'reduced' number then the feature goes haywire again!
I still think this needs attention - does this happen in other solid modellers - anybody know?
I too have not tested 9.1 yet.

 

[HaroldL]

jhiker,
Just for grins, I tried your spiral spring on my laptop to see what my results would be.
It took a while to create the spiral spring, but the preview is Uuuuugly.

I too have just upgraded to 9.1 Expert. I also downloaded and installed the latest version of DirectX and the latest version of my ATI Mobility Radeon 7500 video driver I could find. (I really should upgrade to something with a little more power than Dell Inspiron 8100.)

I had some of the same issues with the preview all screwed up in sheet metal.

BTW, the Minimum Circular Facets is set to 20.

I did some other test parts where I started with a rectangular sketch and was able to extrude a proper spiral with no display problems. But when I changed the skectch to a circle it displayed garbage. (With the same extrude settings.) However I had the Minimum Circular Facets set above 60 for those parts.

HaroldL

 

[jhiker]

Indeed. What puzzles me is that you can get different results depending on whether the helical boss feature dialog is set to "Revolution & Pitch' or 'Height & Pitch'.
Furthermore, when set to R&P and the no of revs to 5 it came out OK but when I increased the no of revs to 50 it failed!
I am still bewildered.

 

[HaroldL]

I submitted an incedent report on this problem along with the screen shots.

And the reply from support is:

This issue seems to be a problem with ACIS (the kernel that Alibre is written off of) and your system. This is an issue that the dev team will have to work on for a future release.

lt will be interesting to see how long it takes for the dev team to come up with a fix. And will it be a "hot fix" for download or will it get rolled into the next or subsequent release of AD.

HaroldL