メインコンテンツへスキップ

Knocking the corners off.

コメント

12件のコメント

  • Tim Heeney

    Buk.

    Edit: Try this: select solid>detach all the combine them back ( i'm trying to remove the 'pattern' knowledge ) then try again.

    I have often thought the below...

    That alteration should be no problem at all but i have seen similar examples of massive stubborness behavour.

    I think those teeth are made up with complex nurb surfaces and operations upon many of those in a single go causes massive problems...

    As you know i'm using V4, my machine is a high end from 2014, 64Gb ram etc and yet this sort of alteration also massively disrupts play.

    I'm suspicious DSM doesn't have the full 'power' that SC does in hefty operations. My pc doesn't allow investigation into what's going on with RAM whilst DSM is running...

     

     

    0
  • Tim Heeney

    Hi Buk.

    Try this: Turn ( off ) Pull tool animation in settings helps massively - approx 500% faster.

    The gear example is only straight pulled, no helix but similar in surface count. Maybe takes 1 to 1.5 secs.

    I'm still suspicious of hidden pattern 'Knowledge' holding / killing operations. try 'detaching all' then combining them, or try detaching all, then selecting all, then apply pull cut to all the individual surfaces.

    0
  • Buk

    I found a way.

    Turns out that the math of constructing a 0.25 bevel on the disc, meant that the cut edge would be exactly coincident with the innermost extent of the grooves:

     

    And that appears to be why the rotate cut failed. Moving the cutting surface in (or out) by just a tiny amount:

    allowed the rotate cut to work:

    0
  • Jacant

    I can't find a problem doing this.

    0
  • Tim Heeney

    Buk, Jacant...

    Am i doing this right ?

    Do you both find pattern responsiveness can slow down operations?

     

    0
  • Tim Heeney

    Not a problem with chamfer coincident with gear trough.

    I suspect the clue is in your picture , near the red circles, the cutting profile appears to disect what appears to be upper and lower scallops. In your successful modified cutter position, the (so called) scallops are no longer dissected and it works...

     

    0
  • Buk

    I guess we'll have to put it down to a "v2 thing". It is utterly consistent. I tried a bigger size to see what happened and once I click Full Pull, it spins for a second and then the surface disappears -- exactly as it would if it had done the cut, but no material has been removed.No errors or warnings. Just no change.

    But if make so that the edge of the cutter surface is even lightly one side or t'other of the edge of the knurling, it works fine:

     

    I've noticed similar effects with other cut/intersect/split operations where if edges or points are exactly coincident, the operation fails; though usually noisily.

    And while I was constructing this post, I was experimenting to see what was and was not possible, and ... one of my experiments resulted in this:

    Holy Solids Batman!

    Note the structure tree!

    It's a bit optical illusion-ish.The diamond shapes are high points; the cylindrical surfaces go in, not out. I pulled the half circle helical cuts left and right handed and then bevelled the edges and rounded the sharp edges.

    Gonna save now and re-boot see if it solidifies! IT DIDN'T!

    If I section it, it hatches it as a solid; but If I move the section plane to the visually "badly rendered" portion at the top; it shows the anomaly is more than "skin deep" [sic].

    Note that you can see the insides of this solid in the areas boxed in orange; and see thru it in the red boxes. That's after saving, exiting, re-start and reload.

     

    0
  • Tim Heeney

    Buk,

    Lets start off by saying you're inspiring me into knurling patterns.

    Lets follow that up by saying patterns do seem a bit tempermental !

    I'll go on to add it's bonkers to see a solid ( i'm assuming 'corrupt' ) not acting like a solid with missing surfaces / additional touching surfaces etc, all displayed in structure as a single solid!

    Perhaps v4, ( sorry that naughty phraise again ;) it might not pass the integrity check.

    Anyways, i ask in the picture below if you set your modelling accuracy to level 6 - but i'm very sure you do as i can't imagine how you'd do what you do without it being at 6!

    As for ACIS/DSM accuracy, i did a test comparing DSM 3D to Solid Edge 2D drafting ( Parasolid and D Cubed) - just a fillet curve fitting between a sphere and a cyclinder - DSM result gave a result identical but one order of magnitude lower ( higher precision). I stopped questioning ACIS precision after that .

     

    high regards...

     

    0
  • Buk

    Tim,

    It is definitely a bizarre effect :)

    I suspect that the invisible faces are "inside out" -- their normals pointing the wrong way, and thus the render engine is culling them because it doesn't think they should be visible,

    I have my render set to 7:

    I'll try converting it to another format and see what happens.

    0
  • Tim Heeney

    Buk,

    Not graphics quality, although i set mine to rendering 4 and medium for anti-aliasing. The above part was at those settings. When the model gets slightly slow, i start turning down it's graphical representation. Note this can be done globally under settings or individual surfaces can be altered under it's 'properties'.

    I really meant 'Units' - primary precision for linear and angular, set to 6

    And i think you're right - some face normals probably inverted somehow.

     

    ps. i found an answer to my problem with disappearings steps - where a altered round swallowed up another round...i posted it under it's thread yesterday. I was quite surprised ! I just haven't fully got used to this modeller yet!

     

    0
  • Buk

    Sorry Tim. I should've zoomed in on your pic.

    I also have Units->Primary precision set to 6; and angular to 3.

    i found an answer to my problem with disappearings steps - where a altered round swallowed up another round...i posted it under it's thread yesterday.

    I have seen/can't find what you are referring to here?

     

     

     

    0
  • Tim Heeney

    Buk,

    Set the angular precision to 6 as well - sometimes it isn't 'sticky'.

    Answer to the disappearing step/rads - it's here...

    https://designspark.zendesk.com/hc/en-us/community/posts/360008665618-Disappearing-Step-unwanted-action

     

    0

サインインしてコメントを残してください。