array of mirrored feature

[Anonymous]

Can someone explain why I get the errors in 'with mirror.ipt'? It has to do with the mirror feature in the arrays, but I don't understand exactly what the problem is.

[Anonymous]

Hi Kevin, The problem is that the way the Rectangular Patterns are defined you end up with multiple overlapping occurrences. Rectangular Pattern1 is 5x5 with direction2 parallel and opposite to direction1. Inventor computes 25 occurrences, but only 9 of those are unique - the overlapping occurrences complain because the nested mirror did not have any affect on the part. If you expand the pattern in the browser and highlight the occurrences individually you'll understand what's going on. I think the best way to create a bi-directional "centered" pattern like that is with two separate patterns - i.e. a 1x5 for each direction. Hope that helps. Tom Sturtevant Inventor Part Modeling "Kevin Wehner" wrote in message news:40f7cf58_3@newsprd01... > Can someone explain why I get the errors in 'with mirror.ipt'? It has to do > with the mirror feature in the arrays, but I don't understand exactly what > the problem is. > > >

[Anonymous]

That doesn't make sense to me, because I only get the error when arraying the mirror. There are also ovelapping occurrences of the hole patterns. "Tom Sturtevant (Autodesk)" wrote in message news:40f80ebb$1_2@newsprd01... > Hi Kevin, > > The problem is that the way the Rectangular Patterns are defined you end up > with multiple overlapping occurrences. Rectangular Pattern1 is 5x5 with > direction2 parallel and opposite to direction1. Inventor computes 25 > occurrences, but only 9 of those are unique - the overlapping occurrences > complain because the nested mirror did not have any affect on the part. If > you expand the pattern in the browser and highlight the occurrences > individually you'll understand what's going on. > > I think the best way to create a bi-directional "centered" pattern like that > is with two separate patterns - i.e. a 1x5 for each direction. > > Hope that helps. > > Tom Sturtevant > Inventor Part Modeling > > "Kevin Wehner" wrote in message > news:40f7cf58_3@newsprd01... > > Can someone explain why I get the errors in 'with mirror.ipt'? It has to > do > > with the mirror feature in the arrays, but I don't understand exactly what > > the problem is. > > > > > > > >

[Anonymous]

Hi Kevin, I understand your point - the patterns with and without the nested mirror are very similar, why does one generate an error but not the other? The answer is that Inventor handles patterns of features (e.g. pattern of Hole) differently than patterns of patterns (e.g. your pattern containing the Mirror). I'm not trying to defend how Inventor does it (well, OK, I am doing that a little). At a minimum it is inconsistent. In order to compute your 5x5 pattern of the mirror, Inventor basically recomputes the mirror in 24 different locations. All but 8 of these overlap a previous occurrence, and errors get generated because the part did not get modified. When the other features are patterned, we can use a more optimized approach which doesn't warn about the overlapping occurrences. Again, I really recommend that you model this as two 5x1 patterns. This will give you a smaller, faster, more stable part. Tom Sturtevant Inventor Part Modeling "Kevin Wehner" wrote in message news:40f8208c$1_2@newsprd01... > That doesn't make sense to me, because I only get the error when arraying > the mirror. There are also ovelapping occurrences of the hole patterns. > > > "Tom Sturtevant (Autodesk)" wrote in message > news:40f80ebb$1_2@newsprd01... > > Hi Kevin, > > > > The problem is that the way the Rectangular Patterns are defined you end > up > > with multiple overlapping occurrences. Rectangular Pattern1 is 5x5 with > > direction2 parallel and opposite to direction1. Inventor computes 25 > > occurrences, but only 9 of those are unique - the overlapping occurrences > > complain because the nested mirror did not have any affect on the part. > If > > you expand the pattern in the browser and highlight the occurrences > > individually you'll understand what's going on. > > > > I think the best way to create a bi-directional "centered" pattern like > that > > is with two separate patterns - i.e. a 1x5 for each direction. > > > > Hope that helps. > > > > Tom Sturtevant > > Inventor Part Modeling > > > > "Kevin Wehner" wrote in message > > news:40f7cf58_3@newsprd01... > > > Can someone explain why I get the errors in 'with mirror.ipt'? It has to > > do > > > with the mirror feature in the arrays, but I don't understand exactly > what > > > the problem is. > > > > > > > > > > > > > > >

[Anonymous]

Until you mentioned it the other day, I never noticed that the pattern overlapped itself. I just assumed it wouldn't create thos instances when the directions were parralell. Now that I realize how it works, I'll definitely use 2 5x1 arrays. It would be nice to see a check box added to the array window that would eliminate overlapping instances, or just create instances along the direction axis. Creating two arrays in this case makes the part harder to manage. "Tom Sturtevant (Autodesk)" wrote in message news:40f84bf2_3@newsprd01... > Hi Kevin, > I understand your point - the patterns with and without the nested mirror > are very similar, why does one generate an error but not the other? The > answer is that Inventor handles patterns of features (e.g. pattern of Hole) > differently than patterns of patterns (e.g. your pattern containing the > Mirror). I'm not trying to defend how Inventor does it (well, OK, I am > doing that a little). At a minimum it is inconsistent. > > In order to compute your 5x5 pattern of the mirror, Inventor basically > recomputes the mirror in 24 different locations. All but 8 of these overlap > a previous occurrence, and errors get generated because the part did not get > modified. When the other features are patterned, we can use a more > optimized approach which doesn't warn about the overlapping occurrences. > > Again, I really recommend that you model this as two 5x1 patterns. This > will give you a smaller, faster, more stable part. > > Tom Sturtevant > Inventor Part Modeling > > > "Kevin Wehner" wrote in message > news:40f8208c$1_2@newsprd01... > > That doesn't make sense to me, because I only get the error when arraying > > the mirror. There are also ovelapping occurrences of the hole patterns. > > > > > > "Tom Sturtevant (Autodesk)" wrote in message > > news:40f80ebb$1_2@newsprd01... > > > Hi Kevin, > > > > > > The problem is that the way the Rectangular Patterns are defined you end > > up > > > with multiple overlapping occurrences. Rectangular Pattern1 is 5x5 with > > > direction2 parallel and opposite to direction1. Inventor computes 25 > > > occurrences, but only 9 of those are unique - the overlapping > occurrences > > > complain because the nested mirror did not have any affect on the part. > > If > > > you expand the pattern in the browser and highlight the occurrences > > > individually you'll understand what's going on. > > > > > > I think the best way to create a bi-directional "centered" pattern like > > that > > > is with two separate patterns - i.e. a 1x5 for each direction. > > > > > > Hope that helps. > > > > > > Tom Sturtevant > > > Inventor Part Modeling > > > > > > "Kevin Wehner" wrote in message > > > news:40f7cf58_3@newsprd01... > > > > Can someone explain why I get the errors in 'with mirror.ipt'? It has > to > > > do > > > > with the mirror feature in the arrays, but I don't understand exactly > > what > > > > the problem is. > > > > > > > > > > > > > > > > > > > > > > > >