Helical fill pattern of small holes changes with parent material diameter

[awberke]

I am using the fill pattern function to create a sprial fill of a constrained circle on a simple circular plate. The holes are very tiny, .026 and the are spaced .129 hole to hole and .129 radially. The parent material is just a circular plate with a diameter of 14.4". When i change the diameter of the part to 16.2" the spiral pattern seems to rotate and change the pattern completely. I have made the original hole and pattern completely independent of the geometry of the part and yet the pattern still changes with diameter change.


Link

Link

Link

If anyone wants to see it for themselves, just go into this part and change the spacing to .129 between holes and .129 radially. I don't recommend regenerating the part as it will take forever, but the black dots will show the issue. If you cancel out and then change the diameter, go back into the pattern, you will see what my pictures show.

It seems that the helix slightly rotates when the part geometry changes, but i have no idea why.

 

[dgallup]

Fill patterns are not meant to be very precise. It looks like in the second example it just did not make the first hole. I did not download your file. Could be an accuracy issue or maybe the distance between the pattern center and the fill boundary changed? I would not say the pattern changed completely, I would say it changed very little. What is one hole out of hundreds or thousands? I'm assuming this is some kind of a filter, probably not orientated. Are you going to count every hole in receiving inspection?

----------------------------------------

The Help for this program was created in Windows Help format, which depends on a feature that isn't included in this version of Windows.

 

[awberke]

The pattern does change completely. The first row in one image and the second row in the second are off by a hole, its not just a matter of two holes added. This plate controls flow to a wafer for electroplating interconnects for microchips. The pattern is important. Thank you for taking the time to post.

 

[dgallup]

If you need precise control, don't use a fill pattern. Use curves and pattern along the curves. It is going to be much more work but you will have much more control.

A couple of holes on the edge out of what must be hundreds or thousands of holes is not going to make any significant difference to your flow. The tolerance on the hole size and position will swamp any effect of one or two holes.

----------------------------------------

The Help for this program was created in Windows Help format, which depends on a feature that isn't included in this version of Windows.

 

[awberke]

The two holes is just a demonstration that the pattern is not the same, the whole pattern is rotated. Also, it is not just flow, it is the field lines that travel though the holes in this plate. For some scale, the original pattern landed a hole directly in the center of the plate, this caused a uniformity issue at the center of the wafer. The pattern had to be changed to allow for the holes to be eccentric by .013, small differences do affect the performance.

Agreed, perhaps the fill pattern is not the best way to model this, but I haven't looked into the method you're talking about. I understand that you can say: "don't use fill" but it isn't a random generation of holes, should it be reproduceable?