Sorry, another LONG one, but I hope it flows easily...
I was looking around the YouTube vids and Google pages to find vids or such on parametric constraints. I ran across one for Inventor:
http://www.youtube.com/watch?v=...N7JKnS70&feature=relatedSomething like this would be awesome for creating brackets and sideshell stiffeners (webs and flanges) and stringers in a ship's sideshell. Problem is is that except in a ship's parallel midbody and and where long expanses of bulkheads exist, the stiffeners and stringers will change in y and z at or between each station, not all that unlike the wing, rudders, and fuselage of an aircraft.
To create stiffeners using a simple T, I:
-- created a web of 458 mm depth and a flange of 288 mm
-- attached the free end of the web to a particular station
-- swept web surface using the web line and the station curve
-- swept a flange surface using the flange line and the station curve
-- thickened each the web and flange surfaces to 15 mm
When the surfaces would not thicken (due to ACIS errors), I rebuilt the surfaces and then usually was able to produce a thickened plate.
If I decide that the 15mm plate of the web is too thick or too thin for a given area, I can adjust the thickness in the data or re-thicken the surface and discard the incorrect solid.
But, it would be very VERY nice if VC, VCP, and Shark had a table input or an import facility so that using .dbf, 123, xls, or open office spreadsheets I could externally enter the station number where flanges and webs would be swept, thickened, and color coded so that view changes won't obscure the parts if they are adjacent and share the same colors or the lighting makes them ambiguous.
It would be really awesome if this input table were portable so it could be extracted, exported, and even manipulated via a real database interface (or, if one is inclined, via a spreadsheet). This would allow real-time editing and graphing and material CGs distribution eye-balling.
Also, as parts are created, they could automatically be appended to that table, or into a separate table. In any case, parts would be flagged in a field as user-created independent of structure contact, user-driven by imported tabular data or constraints, and user-vc/shark produced (where intersections of inserted parts occurs.)
The table might work really well for civil architecture purposes, too, where profiles being swept exist in steel buildings and naval architecture. The ability to define and constrain local curvature (to simulate an awareness of FEA/FEM and to make things easier for external programs) to mathematically smooth and computationally forgiving lines.
I'm suggesting this because I have a feeling that somewhen someone will tell me that certain regions of my model have too weighty or too thin stiffeners. With about 250 in the hull alone (~125 starboard, 125 port), I could just individually click on or in selected groups click on the construction web and flange lines, but there's no grand overview to just do it by rows and columns. In the 15 major compartments (watertight subdivisions) some might be 10 mm, some might be 15mm, some might be 12 mm... For now, for visualization, i vary them so that even if changed, the average weight of installed material will hopefully be closer to real than otherwise.
