Complete Gear Train Sample (In .stl format)

[Greg Smith]

It seems the old problem exists that has been troubling us all for decades:  importing and exporting assets from one environment to another.

I've been able to import 2D .dxf's for a long time, into my CAD program, and explode them, enabling me to modify them in 2D for adding other 2D parts - but 3D seems impossible.

Without 3D import, a program like Gearotic, for me, only serves to generate 2D outlines for CNC machining.  It does have a nice set of odd gears and the escapements and ratchets and plates - which may be enough to justify its cost. 

But, I was hoping for a miracle - the ability to communicate and exchange, flawlessly, 3D data with a CAD program for testing - enabling me to visualize a mechanism in its entirety, before resorting to trial and error physical prototyping.

I may never see it.


Greg Smith

[David Sohlstrom]

Greg

I have had good luck importing Iges files into Geo. from other programs. One of the annoying things about Geo. is you can not constrain gears that are not 1 to 1 ratio. If they solve that I will be a happy camper.

Dave

[ArtF]

Greg:
 
    One of the reasons you may never see it is that shape and form alone are not enough. Things like center of gravity, pitchlines and the relative angualr velocities are just not in the information that makes up a gear. Each type of 3d object also has its own limitations. And the formats you see now for interchange are not really objects but lists of primitive commands that woudl generate such a shape. A modern intercnage format is more like a series of lines like "Take a cylinder, cut it at such and such an angle, now weld it to a box that is cut at this angle..etc... Makes it very hard for something like gearotic to put out such files, and even if it did, youd have to fiull in information like
the formula of the pitchcurve, ( which for noncircular gears would be far out of the scope of most cad programs..)

Its why its an age old problem... :)

Art

[Greg Smith]

Art:

I guess the next question would involve the formats that do import and export properly (3D .dxf).

Do these output files (involving an entire gear train) - though they are output one by one - maintain their position in space - so that when imported into a CAD application (one by one), they "assemble themselves" into the arrangement of this train?

Maybe we could take TurboCAD or ViaCad as a case for the host application.


Thanks,

Greg Smith

[ArtF]

Greg:

No. The dxf's are all at 0,0,0 as center. This is because in essense that is that object. The physical placement in space is data in the simulation engine so no form of output takes it into account. Im not opposed to adding it as an optional output , pretty trivial to just add the center to all vertices on output. Not sure that woudl help though. Lets say you output a 10
tooth gear, and a 20 tooth gear in tandem in proper space. Will alibre kow the ratio is 2:1 for rotation? Or can you easily tell it that? If so it may make sense to have the option
of adding a center location to the data. BUT, it would be very difficult with other than round gears even if you CAN tell alibre the ratio. There is no ratio for an escapemnt or noncirclular
gear. These are algorithmic motions that are probably impossible for alibre to do. While I could tell you the formula for the elliptical gears rotation for example, its a complex calculus based algorithm to actually figure it out, I cant see any way alibre could possibly deal with the data even if it knew the centers of objects. Cad programs normally just arent great simulators except in usually linear ways unless its a very expensive simulator like solidoworks etc..and even then youd have to konw the calculus involved in the relationships..

Art