Barrel Cam

Forum|Forum|8 years ago
November 7, 2017
2 replies
5493 views

Hi,

I have tried to make a barrel cam. I have seen previous threads about this topic on this forum. I know that this is tricky geometry and I have tried different methods. I have attached files below with the cutter tool assemblied with a slot constraint to check for interference. There is very little interference but it is there and I want to see if I can model it with no interference. This part was previoulsy made by pattern a revolved cut in increments of .005. This way is not wrong but I wanted to see if I could do it another way. I had seen a previous response by Tim McLellan but I couldnt quite get his method to work. I think the sharper the corner the more the geometry gets distorted.

1_test_cam_3-asm.zip

General

T

TomU

23-Emerald IV

Multiple separate cuts will give you a pretty accurate slot, but the model will be very heavy and useless for actual machining. I've been unable to find any other method that can accurately reproduce what is really happening between the slot and the follower (or cutter). At this point I've given up trying to produce a machinable model in Creo Parametric and have shifted to using software that can generate G-code based on the cam profile without needing an actual cad model.

Fundamentally, we need something like 3D sweep that can accurately create geometry from all interferences between the model and the cutter (or cam follower.)

I have done some initial testing of a software called MechDesigner from PS Motion. It does appear that they can produce geometrically correct models in SolidWorks, as long as the transitions are not so sharp that the cam becomes undercut. For example, they can't accurately reproduce this area: (shown here in Creo with separate cuts every 0.5 degrees)

T

TomD.inPDX

17-Peridot

From what I can tell in Solidworks, they have codified the operation much the same way as I did in the linked video above and discussions. It is still an iterative series of revolves. How they actually finish up the operation is not immediately obvious. There is also the motion profile in Mechanism, but that is very crude and imprecise. The resulting element is a faceted body.

T

tkocot

Author

1-Visitor

Thank you guys for the replies. I did check out those videos and they were really neat. I had seen the thread on the globoidal cam modelling. I tried to make a boundary blend between two curves and then offset from the mid boundary plane. It is close but not a 100% and I am unsure as to why.