Re: [CAD_CAM_EDM_DRO] Re: Question about angular and linear feedrates

Dan Statman wrote:

> Fred,
>
> Of course you are absolutely correct when taking into account true 4-axis
> 3-D work. But given the fact that low end controllers are not handling the
> mixing of linear and rotary axes very well by using the linear axis feedrate
> exclusively and then spinning the rotary axis at a value to keep the move
> synchronous, which may be far in excess of a reasonable value, don't you
> think it would make sense to limit the speed to the SLOWER of the rotary
> feedrate or the linear feedrate.
>
> This obviously assumes that the feedrates are selected with knowledge of the
> radius of the rotating part. By selecting the feedrates with this
> knowledge, I then do not have to have the control program know the radius.
> and the only downside is that sometimes the part moves slower than I
> intended. This is not nearly as detrimental as moving faster than I
> expected. If it moves slower then the part takes longer, if it moves faster
> the cutter breaks and the part is potentially ruined.

The place this gets messy is in complicated surfaces, where the movements
require several axes to move simulatneously, and where you want to get smooth
motion at a constant feedrate across the surface being produced. If the control
knows the radius, you can set one feedrate, and then just give the 3-D (4-D?)
(mixed-D?) coordinates in a stream of G-code blocks, and the control should
keep the feedrate constant, and the motion smooth.

If you have to do this in the CAD or postprocessor stage, you will have feedrates
varying in each block. I don't know whether any particular CNC control program
will give smooth motion when the feedrate is going up and down with every block.

Jon