On 2021-08-05 06:13, John Dammeyer wrote:

How does one know what size tabs and how many?

I view this as a mystical black art.
For me, the basic idea is to
(a) support the part to prevent damage when it separates from the parent, to stop it catching the cutter for example. In that case, position the tabs to provide enough support to prevent vibration or movement of any element. (b) prevent the parent material from moving or, as in this case, collapsing as the part separates.

Tabs often serve purposes (a) and (b). From the point of view of clean-up afterwards, the fewer tabs the better, and the more accessible they are for removing and for subsequent finishing, the better. I would think first about whether the position of the tabs affects the accuracy required of the part. After that, I think about the aesthetics after clean-up. Tabs on a straight face are easier to clean up than on a curve, and tabs on an external curve are easier to clean up than on an inside curve, just because my flat files are closer to hand than the half-round files. There is an argument that cleaned up tabs on a curve tend to catch the eye less than on a straight face.

In the case of this particular part, I reckon you have this about right. The main problem is an inward collapsing due to the pressure from the vice jaws as well as possibly stresses in the tube which will naturally make it want to either spring apart or collapse inwards.

I quite like the triangular tabs because I find they can be a little easier to cut with a knife or a very fine saw blade, but the rectangular tabs are stronger and easier to program manually if necessary. I like to make the tabs as thin as possible, to aid clean-up, but the over-riding factor must be the strength they require to do the job.

That's maybe not a clear answer, but those are the thoughts in my mind every time I have to place tabs.

Marcus


_______________________________________________
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Reply via email to