HarryGeier wrote: ↑
Wed May 08, 2019 10:32 am
Well, it looks a bit like a missing feature of techdraw, but i would rather say, it´s mistaken use of fillet. One could not precisely manufacture this as the radius of the fillet never touches the extends of the original block.
I don't claim to know much about technical drawings, so I don't know if it's right or wrong, but I've noticed that professional engineers sometimes use the "continuations" of the fillet edges for dimensioning, and it's usually quite obvious what is meant by the design, even if it's impossible to measure the dimension with caliber. See the clipped example I took from a drawing made by a professional. I do agree that one should usually only define dimensions physically measurable with a caliber, so that the two surfaces under the measurement points would be parallel, but it seems that the pros break this rule for convenience every now and then. Possibly because parts are made with CNC, and the dimensions are just for reference..? Note on the quoted drawing, the center of the arcs are not at 21.2 mm from each other, due to the angled cone; so using center-of-arcs causes ugly values with infinite numbers after decimal point.
I'm struggling with a similar situation, but in the end decided to define the dimensions differently. In this case using FreeCAD forced me to find out a solution, where every dimension could be measured with a caliber or a protractor. Maybe that's a good thing.