Greetings
In the CAD class we were assigned the reverse engineering of a molding assembly, composed to two plates, two piece of channel iron, two lengths of 2x4 tubing, and the washer nuts and bolts to hold it all together. [insert picture1] Much fun - as each of use also had to get the same weight and center of gravity. Worth up to 350 points and I got all of them. Even if they were both off by a couple thou. But! There were some problems. E.G., the slots in the plates were .300 wide. Now, I don't know about you, but getting a 3/8 inch bolt to fit through a .300 slot seems a bit tricky. In CAD "it can be done!" (ignore that clash report ...) So, after the assignment was turned in, the class ended, and the grades issued - I made revisions. Lots of revisions, incorporating some of the things I learned after I stopped working on the assignment. This is going to go in the portfolio, after all. One item which the group had discussed was some redesign work - "the plates should have holes on one side and slots on the other." Yada, yada, and so forth. Anyway, part of the assignment revision (we got changes made to assignments right up to due dates) was to add Functional Tolerance and Allowancing. That is a "new and improved" nifty means of adding dimensions, surface finish, GD&T Feature Control Frames and locating the A, B, C (etc) Datums, and so on to the actual 3D modeling. Wee. Be still my beating heart.
Now, here is my question for the collective wisdom of the interwebs, locating Datums (for those not In The Know - Datums are the start points from which all measurements flow, the (0,0,0) spot in the
3D matrix. When making the rough sketches of the parts, and then the actual Drafting of our results, Datums got assigned to the edges. Simplest place to find. Yet as I was 'catching up' Friday, I came to the realization that there were some critical dimensions which had to be 'the same' in two parts, in order to get them to fit together. Namely, the slots in the plates, and the tapped screw holes in the channel iron need to be spaced the same amount in "X" (as well as being the right distance in the "Y") so that they 'fit' . Now, one solution is to make the slots wide enough to accommodate any "drift" in the screw holes' position. But, well, I can be very meticulous, and this is a CAD class and an exercise in precision location and, blah, blah. In a different exercise, the advice was given to make Datum A & B the axis of the round part. What I am thinking is to make the axis of the inner radius (the one furthest from the edge) of Slot 1 to be Datum A/B - that is the end marked "pt1" would be where the inner radius is and that would be the point from which height and width would be set.| ___ | / \ pt1 | \___/
Because N inches to that ----> direction is another "pt1" for a second set of slots, to bolt the plate to the other channel iron, so that it all are in the correct locations. And it would seem "reasonable" to me (bwahahah) to locate pt2 relative to pt1 and then pattern the slots along the edges at the distance "Slot_Spacing". Now I realize that for something this simple, such precision is ridiculous. But there is this bit of my makeup which says "until that first pass removes some metal, the part is undefined in terms of its actual location in space." Some times, where the edges of the plate actually are is of less importance than the precision of the hole pattern _in_ the plate. Not to mention, this was a class where all dimensions are to four decimal places. So yes, locating a hole with a true position of .001 is insane, but that's Dave for you. [now, what was the question I came here to ask for advice? Oh yes.]
So, when the spacing of a hole pattern is more critical than the precise distance from the edges, is it better to 'design' a plate (or whatever) from the critical hole outwards, and dimension the plate from the center point of the hole, or locate the hole from the edge of a plate?
Either way, I know I am going to have to burrow into GD&T a bit more to know which way to make something "sloppy"
tschus pyotr
-- pyotr filipivich. Discussing the decline in the US's tech edge, James Niccol once wrote "It used to be that the USA was pretty good at producing stuff teenaged boys could lose a finger or two playing with."