Slots as Datums

[Jieve]

Hello guys,

Question: I have a series of 5 slots side by side (width-wise) cut into a large flat plate. The slots are symmetrical about the center slot, which is the largest.of the 5. There are a total of three different slot sizes. I want to locate slot two with respect to slot four with a position tolerance, and slots 1 and 3 with respect to slot 5 with a different position tolerance. In other words, i want slot four and slot five to be Datum slots. What is the best way to indicate this On the drawing, and how would a slot be used practically as the datum during measurement?

Working to ISO. Thanks!

 

[fsincox]

Pmarc,
Can you explain your use of the word "physics" in more laymans terms, please?
Frank

 

[Jieve]

Ok, next 3 posts are the drawings and the pics. Have renamed them to something without & in the name.

Drawings

 

[Jieve]

Pic Assembly

 

[Jieve]

Pic Underside of plate

 

[pmarc]

Frank,
I will use two examples to clarify my point:

1. Already described issue with CMM verification and reporting of positional tolerance of a hole/pin: It is simply incorrect to boil down to 2-dimensional case a tolerance which is 3-dimensional by "physical", or maybe I should say by "geometrical", nature (has certain depth). Of course there are situations when such oversimplification can work without risking too much, but in general such approach does not take any axis orientation error into account. So if you imagine a hole with severe actual perpendicularity error to primary datum plane, do the reported coordinates of hole's center [x,y] measured at one plane only (usually located somewhere near to the top of the hole, assuming bottom is the datum plane) really give you something? Do you think such info is useful to assess hole's ability to mate with counterpart's pin?

2. Your example of surface flatness error verification on a CMM: As your collected data will be a set of relatively small finite number of probed points, the only conflict with theoretical approach will be in the amount of points checked, but the "physical" nature of verified tolerance will not change - you will still be probing a surface to check one of its geometrical properties, just like you would do if you had a possibility to probe absolutely every portion of this feature.

Does this explanation make sense?

 

[fsincox]

Pmarc,
Yes it does.
Thank you,
Frank

 

[Jieve]

So what is then the proper way to measure hole perpendicularity without a CMM machine at RFS? Especially for a hole that has a very tight position a.k.a. perpendicularity tolerance? Is measuring edges in two planes truly enough, or should a pin be used? And if so, how is this done?

Our in house shop that is making my parts does not have a CMM machine and in my discussions with the machinists, the syndrome that Frank pointed out above seems to be prevalent - "the process naturally produces". I've also been asked more than a couple times the question "how are we supposed to influence that?" and "how are we supposed to measure that without a measurement lab?" regarding some of my GD&T callouts. Albeit this is a mostly "academic" setting and the guys machining the parts are in their 3rd year of machinist training, based on a lot of your posts here I'm thinking this is a real world problem. I'm trying to impose on the students the importance of GD&T in design, as they are all working on a dual study program to become design engineers. While they're pretty good at turning out good parts with practice and they do have decent understanding of GD&T, it bothers me that I can't always answer the "how to measure" questions. However, thanks to this board, I can give examples of why I spec what I spec and I can often see a light go on with some of them during the discussions.

BTW if you guys have time to kill, it may be a lot to ask, but I would be extremely appreciative if you could just take a quick look at the drawings I posted and tell me how you would recommend dealing with the slots in the question I posted earlier and if there are any ways I can simplify or make the drawings better. I'm posting the drawing of the shafts/posts that I didn't include before so you have all of the drawings of all parts in the assembly.

Thanks!

 

[fsincox]

In your shaft drawing, datum "B" as it is stated (a secondary datum) adds no real value?
Frank

 

[Jieve]

B is specified as a datum for control of the perpendicularity of the hole. Is this not correct?

Thanks.

 

[powerhound]

The hole is controlled centered and parallel to datum axis A first. If it is more critical that the hole be perp to B than parallel to A then the datum precedence should read B|A| instead of A|B|.

Powerhound, GDTP S-0731
Engineering Technician
Inventor 2010
Mastercam X6
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[fsincox]

No it does not, Exactly Powerhound! This is what I call throwing datums at features hoping you have enough, the ASME standard goes to great lengths now to show the effects datum specification has on the reference framework. 2009 even allows for overriding that control. I do not believe the ISO is different in this way.
Frank

 

[Jieve]

The better question would be this: if the top surface of the post is not perfectly perpendicular to the axis but the screw hole is perfectly parallel to the axis and the screw is tightened down, what will happen? Will the surface of the post tend to rest flat against the angle bracket with the screw at an angle through the clearance hole in the angle bracket, or will the screw washer surface sit flush with the top surface of the post remaining at an angle to the mating surface? That would answer the question which datum is more important. I'm actually no longer sure what would happen.

Rather than calling this throwing out datums and hoping i have enough, I'd call it realizing that my original reasoning may be incorrect because I didn't have a füll understanding of what would happen during assembly. It seemed to me in this case that the priority is parallelism with the axis, although i didnt realize that there was a difference in interpretation when specifying which datum comes First in this case.

 

[fsincox]

Jieve,
Sorry, I did not mean to pick on you. It is something I see and have heard that description used for it before. That was all I was trying to point out. I can actually see other reasons like maintaining a simultaneous requirment on multiple features, also.
As you see it depends on how it functions, if it only mounts against a face or is it actually piloted in a long bore, function matters.
Frank

 

[powerhound]

Yes, the order of the datums in the datum portion of the feature control frame absolutely make a difference. They not only communicate part function, they communicate setup for inspection.

As Frank said, as you have the drawing shown, B as a secondary datum adds no value, thus A is the only datum necessary. If, however, you need the hole to be perpendicular to the datum B surface, then it should be called out first to control perp, then A as a secondary controls the location.

Powerhound, GDTP S-0731
Engineering Technician
Inventor 2010
Mastercam X6
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[MechNorth]

You NEVER measure from an edge and subtract half the width. You could be measuring from a surface deviation which will throw things off.

When I get this question in class, I bring up the mousehole and the electrical box in the wall. When checking the location of the electrical box, where do you start? Typically people will say "bottom of the wall, directly below the electrical box". If the electrical box is directly above the mouse hole in the base of the wall, would you not measure from the top of the mousehole then? Yes, if there is a +/- dimension used. If a datum reference frame has been called out, you measure from the datum simulator instead. In this silly example, the floor would be the datum simulator because it effectively mimics the net bottom of the wall which would be the datum feature.

Jim Sykes, P.Eng, GDTP-S
Profile Services

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com