asking for help on how to define G,D &T for this part

[Spurs]

I have a part that I am making which is the cover of a housing.

The cover looks like a box with one side open.

The inner walls of the cover must all fit over its mate which at its maximum size is .2 mm smaller in both the length and width directions than the basic size of the Length and Width of the cover.

The cover also has 4 simple screw holes in it.

The gage I would like to build would be built to the maximum material condition of the cover - i.e. a rectangular shape gage built at the basic dimension of the cover less .2 mm in both length and width.

If the cover fits over this gage, then I know it will fit over the housing it is intended to mate with.

At the same time, the gage will have 4 pins coming out of the surface of the gage hopefully going through all 4 screw holes. The pins will be .1 mm smaller in diameter than the basic diameter of the screw holes.

How should the GD&T control frames look to allow for the use of such a gage? Are the 4 surfaces each datums A to D with the edge of the cover as datum E?
I assume that one of the edges interior sides of the cover is Datum A, then to qualify datum B needs some sort of relationship to A. Then C would need some sort of perpendicular relationship (or maybe profile) to A-B.

Any help would be appreciated.

 

[drawoh]

Spurs,

How are you going to fabricate this thing? Machining? Sheet metal? Moulding? It matters.

Have you a sketch?

--
JHG

 

[KENAT]

You seem to be a bit backwards, normally you look at the design functional requirements and create the drawing (including dimensions & tolerances) based primarily on that (while keeping manufacturability, inspectability etc. in mind) and then look at how you verify parts meet the drawing i.e. gauging if appropriate.

So, from a functional point of view what is the more logical datum structure. Is it really the edges, or maybe a centerplane derived from width/length of part? Should it really be the mounting hole pattern - or an approximation there of?

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[Spurs]

The part is injection molded plastic

@Kenat
The design function is exactly how hi described. The cover must fit over the housing on the surfaces I indicated. I have no idea why you think this is backwards thinking

 

[Spurs]

@Kenat

Its really not much different than a bushing over a pin - except in this case the bushing has 4 sides instead of one cylindrical curface

 

[Tenkan]

If the cover is supposed to sit over the housing, then it needs to be centered on it... the centerplane would be a better way to locate from. I agree you have it backwards because you should want to build your gauge to the MMB of the housing, not the cover itself.

lightweight, cheap, strong... pick 2

 

[Spurs]

Koda94
But a center plane cannot be a datum. The center plane is a theoretical plane, not a real one.
I need a practical datum. The practical datum to measure the holes is the perimeter of the cover

The photo shown here is my drawing as it is now

 

[KENAT]

A center plane derived from feature is entirely legitimate per ASME Y14.5 if that's the standard you're working to. Take a look at section 4.5 in 95 version & figures 4-11 thru 4-17.

If you're actually using the mounting holes to locate the part you could consider section 4.5.8 'pattern of features' as datum. However, this is not always well understood.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[Tenkan]

But a center plane cannot be a datum.

A center plane can most certainly be a datumn and in many cases is certainly practical. (I cant see enough detail in your picture attachment to really comment but it almost looks like what appears to be Datum D is placed like its a centerplane datum... be careful where you place your datum symbols. )

I don’t know enough about your project to really comment but the point of the drawing for your cover (for any part drawing actually) is to fully define the part by capturing the design intent... Another thing to consider is how the part will be inspected, what appears to be a draft on your datum surfaces will make it difficult to origin from.

Looking at what little I can read in your attachment my take is it appears you want the cover to center itself over the housing and align itself with the 4 screw mount features. The surfaces your using as datums right now need to clear the housing, not gauge from it. The mounting holes vertically (as shown) need to be positioned about the center of the horizontal centerplane (do not use symmetry). The centerplane could possibly be defined by the distance between the two surfaces that clear the housing, (what appears to be Datum A and B on your attachment). Horizontally, you can locate the features from the mounting holes on the left end with linear/ordinate dimensions to assure the features do not interfere with the housing when installed.


lightweight, cheap, strong... pick 2

 

[pylfrm]

Spurs,

In the image you posted, the top right datum feature symbol looks like it has an upside down L in it. Is this actually suppsed to be the datum feature E that you mentioned in your first post? I will assume so.

Further, I will assume that datum feature E is a single flat surface that extends all around the perimeter of the part, and that this surface is clamped against a corresponding flat mating surface at assembly. I'll also assume that you are working to ASME Y14.5-2009.

If the above is true, then E should probably be used as the primary datum feature reference. You can use a profile of a surface tolerance for the inner walls and a positional tolerance at MMC for the screw holes. If both tolerances reference datum feature E (and nothing else) then a simultaneous requirement will apply, and no other datum features are needed.

Consider allowing or requiring restraint for inspection if the part is flexible enough for it to matter.


- pylfrm