Establishing datums for machining a die cast box

[Tunalover]

Folks:
I have a very simple die cast box with a 3.5deg drafts on four sides. The opening flange and the bottom are parallel. The problem is that, to machine any features into any side of this box requires three mutually perpendicular datum planes. The primary datum plane is obviously established

Tunalover

 

[MikeHalloran]

Think of the datum planes as features of the FIXTURE, not features of the PART.
Note also that planes are really not features anyway; they are theoretical constructs.

Or, think of it this way:
In order to uniquely locate a part, your fixture needs only:
1. A planar surface.
and
2. A linear feature, or two points, not coincident with the planar feature.
and
3. A point feature, not coincident with (1) or (2).

;--- other thoughts:
A 3.5 degree draft is significant; the tangent is around .049" per inch of depth.
If you want to machine, e.g. bosses perpendicular to the mounting flange,
you will need extra thickness to provide machining stock,
or you will have to tilt the bosses and the machining spindle by the draft angle.

If you are really desirous of having some mutually perpendicular features on the part,
you can probably pull a couple of blunt fins that have zero draft in one direction.



Mike Halloran
Pembroke Pines, FL, USA

 

[3DDave]

I've tried to do this before, but the 'primary datum must be the penetrated surface' guys spoiled it.

I wanted to use the bottom surface as the primary datum and the penetrated face as secondary with one side or the width as a tertiary. This would make it easy to machine and inspect the part and result in a predictable outcome, given that the bottom surface located and oriented the box in the next assembly and the connector body needed to be oriented and located from that same installed location.

The difference for relatively thin walls between being not perpendicular is small, like .008 inches over a .125 thick wall, which is easily made up for by making the opening a bit larger. Lack of perpendicularity under a reasonable sized screw head is about the same and is well within the flexibility of screw heads. The only concession to the draft angle was setting up the back-spotface which needed to match the draft angle to get the right thickness. This could be indicated off the draft surface and centered on the previously made hole.

Targets are also helpful to make discussions go more easily. That way there's no need for tapered jigs that might not match the actual draft angle.

In my case, the amount of non-committal naysayers on every proposed dimensioning scheme, plus significant feature creep for which the box was unsuitable, was more than the cost of milling the box from raw stock, but that wasn't "COTS".

 

[Tunalover]

Guys-
Due to a poorly-timed internet outage I was unable to get in my whole post. Tommorrow at work I will repost. Sorry for the hassle. I wrote this with my smart phone.
Bruce Jackson aka Tunalover


Tunalover

 

[KENAT]

Take a look at ASME Y14.8.

Consider datum targets.

Posting guidelines faq731-376

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

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

 

[Tunalover]

OK, here's what I wanted to post when I ran into the internet problem:

Folks:
I have a very simple die cast box with a 3.5deg draft on all four sides that I want to machine features into the sides. The flanged opening at the top and the flat bottom are parallel. The problem is, to machine any features into any side of this box requires three mutually perpendicular datum planes available to each side. Only the primary datum can be established without using datum targets. If features are machined into all four sides of the box then obviously the primary datum planes are established on each of the four sides of the box. That's four datums. There could be two tertiary datum planes passing through the center, one bisecting the short sides and one bisecting the long sides. That's two planar datums each requiring one datum target, probably on the outside surfaces. Or should I use equalizing datums per ASME Y14.8-2009 which calls for FOUR datum points for each bisecting datum? Won't the four necessary secondary datums then require 8 more datum targets? This seems so complicated for such a simple part! Any tips on simplifying the D&T scheme? This is not a high volume product!



Tunalover

 

[KENAT]

Why not just simple datum target structure and then work around features which are perpendicular to the faces not being perpendicular to the datum structure?

Same as when you have features on a sloped surface on a regular machined part or similar.

Posting guidelines faq731-376

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

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

 

[JNieman]

Do your parallel flanged openings have any fastening features such as bolt holes or a shoulder/lip for seating/locating a cover?