Casting Datums

[Keveo]

I am currently preparing to do a machining drawing from a cast housing and trying to establish the datums. I have taken a class on GD&T some time ago and know enough to be dangerous, but do not use is that often, unfortunately so does my boss. I would like to set up the primary, secondary, and tertiary datums as it would mate to other parts, via parrallel jaws, etc. Im being told that it is better to cast datum features and basic dimension to datum targets/points because of the draft and that the machinist can fixture it better and that you have to start somewhere so the machinist can machine an adequate surface for the datums. Can you share some points on how to establish datums on a casting when there is only one flat surface (housing bottom) on the part. Any input will be appreciated.
Thanks

 

[Heckler]

First part of establishing cast datums is to think about how this is going to be fixtured to establish the machine datums. According to ASME Y14.8M - Castings & Forgings section 4.4 A machining atum reference frame should be established from a cast/forged datum reference frame for subsequent machining.

Heckler
Sr. Mechanical Engineer
o
_`\(,_
(_)/ (_)

This post contains no political overtones or undertones for that matter and in no way represents the poster's political agenda.

 

[Heckler]

Then to take it one more step is to see how the manufacturing datum structure relates to the parts functional datum structure. So the relationships between the cast datums, manufacturing datums and functional datums needs to be analyzed....all part of the design process

Heckler
Sr. Mechanical Engineer
o
_`\(,_
(_)/ (_)

This post contains no political overtones or undertones for that matter and in no way represents the poster's political agenda.

 

[Keveo]

Thanks for your reply...you cleared the cast, manufacturing, and funtional datums up for me. Is it common practice to put a manufacturing datum in the center of the cast part, which you would have to use a parrelel jig to find center or is it better to butt up against the edge of the casting using datum targets (cast datum?)to indentify the datum. Im a little hesitant to use a cast, as well as drafted surface for this? Whats your experience?

 

[Heckler]

If you can post a picture that would help. It really depends on the complexity of the cast part and how it will be machined. The cast part has to be completely constrained using the cast datums before you can establish a manufacturing datum (in a perfect world). Here is a picture of a cast part.

Heckler
Sr. Mechanical Engineer
o
_`\(,_
(_)/ (_)

This post contains no political overtones or undertones for that matter and in no way represents the poster's political agenda.

 

[Keveo]

that is very complex....it appears there are nice flat datum target points casted in via slides. This particular casting doesn't have slides. Im going to try to post a picture.

 

[Keveo]

Heres the housing, not very complex, but im looking for reasonable places to put cast datums/manufacturing datums. Are cast datums just for casting purposes or are they there for manufacturing as well?

 

[Heckler]

Yes, it's a very complex casting. It is manufactured in Canada by Howmet

http://www.alcoa.com/howmet/en/info_page/inv_cast.asp

using their Investment Casting Process. The part is fully supported on those cast datums then that large flat surface is Primary Datum. This part is $$$$$$$$$

Heckler
Sr. Mechanical Engineer
o
_`\(,_
(_)/ (_)

This post contains no political overtones or undertones for that matter and in no way represents the poster's political agenda.

 

[Keveo]

I should state that right now this is a sand cast and will eventually be converted into a die casting. I apologize that im such a newb but my company doesn't usually produce high volume production products and in the past no one has really implemented GD&T so its hard to reference other parts. Im am confused on the difference between cast datums and manufacturing datums. Thanks for all your help.

 

[PaulJackson]

Keveo,

Sand or other prototype castings are usually balanced manually to prevent thin/thick walls from inhibiting function. That does not mean that the eventual process "say die casting for instance" should not be supported by the product design specifications that you are specifying...it just means that alternate or special exceptions will be requirements will govern the alternate prototype casting processes.

If the "oil pan, cover, or whatever" is sand casted it will naturally exhibit blow, shift, and contour variations that are less probable than those in production die casting processes. The die-cast targets are typically ignored so that critical clearances and material thicknesses can be maintained in the prototype process .

Considering the eventual production process the casting datum targets or "3-2-1" registry “if I may speculate about the function” in this case should primarily reflect the functional considerations of the assembly. If the sealing surface could be cast flat enough or the clamp load could be such that the gasket material would maintain compression I would not even machine the gasket surface and risk exposing porosity… furthermore I would include a channel or inner chamfer could to retain excess RTV. I would establish three or possibly “4 clamped" datum targets "as primary" on the gasket surface…acknowledging that this surface would bend under clamp force …naturally it would be machined for early sand-cast prototypes. Next I would establish three "two on one side...fore and aft...and one on the opposite side middle… movable equalizing targets as the secondary datum targets and finally one or “two equalizing” targets longitudinally as the tertiary.

The clamp fixture would cycle rapidly with minimal force -to- maximum force to precisely constrain the “die casting” for the installation of all machined features including dowel holes that mated with expandable split roll pins, all fastener clearance holes and all other solenoid or electrical lead connection holes.

The point is that the casting target designations should reflect/insure both functional and process registry design considerations to insure minimum variation in functional clearance and min/max material outcomes.

Paul