2 separate questions - establishing hole datum axis and datum pattern

[Daewoo200]

I have two separate questions. They are the following:

1. I am confused about how a datum axis is established on a CMM. Is the axis established wherever the hole ends up being -whether the hole is in position tolerance or out of position tolerance? Or is the axis established based off of the basic dimensions and the axis is perfectly positioned? If the axis is always perfectly positioned based off of the basic dimensions, this means that the hole will never be perfectly coaxial/concentric with the axis since there will always be some inherent error in the position of the hole. Can you help clarify this for me?

2. Myself and a coworker are curious about using a hole pattern as a datum. I say that it can be done (I showed him my textbook and that it does satisfy our design intent), but he says that he doesn't understand what a hole pattern datum actually is so he doesn't want to use it. Now I am curious - how should an inspector interpret a hole pattern that is called out as a datum?

Say for example that the hole pattern is called out as datum E. Is there a datum axis E that goes through each hole in the pattern? Or does a single axis go through the middle of the entire pattern? What if some holes are not straight/true to one another within the pattern?

I also want to treat the holes in the pattern as RFS. Will a hole pattern datum still work at RFS?

For other features that are referencing a hole pattern datum in a FCF, I feel like it would be hard for them to pass inspection since that feature would have to be gaged to that entire pattern. Is this correct?

Help clearing up my confusion for my two questions is greatly appreciated.

Thanks,

Chris

 

[Burunduk]

1. Datums are derived from datum feature simulators. Datum feature simulators are practically the fixture elements, ideally also mimicking the mating part in assembly. The default is that the datum feature simulators are at fixed locations relative to each other according to the basic dimensions that tie them. So yes, a datum axis associated with a secondary or tertiary datum reference in a feature control frame will have to be oriented/located/both basically from any preceding datums. Yes, the actual hole will not be perfectly coaxial with the datum but it's principally not different from a secondary datum surface in the classic orthogonal datums scenario where the secondary datum plane is perpendicular to the primary but the actual secondary datum feature is not

 

[Daewoo200]

Hi Burunduk,

Say I have a block. Three sides of the block are defined as datum A, B, and C. A hole in the block is datum E and datum E is positioned off of A, B, and C.

Now, I have another random feature (boss, hole, slot, whatever) on my block that I want relative to datum E. I would relate it to datum E in the feature's fcf. I assume there would be no additional tolerance stackup for the feature (since datum feature E also has a tolerance relative to A, B, C) because datum E is considered to be a perfect entity, correct?

In other words, there is no additional tolerance stack up from referencing off of datum E rather than referencing the new feature off of A, B, and C, right?

Also, if anybody has an answer to my 2nd question:

2. Myself and a coworker are curious about using a hole pattern as a datum. I say that it can be done (I showed him my textbook and that it does satisfy our design intent), but he says that he doesn't understand what a hole pattern datum actually is so he doesn't want to use it. Now I am curious - how should an inspector interpret a hole pattern that is called out as a datum?

Say for example that the hole pattern is called out as datum E. Is there a datum axis E that goes through each hole in the pattern? Or does a single axis go through the middle of the entire pattern? What if some holes are not straight/true to one another within the pattern?

I also want to treat the holes in the pattern as RFS. Will a hole pattern datum still work at RFS?

For other features that are referencing a hole pattern datum in a FCF, I feel like it would be hard for them to pass inspection since that feature would have to be gaged to that entire pattern. Is this correct?

 

[Burunduk]

Daewoo200,
If you reference to datum E only in your newly described scenario, then datum axis E is the axis of the "unrelated actual mating envelope" of the datum feature E hole, which has no relationship with datums A and B. This is regardless of any basic dimensions from A and B locating hole E. These basic dimensions will not be considered because the position that you apply to the boss with reference to E as the primary and sole reference, only recognizes E in isolation from the rest of the world. So to answer your question on the tolerance stack - if your origin for the stack is E - no stack up. You WILL get stack up relative to A and B per the location tolerance of E relative to A and B.

I will address your second question later unless someone else covers it before me.

 

[Burunduk]

Regarding your question no. 2,
Generally, datums are used in order to establish a datum reference frame - 3 mutually perpendicular planes defining an origin to relate the geometry to. The measurement coordinate system is either the simulated datum reference frame or it is based on it (with some fixed translation and rotation). With a pattern of holes, you want to achieve the same outcome, while the datum itself is just to give you something to base your datum reference frame on. Typically a pattern of parallel offset holes will provide a datum axis at the center of symmetry of the true positions of the holes in the pattern (if such exists). The true positions will also be used to orient the datum reference frame planes. The image below is an example from the ASME Y14.5 standard of a datum axis derived from the datum feature simulators of a pattern of holes referenced at MMB. The difference for an RMB datum reference would be that instead of being able to use a gage with fixed size pins as datum feature simulators, you would need a group of simultaneously expanding pins.