Basic Alignment Query.

[EdenG]

Hi guys,
My names Glen, I'm operating a DEA with DMIS 2015, I'm UK based and very fresh to the quality department (Apprentice). Apologies for this not being DMIS specific, should be an easy one for someone no doubt.

The 1st image shows the drawing plan view of the part I'm working on, 2nd cross-section A-A (Plan view from the machines perspective when the part is on bed). I'm going to be working on the part with it laying along the X axis (On one of the faces which calls for a flatness tolerance), which means I intend to be working in the Y minus workplane during alignment; I'm doing this so I can hopefully achieve all measurements in one program. So, my first thoughts for alignment were the following:

Take hits on the front face, level Y minus, translate Y plus.
Take hits around Datum A, translate X, Z.
Measure and construct a line between top holes, X plus direction, so rotate X plus to line about Y minus.

The problem and question for me is then in doing so I'm not including Datum B. It's necessary to include it, but in doing so I would have to include probe rotations...? Would I be better to use the above alignment in the manual alignment to establish the part, and go ahead and use Datum B and rotations in DCC. I feel like I don't fully understand, perhaps being able to visualise, what I'm trying to achieve with alignment, the resulting trihedron etc. YET (Despite reading and seeing many useful posts, articles etc).

Thanks a bunch,
Glen.







Untitled 3.pngUntitled.png

[Schrocknroll]

To answer at least part of your post, you don't need to use the primary datum features to do the manual alignment. Some manual alignments are a single point or a read point to establish X,Y,Z zero. Then in DCC mode you measure the features needed to constrain all 6 degrees of freedom. Whether the datums are required in the alignment will depend on if you are using legacy position or XactMeasure position. With legacy the position is related to the alignment so you will need to align to the datums. With XactMeasure, you define the datums and create the FCF in the alignment window so it's not as important to align to the datums, but it's not a bad idea.

[EdenG]

Hi Schrock,
How do I switch between the two positioning modes? I will take a look into which I'm using and experiment at some-point- Thanks for your help

[Schrocknroll]

EdenG In the menu select Insert | Dimension, then at the bottom of that menu you can click "Use legacy dimensions". When it's not selected you will be using XactMeasure dimensioning for GD&T dimensions.

[BKulpa]

With Datum A being a 1.5mm cylinder your not going to get repeatable results using that as a primary datum feature.

[EdenG]

Hi Kulpa,
I think Datum A is Ø131mm roughly, unless you were trying to indicate the image I have provided is not large enough- which it isn't! I was hoping to use a constructed line between the small holes as a tertiary/rotary datum in my FCF.

[louisd]

Welcome, Glen.
the posts above are dead-on.

Prior to ultimately measuring anything with a feature control frame (FCF) relative to B; you will eventually have to get down there and take hits to establish your actual measured B plane.

I don't see a tertiary/rotation datum defined. If a rotation datum is not defined, technically it gives you the freedom to rotate to whatever feature you want, about A datum. When I have this opportunity, I typically select the features which will produce the largest possible length/size value, like a diagonal line between top left and bottom right holes.

[EdenG]

Hi Louis,
You're correct, many of the parts I measure are like this. That's a great tip. Silly question- the diagonal line once created will not have a clear direction?

Thanks, Glen.

[louisd]

The diagonal line will have most certainly have a clear direction. A line is literally one big vector arrow. The benefit to using a larger set of rotation features is it should -in theory- minimize the amount of error your rotation will produce, if one hole of the ~230mm diagonal distance is off in location by 0.05mm for the longest 2 holes, your rotation datum will for the whole part only have uncertainty/error of 0.05mm.

If in contrast if you produce a rotation line that's those two bottom 13mm holes, and one hole is off let's say 0.05mm, it will project that error in rotation by a factor of 10x, making the rotation uncertainty/error 0.5mm!

If you rotate to a line that's not normal to your part, it's not hard to calculate the theoretical angle, which you can then align/rotate by that basic theo angle, to keep your tetrahedron normal to the part coordinates.