Need for new alignments for results' section

[constadin]

Hello all,

First time posting here, although these forums have helped me a ton. I recently changed work and in here, my new colleagues do things differently with their CMMs PC-Dmis version 2011. After they have finished measuring a part and progress to creating the results section, before controlling a true position feature they always make a new alignment using as alignment elements the datums requested in the Feature Control Frame of the drawing. This is new to me, having worked with PC-dmis for 3 years (version 2010) I never realigned for creating a TP control frame. I mean if my colleagues method is correct and a new alignment is indeed needed then what is the point in choosing primary, secondary and Tertiary datums in the TP tab in the first place? I am mainly referring to US drawings with more than 5-6 Datums even for the most simple designs.

Thanks in advance for your kind feedback.
C.

[brian.freeman]

Realigning to the datums is not required. But they might be doing this to 'verify' the measured results. Ensuring correct dimensions and measurements because sometimes PC DMIS gives weird results.

[Sora5]

Congrats on your new job!

If you're using legacy, then yes, you have to create the alignment before you dimension. Legacy uses the current alignment for its calculations.
If you're using Xact measure, then no, you do not. Xact measure will perform its own internal alignment using the FCF when it calculates. However, in the advance window, if you change the alignment selection to Current alignment, it will show nominal values based on the current alignment but will not affect the actual calculations.

[constadin]

Thanks for your answers. Thing is that since the new alignment is based on the Datums used in the Datums Tab in True position, the "current alignment" feature and its alternatives have no apparent difference. The nominal values never change no matter the choice. Now... as far as weird results are concerned, in my previous job, if I tried to realign before making results, I was always getting different results. This IMO should not happen. Anyone to share his experience?

[Schlag]

You should be creating your alignment to the datum ref frame and then checking the features. I know this cant always happen but that is the ideal situation. If you are checking some kind of composite position, I believe they have to be checked in the same alignment other wise your data set will be skewed.

[constadin]

You should be creating your alignment to the datum ref frame and then checking the features. I know this cant always happen but that is the ideal situation. If you are checking some kind of composite position, I believe they have to be checked in the same alignment other wise your data set will be skewed.

But if this is the case, then why have the option to choose datums in the datums tab if we are to get skewed results? We should always align to said Datums and the just call a true position without defining A B C for example.

[Schlag]

Datums Tab ? What exactly are you talking about ? Are you talking about selecting the datums when you create the position ? I never do this. The ALIGNMENT itself is A-B-C ( or whatever Datum / order ). Create the position dimension and that's it.

[KIRBSTER269]

It's actually a good habit. For example taking in consideration of your basics from ABC. If you shift your basics in an alignment. and my example rotate 45 degrees. Then the machinist will know which way to shift his hole in X and Y. If need be. Remember it isn't about you, It's about the product and the machinist, you're there to help them, not judge them, and what I mean by that is giving them correct numbers so they can make adjustments. If you just pick the Datums then those X and Y numbers don't turn out completely right, because the Demon don't show you everything it does.

Untitled.jpg

[constadin]

God no, I don't judge them. The thing is that creating a complex report with more than 30 pages with most of them being true positions for a design that has more than 10 datums A,B,C,D,E,F,G,H,I etc realigning every time is not only inconvenient imo but also confusing and can cause chaos between multiple programmers.

[KIRBSTER269]

The way I showed you above, Doing it this way takes you to X=0 and Y=0, more work for you, less work for the guy or gal making the part. "Hey BOB, Your 3/8 hole is out of position" "Which Way do I need to shift it to make it right?" "I don't know because I didn't make it easy for you I made it easy for me, but good luck BOB"

POOR BOB

raw[1].gif

[Kp61dude!]

Poor Bob...

[NinjaBadger]

It shouldn't really.

If you create the alignment to measure the feature in (which as Kirbs has shown can be useful anyway) then you can simply recall it when it comes time to dimension things.

If you name your alignments ABC, ABD, EGH etc, then it's a piece of ****! Or in Kirbs example I'd might have an ABC (aligned strictly to ABC) and one called ABC_ROT45DEG.

Whilst you're right and it's not exactly necessary I like to see it. I can tell quite quickly from looking at program if the programmer knew his ****!


Where possible (and it depends on how it's been drawn) I like to rotate coord systems by basic angles but not offset then, so in the report the machinist can see the 'nominals' of the True position command where they tie up with the basics on the print.

[KIRBSTER269]

B i n g o

[dph51]

Absolutely the way to name alignments. Easy to understand.

[constadin]

Ok I will speak with an example. Imagine a plate with 5 holes. 4 Holes in each corner and 1 in the center. I make an alignment using holes on the top left and top right corner and the horizontal plane. I measure the rest 2 holes in the bottom left and right corners and the 5th hole in the center. I assign the 2 holes on the bottom datums B and C.

Let's say I create a true position for center hole using A | B | C and then I make a new alignment based on datums A B and C and repeat the process. Both processes examine center hole in respect to A B and C datums with the the only difference that in the first case I use alignment on the top corners of the plate while in the second I use the alignment that I am also examining (A B & C datums).

Now the questions are... will the results be the same or will they differ even if it is less than a micron? Why is there a difference if this is true?

[KIRBSTER269]

First of all. No matter how hard you try, no matter what you do. There is no such thing as the perfect part. Nothing will ever repeat exactly the same, it will be close. Drills walk, machinist load part, maybe there's a chip, maybe he overtightened the vise. Maybe there's a hair more material on one part, then the next. Datums are for critical starting points. What ever this part you are making MUST line up for it's Mating part. If you use the corners of the part for an alignment, you are mocking the machinist, but this isn't the correct way to check the part, if this isn't ABC. This would be an engineer trying to take a Shortcut (one less setup in the machine) Technically, they should establish A, B, and C. Then make a new fixture, locate off of those Datums and establish, everything else. A lot of times these guys, believe the accuracy of these machines are so precise, that they can skip a process and do everything all in one shot, depending on the situation, sometimes it can be done, or if you have a lot of tolerance, that's ok too, But you need to know if the part functions according to your Datums, not the corner of the part

[constadin]

First of all. No matter how hard you try, no matter what you do. There is no such thing as the perfect part. Nothing will ever repeat exactly the same, it will be close. Drills walk, machinist load part, maybe there's a chip, maybe he overtightened the vise. Maybe there's a hair more material on one part, then the next. Datums are for critical starting points. What ever this part you are making MUST line up for it's Mating part. If you use the corners of the part for an alignment, you are mocking the machinist, but this isn't the correct way to check the part, if this isn't ABC. This would be an engineer trying to take a Shortcut (one less setup in the machine) Technically, they should establish A, B, and C. Then make a new fixture, locate off of those Datums and establish, everything else. A lot of times these guys, believe the accuracy of these machines are so precise, that they can skip a process and do everything all in one shot, depending on the situation, sometimes it can be done, or if you have a lot of tolerance, that's ok too, But you need to know if the part functions according to your Datums, not the corner of the part

Seriously, I think you don't know what you are talking about no puns intended. There is no machinist and there is no directions I need to deliver for them or even an engineer taking a shortcut. I made a question that it might as well be hypothetical. In any case, me and a colleague of mine, both tested what I was questioning before. Our conclusion for whomever is interested is that no matter the alignment, if Datums are defined in the True position tab (A B C for example) the results are gonna be the same. This means if I align in datums D, E, F and I ask for a TP using A B C it will be exactly the same result if I had aligned to A B C prior to the TP call out. Bottom line, my question is answered, thanks everybody for your help.

[KIRBSTER269]

Boeing 737 MAX engineer? ^^^^^^^^^^^^^^^

[Quality ish]

Wow, did you shoot yourself in the foot with this statement. Kirbster has proven time and again that he's willing to be helpful in both insight into the relationship between the inspector (you) and the machinist (the people you have to deal with) as well as the relationship you have to have with engineers (who generally don't understand everything the Machinist and you have to go through to make sure the part they designed conforms to their print.)

My advice is to rephrase.

[constadin]

I didn't mean to insult, I just think he is out of subject (to the best of my understanding at least). Kirbster269, I didn't mean to insult or seem disrespectful in any way.

[aliptpatel]

Hello Constadin, I was just checking your question on one of my program to see if that makes difference if done in datum structure alignment and doing in different. I don't see any difference in the deviation of output. However I did try it on one of my true position controlled by two datums only. they are 3 holes and one hole is being checked from rest of the two holes. See in the image below.

Now if I try to measure them in two different datum structure, it shows different results. I do not know why and would like to know.

[KIRBSTER269]

Seriously, I think you don't know what you are talking about no puns intended. There is no machinist and there is no directions I need to deliver for them or even an engineer taking a shortcut. I made a question that it might as well be hypothetical. In any case, me and a colleague of mine, both tested what I was questioning before. Our conclusion for whomever is interested is that no matter the alignment, if Datums are defined in the True position tab (A B C for example) the results are gonna be the same. This means if I align in datums D, E, F and I ask for a TP using A B C it will be exactly the same result if I had aligned to A B C prior to the TP call out. Bottom line, my question is answered, thanks everybody for your help.

Have you ever been a machinist? Have you ever drawn out a part, by hand on an Easel, or even in Solidworks? Do you have a GD&T certificate? Can you Weld Mig, Tig, or even Arc? When you make a program, do you lay the part out and check every dimension to proof out your program? If there is a Part to be made, There is a machinist somewhere who made it. There is no little magical fairies that make your part. Besides I'm not the one asking questions because I'm confused how to report out true position. I was trying to help, so here's the best I can do for you. This will get you started

https://store.hexagonmetrology.us/course-catalog