CR4® - The Engineer's Place for News and Discussion®

 Previous in Forum: Terry-Quality Engineer Next in Forum: E 8010-P1 & E 7018-1
Participant

Join Date: Dec 2008
Posts: 4

# I,J,K vs X,Y,Z measurements

12/04/2008 7:35 AM

What is the benifit of using I,J,K vs the standard X,Y,Z measurements on the CMM machines?

Pathfinder Tags: measurement quality
Interested in this topic? By joining CR4 you can "subscribe" to

### Comments rated to be Good Answers:

These comments received enough positive ratings to make them "good answers".

### Comments rated to be "almost" Good Answers:

Check out these comments that don't yet have enough votes to be "official" good answers and, if you agree with them, rate them!
Anonymous Poster
#1

### Re: I,J,K vs X,Y,Z measurements

12/05/2008 7:22 AM

its depend on the requirement on the part drawing

especially the point taken is a profile

the i, j, & k are use to calculate the direction of measurement for the stylus to touch a point which will affect the reading of x, y, & z

Score 1 for Good Answer
2
Guru

Join Date: Mar 2008
Location: Bridgewater, Va.
Posts: 1756
#2

### Re: I,J,K vs X,Y,Z measurements

12/05/2008 9:26 AM

X, Y, Z is a location point in cartesian space.

I, J, K defines a vector to a point or a plane.

If I recall correctly (I'm a long time ago CNC, robot and CMM programmer) IJK are developed from the sine of the angle relative to the plane being used by the XYZ point.

ie: X0 Y0 Z0 I0 J0 K1 would be looking straight down on the plane along the Z axis at the point at 0,0,0.

Note: it used to be that CNC vectors and robot/CMM vectors were defined with opposite signs. In other words, K1 looks down the vector TO the point for CNC, but looks away from the point in the robot/CMM world. I don't know if that difference has ever been reconciled.

FWIW

Hooker

Good Answer (Score 2)
Guru

Join Date: Mar 2008
Location: Bridgewater, Va.
Posts: 1756
#3
In reply to #2

### Re: I,J,K vs X,Y,Z measurements

12/05/2008 9:37 AM

Something else just tickled the back of my brain...

I think some systems use IJK as a supplementary (sometimes called local) cartesian coordinate system that is defined relative to the primary coordinate system.

In this case the supplementary system is most efficiently used as a container for macros that repeat a process multiple times in various locations relative to the primary coordinate system.

My apologies for my long winded initial response if this is what your question actually referred to.

Hooker

Score 1 for Good Answer
Participant

Join Date: Dec 2008
Posts: 4
#4
In reply to #3

### Re: I,J,K vs X,Y,Z measurements

12/05/2008 10:58 AM

Thanks Hooker for the reply, it was helpful

Participant

Join Date: Dec 2008
Posts: 4
#5
In reply to #2

### Re: I,J,K vs X,Y,Z measurements

12/09/2008 2:14 PM

I have another question for the IJK vs XYZ. I understand if doing a profile of a surface but what about just a point on that surface and the print calls out X,Y,Z. How does the IJK work here? Also how would you tolerance this IJK?

Guru

Join Date: Mar 2008
Location: Bridgewater, Va.
Posts: 1756
#6
In reply to #5

### Re: I,J,K vs X,Y,Z measurements

12/10/2008 11:51 AM

Hi QE,

I'm not sure if I understand the question. What is the context of the IJK? Is it being used as a machining (CNC) vector or a CMM vector? Or a localized position point?

If it's for multi-axis CNC vectors there is no IJK "tolerance" applied beyond the machine's limitations. The design engineer may use degrees and fractions of a degree to specify the vector of a hole, for example, but it's up to the CNC programmer (and his software) to translate that to the machine's code of IJK vectors, with foreknowledge of the machine's limitations.

If your question is relative to CMM, I'm afraid my knowledge of that stuff is 20 years out of date. In my time, CMM IJK's were just a convenience to get somewhere on a part where a vertical probe couldn't go.

Here's an interesting definition I ran across though, from Geomet:

Vector Points - Surface 3D Points. Compensated by one probe radius along the defined IJK surface vector.

Hope this helps. I didn't know you can comp the radius by using the vector. I'm kinda presuming this is independent of the probe centerline.

Maybe someone else more up to date can chime in.

Hooker

Power-User

Join Date: Mar 2007
Location: Nothing sucks more than that moment during an argument when you realize you're wrong.
Posts: 304
#7

### Re: I,J,K vs X,Y,Z measurements

12/15/2008 9:59 AM

What is the benefit of using I,J,K vs the standard X,Y,Z measurements on the CMM machines?

X,Y,Z is the nominal "perfect" point in space. The I,J,K vector further defines how to get there, like flying a plane. Go from here...to here....about this path. Without it on a CMM you set 2 axis and measure "or fly" in 1 axis. The end result of vectoring vs straight line is different on a contoured "non-flat" surface no matter if your talking about a cutting tool or an inspection probe because you not hitting square to the tool/probe so you need to calculate and adjust cutter compensation/ball error.

I have another question for the IJK vs XYZ. I understand if doing a profile of a surface but what about just a point on that surface and the print calls out X,Y,Z. How does the IJK work here? Also how would you tolerance this IJK?

Your typically not using vectors when you inspect an item on a CMM from a blue print unless you have a "reliable" model to use as well. If you do, you can extract the vectors from the model or use model based inspection if your software has that function. Vectors are not toleranced as stated earlier. Form, Fit and Function dictate tolerance which should all be predefined by the inspection/manufacturing stage.

__________________
You never know when it will strike, but there comes a moment when you know that you just aren't going to do anything productive for the rest of the day.
Score 1 for Good Answer
Participant

Join Date: Dec 2008
Posts: 4
#8
In reply to #7

### Re: I,J,K vs X,Y,Z measurements

12/15/2008 10:33 AM

Thank You for the information. It was very helpful.

Interested in this topic? By joining CR4 you can "subscribe" to

### Comments rated to be Good Answers:

These comments received enough positive ratings to make them "good answers".

### Comments rated to be "almost" Good Answers:

Check out these comments that don't yet have enough votes to be "official" good answers and, if you agree with them, rate them!

### Users who posted comments:

Anonymous Poster (1); Hooker (3); markar (1); QEforever (3)

 Previous in Forum: Terry-Quality Engineer Next in Forum: E 8010-P1 & E 7018-1