Center offsetS?

[cumminsdesiel]

I have an engine block in which crank#5 and crank#1 should be align on its centers of the cranks (cranks#1 is 23"inches away from cranks#5). I'm getting .3365 offset and cannot understand what I'm missing. The block is a master block so it should be right and on centers. Any ideas would be helpful.
Thanks




A0 =ALIGNMENT/START,RECALL:STARTUP, LIST= YES
ALIGNMENT/LEVEL,XPLUS,PLN1
ALIGNMENT/TRANS,XAXIS,PNT1
ALIGNMENT/TRANS,YAXIS,PNT1
ALIGNMENT/TRANS,ZAXIS,PNT1
ALIGNMENT/ROTATE_CIRCLE,YPLUS,TO,PNT1,AND,PNT3,ABOUT,XPLUS
ALIGNMENT/ROTATE_CIRCLE,ZPLUS,TO,PNT1,AND,PNT2,ABOUT,XPLUS
ALIGNMENT/END

[Paul Sarrach]

you can not rotate twice. I don't think. Once you level, then rotate. and rotate agian, somethings gotta give, right Matt?

[Underspec]

Follow the 3-2-1 alignment. Looks like you have points for the X & Y. Use lines if possible or where it's called out on the print / Cad. And rotate once only.

[cumminsdesiel]

I remove the 2nd rotation and still get the offset..

[Paul Sarrach]

Take out the first rotation and add in the second, see what happens?

[Don Ruggieri]

rotate away

You can actually rotate as many times as you want. It gets confusing to use, but the software doesn't care. (Level is a whole 'nother story!)

I suspect a DISTANCE issue here. Check your workplanes, your distance qualifiers (PERPTO, PARLTO, etc.) a 2D distance is based on certain other factors, how it is vewed, etc.

The problem may or ay not be in the alignment, but we need to see more of the program.

[Don Ruggieri]

then again

What I said before was true, but re-visiting the alignment code I see that the two alignments in question are in the same plane, and the 2nd effectively would void the 1st. Its like saying rotate TO 45, then rotate TO 30. When you're done, the only rotattion is the 30.

If they were incremental rotations, then they would stack, and both would be significant, but that's not what I see here.

Again, the alignment may or may not be the problem, but we don't have enough to go on.

[Paul Sarrach]

You can actually rotate as many times as you want. It gets confusing to use, but the software doesn't care. (Level is a whole 'nother story!)

I suspect a DISTANCE issue here. Check your workplanes, your distance qualifiers (PERPTO, PARLTO, etc.) a 2D distance is based on certain other factors, how it is vewed, etc.

The problem may or ay not be in the alignment, but we need to see more of the program.

I did not know that. I always make a new alignment with every rotation. And probly still will, because you are right, it is less confusing.

[cumminsdesiel]

program here

Here is the program attached. Did a 321 alignment..

[craiger_ny]

Could be a problem with the leveling feature. You can rotate as many times as you want and it will do just that, rotate in the order you specify. I'd want to know more about your leveling feature.and what is the point you are translating to and how are you dimensioning (location, concentricity TP?). If you are leveling to a plane and your hits are not spaced very far apart and then you go measuring at long distances from that plane I'd be skeptical, especially if the face of the "master" you have leveled to is not certified perpendicular to the shaft diameters you are measuring and dimensioning later. Post some code and a pic if possible. I am new to this forum but I can tell you that all of the folks here are great about wanting to help out. Posting some more code would be a big help.

[djayne]

Do your rotation last. Or at least after you establish the X-axis. The error could come from the squareness of your part to the CMM axis so, you'll need to teach the part axis before you can rotate.

[cumminsdesiel]

pictures

back side contain crank#5 and cam#5 and the plane. The front side contains the crank#1

[craiger_ny]

Meauring one circle in one leveling scenario (Startup) and then measuring another circle in a different scenario (A0) is not helping you, it gives your circles vectors based on the level of your current alignment scenario, so try getting an alignment with at least a level going on before you measure the two circles that you wish to dimension relative to each other. I also find it curious that all of your deviation is in the Y axis only. You are sitting on a machined surface so this squares up your part to the machine Z. Your error may be comming from the fact that you are measuring this 4 hits manually on circle crank1 while in A0 alignment when crank5 was measured in startup, although I have had better luck with manual measurements than this. If your part is square to Z but not to Y then this will show up the way it has when measuring in two different alignment scenarios. I am leaning this way as all of your deviation is in the Y and the way the part is set up it looks like there is the most chance of being out of square to the machine in that axis. Could you try manual features strictly for alignment then go DCC on the rest? If you must go manual then perhaps square the part up to the machine as best as you can and take more hits on crank1 or at least vector your hits as best as you can relative to the part, and get a leveling alignment before measuring the circles to be dimensioned.

Craig

[Matthew D. Hoedeman]

From looking at your program, the second circle has a wrong vector. If these should be in line, then they bnoth should have the same vector. Now I realize, there isn't much difference, but it does not take much.

[cumminsdesiel]

I align the block agin. level the plane before I took the circles and got .18 off center instead of .33. How do I make sure that the vector is going in the same direction. Novice User Thanks
Kenn