5 Axis Alignments

[cincy k]

Hi,

Does anyone have a process they follow for checking axis alignments on a 5 axis? How do you manually find center of rotation of the trunnion? Table, table. Any documents with steps out there floating around? 

Is a granite square necessary? I would certainly think it would be. 

I'm trying to educate myself on how this is done at the factory and how I should be doing this on an ongoing basis as part of our preventative maintenance/ machine health check. 

I have ez-5 on our Matsuura MX-520 but am finding it may not be the end all be all for finding center of rotation, or maybe it wasn't set up properly. 

Thanks,

TK

[Leon82]

From what I have seen you Mount a tooling ball on the outer most edge of the rotary head and indicate in varius positions, I've never done it though.

 

 

[RaiderX]

there is some really cool laser tracker options out, usually done by an outside service if you want to really get serious. otherwise its a tooling ball and indicator as said.

[gcode]

We bought a 5X alignment package for our Okuma,

you mount a tooling ball in one corner on the table, tag it with the probe a couple of times

launch the app and go get a cup of coffee.

[cincy k]

With a tooling ball and indicator- What steps do take to check the machine? Any documentation or steps to follow? This is what I'm looking for. 

[cincy k]

8 minutes ago, gcode said:

We bought a 5X alignment package for our Okuma,

you mount a tooling ball in one corner on the table, tag it with the probe a couple of times

launch the app and go get a cup of coffee.

Was this a Renishaw add on or Okuma's?

I have something similar on our Matsuura. I just don;t believe it or the probe is potentially setup properly. 

[Colin Gilchrist]

I'm pretty sure Karlo Apro's book "The Secrets of 5 Axis Machining" cover this procedure in detail...

[Colin Gilchrist]

From what I remember, you follow a procedure like this:

• Tilt A-Axis to +90 degrees. Use indicator to indicate the platter "vertical" record the "A" position.
• Tilt A-Axis to -90 degrees. Use indicator to find vertical.
• Use Edge Finder to find edge of platter at A -90. After finding "kick off point", move Z Axis up and tram spindle to be "Centered" over the edge of the platter. Record this position. (Usually Machine Y Axis, but could be X, depending on your configuration.)
• Now, rotate A-Axis back to +90, and perform Edge Find procedure again. Record the "Center" position when the centerline of the spindle is now exactly aligned with the edge of the platter.
• Take the Absolute value of both numbers, and subtract the smaller number from the larger value, and divide by 2. (Example: Y-11.8211 Y-12.2214 = .4003 "total distance" between platter faces. So the C Axis Platter should be .20015, below the Center of the A-Axis rotation. (Also, record this "Y" position as the Centerline of the A-Axis. Ex: -11.8211 - .20015 = Y-12.20125)

So that gives you the distance from A-Axis Centerline, to the face of the C Axis platter.

Next, rotate the A-Axis until the C-Axis Platter is "Flat", then use an indicator to find "true flatness". This position (Example: A 0.029), should be set for your "Work Offset" to bring the platter "truly flat" to the Spindle Axis.

Once you've got the A-Axis rotated so that the C-Axis is "flat", change the orientation of your Indicator, and now "sweep" the O.D. of the platter, until you've got the Spindle exactly centered on your platter. If you've got a "precision bore" in the center of the platter, you can use that for indicating, instead of the outside of the platter.

Record the XY position, this should put you exactly on "center" for both the C-Axis and the A-Axis. NOTE: your Y-Axis position may not exactly match the Y-Axis position you recorded above. This is important! You may have an "offset" in Y between the Center of the C-Axis platter rotation, and the true "Y-Axis Center of rotation for the A-Axis". If that is the case, you can make adjustments inside your Post Processor to "compensate" for this slight misalignment... (And you probably know a Post Guy that could do that for you at no cost!...)

 

[huskermcdoogle]

5 minutes ago, Colin Gilchrist said:

If that is the case, you can make adjustments inside your Post Processor to "compensate" for this slight misalignment..

 

Also, don't forget that these kinematic offsets will need to be adjusted in the machine parameters for those that use TCP for everything.

[Colin Gilchrist]

You should be able to do the whole procedure with a Test Indicator and a stem/bar for adjusting the arc (sweep) of the indicator. The only other thing I would recommend is buying a good Tool Holder Test Bar.

Example: CAT 50 Test Bar  http://www.commandtool.com/online-catalog/precision-test-bars-accessories/ct50-test-bar

https://us.bigkaiser.com/products/accessories/dyna-test.html

http://www.lyndexnikken.com/products-viewprd-products-CAT50---Test-Bar-2in--x-14in------------------------------------1070p1707.htm

That way you can mount your indicator to the platter using a magnetic base, and you can indicate the Test Bar in the Spindle, while moving the Z Axis to test for perpendicularity of the spindle to the table.

Also, most machines have a small gap between the lowest Z Position of the spindle face, and the machine's table/platter. Using a Test Bar of known length means you have a reference tool that can be used on the machine, or to calibrate your tool pre-setter. If a Test Bar is not available, then you can use a Gage Block Stack to get the distance between the spindle face, and the machine's C-Axis Platter or the machine's table. This part of the procedure isn't necessary for finding "center" of the A-Axis or C-Axis, but is used for Tool Length Measurement checking...

[Guest]

3 hours ago, Colin Gilchrist said:

If a Test Bar is not available, then you can use a Gage Block Stack to get the distance between the spindle face, and the machine's C-Axis Platter or the machine's table. This... is used for Tool Length Measurement checking...

Be forewarned, if these values don't match up to the Center of Rotation parameters you're in for a world of grief and endless troubleshooting.

Get a test bar. It's money well spent and will save hours and hours of hair pulling. 

[Guest]

4 hours ago, cincy k said:

I have something similar on our Matsuura. I just don;t believe it or the probe is potentially setup properly. 

 

PM Sent

[BoscoDrummer]

If you have a MX-520 then you probably have CAMPLATE and all the machine parameters should be in there to view.

[?Mark]

Even with EZ5, you don't necessarily will get good number, as it depends on your probe being setup perfectly.

We run a test part for that purpose to verify the parameters.

 

James, can you share that with me as well? I'm after any and all info/tricks that can get me to as perfect numbers as possible.

 

TIA

[cincy k]

2 hours ago, Mark @ PPG said:

Even with EZ5, you don't necessarily will get good number, as it depends on your probe being setup perfectly.

We run a test part for that purpose to verify the parameters.

 

James, can you share that with me as well? I'm after any and all info/tricks that can get me to as perfect numbers as possible.

 

TIA

Can you share the test part you cut? I've got something is mind, but would like some feedback on what others have used. 

I don't believe at the time of install that our probe was setup perfectly. I know it is now but not sure what I need to change on the EZ-5 side of the equation to compensate for a now perfect probe.  

I would also like to check the physical machine alignments to make sure they are at factory spec. Still searching for info on this...

[cincy k]

3 hours ago, BoscoDrummer said:

If you have a MX-520 then you probably have CAMPLATE and all the machine parameters should be in there to view.

Where? I do not think all the machine parameters are setup in Camplete. 

[cincy k]

Test bar on the way... Figure it will be money well spent. 

[crazy^millman]

I will speak from some recent experience that almost cost us a customer. If the machine doesn't have the correct TCP parameters in it you can make the best program you could ever think to make, but the parts will come out bad. I was real glad when someone spotted a customer machine wasn't and got it dialed in and then all of a sudden our bad programs were good. I was like magic. 

[?Mark]

1 hour ago, cincy k said:

Test bar on the way... Figure it will be money well spent. 

I have one, but the adjustments might be tricky for you. I know Sean created his own part together with spreadsheet to calculate the adjustments (after fighting thru our test and making adjustments) . I suggest you contact him as well. PM me your contact again.

We got everything to within .0005 or less at any rotation/position. I need to get his part too. We just bought another matsuura 5 axis.

[Leon82]

12 hours ago, cincy k said:

Where? I do not think all the machine parameters are setup in Camplete. 

Perameter19700 has the rotary position in the 31i

And the next one has the z value

 

[crazy^millman]

19696 is an important also how is the WKP setup on the control bit 5 is the important one.

P19696#5.

0: Table Coordinate

1: Work Coordinate

 

Fanuc helped me look into an issue I was having with an OKK-HMX6000 this must be set to Zero and the post most be using un-rotated coordinates for that machine.

[Guest]

On 7/19/2017 at 5:54 AM, Leon82 said:

Perameter19700 has the rotary position in the 31i

And the next one has the z value

 

#19700 - Center of Rotation X

#19701 - Center of Rotation Y

#19702 - Center of Rotation Z

#19703 - 1/2 Offset X (Only Effective for B/C Kinematic Machine)

#19704 - 1/2 Offset Y (Only Effective for A/C Kinematic Machine)

#19705 = Center of Rotation to Top of Table

[Guess_who]

I wrote a program to do this on our D500 before we purchased Axiset. Which I highly recommend. I didn't read through this entire thread so I apologize if it's already been mentioned. 

I can show you the entire program if you want, but basically, we take 3 points about the 'X' axis. Then we do the below calculation and voila, it will return the Y axis Centerline and the Z axis centerline. 

Variables;

#750= Y 1st point
#751= Z 1st point

#752= Y 2nd point
#753= Z 2nd point

#754= Y 3rd point
#755= Z 3rd point

 

(MATH TO CALCULATE CL OF Y IS WITH 3 POINTS) 
#101=[[#750*#750]+[#751*#751]]*[#753-#755] 
#102=[[#752*#752]+[#753*#753]]*[#755-#751] 
#103=[[#754*#754]+[#755*#755]]*[#751-#753] 
#104=#101+#102+#103
#105=#750*[#753-#755]
#106=#751*[#752-#754]
#107=#752*#755 
#108=#754*#753 
#109=2*[#105-#106+#107-#108] 
#763=#104/#109( CALCULATE Y CENTERLINE)
#763=#763*100000 
 
#101=[[#750*#750]+[#751*#751]]*[#754-#752] 
#102=[[#752*#752]+[#753*#753]]*[#750-#754] 
#103=[[#754*#754]+[#755*#755]]*[#752-#750] 
#104=#101+#102+#103
#105=#750*[#753-#755]
#106=#751*[#752-#754]
#107=#752*#755 
#108=#754*#753 
#109=2*[#105-#106+#107-#108] 
#764=#104/#109( CALCULATE Z CENTERLINE)
#764=#764*100000 

[Zbuilder]

Hey Cincy K, did you ever figure it out? We also have the MX520 but got the EZ-5 Retrofit kit from Matsuura. What we do is set the gauge ball at the 1:30 position on the table (back right corner) as far away from center (for best precision) without over traveling. We have the Marposs probe and not the Renishaw so Matsuura had to make/modify macros for us so the Marposs probe talked to the EZ-5 part of the control correctly. The program checks the ball 3 times (120° apart) on the C axis while A0.0, Then it tilts at A-45 (I think) and does it again at 3 positions around C axis.

[Zbuilder]

I also have all the macro variable descriptions if you need it.