[OK_Corral]

Hello all,
I am a new member here that is trying to get feel for how other people in the machining world tackle what I will describe. So currently we are running (4) Okuma MU5000V mills and are looking to start doing some on-machine probing for inspection. I realize this is a controversial topic but due to the nature of our work we will need to have it right and cover our bases. So, basically what I am wondering is, for those of you that do any kind of probing past finding the stock, how do you get the data from the probe. We have Renishaw probes and are running them through Inspection Plus. All that produces data wise is a text file, and we need something more robust to track trends. Is Renishaw CNC Reporter my best bet or are there other options? I haven't found many people talking about CNC Reporter so I was wondering if there is something else. Thanks

[deadlykitten]

hello, i will go fast through some stuff

probes varies ... there are some that work on metal / conductive and/or plastics / non-conductive

some have limited behaviour : if they move across a direction, and the surface that comes in contact with is perpendicular on their movement, thus "tangent vector" at contact point makes 90 degress with approach direction, than they work fine in a cone of ±2..5degrees ... such probes don't trigger erorr signals if deviation is out of 90±2..5degrees, so in this case, absolute position is not to be trustfull

there are others that always make contact on same point, and can be used on a polar angle, as X&Y syncro ... this can be trusted on absolute position

so 1st, check how much can you trust that probe ; renishaw delivers different probes, with diferrent software

they write data into files ... from here, you may use specific statistic software; some probes come with aferent software, so they deliver instant data

i also use a taster for out of 90±2..5degrees cases, and once it is in position, i run a program to write data to a file; thus, i was able to correct a part with complex geometry, used in leather manufacturing

[deadlykitten]

hello again

before analyzing output data, is important to know how accurate your probe is, depending on tangent @ contact point

please take a look at attached image : take multiple points into an internal diameter, respectively, into an outside diameter (try 100 .. ). Surfaces should have a small rugozity, so craft them on the machine, or grind them and after put them inside the machine.

Locate their center with other device, not the probe ... run a flow of paralel measuring, at 2.5 .. 5 mm, and after analize the output data ... see if it fits on a circle i can help you with that analyze, if you provide real diameter, center before measuring and output data from the probe

Paralels should be aligned with X or Y axis ... is enough to test one direction.

Thus, you can determine the precision cone of your probe. After, you can decide what that probe is it good for If this test proved ok, and was on X, do it among Y, and after, do it at a random angle

Otherwise, you may fall into the case described at image 2 > probe outputs a smaller diameter during testing measuring on a inside diameter ; if you skip this test, but the probe has this behaviour, than will be a point when all measures on the machine will be wrong, and nobody will know what is happening ...

During calibration, probe was oriented somehow ... it may be necessary to always use it with the same orientation, so if you take it out from the pot, put it back in the same way

If you wish, consider to block a pot into ATC to be used only for the probe; neighbours may be empty and also blocked kindly !

[deadlykitten]

So, basically what I am wondering is, for those of you that do any kind of probing past finding the stock, how do you get the data from the probe.

hello, i think you still struggle with this

i will go through a few variants, because i don't understand exactly what you wish ( be more specific please ) :

1) your probe does what you wish :

.... but you can not find outputdata > probe software may display on screen the final result; also, it may be written inside a file on your machine

.... you wish to extract specific values from the output file > look inside the file

.... you wish to analize specific data from the output file > you need specific software, to handle your geometry ; i don't use "Renishaw CNC Reporter", but i can analize output data for my parts

2) your probe does not what you wish : so you get points randomly, and maybe you wish to create a surface with them ? extract data from the file, input it inside a CAD that can handle what you wish ( render surfaces from a cloud of points ), and check

ps : because of "direction cosines" i don't use the probe for complex geometries, but the taster, and once i am in position, i run a program that writes coordinates into a file ; after that, i analize it

[deadlykitten]

because of "direction cosines" i don't use the probe for complex geometries, but the taster, and once i am in position, i run a program that writes coordinates into a file ; after that, i analize it

once i write data into the file, i use "tabs", thus making it possibile to be "copy-paste" inside an excel file

to do the same with data from probing file, you need a convertor a programmer can create such a small aplication, that reads probing file and outputs into excel kindly !

[Superman]

once i write data into the file, i use "tabs", thus making it possibile to be "copy-paste" inside an excel file

to do the same with data from probing file, you need a convertor a programmer can create such a small aplication, that reads probing file and outputs into excel

I don't think that would be the way to record data
- there is no function inside the measure cycle to send to a particular file.....
so you would need to create a .TAP file that probes a feature, then send that data to a file, probe the next feature, then add it to the data file & so on
- each probe feature would write data to a specific addresses, so the 2nd probing would overwrite the 1st results

What controls are used ?

Look at attached manual for OSP E100M Autoprobe
ref....Section 3 for Output Functions

[deadlykitten]

each probe feature would write data to a specific addresses, so the 2nd probing would overwrite the 1st results

this is rewriting, so old data is lost ... to avoid this, simple writing is required, like a "data log" ; i log the 3dtaster

of course, more stuff is required to log probe data in a specific format ; some (my) probe logs automatically, but if format is not ok, than a converter is required, so to avoid direct/manual edit

if someone can afford it, than he may get a custom probe software, doing directly what he wishes ... question is what exactly needs Ok_corral ?

[Superman]

Hey Puddy Tat
"this is rewriting, so old data is lost" ... no....it overwrites, use the correct terminology
"to avoid this, simple writing is required, like a "data log" ; i log the 3dtaster " ..... we are not talking about a 3Dtaster.......he's using a Reinishaw Probe & the associated cycles....so do NOT add confusion

"of course, more stuff is required to log probe data in a specific format ; some (my) probe logs automatically, but if format is not ok, than a converter is required, so to avoid direct/manual edit".....he's got to do it, ... not you

"if someone can afford it, than he may get a custom probe software, doing directly what he wishes ... question is what exactly needs Ok_corral ?.".... He's already got the probe software, why does he need to buy more ?... what he needs is a method to extract it to a file

[deadlykitten]

hy

"this is rewriting, so old data is lost" ... no....it overwrites, use the correct terminology

if a program outputs a file, through "save as window" at a same path as an existing one, the user is asked if he wishes to overwrite/replace/etc

if user agrees with it, than the program rewrites the file ....

"overwrite" is most often used, is more common language ; rewrite is used more when developing software, is "under the hood"

also, my native language is not english i am just tring to find my words so be kind ....

"to avoid this, simple writing is required, like a "data log" ; i log the 3dtaster " ..... we are not talking about a 3Dtaster.......he's using a Reinishaw Probe & the associated cycles....so do NOT add confusion

i was talking about the fact that "log" can be done on probes, on 3d taster, or wherever needed ... after, the "log file " can be sent to statistical software

"of course, more stuff is required to log probe data in a specific format ; some (my) probe logs automatically, but if format is not ok, than a converter is required, so to avoid direct/manual edit".....he's got to do it, ... not you

of course he has to do it / if he still wishes to however, if he will explain a bit more what he needs, i may help him with specific formats

far as i have understood, he is not comfortable with the output format of probe data :
... he may rewrite probe codes
... he may wish for an output format compatible with a specific statistical software
... he may re-arange output file, to be easier to read

i do not insist on any of this ... i am just saing, maybe something will ring a bell

"if someone can afford it, than he may get a custom probe software, doing directly what he wishes ... question is what exactly needs Ok_corral ?.".... He's already got the probe software, why does he need to buy more ?... what he needs is a method to extract it to a file

maybe he is not satisfied about how his software behaves

same probe may come with different software/functions, in respect to the machine that will serve

there are probes producers, that can implement special functions ... just say what you need, and pay

maybe his software does not compute trends the way he wants .... i don't know

[tea hole]

Hello all,
I am a new member here that is trying to get feel for how other people in the machining world tackle what I will describe. So currently we are running (4) Okuma MU5000V mills and are looking to start doing some on-machine probing for inspection. I realize this is a controversial topic but due to the nature of our work we will need to have it right and cover our bases. So, basically what I am wondering is, for those of you that do any kind of probing past finding the stock, how do you get the data from the probe. We have Renishaw probes and are running them through Inspection Plus. All that produces data wise is a text file, and we need something more robust to track trends. Is Renishaw CNC Reporter my best bet or are there other options? I haven't found many people talking about CNC Reporter so I was wondering if there is something else. Thanks

(regarding what others are doing)
What others do when you get this deep into probing, in my experience, is nearly irrelevant to what you are after. What I mean is, getting the data from the machine is easy. You can use Renishaw subs, Inpsection Plus software or write your own, though in the end, it is going to depend on what information you want to gather. The renishaw cycles that write to a text file output based on the "READ/WRITE GET/PUT" functions. They can be altered if they're machine sub programs, or Renishaw can modify the ones going through Inspection+. You can also track data yourself. What you do after that is going to depend entirely on what you and/or your customer want for CYA methods. I see a lot of people who grab data from the text and import into a .xlsx and use the data how they want.
I think before you spend too much time asking others what they do, the more pertinent question is for you; What do I(you) want?

Unless you know what you want to do with the data, there's nearly nothing we can do or say to help guide you to anything meaningful. IMHO.

I would personally think that gathering the data en masse and occasionally analyzing it for SPC comparisons to prior audits would be about as useful as the accumulated real time data can be. Otherwise, if there's a problem, I want the machine to find it with the probe and fix it long before I see it in my data; otherwise I've got a costly pile of scrap. As for long term inspection reports, I want those coming from inspection anyway since that's how the customer will be measuring the parts anyway. Too many things can change after you release something from a fixture to distort or relax a part to use that data as anything but a reference to track machine performance. In my opinion.

(regarding inspection plus)
I truthfully think if what you're after is somewhere to collect the data the renishaw cycles is outputting, the obvious answer is to use the renishaw cnc reporter. I think if the machine is configured for ren. probing, and the software is collecting ren. output data and creating reports in a format created by Renishaw, that is the best answer. At some point too many hands get into the mix and multiple softwares from multiple companies will only cause headaches not outweighed by the benefits, ultimately frustrating you and nullifying the gains the act of collecting the data offered in the first place. I'm certain i'll get flack for that, but at some point, you have to ask yourself if you want to take on the endeavor of reinventing the proverbial wheel, or if you want to just go buy one. It may not meet all your needs, but software can be altered so picking something that melds well and sticking with it and working with your supplier to make it what you need, or at least closer to it, will be way less stress than trying to put a bunch of 3rd parties in there. even if one of the 3rd parties is much better than what the Renishaw solution would provide, your'e limited by the least common denominator. So whatever the weak link is will be the best the solution can be. Again, my opinions.

[Superman]

hy
if a program outputs a file, through "save as window" at a same path as an existing one, the user is asked if he wishes to overwrite/replace/etc

if user agrees with it, than the program rewrites the file ....

"overwrite" is most often used, is more common language ; rewrite is used more when developing software, is "under the hood"

also, my native language is not english i am just tring to find my words so be kind ....

You missed the f***** point..... you jumped to the end, without considering the steps at the beginning
When you probe, the data from that cycle is stored in fields ( say the result of a diameter, and the calculated centre point at XYZ) , then probe the width of a slot,....... the 2nd probing results OVERWRITES the previously stored data without asking ( other fields would be "cleared" )
you need to do something with the data after EACH probe cycle

What then needs to be worked out is, what data is needed to be recorded into the external file. Yes, you could dump a gauging result back to a file. You possibly could also extract the data field to a file in a tabular form using VS variables ( you need to copy to a user variable [VC#] to extract) .... ref Section 5-3-3

The other contentious issue is
- using an un-calibrated machine to verify size..... unless the machine is certified back to a measurement standard, it cannot be taken as being accurate. Only to be an indication that you are in the ballpark
( this calibration is NOT a cheap process, and it has to be done regularly )
ie. - the machine is programmed to create holes at a 4.0000" pitch, it cuts them at what the machine says is 4.0000", your probing would say 4.0000".... but external measurement may give a different size....... do you accept the machine results ? .... or the independent measuring result ?
- the inspecting of your own work....... will always come back to haunt you

[deadlykitten]

hy superman

You missed the f***** point..... you jumped to the end, without considering the steps at the beginning
When you probe, the data from that cycle is stored in fields ( say the result of a diameter, and the calculated centre point at XYZ) , then probe the width of a slot,....... the 2nd probing results OVERWRITES the previously stored data without asking ( other fields would be "cleared" )
you need to do something with the data after EACH probe cycle

my probe software auto-logs, so i can inspect results history / so, if i don't delete them, than results will be there till next year

if there would have been "no auto-log", i would have fixed it

The other contentious issue is ... will always come back to haunt you

yup, very true

there is a case when machine accuracy is ok, but probing/measuring is not, because of "direction cosines" ; in this case, probe outputs wrong values, without trigering erros, and this is not good ... this is what i was talking about at begining of this thread, so is more important to know how much "a probe can be trusted" before data export

i will be back on this aspect soon, because the way that i have described it is not clear

[deadlykitten]

Hy teahole

...import into a .xlsx and use the data how they want

arround here *.xls is more comon ... i really hate that Microsoft delivered their latest packages with " *.*x " file format

I want the machine to find it with the probe and fix it long before I see it in my data

that's a nice one : using the probe to auto-correct the machine, without external intervention ... i know is possible, but i have never used it

Too many things can change after you release something from a fixture to distort or relax a part to use ...

a few days ago i was asked to craft some lathe parts at 0.002 tolerance, and release a quality certificate / so to measure a single part at each 100

i asked how is it possible to wish 2um precision, and inspection 1% ? well, i was not going to reject it, but to be sure, i asked about parts destination/functionality, so to be more relaxed when crafting them ... it proved that who i was talking about was an intermediate, more like a trader there are problems when crafting single parts or small series if the customer is an intermediate, so i could not risk going on cnc's if things are fishy

At some point too many hands get into ... whatever the weak link is will be the best the solution can be

someone inspected my renishaw software, and said that he could use my probe by writing his own code, and thus, adding custom functions ... sounds nice

also mr. Wizard recomend me a guy, when my probe failed at some tasks ... [email protected]

and there are "direction cosines" which limits most probes / i will be back on this soon

thus, i found my way through all this, and i have some contacts, to call for help, just in case

IMHO. In my opinion. Again, my opinions.

stop saing that it fills your words with uncertainty, and what you say is not wrong

i was also told so ( by a woman, when i was little ) , and i realize that is true ... such words decrease your value ; no one needs an "excuse me, may i ?" ... think about all guys talking s****, but with guts

be technical, be sure

[deadlykitten]

if there would have been "no auto-log", i would have fixed it

i just hit into such a thing : part is clamped arbitrary, and i needed to determine rotation

there is a function for this, as shown into attached images, but it works manually, so it does not care about the probe

also probe software is limited; i am glad at this point that actually there is a software

so :
... i measured a position among Y with a program delivered by probe software
... i inspected the log file, to check output data
... i searched key words from the log file/output data, and i find them inside a *.msb
... the *.msb file has variables that handle this type of measurement
... i modified the initial program to deliver measurements at 2 coordinates, and i recorded the output value ( thus the variable from the *.msb that performed global ), from the probe, for each coordinate
... thus, having this, i can determine rotation
... when i saw that it works, i disabled the log

because part is rotated, output values from the probe are not real, but the slope between them is, because bought measurements are affected by same error