New DNC System

[JakeL]

6 months later than promised, but better late than never! We've finally almost got CIMCO DNC up and running.

We've got a meeting tomorrow to go over the details, but it sounds like engineering is pushing to have all programs be O0001. Does anyone else exclusively use O0001? Seems like a really dumb idea to me.

They're argument is this will force the machinist to keep only one program in the machine at a time. We've had it happen a couple times where a program stays in the machine, the part gets a rev change, production starts up again with the old program and we scrap a bunch of parts because they're the old rev.

I'm a big believer in not complaining without proposing a new solution. But I can't seem to come up with any good ideas on this one, other than telling the guys on the floor to save and delete old programs, but that's been done and is not always feasible.

Curious how other shops handle this, and any other thoughts on the matter.

[JParis]

We've got a meeting tomorrow to go over the details, but it sounds like engineering is pushing to have all programs be O0001. Does anyone else exclusively use O0001? Seems like a really dumb idea to me.

This sounds like a solution in search of a problem....I am guessing a mish mash of different controls are in use and you're still locked into 4-digit program numbers?

We've had it happen a couple times where a program stays in the machine, the part gets a rev change, production starts up again with the old program and we scrap a bunch of parts because they're the old rev.

This is a "Process control" issue....they should not even be allowed to hit cycle start until; the Rev level of the paperwork and program match...

We have dropped program numbers and gone strictly alpha-numeric...

All information is in the header and the Cimco NCBase queues on that info to keep track.

<59904V02-SLIDE-VERTICAL-OP1>(59904V02-SLIDE-VERTICAL-OP1)
(PART NAME - SLIDE-VERTICAL)
 
(PART NUMBER - 59904P02)
(REVISION - 3)
(SOW# - 599)
(OPERATION # - 1)
 
(MATERIAL - ALUMINUM INCH - 7075)
(STOCK SIZE - .400 X  1.500 X  3.600 LG)
 
(PROGRAMMER - JOHN PARIS  -  06-13-24)
 
(XY ZERO LOCATION - G54.1 P1 - CENTER OF LEFT SIDE PART / -.830 FROM STOCK CENTER)
(Z ZERO LOCATION - G54.1P1 Z0 IS TOP OF PART)
 
(FIXTURE # - HOLD IN VISE JAWS)
 
(PROGRAMMER COMMENTS - XY ZERO - G54.1P2 RIGHT PART / +.83 FROM CENTER OF STOCK)
(ADD'L COMMENTS - Z ZERO IS THE TOP OF THE PART)
 
 
(****************NO EDITS OR COMMENTS ABOVE THIS LINE****************)
 
(CYCLE TIME EST - 18:58 MIN:SEC)
()
(CYCLE TIME ACTUAL)
(**************************SHOP REVISION LOG*************************)
 
(********************SHOP REVISIONS ABOVE THIS LINE******************)

[SuperHoneyBadger]

Each part here has a unique 4 digit number, it's on the physical folder for the job, it's the last 4 characters of the file folder name on the network, it's on the PCD, it's the first 4 digits of the NC file name, and it's the number when sending data to the controls. It also ends up being the colloquial name of the parts around the shop, "Hey, bahd, you got the folder for 3320?".

The rev # or letter at the end of the file name should be what differentiates the revs from one another, and IMO parts should have unique ID#'s.

We usually load an NC file each time a part is made, as we have very limited memory on 90% of our machines.

I have taken the stance that it is my job as programmer to ensure that the directory where the set up guys pull NC files from has the current rev of all programs. When I update a part the new files are stored on the "hot" drive that the shop floor can see, and the old rev NC files get stored in a hidden folder called OLD inside the part's folder elsewhere on the network.

It is on the operator to load the files they can see, check the rev matches the set up documents and PCD, and make the part accordingly. If they DO notice a rev mismatch, it should be reported to someone, lest it is not caught in 1st inspection - which it usually is.

Each person has a part to play in making the correct part, from the correct material - the process needs to be in place though. If you shoot from the hip as an organization, you'll probably end up with your equipment up for auction.

[JakeL]

This sounds like a solution in search of a problem....I am guessing a mish mash of different controls are in use

Correct, for the mills we have OSP, Fanuc, Yasnac, Siemens. We've also got lathe and EDM which I don't know much about.

you're still locked into 4-digit program numbers?

What's the easiest way to find out if a control only accepts 4-digit program numbers? I'd guess we can't get away from the 4-digits.

With the old DNC we had a 5-digit DNCID and the 4-digit program number was derived from that. Sounds like the filename will be taking place of the DNCID, like your example.

[Jeff]

Does anyone else exclusively use O0001? Seems like a really dumb idea to me.

It depends on whether or not the control can accept actual program names like OP1.NC

When I was doing this for Fanuc 0-M and 0-T controls that would only accept 4 digits, I always made my programs O1234.NC
This way if the operator had made some program for something like soft jaws or whatever, they could make their program O0001
They knew my number would never change, so when they downloaded mine, they could easily change it to whatever number they wanted if O1234 was already in use.

[JParis]

<59904V02-SLIDE-VERTICAL-OP1>(59904V02-SLIDE-VERTICAL-OP1)

By the way, that "V"....that represents that this is a vertical program...  59904 "V" 02

V = Vertical
H = Horizontal
X = 5 Axis
L = Lathe
T = Integrex

SO we can have the same part programmed for different machining centers and not have them trump the other program

[CNCAppsJames]

Siemens, FANUC 0I-F, FANUC 30 Series can all take alpha-numeric. 

I haven't used an Oxxxx in 15 Years for a part program. That's arcane.

[JakeL]

Siemens, FANUC 0I-F, FANUC 30 Series can all take alpha-numeric.

I haven't used an Oxxxx in 15 Years for a part program. That's arcane.

The idea of alpha-numeric is very appealing. Unfortunately we've got a handful of early 2000's and late 90's machines still kicking around, and I'm almost positive they need Oxxxx.

I think I'm just going to say we're better off sticking with something similar to what we're doing. Have a list 0001 thru 7999 and just use those program numbers in a loop. That way it is very unlikely (albeit possible) to have overlapping program numbers in a machine.

[SuperHoneyBadger]

I haven't used an Oxxxx in 15 Years for a part program. That's arcane.

Master of the Arcane, reporting in. Working with a stable of Yasnac i80's, so my only option for 80% of my spindles is Oxxxx. Honestly, I love those big green Matsuuras to death - I bet they paid for themselves 50 times over. And it will get hairy if I start using alphanumeric just for the 2 new machines.

[JakeL]

Current plan is to have an excel list numbered O0002 thru O7999 and grab numbers as needed for new programs. A few reasons we went with this. Legacy programs don't need to be changed. File management at the machine will be much easier with "unique" program numbers (we'll have to start at O0002 again when we reach O7999). Much lower risk of accidentally overwriting a program at the machine.

The engineer's plan was to have the machinist change the program numbers to whatever they wanted. The flaw with this was the program number would never be recorded anywhere.

The other big plus to the new DNC that I had forgotten about is we now have the full version of CIMCO Edit. I was using CIMCO edit 2022 so I'm also excited to be using 2024.

[JakeL]

As always huge thanks for everyone's feedback.

I'm sure there will be a couple more bumps in the road before everything smooths out, but I like the direction we're headed. CIMCO is miles better than our old system (Focal Point)

[mkd]

This sounds like a...... shitty Fanuc problem

fixed

[Thee Byte]

You're nc output would ideally have job numbers and rev numbers in the program header imo, would be easy for you to set up I think, seeing what you've accomplished so far.

[JakeL]

You're nc output would ideally have job numbers and rev numbers in the program header imo, would be easy for you to set up I think, seeing what you've accomplished so far.

part#_OP#_version#_machine#.NC
This is what we are using for the file name, which will also be the first line of the g-code header.

Someone else proposed this naming convention and my first question "is someone updating all our posts?" (22 individual posts). So far I've done the header updates on 3 of them (as I needed them).

[JakeL]

22 posts?!
How many different machine types do you have?

You don't want to get me started on this. One of my biggest complaints about our shop, but I don't cut the checks so only so much I can do.

We have about 35 mills, Okuma, OKK, Matsuura, Yasda, Smart, Quaser. We have mostly Fanuc with some OSP, Siemens, Yasnac. And I'm sure I'm forgetting one or two. Basically it's hard to find two of the same machines in our shop.

Recently they found a "really good deal" on an OKK HM1250 and we almost weren't able to talk them out of it. Couple years ago they bought an OKK HM-X8000 5-axis head table machine. For reference, our largest machine before the 8000 was Matsuura HPLUS-630