O/T "Standard Industry Practice"

[chris m]

Brothers and Sister of the forum

 

This is, more or less, an adjunct to my Integrex saga but I am in need of input

 

The time for the final acceptance runoff for my machine is at hand and the program does something that I do not agree with, but the Mazak people are trying to sell as "industry practice."

 

The situation is this:

 

One of the features being turned is +.0000/-.0005 and the machine is programmed to turn to a known oversize condition and then stop so that the operator (in this case a Mazak apps guy) can measure the size and tweak his offset as necessary to hit the final dimension, then the machine turns to finish size.

 

I personally feel that this is weak method but Mazak is trying to sell it as a "standard industry practice." The rub is that the acceptance was based on "general good machining practices" being used. I personally don't feel that this qualifies, but they do.

 

Please weigh in

 

Thanks in advance

 

C

[STEELTHOM62]

Chris,

 

Usually, on anything tight we only do this on the first piece and then adjust accordingly for tool wear after that.To do this on each part,in my opinion, is a waste of time.

 

Just my $.02

[godzillia_dude]

This would be an "acceptable practice" IF I were running Inconel 718 and I was only running 10 or 20 parts. This seems to be a much higher volume and, as I do not know what your criteria for inspection is (ie. what dimensions, how often, freq. of 100% inspect, etc.) and also not knowing the run time or quantity, I would press for more of the "expected and promised" performence in this machine. My curent question is are you confident that your group will be able to produce the part as consistantly as the machine applications techs? Renmember, at some point (hopefully soon !) the machine techs and service engineers will be out of the building and it will be you on the line everyday from here on out. HTH and hoping the very best

[Tony]

Industry standard for "One zee's and Two zee's"

Production :Let them fly and use SPC.

 

[ 01-22-2003, 05:22 PM: Message edited by: Tony ]

[chris m]

Thanks for the input guys

 

Keep it coming

 

We turn parts +/-.0002 all day long without doing this, but they say "...but those other machines are much smaller..."

 

C

[kathy]

You can always close your eyes, push the button and pray...

 

There are a lot of factors that can affect machining tolerance besides program and machine - material and/ alloy, tempature, chip buildup on tool... Personally with that kind of dimension I would check every part. You can always take off material but it's difficult to machine it back on!!

 

just my ho

 

Kathy

 

[ 01-22-2003, 06:05 PM: Message edited by: kathy ]

[Guest CNC Apps Guy 1]

I'm with Tony. They need to give you a stable enough process that you can "run with confidence" so you can predict (with relative ease) tool wear, etc...

 

quote:

---

general good machining practices

---

Yeah, this statement would never hold in court. LOL!!!!

 

How many good ones do they have to run before you buy off the manchine? I certainly would not agree to less than 100 pcs. Check them all and they shoudl all fall within a certain range. Your QC guy should be able to give you a good set of parameters.

 

You need to grab these guys by the nuts and twist if you know what I mean. Give 'em an inch and they'll take a mile!

 

JM2C

[Gary]

Yes i agree with that on the first part or two check each one, after that (if running a seperate finish tool) you should be able to at least check the last part while the current one is running. A good machine will not creep out of tollerence that quickly.

 

just my .02$

Gary

[ssibiski]

I agree with Tony. I will check before i make my final cut when it is one or two parts. But if it is production. I don't want to see the machine stop untill the part is done.

 

This is the only way i know to make money in this industry.

 

Steve Sibiski

 

Ps ask what industry they are talking about

[Scott Bond]

quote:

---

One of the features being turned is +.0000/-.0005 and the machine is programmed to turn to a known oversize condition and then stop so that the operator (in this case a Mazak apps guy) can measure the size and tweak his offset as necessary to hit the final dimension, then the machine turns to finish size.

---

Hi brother Chris

Unless you had some agreement with Mazak that the machine would scan the part ,then adjust the tool compensation un-manned .

 

I.M.H.O. It would be standard to put this in your program for say 2 to 10 pieces. For a longer run I think the standard would be to floor inspect and dial in the next part from the last part cut.The Mazak would be expected to hold tolerance, so that the shop floor could consistently predict tool wear on a common part.

We pay for the machines--- we don't have to pay a guy to watch it run also.(it's a robot)

 

[ 01-22-2003, 09:11 PM: Message edited by: Scott Bond ]

[iowajim62]

Hi Chris,

 

Time for a Mazak "horror" story.

My current employer setup a FH-680 Horizontal Mill Cell with 3 machines, 16 pallets and robot.

These are '96 machines that I started working on in '97. The "Z" axis way covers started coming apart 3 months after we started "lights out ops". To make a long story short over the next 2 1/2 yrs we went through 67 left and right covers before the engineers from Oguichi,Japan came over and took pictures. They then redesigned the inner support plates with Teflon bushings. Throughout this whole ordeal our owner stuck to his guns and made them cover it all under warranty. I can tell you also that we needed these machines to run so bad that we kept spare covers on hand and could change them in 30 min. Not cool but necessary.

 

Not to keep rambling but "standard practice" depends on where your from and what you do.

 

Tell your owner/boss to stick to his/her guns and don't cough up that check until he decides the machine passes his/her criteria. Mazaks are good machines but it doesn't mean they are perfect.

HTH

[Jack Mitchell]

Chris,

 

Reading between the lines, I gather you are not happy with the approach. This is a good thing. Be wary of the application experts – they are replaced quite regularly.

 

Because an individual knows a lot about Mazatrol or Mastercam for that matter, does not make him an expert. Many of us in here have a lot of experience, in fact tons of experience. Yet we do not ever refer to ourselves as experts – this is why we call it good machine shop practice, sort of like how a doctor calls his trade a medical practice.

 

Realistically, your quandary is unique to the application – which is to say, you are the expert for the moment – use your own judgment with this application, and do not fear the argument – cherish it; it’s what makes ChrisM what he is.

 

The final sign-off will probably rest with someone other than yourself; if this is the case then perhaps you might accept that the machine tool is performing as it should. We cannot control the impressions of others nor can we control the things that we do not agree upon.

 

Funny thing about the unemployed application expert when he arrived and applied at our domain, I actually interviewed the man, who felt that it was quite indignant that I, the operator, was asking him these questions instead of the Engineering Manager or Human Resource Manager. (You just never really know the ability of the guy asking the questions) needless to say, I dismissed the expert’s ability in about 45 seconds – tastefully, I might add. This man is a white collar worker that did have any of his own precision measuring tools for I’m sure he knew exactly how to use them.

 

Mazak does indeed make a fine machine tool; hardly the best – nobody owns that trademark it this industry. If I were to pigeon hole Mazak as a machine tool builder I would say that they guarantee their products above and beyond many others – go with your instincts with this one.

 

Regards, Jack

[kkominiarek]

If pryor to machine purchase, the rep stated that you would be able to hold this tolerance and produce parts in an acceptable time frame.....I would say this is not an acceptable machine practice.

 

This sounds like a production job. Production means "produce parts" not stop and check!

 

MHO

Ken

[chris m]

It seems that most of you are of similar minds on this (as am I) but please keep the information flowing regardless of which side you are on. I am not looking to start a flame war or to discount anyone's opinion so if you have read the posts in this thread and hold a dissenting opinion but don't want to post it because you feel you'll get abused, please don't hold back.

 

So all of you know a little about where I'm coming from we manufacture precision reduction gears, gear heads, and actuators so we machine parts with .0005 total tolerance all day long. We do mostly turning, a lot of drill/c'bore/tap operations and a limited amount of contour milling.

 

In our typical turning operations the operator will inspect every piece as it comes out of the machine if there are fussy dimensions (which there usually are) to keep an eye on tool wear and thermal growth (we have thermal comp in all of our machines but nobody's perfect) while the machine is running the next piece.

 

The only job that I have that we use a "gage cut" is a job we turn that has a bore that is +/-.0001 and we run this in a 1984 Mori Seiki SL4. Yesterday I asked the Mazak guys why my 20-year-old machine could hold +/-.0003 all day and their $480,000 new machine can't. They were not smiling.

 

Thanks for input.

 

C

[jeff]

I couldn't have said it any better myself....mr bond.

[roy g. biv]

quote:

---

We turn parts +/-.0002 all day long without doing this, but they say "...but those other machines are much smaller..."

---

I can smell the steaming pile they are shoveling from here.

 

Let me ask you this Chris. After he makes his rough pass "to a known size". is the finisher going to the same offset every time?

If the offsets are varying more than +/-.0003 from each other I would have to say that there is a problem with their machine or process.

 

I typically run parts with +.0002/-0.0 to +/- .0003 all day on my Mori SL150 and SL15's without adjusting for anything other than tool wear. And this is on 17-4PH ,416,and 440C all 40-60 Rc.

 

Good luck

[Guest CNC Apps Guy 1]

quote:

---

Yesterday I asked the Mazak guys why my 20-year-old machine could hold +/-.0003 all day and their $480,000 new machine can't. They were not smiling.

---

Of course they were not smiling. They just got put down hard!

 

While I worked as an AE, one of our customers (who ran their LPP literally 24/7/365) was not pleased that after 3 years they started having issues with spindles (and they were running spindle load up around 100-125% most all of the time as well and several large documented crashes). So the customer started courting Mazak as an "alternative". Mazak said "Oh yeah, you can plug that Mori program(after some format changes but not big) right in there and we'll outperform the Mori everytime.." To make a long story short. They pretty much destroyed the machine. Spindle - TOAST! They just another Mori LPP two years ago.

 

JM2C

 

Make 'em step up. Don't let them off the hook. They made a promise, make 'em keep it.

[Quickmike]

I think they should set up a probe looped to a macro program that would control the offset. Do you have any open tool positions where a probe could be used? How many parts and what kind of cycle time are we talking? If this is along run implement spc possibly using the probe to provide the data. If the probing is an option the parts could be probed before the finish pass to get the process going. After the process is running good use spc by probing the finished part on the machine and using this data to control the process. When probing on the machine be sure the chuck pressure is not affecting the readings. If it does, hopefully it will be by a consistent and repeatable amount. If it is adjust by this amount.

[McRae]

Acceptance Criteria for a new machine should be referenced on a purchase order. When specifying a new purchase, with the benefit of a Turn Key Solution, there should have been minimum process capabilities established that takes into account Machine, Program and Tooling. An accetance of 1.33CPk is the old standard - I would use a tighter figure of 1.66. AIAG has a nice PPAP document that outlines what the automotive guys do when evaluating and certifying a process (to 6 sigma). These data analysis calculations can be found in any SPC reference manual or textbook.

 

If this wasn't specified on the PO - Good luck getting buy in from the machine guy.

 

As far as the trail cut - This is an excellent practice for producing top quality products. I agree that with a stable process and a caable machine, the sample frequency could be reduced from every part to something in the order of 1 of 15. After the machine is yours, then use a little bit of trick code to jump around the stop after the intermediate cut. Send me an email if you want the code.

 

If the machine will make good parts, within the given takt time, then you are off to the bank with all the profit! Any extra productivity improvents are yours to keep.

[jeff]

I'm with James,

When it comes to turning,it's tough to beat a Mori Seiki!

We have an SL35 we bought back in '89,we always run the load to about 100%-125%,it's been crashed hard many times,but it still runs strong!

[McRae]

To skip over section using the Fusion 640

 

/IF(#100 EQ 15) GOTO 9

M00

(CHECK DIAMETER AND ADJUST OFFSET)

#100=0

 

**INSERT RESTART BLOCKS HERE AND ENSURE**

**SAFE START POSITION **

N9

#100=#100+1

 

Place this bit of code around the M00 and restart blocks so that you don't need to measure every part. To override and force a program stop - just use block skip and the conditional will never be evaluated and the program will reach the M00 and also reset the counter back to zero!

[Rob B]

I have ran Mazak machines most of my 15yrs in the business. First 6 yrs I was running QT 25 in production shop(cast iron housings). They were small parts. The machine would move .002 in the first couple of parts(didn't run warm on machine). Then it would settle in and run within .0001-.0002 all day.

I run Mazak mills in tool and die shop now. They run great(AVJ-25 & 414). I hate the M32 controller Using EIA programming.

Now for the point to be made. Our first lathe in the job shop was a Hyundai. Same as a mazak (except the controller) and about $30,000 cheapier. Start the machine up in the morning holds within .0001 on first part. We decided to try to run some production parts on the side. We bought 2 Mazak QT20 Lathes. After about 6 months 1 of the lathe began to jump around on tolerance(.002-.006). It would move again an hour or two later. Mazak Tech guy was an xxxx at the time an didn't want to help. The Tech guy finally sent a service man in, he said there was no problem with the machine. His statment was: It is thermal expantion, it's been around for years. What an idiot. Long story shortwe traded for another Hyundai. Two weeks before the Mazak was to be shipped out another Mazak service guy came through spent 4hrs baby setting the machine as it ran. He seen that the machine just began to move radically all at once. This service guy found a bearing out "X" lead screw. He fixed the bearing and the machine ran great after that.

As said before, you run the operation everyday. You now what you need the machine to do. If it does happen DO NOT SIGN OFF ON MACHINE!!! It is your baby then. When the parts are not coming off the machine as fast as Owner or Supervisor wants. The xxxx Peeeeeeeling Comes Home To You. The Mazak sales guy will be trying to help you all he can(as in it is your fault). He sure doesn't want the Peeeeeling to come his way. Give your thoughts on the deal to the Owner- Supervisor.

When buying a machine for turn key production job. Always get it in writing how manty parts they guarantee the machine to prduce in a set amount of time(8hr shift).

[chris m]

quote:

---

The Mazak sales guy will be trying to help you all he can

---

Actually, our sales guy doesn't seem to want much to do with us. If that's the service you get for a $480,000 machine; I'd hate to see what the $100,000 machines get you

 

quote:

---

When buying a machine for turn key production job. Always get it in writing how manty parts they guarantee the machine to prduce in a set amount of time

---

We did not spec a cycle time for our parts, which was a mistake but not the big deal here as the cycle time is reasonable. If the machine will run correctly we can always try to wring some time out of it; right now it just needs to make good parts. We are currently on Part Six of our fifteen-piece acceptance runoff and the roughing mills are taking a beating. The machine will have to run all day today to meet their "close of business" deadline; we'll see tonight what the deal is!

 

C

[chris m]

Just bringing this back up to the top to say thank you all very much for your input. The runoff is done and next week our QC will perform a 100% inspection on all parts and we'll find out what's what.

 

Just to update the actual topic of this thread, after the 2nd part I think the machine would've maintained the size within tolerance without any of the gage-cut rigamarole.

 

C

 

[ 01-27-2003, 09:46 AM: Message edited by: chris m ]

[Part Factory]

Chris M

 

I am new to this forum.

 

Recently a industry friend directed my attention to this forum because there was some interesting conversation regarding the Integrex.

 

I have read as many of the postings as I could regarding your "Part acceptance purchase order requirements". My suggestion is this!!! Call your dealer and have them remove the machine by the end of the monday business day. I realize this is a harsh comment but I support this action with a Integrex journey of my own that cost me a business failure.

 

Don't be whoed by the machines pretty looks and imagined versatility. It is nothing but I candy. (A super-model without a brain!!!) The software in this machine is a horrendous collection of band-aides and patches that Mazak has difficulty understanding.

 

Why is it so difficult getting the CAD-CAM post processors to work with this machine? Answer. The control code in this machine was never engineered to talk to the outside world. Mazak's answer is MAZATROL which is a 25 year old piece of junk. Consider this!! If your Integrex has a sensor failure, loose bolt, or a faulty connection then the warrenty can be used to get it fixed. The warrenty won't fix the software problems in this machine. The software problems are endless. I composed a thick document that details many, but it is so lengthy that posting them in this forum would be tedious.

 

Mazak is a very difficult company to work with after the sale. I waited almost a year to get a list of M-Codes that addressed the sub-spindle functions.

 

Get rid of this machine. Please.