Axis offset constant deviation, were to search the causes?

ricardo_gt · Jun 1, 2018

Hi people,
I am having a problem at a Gildemeister fabricated at year 2002 lathe:
After machine and correct the first part at the machine it is suposed to machine 50 parts continuously, without any correction on the offsets and the part staying inside the tolerances, even after part number 50...No way i can get that at OD and ID... all diameters go out of tolerance somewere from 11 to 13 part and continues to go away from nominal values. This happens as this: OD X axis comes down and X values decreses and ID values also comes down, diameter comes tighter and tighter. First i tried to warm up machining at least 50parts, correct offsets after and machine more 50 without touching offsets... did not worked. I changed the tools to new ones, all new, inserts and irons...the same.
It uses mist coolant and it are stainless steel cast parts, just needing finishing, 2.0mm (metric) plus all around.
The part have 60mm outside and 30mm inside, +-0.015 tolerance (metric) and 40mm long.
I am using carbide inserts.
It is the first time i have a machine that can not keep tolerances for long and i do not know were to search what or why.
I accept any advice. Thank you.
Best regards.

sinha · Jun 1, 2018

Both OD and ID decrease?

DMF_TomB · Jun 1, 2018

ricardo_gt said:
Hi people,
I am having a problem at a Gildemeister fabricated at year 2002 lathe:
After machine and correct the first part at the machine it is suposed to machine 50 parts continuously, without any correction on the offsets and the part staying inside the tolerances, even after part number 50...No way i can get that at OD and ID... all diameters go out of tolerance somewere from 11 to 13 part and continues to go away from nominal values. This happens as this: OD X axis comes down and X values decreses and ID values also comes down, diameter comes tighter and tighter. First i tried to warm up machining at least 50parts, correct offsets after and machine more 50 without touching offsets... did not worked. I changed the tools to new ones, all new, inserts and irons...the same.
It uses mist coolant and it are stainless steel cast parts, just needing finishing, 2.0mm (metric) plus all around.
The part have 60mm outside and 30mm inside, +-0.015 tolerance (metric) and 40mm long.
I am using carbide inserts.
It is the first time i have a machine that can not keep tolerances for long and i do not know were to search what or why.
I accept any advice. Thank you.
Best regards.

.
.
temperature of machine changing. some machining cycles create higher heat. for example high rpms
.
temperature change of tooling and eventually parts from cold coolant. and more importantly if parts are cold and measuring tools are room temperature the part will measure different when they get to room temperature. for example measuring a test cut and on a long bored hole the hole gets gradually smaller like .0005" smaller at end of bore from cold coolant. often effect stops near immediately if coolant turned off
.
tool wear from material being machined. i often see a wear in period on tooling, some stability, then a going dull progression. sudden tool failures can be from hitting slag or hard spots in metal. bue or built up edge on cutting edges creates false cutting edge and cuts different size. sometimes bue is semi stable or takes minutes to form and break off
.
i have also had parts change size when unchucked cause they were being distorted from the variable pressure of chuck pressure
.
obviously you hold a measuring gage too long in your hand and you will get greater calibration error. and if you set it down on a warm spot you get variations too
.
your coolant temperature can be changing during the day unless its temperature controlled. coolant that gets 5F or warmer during the day will create changes in size. coolant mist creates a chilling effect from evaporation.

ricardo_gt · Jun 1, 2018

sinha said:
Both OD and ID decrease?

Yes. Not the same amount, OD go off 0.060 and ID goes off 0.025, allways at 50parts. I have sent near 600parts to garbage trying it... i have to find a valid solution... the only solution i am thinking at the moment is to create a variable to up the values part by part, but that is not an acceptable solution to the client.

sinha · Jun 1, 2018

Then, it is not because of tool wear. Temperature rise can be a reason.
When the temperature becomes stable, no further change is expected.
One method can be to adjust offset every 30-40 parts.

DMF_TomB · Jun 1, 2018

ricardo_gt said:
Yes. Not the same amount, OD go off 0.060 and ID goes off 0.025, allways at 50parts. I have sent near 600parts to garbage trying it... i have to find a valid solution... the only solution i am thinking at the moment is to create a variable to up the values part by part, but that is not an acceptable solution to the client.

could be just tool going dull

Ox · Jun 1, 2018

"Thermal Growth" is the term.

Some machines are better/worse than others.

Not uncommon to have to babysit during warm-up.

-------------------

Think Snow Eh!
Ox

DMF_TomB · Jun 1, 2018

Ox said:
"Thermal Growth" is the term.

Some machines are better/worse than others.

Not uncommon to have to babysit during warm-up.

-------------------

Think Snow Eh!
Ox

.
cold coolant or cooling from coolant evaporation creates size changes literally every few minutes on many machines

ricardo_gt · Jun 1, 2018

DMF_TomB said:
.
.
temperature of machine changing. some machining cycles create higher heat. for example high rpms
.
temperature change of tooling and eventually parts from cold coolant. and more importantly if parts are cold and measuring tools are room temperature the part will measure different when they get to room temperature. for example measuring a test cut and on a long bored hole the hole gets gradually smaller like .0005" smaller at end of bore from cold coolant. often effect stops near immediately if coolant turned off
.
tool wear from material being machined. i often see a wear in period on tooling, some stability, then a going dull progression. sudden tool failures can be from hitting slag or hard spots in metal. bue or built up edge on cutting edges creates false cutting edge and cuts different size. sometimes bue is semi stable or takes minutes to form and break off
.
i have also had parts change size when unchucked cause they were being distorted from the variable pressure of chuck pressure
.
obviously you hold a measuring gage too long in your hand and you will get greater calibration error. and if you set it down on a warm spot you get variations too
.
your coolant temperature can be changing during the day unless its temperature controlled. coolant that gets 5F or warmer during the day will create changes in size. coolant mist creates a chilling effect from evaporation.

Thank you for your reply. Yes, it is true. And i have a major problem with it: if i say all those to my boss he wil argue that it is me who do not know how to do the work lol Even, all those variables are controlabe, also the machine, just need to know how.
Is there anything specific for looking at temperature changes on the machine? Busbars? Machine axis engine guides? other? Thank you.

ricardo_gt · Jun 1, 2018

Ox said:
"Thermal Growth" is the term.

Some machines are better/worse than others.

Not uncommon to have to babysit during warm-up.

-------------------

Think Snow Eh!
Ox

I can not send the machine to the junk yard! I wanted to, but i cant...lol Is there anything i could do about it? Thank you.

dstryr · Jun 1, 2018

ricardo_gt said:
Yes. Not the same amount, OD go off 0.060 and ID goes off 0.025, allways at 50parts. I have sent near 600parts to garbage trying it... i have to find a valid solution... the only solution i am thinking at the moment is to create a variable to up the values part by part, but that is not an acceptable solution to the client.

So you have nailed down to the machine will run about 50 pcs before moving enough to be out of spec... and instead of adjusting every 40 pcs you keep doing same thing and getting same result of bad parts? (600 pcs ??)

DMF_TomB · Jun 1, 2018

heck you ought to try precision machining when somebody got outside door open in the middle of winter
.
fancy cnc with temperature compensation and coolant system with 1F temperature control cost more money
.
$15,000 car not going to perform like a $150,000 car. why would shop equipment be any different ?

ricardo_gt · Jun 1, 2018

dstryr said:
So you have nailed down to the machine will run about 50 pcs before moving enough to be out of spec... and instead of adjusting every 40 pcs you keep doing same thing and getting same result of bad parts? (600 pcs ??)

Yes, because the machine needs to be validated to this job, without touching it, 50 parts in a row. Toyota uses the same method. This is an important client and they made the rules... i am just trying to run with them, some how.

DMF_TomB · Jun 1, 2018

ricardo_gt said:
Yes, because the machine needs to be validated to this job, without touching it, 50 parts in a row. Toyota uses the same method. This is an important client and they made the rules... i am just trying to run with them, some how.

.
sometimes boss get upset at worker as he thinks worker needs to try harder. but often its best to let somebody else try the job and fail to do better it is best way for boss to see the first worker was not as bad as he thought.
.
sometimes you have to leave to be appreciated. thats why journeyman machinist needs to journey or leave or work on other machines or places. quite often you are welcomed back after a year and then they realize you aint as bad as they thought
.
same as programmer running machine for a day to find he is having just as many problems as regular operator. sometimes they think worker doing something wrong til they try and fail for themselves. i try real hard not to laugh at the programmer when he fails

dstryr · Jun 1, 2018

ricardo_gt said:
Yes, because the machine needs to be validated to this job, without touching it, 50 parts in a row. Toyota uses the same method. This is an important client and they made the rules... i am just trying to run with them, some how.

Sure 50 pcs in a row on a 2018 machine... should be no problem.
You are working on 2002 Gildemeister ... lol

DMF_TomB · Jun 1, 2018

dstryr said:
Sure 50 pcs in a row on a 2018 machine... should be no problem.
You are working on 2002 Gildemeister ... lol

stainless steel cast parts.......yeah right no problem. he should easily do a million parts before changing tool. next some will say at 3000 sfpm too

dstryr · Jun 1, 2018

DMF_TomB said:
stainless steel cast parts.......yeah right no problem. he should easily do a million parts before changing tool. next some will say at 3000 sfpm too

Rough and finish tools... I dont see the problem?

Bobw · Jun 1, 2018

It uses mist coolant and it are stainless steel cast parts

You might be pissin up a rope... 600 scrap castings would have paid for at least a down payment on a
machine that has flood coolant.

I can see the ID and OD both going off in the same directions due to thermal growth, but unless
something is really whacky, I don't see them moving different amounts.

ricardo_gt · Jun 1, 2018

It will be more important the 250000 parts order than the 600 or 1000 scraped... rigth now i need to solve the problem at this machine to guarantee the order.
Ok, so thermal growth could be the problem... the machining method is very fast, mitsubishi tech came here and the values to machine this part are (by mitsubishi tech), 3000rpm and 0.0154ipr for roughing and finishing with 3000rpm and 0.0078ipr. This produces a lot of hot, but each part is finished on 33seconds included load and unload from machine. Machine have a high pressure coolant pump to feed the coolant exit at tools and a low pressure pump to feed 2 other "showers" near the chuck... not enough? How can i solve it better?

Ox · Jun 1, 2018

ricardo_gt said:
I can not send the machine to the junk yard! I wanted to, but i cant...lol Is there anything i could do about it? Thank you.

Yes, if you need to doo a 50 pc run w/o adjusting - just for the cert, then what you doo is to let the machine cut air for 2 hours (?) before you start. Doo not use a "warm-up routine", but rather your actual part program. The warm-up routine could overshoot your "in process" temps.

Then you can run your 51 pcs (the first one will likely be off) and you may have to battle tool wear, but thermal expansion should be OK.

I would NOT expect to run production this way, but if you need to doo it for the paper jockies, this should work.

Doo not use HP coolant on this if you don't need it. The HP pump will heat up the coolant and turn your machine into a sauna, and will add to your thermal growth issues.

---------------------

Think Snow Eh!
Ox

Axis offset constant deviation, were to search the causes?

Aluminum

Titanium

Diamond

Aluminum

Titanium

Diamond

Diamond

Diamond

Aluminum

Aluminum

Diamond

Diamond

Aluminum

Diamond

Diamond

Diamond

Diamond

Diamond

Aluminum

Diamond

Similar threads