thermal expansion coefficient

[Code_Breaker]

How many people are using thermal expansion coefficient to modify programs? What methods are you using?

 

For example, on a aluminum part 300 inches long, 36 inches wide and about 6 inch deep, at 68 degree the part is nominal, at 55 degrees, it is .047 smaller, or 299.953 oal ... Part are inspected at 68 deg.

 

We are thinking about heating/cooling the coolant either in the tank or by inductions ... has anyone done this? What is the best way to go?

 

Looking for information that has been proven by you ...

 

Thanks

[Guest CNC Apps Guy 1]

Worked in a shop where I had some 22. x 1. x .06 aluminum parts... length was ±.0001 @ 68˚.

 

The ONLY way we were able to hit it and hold it consistently was after we controlled coolant temperature. I was running a small enclosed mill which is orders of magnitude easier to do than on your machine.

 

What I've seen at Acramil was they enclosed (via plastic) the machine and used a swamp cooler to cool the machine's environment.

[Matthew Hajicek - Singularity]

Whatever engineer spec'd that tolerance better have had a darn good reason.

[Dave Ball]

I have used custom macros to comp for temp variations, but only on a few simple shaped features...

[Dave.L]

22 x 1 x .06 +/- .0001 on length, you probably had a .020 bow, how in the world did you measure these with any accuracy?

[gcode]

Temperature control is the only way to do it with a part that big..

The stock will have to be stored for a couple of days at a stable temperature before you even start cutting.. and you'll have to keep it at a

stable temperature between operations..

If you run one side at 75 degrees and the other

at 65, you're going to get hurt.

The simplest solution is to build an enclosure

around the machine. That's not cheap and not practical unless you've got some volume to

absorb the cost.

A cheap and dirty way would be to comp your X and Y gcode with a mass edit in CimcoEdit.

That could work in theory, but it would be real

easy to screw it up.

[Code_Breaker]

At the present, we are using SCALE X.xxxxx Y.xxxx

with .xxxx representing factors based on the thermal expansion coefficient of the part at the time of machining ... part temperature is monitored before, mid and end of operation ...

 

I have suggested a covering over the gantry, but Barnum and Bailey won't be in town until summer ... too late ...

 

Feel free to email me info on equipment, systems, etc at [email protected] ... I will present my findings after the first of the new year ...

 

PS ... we use Verisurf to inspect our parts and been in contact with Ernie ...

[Code_Breaker]

Does anyone know of a company that makes these heater/coolers for keeping the coolant temperature constant?

[pronc]

We use one from CoolJet and having great success.

[chris m]

Daikin makes all kinds of oil / coolant coolers; we use Daikin chillers on all of our gear gutting machines

[Guest CNC Apps Guy 1]

I've used Cool Jet with success.

 

quote:

---

...you probably had a .020 bow, how in the world did you measure these with any accuracy?

---

No bow, they were double disk ground on 4 sides. I got about a 10-12 finish on the ends (length) They were measured on a Zeiss, and the best part... out of the lot of 15, I used LESS than 50% of the tolerance on ALL of them.

 

quote:

---

Whatever engineer spec'd that tolerance better have had a darn good reason.

---

I know what they were for, who they were for, and why they were VERY important and I'll leave it at that.

 

[ 01-04-2010, 12:29 PM: Message edited by: CNC Apps Guy 1 ]

[Code_Breaker]

both CoolJet and Daikin maked chillers/coolers ... I could not find on their site that their make "heaters" ...

 

I am sure someone in the "snow country" or Canadians must use something the "heat up" their coolant ...

 

quote:

---

I know what they were for, who they were for, and why they were VERY important and I'll leave it at that.

---

I've had CNC Operators apply for a position saying they could hold .010 or even, .005 tolerance ... How about .0002? Some folks never workedd in an envirnonment where the temperature MUST be controlled ...

 

Once I had a GD&T .0000 true position on holes alignment for a "hinge-like" part that went on a fighter jet ...

[Ron Branch @ 5th Axis]

Most shops in the colder parts have climate control and keep their shops warm so they do not need coolant heaters. That is the problem with So. Cal it really never gets cold enough to warrant keeping a shop a certain temperature until you got work that does then most times it is hard going in that direction. I have used chillers on coolant, but never used heaters and think it is really going to come down to enclosing the machine to control the whole environments temperature the manufacturing of these parts is in. Did you get my email??

[Guest]

quote:

---

I've had CNC Operators apply for a position saying they could hold .010 or even, .005 tolerance ... How about .0002? Some folks never workedd in an envirnonment where the temperature MUST be controlled ...

 

Once I had a GD&T .0000 true position on holes alignment for a "hinge-like" part that went on a fighter jet ... [big Grin]

---

Yup and I bet .0000 had MMD attached to the dimension ;o)

[Jack Neelands™]

quote:

---

 

I am sure someone in the "snow country" or Canadians must use something the "heat up" their coolant ...

 

 

---

We usually set up a scrap block of steel for the next morning and take a quarter inch cut with the face mill and pile the chips up and huddle around them to keep warm..You can take some of these and break the ice layer over the coolant tank and drop them in to warm the coolant up....make sure you dont drop them near the pump....recommend you put them in the filter part for best results.

 

no seriously ...most of us dont have heaters on the coolant as we dont let the shops get that cold when working to tight tolerances.

[Tyler Robertson]

Most shops here just run CC for coolant anyway

[Guest CNC Apps Guy 1]

quote:

---

Once I had a GD&T .0000 true position

---

I love those. Saw one twice that was RFS... The Engineer and I had a hearty discussion about that. He eventually saw things my way and for the application .0003 True Pos. was agreed upon, fortunately without having to resort to violence.

 

Don't get me wrong, I was not overjoyed about having to hold that tolerance between 3 seperate operations, but the application dictated that necessity. I'm a reasonable guy. I just don't like random/default numbers in tolerances. Let the application dictate what the numbers should be.

 

JM2C

[Code_Breaker]

quote:

---

MMD attached to the dimension

---

+/-.0002 and part was aboiut 9.0" with three "hinges" with holes ... like a common door hinge ...

[Kevangel]

I build gages to 50 millionths (my share) and you have really watch the temp. of coolants contacting gages and the heat from your hand can be a big killer and I don't have all day (gotta get-R-dun) so I keep a pail of coollant at the exact same level of surface plate as I learned a pail on the floor can be cooler than 68 degrees. I check it with a thermometer...same with surface plate, I keep a thermometer on it...I have had gages rejected over 10 millionths and inspectors are wrong alot I have learned.

[Kevangel]

My Cincinatti i.d.o.d. grinder keeps moving for the first 2 hrs. of operation until the spindle finally hits operating temp. My Moore jig grinders have a heater built into spindle area that keeps that area heated to 85 degrees so it never changes overnite as that must be a max. number for it to warm up to. My shop is air condtioned to 68 degrees. If I had to make large parts to close tol. I would flood the entire part and machine base with 68 degree coolant and have extra pumps etc to do it. and attach. thermoneters to various areas to see its getting done and have the head at a constant temp. which might mean using a heater in that area like Moore does..

[Dave.L]

quote:

---

I build gages to 50 millionths

---

Kevin,

 

Are you talking resolution or accuracy? If you are talking accuracy, how are you measuring, what instruments? Just curious.

[Kevangel]

quote:

---

I build gages to 50 millionths

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

 

Kevin,

 

Are you talking resolution or accuracy? If you are talking accuracy, how are you measuring, what instruments? Just curious.

---

Dave,

 

Accuracy, and I use inspection level gage blocks and Mitutoyo height masters also (have 3)..so sometimes I use 2 stacks of blocks and average out the build-up.etc and have 3 height masters and will average out the readings on those when used (not for 50 millionths). Have 50 millionths Shadow gages for reading over the blocks which are nice. A cheapo way to do this also is a Mahr supremness 20 millionths comparator from MSC for around $350 I think. They are awesome. I also have lever type indicators with electronic box that go down to .0001 full reading across table. Nothing real sophisticated all surface plate inspection. Holes are jig-ground in and 50 millionths readout compares to the Moore lead-screw within .0001. as a double check. Plain old seat of the pants gagemaking. I have contiplated going to the next level.

[Dave.L]

Kevin,

 

I'm not arguing, but some would argue that the instrument should be some magnitude of higher accuracy then the measurements. That's why I ask.

 

It seems to be a topic that ends up "what ever you and your customer agree upon is the standard".

[Kevangel]

quote:

---

Kevin,

 

I'm not arguing, but some would argue that the instrument should be some magnitude of higher accuracy then the measurements. That's why I ask.

---

Dave,

 

My blocks are certified to 5 millionths if thats what you mean and the shadow gage easily reads closer than 50 mill as that spacing is around .12 of an inch. as an example. Most of my work has to go thru a certification house and or an fortune 500 co. million dollar plus inspection dept. I do maybe 50 gages a year to 50 millionths and always feel my read end pucker up. (they are all less than one inch thick)...I have almost no re-jections and people love to re-ject gages from what I hear as in "flunking everything" out of some inspection depts. The electronic indicator is graduated 5 millionths also.