machinist@lists.ibiblio.org
Subject: Machinist
List archive
[machinist] Contolling expansion on close tolerance fits (excellent and important thread)
- From: Lawrence London <lfljvenaura@gmail.com>
- To: machinist@lists.ibiblio.org
- Subject: [machinist] Contolling expansion on close tolerance fits (excellent and important thread)
- Date: Fri, 28 Jul 2017 13:50:33 -0400
Contolling expansion on close tolerance fits (excellent and important thread)
http://www.practicalmachinist.com/vb/general/contolling-expansion-close-tolerance-fits-338190/
-
07-27-2017, 10:37 AM #1
Stainless
- Join Date
- Aug 2004
- Location
- Richmond,VA,USA
- Posts
- 1,876
- Post Thanks / Like
-
- Likes (Given)
- 133
- Likes (Received)
- 611
Contolling expansion on close tolerance fits
Let me start by saying that I do not have a ton of experience with very close tolerances. Most everything I do is +- .001". Yesterday I was turning a piston for a friend of mine who is in an aero club and taking some pretty heavy cuts on the bar stock (.125") or better. Needless to say when I got close enough to start making some precision cuts (+- .0002 or so) the blank was damned hot. I know there are charts and what not to calculate expansion but I am wondering what you guys do in the real world. I don't have a very precise method of measuring how hot the blank is and I would not feel comfortable trying to compensate my depth of cut using the expansion formulas anyway. Do you walk away from things and let them cool down on their own before starting your precision cuts? I have a vortex cooler (which I hate because of the noise and air consumption) and if you don't get the core cool the residual heat will cause more expansion while making your final cuts and you end up with a taper. The only way I have found around it is to walk away and do something else and wait for the stock to equilibrate to my shop temperature which can take over an hour or more. Any suggestions? Thanks very much.
-
-
-
07-27-2017, 10:45 AM #2
Diamond
- Join Date
- Jan 2005
- Country
- CANADA
- State/Province
- Saskatchewan
- Posts
- 8,875
- Post Thanks / Like
-
- Likes (Given)
- 917
- Likes (Received)
- 2713
Either use coolant for a quick cool down, or get a fan placed nearby to speed up the cooling rate. Then go do something else, I do this often enough, as I've got lots of other machines to work on and things to do. I hate open coolant on a manual machine, not just so much for the mess, as for the fooling around maintaining the coolant (skimming it, etc). Coolant goes bad fast on a manual lathe doing all sorts of work, sometimes sitting for days undisturbed. I got sick of it, so I prefer just dry machining.
Or take the part out and cool it in a bucket and set it back up. Do this while you've got one light semifinish cut remaining so that it can make setting up again easier to do. -
-
07-27-2017, 10:47 AM #3
Diamond
- Join Date
- Dec 2008
- Country
- UNITED STATES
- State/Province
- New York
- Posts
- 6,965
- Post Thanks / Like
-
- Likes (Given)
- 0
- Likes (Received)
- 1865
my shop has a 2000 gallon coolant system which is temperature controlled to 1 degree F so just using flood coolant will equalize temperature. if you have no coolant the air vortex cooler can help but they actually can make part too cold. you might be better off just waiting but often you would need to wait 1 hour per inch of thickness. waiting 2 hours for 2" thick part is a long wait.
Originally Posted by crossthread
.
i sometimes bore holes to +/-.0001" tolerances. i have to wrap the indicating bore gage with a rag to insulate it from the heat of my hand.
.
different metal alloys expand at different rates as temperature changes. aluminum part will grow bigger faster than a steel micrometer as temperature rises. usually better to measure parts at 70 degrees F. some shops use 68F. that means you need air conditioning in the summer -
-
07-27-2017, 01:22 PM #4
Aluminum
- Join Date
- Apr 2017
- Country
- UNITED STATES
- State/Province
- California
- Posts
- 68
- Post Thanks / Like
-
- Likes (Given)
- 5
- Likes (Received)
- 30
Expansion coefficient of steel is 6.5 micro inches/inch/degree F.
A one inch shaft that is 20 degrees hotter than room temp. has grown 130 micro inches.
.0001 inches is the same as 100 micro inches. -
-
07-27-2017, 01:28 PM #5
Diamond
- Join Date
- Dec 2000
- Location
- Bremerton WA USA
- Posts
- 10,424
- Post Thanks / Like
-
- Likes (Given)
- 40
- Likes (Received)
- 3644
Excellent point to discuss. There is no path to mechanical measurement without consideration of temperature differences and rate of thermal expansion.
Example of rates of thermal expansion in inches per degree F
Most ferrous alloys about 6.5E-6
Aluminum alloys 9.5 to 11.2 E-6
etc. Here's a handy table:
Thermal Expansion of Metals
All linear measurement is referred to 68 degrees F (20 degrees C), a world standard. Multiply your measurement by its rate of thermal expansion times the temperature difference ni degrees F. Metricoids, use the rates for the Metric world..
Post the expansion chart in your shop near near the surface plate or what passes for your inspection station.
You have to consider the temperature of the measuring instrument as well as the temperature of the part. Most measuring instruments (mikes, bore gages, etc are made of steel. If you handle a 4-5 outside mike and warm it 20 degrees F the frame will expand about 0.0006 affecting the accuracy of your measurement.
Another item to keep handy is a thermometer. For may years, I've used a dial type pocket thermometer like this
Taylor 5989N Stainless Steel Instant Read Pocket Thermometer
But now we have these remote reading infra red gadgets
Taylor 5989N Stainless Steel Instant Read Pocket Thermometer
which are very handy but may be plagued with petty emissivity errors.
There is no convenient path to precision in the home shop without consideration of temperature and thermal expansion and the necessary math for calculating its effects in precision measurement, clearances, interferences, etc.
Story: In my chip making days I was a specialist on large machine tools. One day working on a 10 ft vertical boring mill on a ring involving an inside diameter of 5 feet or so, I needed an inside mike to measure it. I had one prepared ahead of time, ready for pick-up in the tools room's calibration lab. I sent my apprentice to fetch it and to handle it using thick wads of shop towels to insulate it from the heat of his hands. I'd already instructed him on the effect of temperature, walked him through the math of calculating expansion, why it was necessary to handle the inside mike to avoid uncontrolled heat transfer into it, etc.
Imaging my shock when he returned down the long aisle not only bare handing a calibrated inside mike but using it like a walking stick, tapping the measuring tip on the concrete with every stride.
Sometimes outrage outstrips the available profanity. I had to tell the boss there would be a three hour delay for re-calibrating the inside mike and why, suggesting a month in the sand blasters or tank cleaners (nasty jobs in a shipyard) as a reward for my errant apprentice. He got a week cleaning the chip troughs of the big shaft lathes which, in fairness, was at least as nasty. My apprentice took his punishment in good heart and in time developed into a fine machinist.
Notice I tell no stories on myself. My apprenticeship was free of similar lapses. Trust me. -
-
-
07-27-2017, 01:37 PM #6
Diamond
- Join Date
- Dec 2008
- Country
- UNITED STATES
- State/Province
- New York
- Posts
- 6,965
- Post Thanks / Like
-
- Likes (Given)
- 0
- Likes (Received)
- 1865
they teach millwright on a motor to pump coupling alignment.
.
each end of motor and each end of pump can be a different temperature. sometimes guessing which end is higher is not easy. basically one end can go up forcing the other end lower.
.
i rarely find expansion calculations worth it. easier to just machine with everything at room temp
.
also i once had to figure out why plastic film machine appeared out of alignment. rollers were in alignment. turned out they installed a air vent on one side of the 60" wide web of plastic and it was causing one side to dry more and it shrinks when it dries. sometimes moisture cause plastics to change size as they get wet or dry too. -
Limy Sami liked this post
-
07-27-2017, 01:40 PM #7
Diamond
- Join Date
- Dec 2008
- Country
- UNITED STATES
- State/Province
- New York
- Posts
- 6,965
- Post Thanks / Like
-
- Likes (Given)
- 0
- Likes (Received)
- 1865
. Originally Posted by Forrest Addy
.
yes i have often had a inside mic change size .0005 to .001" just holding it in my hands too long. also have to watch where you set it down. often the top of a cnc control panel is warm. i usually have a tool cart near machine so all measurng tools are not sitting on something warm -
-
07-27-2017, 02:01 PM #8
Aluminum
- Join Date
- Oct 2009
- Location
- Seattle
- Posts
- 65
- Post Thanks / Like
-
- Likes (Given)
- 0
- Likes (Received)
- 3
I would think the cooling could be significantly accelerated by wrapping the part with one or two flexible ice/gel packs, held in place with something stretchy like an Ace bandage, or perhaps even rubber bands.
Mike -
-
07-27-2017, 02:14 PM #9
Diamond
- Join Date
- Oct 2005
- Location
- erie,pa
- Posts
- 6,677
- Post Thanks / Like
-
- Likes (Given)
- 13269
- Likes (Received)
- 3065
Rapid cooling has to be done carefully, same as when heating parts up, it takes awhile for the center
to soak up (or get rid of ) heat. -
sfriedberg liked this post
-
-
07-27-2017, 02:35 PM #10
Diamond
- Join Date
- Dec 2008
- Country
- UNITED STATES
- State/Province
- New York
- Posts
- 6,965
- Post Thanks / Like
-
- Likes (Given)
- 0
- Likes (Received)
- 1865
. Originally Posted by digger doug
.
facemill at -100F the cold being used to help loosen screw holding facemill on.
.
that tool holder was cold easily for over a hour. notice the frost on it -
-
07-27-2017, 02:54 PM #11
Diamond
- Join Date
- Apr 2004
- Location
- Modesto, CA USA
- Posts
- 5,049
- Post Thanks / Like
-
- Likes (Given)
- 0
- Likes (Received)
- 656
.
that tool holder was cold easily for over a hour. notice the frost on it[/QUOTE]
Is that water ice from the air?
Bill D -
-
07-27-2017, 03:21 PM #12
Diamond
- Join Date
- Oct 2005
- Location
- erie,pa
- Posts
- 6,677
- Post Thanks / Like
-
- Likes (Given)
- 13269
- Likes (Received)
- 3065
So what is your point for/against my "soak time" posting ? Originally Posted by DMF_TomB
-
-
07-27-2017, 03:40 PM #13
Stainless
- Join Date
- Aug 2004
- Location
- Richmond,VA,USA
- Posts
- 1,876
- Post Thanks / Like
-
- Likes (Given)
- 133
- Likes (Received)
- 611
Thanks for the replies. A few of you have mentioned the exact question I have. I can cool the part but how cold is it really? Is the core cool enough so residual heat doesn't expand it further during the final cuts? Is it too cold? Forrest, I hear what you are saying as well and agree in a shop where folks work to tenths on a regular basis that having the coefficients of expansion handy and the instruments to measure the temperature accurately would be invaluable. I think most of us know to handling a set of mics by the insulating pads provided for just this purpose is a good start but dealing with all the other variables must be a nightmare for folks who work in tenths all day long. We live in a rubbery world. Shafts deflect from their own weight, gauges droop, stock grows from the heat of machining etc. In my case the few times I need to work to that level of precision I have the luxury of walking away from the job and letting everything settle down. I was just curious how people who do this on a regular basis deal with it. Thanks again.
-
-
07-27-2017, 03:40 PM #14
Diamond
- Join Date
- Jan 2007
- Location
- Norfolk, UK
- Posts
- 15,629
- Post Thanks / Like
-
- Likes (Given)
- 11037
- Likes (Received)
- 10575
Machining a lot of (mainly) plastics for a living, thermal control is vital, as plastic moves a lot with little heat AND takes for ever to cool down, so for me the only answer is to control the temperature right through the process.
>From allowing at least 12 hours to reach room temp (usually around 20C / 70F) and avoid putting the heat in to the part when machining, which can mean lot's of coolant, and tricks - like cool box gel chill packs in the coolant tank etc etc.
And IME the same goes for metal!
Like DMFTomB I've found calculations, at best unreliable, at worst VERY expensive
-
-
07-27-2017, 08:31 PM #15
Diamond
- Join Date
- Oct 2012
- Country
- UNITED STATES
- State/Province
- Virginia
- Posts
- 6,566
- Post Thanks / Like
-
- Likes (Given)
- 2922
- Likes (Received)
- 3087
ROFL! Originally Posted by Forrest Addy
Horrifying, I am sure, "on the day". Funny as Hell a half-century later, though.
BTW ... he was lucky to have been found "salvageable".
One thing the USWA and Management didn't fight over was the ones we agreed were NOT salvageable - even dangerous fools. They'd be "blueprinted" right out of our zone with nary a grievance filed nor supported, ass-whupping offered in lieu of if they pressed their luck with a pissed-off Shop Steward.
EG: Make the part to a print. Find it failed every spec.
Two different prints.
Of course.
Wasn't what one would exactly class as a "sweetheart" Union. But we all had to stay ALIVE to eat.
Half a ton of salvageable HRS diverted and a bit of machine-time forever-lost was waay cheaper and faster than tolerating perennial f**k-ups and screwing with paperwork in those days.
Back to basics:
Absent Air-Con or enough coolant, that same company wasted a HUGE amount of time and shop air cooling large shafts to check bearing-fits. Then still had to do some experienced guessing.
If that was not good enough, the work might be left in the machine for the rest of the shift, card-carrier assigned another machine, another task.
We were on three-shifts, so it would simply be finished sometime next shift. -
-
07-27-2017, 09:36 PM #16
Diamond
- Join Date
- Dec 2000
- Location
- Bremerton WA USA
- Posts
- 10,424
- Post Thanks / Like
-
- Likes (Given)
- 40
- Likes (Received)
- 3644
Don't know how it is now but when I started my time there were 1600 applicants competing for 130 apprentice slots. Not many knuckleheads survived that kind of selection. A little more than half survived the four years of OJT, schooling, work requirements, and the pressure to perform. Those who survived into their second year when I got 'em were pretty much the cream of the crop and worth a little tough love to bring into the brotherhood. Once in a while there was a Dennis Mallet or a Bob Sogelburg. Drunks and druggies failed the cut sooner or later. Originally Posted by Monarchist
Many graduate apprentices after seasoning on the tools went on to management and technical staff. .It's difficult to quantify the worth of an effective demanding apprentice program but the number of formerly great industrial giants who abandoned their in-house employee development programs later unaccountably bit the dust is astounding, infect its the blueprint of America's latter-day Industrial devolution.
It takes a generation for the benefits of trade and technical development to have an effect on the competitive edge of a company. Once an employee development program is terminated, it takes a generation for the organization to lose its technical vigor. MBA's and finance people will never understand the relationship between competent committed employees and the ability of their company to be competitive. They associate only with their peers, think "worker bees" are interchangeable, live from quarter to quarter, setting themselves up for their Golden Parachute. Suck dry, diversify, and die may not be a good business model but if an executive times it right he's out with his accumulated loot before the final accounting.Last edited by Forrest Addy; 07-28-2017 at 12:18 AM.
-
-
07-28-2017, 12:26 AM #17
Titanium
- Join Date
- Oct 2010
- Country
- UNITED STATES
- State/Province
- Oregon
- Posts
- 3,881
- Post Thanks / Like
-
- Likes (Given)
- 2240
- Likes (Received)
- 1179
To Forrest's point, the value-destruction-by-MBA process was well underway in the US a generation ago. During about three decades, 1960s through 1980s, just about every industrial brand of any significance was purchased as part of a conglomerate portfolio, frequently "just because" and based far more on low-interest financing than on any coherent product planning. In fact, the whole principle of the diversified conglomerate was to get away from a simple, natural family of products.
When the financial wheels came off most of these money-based sacks of miscellaneous product lines and brand names, it was pretty common for the remnants to be simply discarded, not even offered for sale or management buyout, to simplify the accounting writedown on the losses. By the 1990s, the whole craze for the industrial conglomerate was over. But a huge swath of American industry had been devastated (much worse than merely decimated) in the process. I view this as extremely sad, and after a few beers, as downright criminal.
A few noteworthy conglomerates still survive today. The usual examples are GE, which differed from outfits like LTV by not indulging in over-leveraged borrowing. And Berkshire-Hathaway, where Warren Buffett has avoided the temptation to meddle with the actual operations of the companies he owns. -
-
07-28-2017, 12:59 AM #18
Diamond
- Join Date
- Jan 2007
- Location
- Flushing/Flint, Michigan
- Posts
- 5,808
- Post Thanks / Like
-
- Likes (Given)
- 225
- Likes (Received)
- 4646
You seem to have your head wrapped around the biggest problem. Originally Posted by crossthread
And why many have had bad luck using temp corrections measuring the temp with contact or non-contact means.
Your part is hot right after roughing, but how hot is the core?
Cooled on the skin by a fan but how many BTUs sitting inside? Remove the fan and generally the measured part temperature increases for a while and then backs off. Same with ice packs, chilled water, fridge or freezer or such.
Best to allow some serious soak time unless you have large production runs and can build your own comp table from experimental results on your process.
Water pulls out heat faster than air but you still need to wait long after the skin measures room temp.
In fast crazy low temp cooling a part with frost on the outside can be well above room temp on the inside so any "straight" temp/size comp will be wacky. Such a part then exposed to air will be cold to the touch for quite some time but the size will not match the temp coefficients of expansion of the easy math.
The rough part is probably not as hot as you think it is, forced cooling gives a backward sense that the part is cooled.
If you are not making 10's or 100's of thousands or have a lot of experience with this size and shape part and material...... time is your friend.
Bob -
-
07-28-2017, 01:58 AM #19
Stainless
- Join Date
- Nov 2013
- Country
- UNITED STATES
- State/Province
- Utah
- Posts
- 1,334
- Post Thanks / Like
-
- Likes (Given)
- 403
- Likes (Received)
- 687
Curious, what is the formula for calculating the time required for a thermal change to reach the core of a part, or distance, based on area of material? Just a curiosity.
I don't bother trying to comp. for the T' coefficient, but I run low volume compared to most. So I just let it sit until I feel the spirit, or whatever.
R -
-
07-28-2017, 01:01 PM #20
Diamond
- Join Date
- Dec 2008
- Country
- UNITED STATES
- State/Province
- New York
- Posts
- 6,965
- Post Thanks / Like
-
- Likes (Given)
- 0
- Likes (Received)
- 1865
. Originally Posted by Forrest Addy
apprentice programs get cut when the accountants try to cut short term costs ignoring long term costs.
.
its like laying off people who calibrate tools and also throwing out many precision tools cause it cost too much to check calibration on the tools.
.
i remember cleaning equipment and getting rusty water out of the sink for a good 15 minutes and calling to have a water filter installed. i get told we dont install water filters we remove filters. why? cause then we got to pay people to change the dirty water filters. odd logic seems to be a modern management practice
.
its like not paying to have oil changed on a car. sooner or later it will cost a lot more to not have changed the oil. trouble is often a boss figures he wont be around when things eventually fall apart
Reply
Originally Posted by Bill D
Bill D[/QUOTE]
frost forms from air within a few minutes. point is when mill cooled it took easily over a hour to get back to room temperature
- [machinist] Contolling expansion on close tolerance fits (excellent and important thread), Lawrence London, 07/28/2017
Archive powered by MHonArc 2.6.24.