I can remember running Teflon parts that we had to let sit in inspection for at least 30 minutes before determining the difference in size between what we measured on the floor. We would then have to do this about every 2 hours in order to consistently track the discrepancy and make accurate parts. Great info to be aware of; especially if you are machining tight tolerance parts!
@@danielgarst2578 Had the same thing happen with Delrin parts. We had to give our parts an expiration date. We told our customer that after a month on the shelf we were no longer responsible for replacing parts that have changed. We have AQL lot data that shows our parts were in print when they were shipped.
@@brandons9138I use to run a type of delrin that had to sit over night. I think it was delrin 150 that would move. The tolerance weren't tight but it was enough to move them out of spec. Stuff smelled horrible when cutting it and played havoc on coolant pumps.
Thermal Expansion of Spindles and the Machine, can be mitigated by a glas scale (german: glasmaßstab), that runs the complete lenght of the axis. If you have a propperly mounted glas scale in your machine, the only thing you need to care about ist the thermal expansion of your part, because your machine is true.
Each CNC Mill in our shop is equipped with an coolant temparature control system. Usual, when mashines running for 24 hours, all components are getting warmer, the coolant too. The control system keeps the coolant at same temparature, each time. While the mashine is running all parts of the mashine get flushed by coolant, so the mashine itselfe get cooled down. That make sure, we can hold the tight tolerances of all materials we cutting. For example, lot of Bores we have to produce, has tolerances bitween 0,03 mm and 0,0125 mm !
I work as a service technician and had problems with a customer over a year. They were making housings for big gearboxes out of aluminium, the biggest side was around 1.6m long. They had positioning tolerances of +-0.01mm over a 1,2m distance (convert it to your units yourself) in a not temperature controlled workshop. To make it worse there was a roof window through which the sun was shining exactly on top of the machine. Temperature differences at the bottom of the machine were in the range of 10°C up and down over a day at the control panel, at the top it was close to 20°. One of those parts was in the machine for 4 days or so until they were finished with it. The customer constantly claimed the temperature compensation of the machine is not working properly, even though i told him endless times it´s not the machine but the part that is the problem (even showing him the distance variations they had compared to the temperature and they lined up perfectly). This went on for a long time until we finally got to the point to get a guy with a laser in to double check all the machine accuracy. We had deviations of the machine positioning accuracy measured over the course of a day of less then 0,005mm, and that is over a 2.5m travel. Only then the guy finally believed me that it is the part. We mostly solved the problem by installing a proper coolant temperature control unit, and then had them shower the part in lots of coolant the whole time so it would stay at the same temperature over the day.
I did not think much of the effect temperature had on machining , until I went for my 2day trade test to qualify as a tool jig and diemaker. Day 1 I machined some components of the press tool including the die block . My punch was fitting perfectly on the die , with the right clearance to achieve a blank without burrs. The next morning when I came in the punch was not fitting. I had to re-machine to size again.
I work in the plastic injection molding biz.. Automotive instrument panels ect… Our core/cavity steel is the size of an F150 … So many possibilities to get it wrong lol… I’m always impressed we can hold the tolerances we do.
I was in that field also and our inspection area (not a room) was climate controlled whatever the climate was outside it was warmer than that inside ...
Unfortunately, I wish I had seen this video six months ago. I learned this information by the Qtsi method ... I had to dispense with four pieces of aluminum because of the thermal effect 😑 Thank you mr jessie for those priceless informations 🌷
i once did the montage of a galvanized steel roof crown, the engineer of the project did not account for heat expansion on the roof crown. they where laid edge to edge in 40-170m long lines using 3m segments, laid in the late autum. by the following summer we had an emergency repair, since the screw heads started to spring off, the solution was a 3mm gap between each 3m segment that taught me about the properties of heat expansion between summer and winter for the same piece of steel.
We had some aluminum extrusion that was pulled from outside in winter, immediately cut to size but a few days later we started running them and it had expanded an 1/8 inch
I'm working in a shop with a lot of holes in the bulding, and no thermal regulation at all. At this winter, with all those savings on heating, it's so cold inside the shop, that my lathe's door handle, and any other metallic objects can do a frostbite to my hands. And vice versa - when it's 30 degrees by Celsius at summer - a coolant turns into a warm water, if you are doing some non-stop performance cutting during the day. So, parts at this shop can change their size just because of season change xD
Mate. I'm suffering that right now. I have no choice but to work with gloves, long sleeve wool shirt, jumper and a softshell jacket. Pretty damn awful. Thank god the parts I'm doing are nylon with +-0.5mm tolerances haha.
Reminds me of our old lathe operator. Had some low tolerance parts ( H6 ) ~1500mm long. He always roughed em one day and finnished em the other. Our cheffe then once went "reee that not fast enough!" Since he learned that stuff too he thought he's doing the next ones himself and faster at that too. All scrap because around 0,01 to 0,015mm to small in diameter. 2 days of machinetime down the sewer, several thousand € down the drain and a VERY unhappy customer! Thermal expansion learned the hard way. Copper is also VERY interesting material in that regard lol.
Materials dont expand equally on all sides, a shaft will grow more in length than it will in diameter if it is heated evenly. But it will still grow or shrink predictably, but you mostly just learn that by experience. Thanks for sharing, great video. Charles
@@ullrikegablerthat is if you have a cube The bigger your part, the more it will expand. Say you have a 50mm x 100mm x 30mm Block of aluminum Which has a coefficient of 0.0000235 1/°C And let's say a temperature drop tdif of -20°C Ldif = L x a x tdif 50mm x 0.0000235 1/°C x (-20)°C = -0.024mm 30mm x 0.0000235 1/°C x (-20)°C = -0.014mm 100mm x 0.0000235 1/°C x (-20)°C = -0.047mm So your part will shrink .047mm in length, .024mm in width and .014mm in depth.
expansion of the ball screw won't be much of an issue most of the time because almost all modern machines measure the table movement directly with a linear encoder. however, heat spreads and if you have temperature gradients within structural machine parts they can warp in all kinds of ways.
I've been machining for a while but this has me thinking. If we could get the temperature of the part in real time and have the controller adapt the code being read ahead of time to compensate for the expansion over time to ensure everything lines up perfectly. That could possibly eliminate the issue no matter the temperature of the shop/part. Helix cutting out 9 holes before boring, on the 9th hole it would compensate say .0005 on the ID automatically as a example.
Machining a 40' tube when it was 100+ F (summer in Texas and shop is NOT climate controlled). Come in the next morning and the damned thing had shrunk almost 1/8" in length after it sat all night in considerably cooler temps. And yet no one will listen/understand when I try to explain about thermal expansion. These are the same people that think there is no linear movement when interpolating a hole.
Ok so it was machined at 80 and was in spec. But the inspection room was 68 so the part was out so you honed it to size. What happens when the customer gets it and is installing it in another 80° setting? Then your honed hole will be too big
Thought the same thing. Unless all the parts are speced in the same testing facility or there were testing standards across facilities then honing it doesn't matter. I'm not a machinist but it seems like common sense.
Yup, when chasing tenths, you've got to scrupulously dot your i's and cross every "t". If you fail to pay attention to details then the details will fail you and your parts.
Amazing how you guys measured with vernier inches imperial that fast im not familiar with but metrics in NZ I make Termination muzzle brake T1 to T7 Various sizes and threads compared to you guys superior work 😊
I don’t necessarily think heat is a bad thing. I think it’s the drastic temperature changes that can be catastrophic. I mean you can adjust your offsets over time but a sudden temperature change can catch you off guard. This past summer the AC was out in my building so we had to be mindful of our runtimes
Exactly, when my coworker is calibrating the geometry of a machine he always wants to know if someone has to open the rolling gate on that day because just the cold air from outside coming in can make you get inconsistent readings
Since 3 years it´s my biggest pain at work when temperatures outsides are dropping below 10° C because my company is so uncaring that they ignored the leader of the milling group as well as the leader of the turning group and went and placed a DMG CTX Beta2000 TC in a workshop that regularly has the loading door wiiiiide open to get in rawparts for 30 minutes + (longest was 94 minutes of enduring 8° Celsius INSIDE the shopfloor (drop of about 13-15 °C) . I can watch my machine shrink in the temperature diagnostic screen and am forced to compensate awful lots in X & Z-axis (most I had to correct so far was up to 0,18mm´s /0,00708 inches ) on parts that have straightness+roundness+coaxiality+concentricity+flatness tolerances of 0,03 to 0,05mm ( 0,00118 to 0,001968 inches) to not continously make wasteparts and that costs tons of time and nerves . Once that loading door finally closes I need to compensate in the other direction because it slowly gets warm again *sigh* Welcome to machining in Austria :( No change in sight so far ...
The formula is the same for imperial or metric... Your coefficient for each material is what is different depending on which unit system you are using. I.e. if you work in metric, use the metric thermal expansion coefficient.
If I were to order some high-precision parts from you that were made of…oh, let’s say aluminum, which has a really high coefficient of thermal expansion…would it come with a document that says what temperature it was made at? Next question: do you have a way to calculate acclimatization times? Some of us keep our materials in different climate condition than we work them in.
QC rooms should let parts sit until they're at a reasonable measurement temperature (due to being nearly ambient can take days/weeks). Simple parts you can easily do transient heat transfer problem for acclimatization, talking square, rectangles, etc. Otherwise you'd want to do it with a simulation software.
@@ScuffedEngineer I was more thinking of acclimatization of raw stock. I, like a LOT of other people, store my materials in a place that has only heat but no AC. If I was working four-inch plate steel stored in an open-air rack in the Texas heat and I bring it into my air-conditioned machining room, how long is it going to take to get the steel from 120 degrees to 75 all the way through?
@@jmowreader9555 I mean, I'm not doing the math for you, grad students don't have that time. It depends on the type of steel, the volumetric dimension of the stock (never said if it was square, rectangular, thickness just specified 4" plate). Also depends on what you consider all the way through and what temperature variance is negligible. You can look into transient heat transfer and the FTCS method for quick and dirty calculations, there might even be some code or a program out there for free. That being said, with that temperature variation I'd say not quick unless it was say .01" thick copper.
Same with stresses in material, brass part. Faced second op, loosened vise, bent like a banana, god that sucked. I managed to fix it tho phew, some things, never an issue, then boom!! Failure
I remember whilst at college we had a dare to go up to the girls in the local bar and explain the thermal coefficient equation and the winner was the girl that lasted the longest We made math entertainment
Thermal expansion is nothin new to those that have worked with anything that has been through nitriding! Would constantly see 20+" stem valve length's come back at least .01 longer than prior.
i know about it , you trying to make technology to write program , run it make few parts pristines, and then operator came, bang and machine is out of order.... Boom titans. cool video .but iam alwaays smiling when you talking in inches . ! Metric rules man :) .
We installed a 4 inch PVC conduit on a roof 100 (or so) ft long. worked up a BOM. my coworker and I were curious as to how much exactly and calculated the length summer to winter and decided to add a second expansion joint and anchor it in the middle of the run instead of each end and joint in the middle.
They still do that with automotive ring gears on flywheels. Rarely have to heat them red hot, but they still do just fall on when at the right temperature
@@bigboybuilder considering how easily they’re chewed out by the starter motor gear, as well as how easy they are to file, I’d question that. It likely varies from model to model though
I can remember running Teflon parts that we had to let sit in inspection for at least 30 minutes before determining the difference in size between what we measured on the floor. We would then have to do this about every 2 hours in order to consistently track the discrepancy and make accurate parts. Great info to be aware of; especially if you are machining tight tolerance parts!
Teflon cut it today, measure it tomorrow.
2 days on some of the parts for us.
Teflon a pain. Seen some parts sit for a week and check good. Then went out for shipment and a month later came back and they had changed.
@@danielgarst2578 Had the same thing happen with Delrin parts. We had to give our parts an expiration date. We told our customer that after a month on the shelf we were no longer responsible for replacing parts that have changed. We have AQL lot data that shows our parts were in print when they were shipped.
@@brandons9138I use to run a type of delrin that had to sit over night. I think it was delrin 150 that would move. The tolerance weren't tight but it was enough to move them out of spec. Stuff smelled horrible when cutting it and played havoc on coolant pumps.
for every problem there is a solution .... for every solution : a welcoming friend - thank you , great video
Thermal Expansion of Spindles and the Machine, can be mitigated by a glas scale (german: glasmaßstab), that runs the complete lenght of the axis. If you have a propperly mounted glas scale in your machine, the only thing you need to care about ist the thermal expansion of your part, because your machine is true.
Each CNC Mill in our shop is equipped with an coolant temparature control system. Usual, when mashines running for 24 hours, all components are getting warmer, the coolant too. The control system keeps the coolant at same temparature, each time. While the mashine is running all parts of the mashine get flushed by coolant, so the mashine itselfe get cooled down. That make sure, we can hold the tight tolerances of all materials we cutting. For example, lot of Bores we have to produce, has tolerances bitween
0,03 mm and 0,0125 mm !
I work as a service technician and had problems with a customer over a year.
They were making housings for big gearboxes out of aluminium, the biggest side was around 1.6m long. They had positioning tolerances of +-0.01mm over a 1,2m distance (convert it to your units yourself) in a not temperature controlled workshop. To make it worse there was a roof window through which the sun was shining exactly on top of the machine.
Temperature differences at the bottom of the machine were in the range of 10°C up and down over a day at the control panel, at the top it was close to 20°.
One of those parts was in the machine for 4 days or so until they were finished with it.
The customer constantly claimed the temperature compensation of the machine is not working properly, even though i told him endless times it´s not the machine but the part that is the problem (even showing him the distance variations they had compared to the temperature and they lined up perfectly).
This went on for a long time until we finally got to the point to get a guy with a laser in to double check all the machine accuracy. We had deviations of the machine positioning accuracy measured over the course of a day of less then 0,005mm, and that is over a 2.5m travel.
Only then the guy finally believed me that it is the part.
We mostly solved the problem by installing a proper coolant temperature control unit, and then had them shower the part in lots of coolant the whole time so it would stay at the same temperature over the day.
I did not think much of the effect temperature had on machining , until I went for my 2day trade test to qualify as a tool jig and diemaker. Day 1 I machined some components of the press tool including the die block . My punch was fitting perfectly on the die , with the right clearance to achieve a blank without burrs. The next morning when I came in the punch was not fitting. I had to re-machine to size again.
I work in the plastic injection molding biz.. Automotive instrument panels ect… Our core/cavity steel is the size of an F150 … So many possibilities to get it wrong lol… I’m always impressed we can hold the tolerances we do.
I was in that field also and our inspection area (not a room) was climate controlled whatever the climate was outside it was warmer than that inside ...
Very interesting ! What kind of tolerances are we talking about ? Like how much tolerance you have
Unfortunately, I wish I had seen this video six months ago. I learned this information by the Qtsi method ... I had to dispense with four pieces of aluminum because of the thermal effect 😑
Thank you mr jessie for those priceless informations 🌷
i once did the montage of a galvanized steel roof crown, the engineer of the project did not account for heat expansion on the roof crown. they where laid edge to edge in 40-170m long lines using 3m segments, laid in the late autum. by the following summer we had an emergency repair, since the screw heads started to spring off, the solution was a 3mm gap between each 3m segment that taught me about the properties of heat expansion between summer and winter for the same piece of steel.
We had some aluminum extrusion that was pulled from outside in winter, immediately cut to size but a few days later we started running them and it had expanded an 1/8 inch
I'm working in a shop with a lot of holes in the bulding, and no thermal regulation at all. At this winter, with all those savings on heating, it's so cold inside the shop, that my lathe's door handle, and any other metallic objects can do a frostbite to my hands. And vice versa - when it's 30 degrees by Celsius at summer - a coolant turns into a warm water, if you are doing some non-stop performance cutting during the day. So, parts at this shop can change their size just because of season change xD
Mate. I'm suffering that right now. I have no choice but to work with gloves, long sleeve wool shirt, jumper and a softshell jacket. Pretty damn awful. Thank god the parts I'm doing are nylon with +-0.5mm tolerances haha.
Reminds me of our old lathe operator. Had some low tolerance parts ( H6 ) ~1500mm long. He always roughed em one day and finnished em the other.
Our cheffe then once went "reee that not fast enough!" Since he learned that stuff too he thought he's doing the next ones himself and faster at that too.
All scrap because around 0,01 to 0,015mm to small in diameter. 2 days of machinetime down the sewer, several thousand € down the drain and a VERY unhappy customer!
Thermal expansion learned the hard way.
Copper is also VERY interesting material in that regard lol.
If you working with cast parts, take them in at least a day befor work. It takes ages to reach thermal equality. Bigger parts takes a lot more time.
Materials dont expand equally on all sides, a shaft will grow more in length than it will in diameter if it is heated evenly. But it will still grow or shrink predictably, but you mostly just learn that by experience. Thanks for sharing, great video. Charles
Nah, you can assume isotropic expansion. It will expand the same relative amount in every diection.
A 5xD shaft will grow roughly 5 times as much on length compared to diameter
@@ullrikegablerthat is if you have a cube
The bigger your part, the more it will expand.
Say you have a 50mm x 100mm x 30mm Block of aluminum
Which has a coefficient of 0.0000235 1/°C
And let's say a temperature drop tdif of -20°C
Ldif = L x a x tdif
50mm x 0.0000235 1/°C x (-20)°C
= -0.024mm
30mm x 0.0000235 1/°C x (-20)°C
= -0.014mm
100mm x 0.0000235 1/°C x (-20)°C
= -0.047mm
So your part will shrink .047mm in length, .024mm in width and .014mm in depth.
@@samsam.03 i think you dont know what relative means.
@@ullrikegabler perhaps
My point is that it is not the same amount of shrinking everywhere
expansion of the ball screw won't be much of an issue most of the time because almost all modern machines measure the table movement directly with a linear encoder.
however, heat spreads and if you have temperature gradients within structural machine parts they can warp in all kinds of ways.
Being a welder we have to deal with expansion and contraction. Very good video in explaining the effects of heat
I've been machining for a while but this has me thinking. If we could get the temperature of the part in real time and have the controller adapt the code being read ahead of time to compensate for the expansion over time to ensure everything lines up perfectly. That could possibly eliminate the issue no matter the temperature of the shop/part. Helix cutting out 9 holes before boring, on the 9th hole it would compensate say .0005 on the ID automatically as a example.
I work as a CNC setter for jet engine parts of which we have some relatively large tolerances due to those parts thermal expansion
Costs go way up if you want the tight tolerances.
Machining a 40' tube when it was 100+ F (summer in Texas and shop is NOT climate controlled). Come in the next morning and the damned thing had shrunk almost 1/8" in length after it sat all night in considerably cooler temps. And yet no one will listen/understand when I try to explain about thermal expansion. These are the same people that think there is no linear movement when interpolating a hole.
Ok so it was machined at 80 and was in spec. But the inspection room was 68 so the part was out so you honed it to size. What happens when the customer gets it and is installing it in another 80° setting? Then your honed hole will be too big
Thought the same thing. Unless all the parts are speced in the same testing facility or there were testing standards across facilities then honing it doesn't matter. I'm not a machinist but it seems like common sense.
Yup, when chasing tenths, you've got to scrupulously dot your i's and cross every "t".
If you fail to pay attention to details then the details will fail you and your parts.
Amazing how you guys measured with vernier inches imperial that fast im not familiar with but metrics in NZ I make Termination muzzle brake T1 to T7 Various sizes and threads compared to you guys superior work 😊
I don’t necessarily think heat is a bad thing. I think it’s the drastic temperature changes that can be catastrophic. I mean you can adjust your offsets over time but a sudden temperature change can catch you off guard. This past summer the AC was out in my building so we had to be mindful of our runtimes
Exactly, when my coworker is calibrating the geometry of a machine he always wants to know if someone has to open the rolling gate on that day because just the cold air from outside coming in can make you get inconsistent readings
Since 3 years it´s my biggest pain at work when temperatures outsides are dropping below 10° C because my company is so uncaring that they ignored the leader of the milling group as well as the leader of the turning group and went and placed a DMG CTX Beta2000 TC in a workshop that regularly has the loading door wiiiiide open to get in rawparts for 30 minutes + (longest was 94 minutes of enduring 8° Celsius INSIDE the shopfloor (drop of about 13-15 °C) .
I can watch my machine shrink in the temperature diagnostic screen and am forced to compensate awful lots in X & Z-axis (most I had to correct so far was up to 0,18mm´s /0,00708 inches ) on parts that have straightness+roundness+coaxiality+concentricity+flatness tolerances of 0,03 to 0,05mm ( 0,00118 to 0,001968 inches) to not continously make wasteparts and that costs tons of time and nerves .
Once that loading door finally closes I need to compensate in the other direction because it slowly gets warm again *sigh*
Welcome to machining in Austria :(
No change in sight so far ...
Great video!
with that formula for thermal expansion, are the units of measurement in inches or metric, Celsius or F?
just be consistent and it works
The formula is the same for imperial or metric... Your coefficient for each material is what is different depending on which unit system you are using. I.e. if you work in metric, use the metric thermal expansion coefficient.
I had an old machine that refused to work when it was too cold. I had to put a heater on it in the winter.
If your lathe part feels warm then do not measure. Wait until it's cool to the touch. A part at 68 degrees feels a little cold to the fingertips.
Hi I am loving it’s Titans my question is how Can I be part of your team? Please help me sir thank you very much!BOOM!
3:05 except if your machine has linear encoders
This has to be an outlier I machined for over 20 years never seen a no-go fail due to that but then again we had air conditioning.
If I were to order some high-precision parts from you that were made of…oh, let’s say aluminum, which has a really high coefficient of thermal expansion…would it come with a document that says what temperature it was made at?
Next question: do you have a way to calculate acclimatization times? Some of us keep our materials in different climate condition than we work them in.
QC rooms should let parts sit until they're at a reasonable measurement temperature (due to being nearly ambient can take days/weeks). Simple parts you can easily do transient heat transfer problem for acclimatization, talking square, rectangles, etc. Otherwise you'd want to do it with a simulation software.
@@ScuffedEngineer I was more thinking of acclimatization of raw stock. I, like a LOT of other people, store my materials in a place that has only heat but no AC. If I was working four-inch plate steel stored in an open-air rack in the Texas heat and I bring it into my air-conditioned machining room, how long is it going to take to get the steel from 120 degrees to 75 all the way through?
@@jmowreader9555 I mean, I'm not doing the math for you, grad students don't have that time. It depends on the type of steel, the volumetric dimension of the stock (never said if it was square, rectangular, thickness just specified 4" plate). Also depends on what you consider all the way through and what temperature variance is negligible.
You can look into transient heat transfer and the FTCS method for quick and dirty calculations, there might even be some code or a program out there for free.
That being said, with that temperature variation I'd say not quick unless it was say .01" thick copper.
Same with stresses in material, brass part. Faced second op, loosened vise, bent like a banana, god that sucked. I managed to fix it tho phew, some things, never an issue, then boom!! Failure
Ah who needs the piece to be strong and sturdy or in square
The customer smh
@@cyber2526 it’s called sarcasm chief
@@Everythingwentblack69 i know =)
This happened to me with some plastic. Except it was the opposite.....thermal expansion. Parts scrapped.
When it gets cold outside I personally get thermal shrinkage..
I remember whilst at college we had a dare to go up to the girls in the local bar and explain the thermal coefficient equation and the winner was the girl that lasted the longest
We made math entertainment
😂that's been my life for the passed week .0002 +/-
We have that problem. But the owner doesn’t understand that. 🤦🏻
Thermal expansion is nothin new to those that have worked with anything that has been through nitriding! Would constantly see 20+" stem valve length's come back at least .01 longer than prior.
Lets be real, We are going to learn the hard way...
I took ya serious until you threw a tenths indicator on the fuckin door that flops back and forward atleast an 1/8
i know about it , you trying to make technology to write program , run it make few parts pristines, and then operator came, bang and machine is out of order.... Boom titans. cool video .but iam alwaays smiling when you talking in inches . ! Metric rules man :) .
We installed a 4 inch PVC conduit on a roof 100 (or so) ft long. worked up a BOM. my coworker and I were curious as to how much exactly and calculated the length summer to winter and decided to add a second expansion joint and anchor it in the middle of the run instead of each end and joint in the middle.
Boom
Haha we’ve all learned this in the worst way
Back in the steam locomotive days a tire was shrunk fit on the wheel. when red hot they fell on the wheel no force involved other than gravity.
They still do that with automotive ring gears on flywheels. Rarely have to heat them red hot, but they still do just fall on when at the right temperature
@Glenrok yea about 350F red heat would change hardness and temper. Same with bearings
@@bigboybuilder that’s true, although not particularly significant with a ring gear. They don’t receive any special heat treatment
@@Glenrok I believe they are induction hardened hence the blue on the teeth but i could be wrong
@@bigboybuilder considering how easily they’re chewed out by the starter motor gear, as well as how easy they are to file, I’d question that. It likely varies from model to model though