PDA

View Full Version : Temperature difference on 2 sides of a thin copper plate



SirLesPatterson
07-23-2014, 09:14 AM
I have placed a 3/16" thick 4" x 4" copper plate 7/8" horizontally above a IR heating plate calibrated to 300C in a 24C room with thermocouplers touching the bottom and the top of the copper plate. After 15 minutes the temperature seemed stable Top=79C Bottom=139C

I realize there could be other factors but do these readings sound reasonable?

lynnl
07-23-2014, 09:47 AM
I have placed a 3/16" thick 4" x 4" copper plate 7/8" horizontally above a IR heating plate calibrated to 300C in a 24C room with thermocouplers touching the bottom and the top of the copper plate. After 15 minutes the temperature seemed stable Top=139C Bottom=79C

I realize there could be other factors but do these readings sound reasonable?

Certainly they're reasonable, since they are what you observed. Tho it does seem surprising there's that much difference between top and bottom surfaces.
If in doubt about the accuracy flip the sensors.
The first thought I had is how much air movement is in the room?

SirLesPatterson
07-23-2014, 09:54 AM
I did flip the sensors and it was very similar so I didn't mention it. Not much airflow in the room at the time of the test, did it this AM before the AC kicked on. I was just surprised that the copper plate didn't even reach half of 300C.

RichR
07-23-2014, 10:05 AM
I was just surprised that the copper plate didn't even reach half of 300C.
That's probably because the copper is losing a lot of the energy on the top side. Try laying a 4" square piece of cardboard on top and see what happens to
your readings.

ironmonger
07-23-2014, 10:14 AM
I have placed a 3/16" thick 4" x 4" copper plate 7/8" horizontally above a IR heating plate calibrated to 300C in a 24C room with thermocouplers touching the bottom and the top of the copper plate. After 15 minutes the temperature seemed stable Top=139C Bottom=79C

I realize there could be other factors but do these readings sound reasonable?

Is the bottom sensor being irradiated with the ir directly?
Try shielding the bottom thermocouple with a piece of kaowool or moving the plate so that it is not irradiated by the ir.

paul

lynnl
07-23-2014, 10:22 AM
That's probably because the copper is losing a lot of the energy on the top side. Try laying a 4" square piece of cardboard on top and see what happens to
your readings.

Yep. That copper would be very effective at losing heat from the top surface via radiation.

Reminds me of a trick my Dad taught me when I was a kid. Try this. Take a cast iron pan, with an inch or two of water in it, bring it to a good boil, and you can then take it off the stove and set it in your palm. The bottom of the pan will feel room temperature, while the water inside is boiling. (But you better have a plan for quickly divesting yourself of the pan when the boiling stops.)

strokersix
07-23-2014, 10:39 AM
Yep. That copper would be very effective at losing heat from the top surface via radiation.

Reminds me of a trick my Dad taught me when I was a kid. Try this. Take a cast iron pan, with an inch or two of water in it, bring it to a good boil, and you can then take it off the stove and set it in your palm. The bottom of the pan will feel room temperature, while the water inside is boiling. (But you better have a plan for quickly divesting yourself of the pan when the boiling stops.)

Please post the video when you do this.

jlevie
07-23-2014, 10:54 AM
The numbers look screwy to me. Logically, the side facing the IR source should be hotter.

Guido
07-23-2014, 11:03 AM
IR heaters hanging overhead and switched on after an overnight cold soak, will heat metal objects laying on a workbench, like fast. Takes several minutes for the metal objects to radiate their warmth to the surrounding air, which was the main idea to begin with.

--G

SirLesPatterson
07-23-2014, 11:52 AM
The numbers look screwy to me. Logically, the side facing the IR source should be hotter.

It is. You get a gold star! I reversed my readings when typing it up. Fixed.

Paul Alciatore
07-23-2014, 03:17 PM
Read his figures again, the bottom is hotter: "Top=79C Bottom=139".


The numbers look screwy to me. Logically, the side facing the IR source should be hotter.

Another thing to consider here is the degree of thermal contact between the copper plate and the thermocouples. Just because they are touching does not mean that the thermocouples are at the same temperature as the surface of the plate. I have had soldering iron tips in contact with copper wires for several minutes but some corrosion or other factor prevented the transfer of heat. Solder would not melt on the wires while it readily did so on the iron tip. Physical contact does not insure good heat transfer.

And then consider that the thermocouple wires have their own physical properties: size, IR absorption properties of their surface, etc. So the bottom thermocouple IS closer to the IR source and it may have better absorption properties than the copper plate. The IR could be bouncing off the plate.

If you want a more accurate reading of the plate's surface temperature, may want to try another method or modify your present method. For starters, you could use some computer style thermal grease between the thermocouples and copper plate. And perhaps install a small shield below the thermocouple junction and the IR source so it is responding only to the heat from the copper plate and not picking up the radiation directly from the IR heater.

Likewise on the top side, the thermocouple is mostly in contact with the air, not the copper. The thermal grease may help there also. And a small cage to prevent the air from cooling it.

Another way may be to use one of those thermometers that you point from a distance and read the surface temperature on a digital readout.

jlevie
07-23-2014, 03:29 PM
There is either a lot of convective or radiative loss from the top of the plate. Or as Paul Alciatore mentioned, a flaw in you measuring technique. Assuming both sides of the plate are is similar finish/condition, I'd try an IR thermometer.

Weekend_Scientist
07-23-2014, 04:10 PM
An infrared heating plate? Not sure I've seen one of those but if it heats stuff by shining infrared light at something I would expect different things to heat up at different rates. You said: "....calibrated to 300C in a 24C room...." What were you heating up when you calibrated it?

I would expect it to heat something non-reflective much better than something that reflects a lot of IR.

Copper may reflect a lot of IR and so not be heated very efficiently with this type of heater. Hence only making it to 139C on the hot side. If that is the case then it doesn't have to reject as much energy for the cold side to end up at just 79C.

If you try the same setup with a material that doesn't reflect much IR (lower IR albedo) will the hot side get hotter?

macona
07-23-2014, 06:06 PM
Try coating the heated side in lampblack and see what kind of difference that makes.

cameron
07-23-2014, 07:21 PM
For that size copper plate with 60 degrees C temperature difference across it, the rate of heat transfer through it would be over 35 kilowatts.

I think there may be something wrong with the numbers.

PStechPaul
07-23-2014, 08:11 PM
I would expect the top and bottom surfaces of a thin copper plate to have virtually the same temperature. But as mentioned, the thermocouple may not have a solid thermal connection to the plate, and one side is exposed to air. The air on the bottom will be hotter because it is closer to the heat source and it is somewhat trapped, especially at the center, and convection will be up from the center and out to the edges. On the top side, although the plate should be the same temperature, there will be more free convection cooling, and the air just above the plate will be much cooler. Again, since the thermocouple is exposed to both the plate and the air, and its heat conduction is much less than copper, it will read lower.

Wrapping the thermocouple in copper, and also shielding it from the IR from the heat source as well as the moving air, should produce better readings. Please let us know what you find.

Don Young
07-23-2014, 09:53 PM
I agree that the temperature difference is actually pretty small, perhaps not detectable with a common thermocouple setup. Something is causing one or both thermocouples to respond to something other than the plate temperature. As Paul mentioned, shields well bonded to the plate and thermocouples should greatly improve the results.

Paul Alciatore
07-23-2014, 10:40 PM
Err, I did not say that a shield should be bonded to either the plate or to the thermocouple. In fact, if you want to measure the surface temperature of the plate, I would keep the shields thermally isolated from both. An air gap should be OK. And keep them small. On the bottom, just big enough to throw an IR shadow on the thermocouple. On the top it could be a little larger as it is to keep the air from cooling the thermocouple from cooling down too much. But not too big as that would create a hot spot on the top surface. Perhaps no larger than 2 times the width/diameter of the thermocouple.


I agree that the temperature difference is actually pretty small, perhaps not detectable with a common thermocouple setup. Something is causing one or both thermocouples to respond to something other than the plate temperature. As Paul mentioned, shields well bonded to the plate and thermocouples should greatly improve the results.

J Tiers
07-24-2014, 08:40 AM
It's time to consider the physics.....

When you have a temperature difference, you have heat energy flow. The amount of flow is related to the temperature difference, and the heat conductivity of the material. Or, if you prefer, as I do for electronics, the thermal resistance.

You have a relatively thin copper plate. Copper is known to have a high heat conductivity. You also apparently have a large temp difference across the thickness of the plate, which is actually not very thick.

So ASSUMING that the temps are real, there must be a large heat flow through the plate from bottom to top. You can calculate teh required heat flow by getting the thermal resistance of copper, and figuring how many watts have to pass through per unit area to produce the difference.

"By inspection", we can make a pretty good guess that all is not as it seems.... nothing is apparently touching the top other than a thermocouple. It seems impossible to remove that much heat from the top through air, assuming the plate really is copper right through.

So, most likely the top thermocouple is telling lies, due to poor contact, or being so routed that a lot of heat is being drawn off through the wires.

ckelloug
07-24-2014, 04:33 PM
I just noticed that 139F is 77C. That's awfully close to the temperature of the other side which is more consistent with a big copper plate.

Paul Alciatore
07-24-2014, 08:10 PM
????? What are you saying? He gave both top and bottom measured temperatures in deg C.



I just noticed that 139F is 77C. That's awfully close to the temperature of the other side which is more consistent with a big copper plate.

Don Young
07-24-2014, 11:07 PM
I certainly did misunderstand your post. For some reason I was thinking of creating something like a temperature well, where the whole thermocouple environment would be maintained at the same temperature by thermal bonding.


Err, I did not say that a shield should be bonded to either the plate or to the thermocouple. In fact, if you want to measure the surface temperature of the plate, I would keep the shields thermally isolated from both. An air gap should be OK. And keep them small. On the bottom, just big enough to throw an IR shadow on the thermocouple. On the top it could be a little larger as it is to keep the air from cooling the thermocouple from cooling down too much. But not too big as that would create a hot spot on the top surface. Perhaps no larger than 2 times the width/diameter of the thermocouple.

jkopel
07-25-2014, 01:47 AM
What happens to the temp readings when you turn OFF the heater?
If physics (and the thermocouples) is working, the temps should start to equalize, right?
And drop of course.

boslab
07-25-2014, 03:05 AM
Bings back one or two memories, steel is cast into a water cooled berylium copper mould, the hot side is about 1450 or more even degrees C, and behind 1/2 inch at its thinnest is water so heat flow was immense, radiation was not a major factor in that case as there was liquid metal solidifying against the copper, it can get fairly involved!
http://www.freestudy.co.uk/heat%20transfer/conduction.pdf
Interesting read
Mark

J Tiers
07-25-2014, 09:21 AM
I suppose the other and even more likely cause is the bottom side thermocouple is being heated more due to the IR source.... and poor contact to the plate....A lot of wire exposed means the IR can heat the wire and be conducted down to the junction.... so the 139C might be the really suspect reading....

An IR source at a high temp is a sort of "thermal current source".... if you don't actively cool something that is exposed to it, the "something" can get pretty hot, almost as hot as the source, theoretically.

Paul Alciatore
07-25-2014, 04:07 PM
Exactly my reason for a heat shield between the thermocouple and the heater. The "temperature well" idea as Don Young suggests is valid and is contained in my suggestion, but it should be separated from the shield. I strongly believe that the biggest reason for this obviously erroneous reading is the IR striking the thermocouple junction and perhaps the wires leading up to it directly.



I suppose the other and even more likely cause is the bottom side thermocouple is being heated more due to the IR source.... and poor contact to the plate....A lot of wire exposed means the IR can heat the wire and be conducted down to the junction.... so the 139C might be the really suspect reading....

An IR source at a high temp is a sort of "thermal current source".... if you don't actively cool something that is exposed to it, the "something" can get pretty hot, almost as hot as the source, theoretically.

garyhlucas
07-25-2014, 06:25 PM
Copper is used for the mirrors in infrared lasers used for metal cutting, because copper is a great reflector of IR. So heating copper with an IR source is not very effective. So I suspect the bottom thermocouple is absorbing IR heat much better than the copper, hence it is warmer.