HomeRoast Digest


Topic: A few more IR tests (3 msgs / 116 lines)
1) From: Ken Mary
My IR thermometer will not measure boiling water accurately. Maybe the water
vapor above the surface interferes. With water near 150 F it compares within
a few degrees with a dial thermometer. Strike off one use for measuring brew
water temperature.
Polyethylene does transmit IR to the sensor with an error due to its
emission at room temp and absorption/reflection at higher object temps. Cold
temps are high by several degrees F. Room temp readings are unaffected. One
test with a warm skillet, the IR read 196 F, with the plastic wrap, 186F. I
suppose a shield could be made from plastic wrap to keep dust away from the
sensor. But in actual use, the wrap would get hot and affect the sensor
accuracy.
Green coffee temperature reading is affected by radiation emitted from the
hotter roaster wall and reflected off the surface of the coffee. The empty
roaster was turned on to warm up, then shut off for one minute. The chamber
bottom read 170 F, the wall, 120 F. Green coffee originally was 62 F, a few
seconds after dumping into the roaster, 67 F. However, this would not create
any accuracy problems during an actual hot air roast where wall and bean
temps would be about the same. In drum roasters where the wall temp may be
hundreds of degrees above the bean temp, there would be a significant error.
--
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

2) From: Tom Gramila
Ken,
	A quick thanks before my comments -- Its a big help, as someone
who has wondered alot about the suitability for IR for roasting
instrumentation , to learn of these results!!  Thanks!
On Wed, 11  Dec 2002, Ken Mary wrote:
<Snip>
	I wonder if its the small water droplets above the water, like in
	a cloud.  I think that is what you see with your eye above the 
	pot, and  not the vapor itself.
<Snip>
	I think that there may be hope here yet.  Your temp difference
would imply a transmission of 92%, which is very good.  The errors
introduced for cold objects are, I think, partly from weak emission from
the plastic, but also from reflection of other room temperature objects
back into the sensor.  Since the radiated power goes as T**4, it does not
take much of either before it makes the apparent temp of the cold object
wrong. But....  the same effect (T**4)  means temperature errors due to
emission from the plastic above roasting coffee should be relatively
small: the plastic emits way less IR than the hotter beans below it emit,
and the plastic is a poor emitter since its mostly IR transparent.  The
assumption here is that the plastic sheet is COOLER than the beans being
roasted, which would be critical in order for the plastic not to melt. The
trick would be, I think, keeping the plastic from melting.
	 I am imagining using a tube, something like 1" diameter by a few
inches long.  One end is toward the roast chamber, and the plastic is on
the other end with the thermometer outside that.  The column of gas in the
tube sould provide some thermal isolation from the hot air (and some
cooling of it perhaps, if it is thin wall metal).  Would anyone venture a
guess as to whether this sort of arrangement could keep polyethylene from
melting????  Or is this nuts??
 > 
<Snip>
			^ Hooray!  At the very least, an IR thermometer
could provide a spot check of the accuracy of my thermocouple in the
roasting beans.....
Thanks again...
		Tom
homeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast

3) From: Paul Goelz
At 10:41 PM 12/11/02, you wrote:
<Snip>
There are several issues involved in these sorts of measurements.  The 
first is the transparency of the polyethylene at the wavelength of your 
sensor.  You likely have a sensor that is sensitive to very long 
wavelengths.... probably something like 8-14 microns.  I do not have access 
to emissivity charts any more (I used to sell these things) but finding 
something that is transparent at those wavelengths is problematic.  The 
window in your instrument may be germanium, for example.  I do not recall 
the transmissivity of polyethylene at 8-14 microns.  And a room temperature 
test is meaningless... a window with a transmissivity of zero will still 
produce a room temperature reading.  You are reading the temperature of the 
window and not the target, but they are both at the same temperature.
A second issue, and a serious one, is keeping the window clean.  There is a 
lot of smoke involved in roasting, and keeping deposits off the window is 
going to be a real challenge.  Any deposit on the window will reduce the 
transparency and introduce an error.  And any decrease in transparency will 
lower your readings, which will be more the window temperature and less the 
product.  In industrial applications, this is commonly dealt with by using 
a purge gas that is compatible with the process.  The gas is introduced 
into the sight tube downstream of the window and gently flows out the open 
end.  This prevents smoke from entering the tube and contaminating the window.
Keep at it, though.  I have always felt that I/R was the perfect way to 
measure bean temperature and thus provide a controllable and repeatable 
roast profile.
Paul
Paul Goelz
Rochester Hills, MI
paul
pgoelzhttp://lists.sweetmarias.com/mailman/listinfo/homeroast">http://www.pgoelz.comhomeroast mailing listhttp://lists.sweetmarias.com/mailman/listinfo/homeroast


HomeRoast Digest