Message Board Thread - "front car window temperature profile--HELP!!"

Back to Threads | Back to Forums

TitleByPosted On
front car window temperature profile--HELP!! thermoimagen 2/29/2004
one manufacter of car windows ask to me to inspected the temperature loss of car front window when this window travels thru conveyor after leaving a furnace in the process..

the window is sandwich type, means that all front cars windows are 2 glass layers (1/8")and in the middle plastic film which prevents that shatter fly away when crash ocurs.

the conveyor is 45 feet long and this guys want to see how temperature drops from begining to end of line. (after furnace and before the next process)

My concern is that the material to be inspected is transparent GLASS.

i will try to determine if it is posible to paint one window from scrap to make test.

any ideas will be apreciate.

Re:front car window temperature profile--HELP!! Doctir bob 3/1/2004
If you are using a 3-5 micrometer waveband IR camera, you need a "glass filter" that has a narrow bandpass that peaks at 5 micrometers. It cuts on above the transmittance of glass. As it is a narrow bandpass filter, and peaks right at the end of the IR camera response, the target must be hot enough for good measurement. Knowledge of the temperature range to be measured and required sensitivity is necessary to answer whether you would have sufficient sensitivity.

Perhaps a better approach would be to use a longwave 8-12 micrometer IR camera. Here glass is opaque and has a high emissivity except for a slight dip in emissivity at 10.6 micrometers due to the Restrahlung effect. Still the average graybody approximation to emissivity from 8-12 micrometers is just below 0.9.
Re:front car window temperature profile--HELP!! thermoimagen 3/1/2004
thanks Dr Bob ..
i actually use agema 470 pro and inframetrics thermacam pm350, i guess i am not able to perform this test due the following, my systems are SW and i have only flame filter, not "glass filter" .. thanks anyway.
Re:front car window temperature profile--HELP!! Gary Orlove 3/2/2004
I would recommend measuring the glass emissivity (we always recommend this for new materials)to make sure you know what it is for your camera and your sample of glass. Different glasses have different emissivities depending on how they are made. I have seen LW emissivities as low as 0.8 on some glasses.

Gary Orlove
Infrared Training Center
Re:front car window temperature profile--HELP!! jdemonte 3/2/2004
I agree with Bob and Gary on this one. Also contact your local FLIR rep to see if they have a glass filter for your Inframetrics camera or could let you try out a LW instrument to test the feasibility of this.

I have, in the past, done this at an automotive front windshield plant. The camera was a LW Agema SC500 and it was put in place after the radiant heater to observe the glass as it exited. The camera's output was then put into a computer and it displayed and saved a scan of each window as it went through. If one of the windshields did not meet their thermal criteria (proprietary info), they would be alerted by a built-in alarm with the software.

Hope this helps,

Re:front car window temperature profile--HELP!! jk 3/3/2004
Would it be possible to place a reflecting medium such as a foil board and read the temp from the reflection?
Re:front car window temperature profile--HELP!! Jim T. 3/3/2004
In my experience, the opacity of glass to IR varies somewhat. In general however, most everyday glasses tend to be pretty opaque in the midwave band. My suggestion would be to go ahead and use your 3-5 camera, and re-establish the calibration between camera image greyscale level and temperature. This would mean that for the first sample windshield you would have to measure the temperature at a number of points to calibrate the camera images. This is not a big deal; I do it all the time with our SC1000. I don't really trust the internal camera calibration, so I always do my own independant calibration for every measurement trial.

The accuracy of your glass temperature measurements are not critical as long as you take enough data points (>10). Thus you would be fine with a contact type thermocouple probe.

My experience with the mid-wave cameras (PM380, SC1000) is that the greyscale is very linear with emittance (W/m^2), and so is an exponential with temperature.

I know that other people have suggested trying longwave cameras, or trying to get special filters. At the end of the day, you are probably better off figuring out how to make due with what you have, which I think in this instance you can do.

P.S. Make sure that you take into account the high reflectivity of the glass. You will want to observe the glass from a position as perpendicular as possible. Also try to keep the reflected background cool and uniform if you can.
Re:front car window temperature profile--HELP!! thermoimagen 3/4/2004
thanks for the feedback to all of you guys, i will make several test with my 2 cameras. regards.
Re:front car window temperature profile--HELP!! thermoimagen 3/4/2004
here are the pict of the process to inspect, please take a look:
Re:front car window temperature profile--HELP!! johanw 3/9/2004
for these kind of problems it's always useful to have some background info of your camera, like the exact spectral response: 3-5 micrometer or MW is just a "normalized range"; ask your supplier for a graph holding the exact spectral response of your detector; it will help you to understand what your camera can see and not see. The same is relevant to the car window: the manufacturer of the glass knows what the spectral transmittance is. Just put the graphs of the glass and the camera in one picture and it will help you to understand the measurement problem. Of course you can do some test yourself, like putting a blackbody behind the window and measure the apparent temperature, convert this to energy by means of Planck's law (this is the sum of transmitted energy from the blackbody and emitted energy from the glass)and do the same with the glass removed: the camera should indicate the right blackbody temp and convert this to energy too. For reflection of non-metals I thought one often uses 4 percent: see what happens if you put the blackbody and the camera at the same side of the glass and let the bb radiation reflect by the glass into your camera; convert to energy and take care again of emitted, reflected and transmitted energy.
I hope my contribution is of some help to you; Success.

  • Back to Threads
  • Back to Forums


  •   Copyright © FLIR Systems, Inc 2012