Oven Qualification - Thermal profile during the OQ of a vacuum oven

[sledesma]

Dear all,
I am writing to ask for some help in analysing the results obtained in a thermal profile during the OQ of a vacuum oven. It was carried out without appling vacuum, only thermal profile, with the set point at 105 centigrades (analog controler). The oven is 16 litres of capacity, and the test was carried out using 12 calibrated thermocouples during 2 hs (stability). The specification of the termal range is 103-107 C because the oven is used for API's loss on drying determination. 3 of the thermocouples have a medium temperature of 120 centigrades, they were near the top, 2 of them that where near the glass door have a medium temperature of 100 C and the other 7 have a medium temperature of 110 C. I want to know if the equipment have a problem or may be the test was not properly designed. What do you think? Is it common to have these differences in temperature profile in this kind of equipment? Thank you very much. Regards from Argentina!

[Kevin H]

Sledesma, I'm not surprised at the temperature variation in the thermocouples at that low temperature. I'm guessing that the oven is heated electrically - and that you're relying primarily on radiation for heat transfer. (Actually with air in the furnace you'll get some heat transfer via convection as well). Radiation only becomes effective at high temperatures - say 1800 degrees F or 980 degrees C. I used to heat treat precipitation hardening stainless steel tensile specimens at 900 F to verify capability. I had to be at +/- 15 F through out the furnace - a furnace without atmosphere, just air, that was heated via electrical resistance. After doing a temperature uniformity study we had to restrict the acceotable working volume to about 1/3 of the actural furnace working volume to meet those requirements.
At the solutioning temperatures, about 1750 F if I remember correctly temperature uniformity was fine. Our final solution to the temperature issue was to purchase a furnace with a builtin fan to circulate the air iresulting in transfer via forced convection n the furnace so that were more uniform and had better heat transfer, not relying just on radiation and natural convection.

You'll probably see even worse uniformity if you test under vacuum, which I believe you should do, as conformity tests should reflect working conditions. And yes, doors, even when insulated are usually cooler. I'm not a furnace designer, just a user, so I can't help with your thermal uniformity problems. - One possibility would be to do what we did, i.e., restrict the usable volume of your oven to that which meets the temperature variation requirements of your test.

The uniformity that you're looking for at the low temperatures you're measuring will be extremely challenging to obtain in a vacuum oven. Best of luck on getting them.

[MIREGMGR]

With only radiation as a heat transfer mechanism, what is the meaning of a reading from a thermocouple located away from the part? What is it measuring?

If the thermocouple is conductively coupled to an oven enclosure element, it's measuring the temperature of that element...which may have practically nothing to do with the spatial evenness of the radiation pattern, if in fact that would result in the desired part temperature profile.

If your heating mechanism is solely radiation and you're using thermocouples to measure temperature, they should be conductively attached to either a test part or a specially engineered object that's located where the production parts will go, and has appropriate characteristics to heat in the same pattern as the known actual heating pattern of the production parts. The thermocouples also should be locally radiatively shielded, so they measure the test part temperature via only conduction, without consideration of their own radiative heating characteristics.

[sledesma]

Yes Kevin H, this oven works with electrical resistences, and the test was carried out without applying vacuum, but the chamber is hermetically closed. So you said that in that condition the primary heat transfer is because of radiation. And so that the homogeneity will be poor at that temperature. Thank you very much for your contribution! I think the problem is being clarified!

[sledesma]

Thank you very much Miregmgr! Now I will only concentrate in the results of the TC that are near the working surface!

[BradM]

Hello there!

First, it sounds to me that your temperature specification is way too tight. Even with high quality Type T thermocouples you have +/-.75 C tolerance on each wire. So, you can have a total of 1.5 C deviation. That is not including any error within the recording device and other uncertainties along the way.

BTW... what kind of specifications were provided with the oven? If the oven was not designed for your particular uniformity, it may not ever be able to achieve it. Ovens have to have very high air exchanges per minute to achieve that type of uniformity.

What kind of oven is this? Is it a recirculating-air type oven, or is it constant temperature? If it is recirculating, there is a discharge and a return air. Generally the discharge side will be the hottest. Also, heat rises, so the top of an oven will be hotter than the bottom.

If this is a constant temperature vacuum oven and your process will involve vacuum, you really should consider running the profile with the vacuum on.

I hope something here helps.

[sledesma]

Thank you Brad! Your comments are really helpful!