Double Thermocouple - Two types, same sheath (AMS2750/Nadcap)

[Sorin]

We are looking to buy a furnace (used) that is to be used for aerospace parts AMS2750 / BAC 5621 / (Nadcap) requirements implied.
The issue is that there is no way for us to add ports (holes) for SAT probe or additional resident thermocouples (for SAT).

The process sensors are Type K and I was thinking about installing double T/C (dual element - installed in the same metallic sheath). One type K which is the process sensor and one type N which will be the SAT resident.

I have to install 3 - control, hot, cold.

I looked through specifications and did not see anything that would prohibit me from doing so.
See picture attached - sample
What do you think?

 

[Ninja]

Not so very familiar with Aerospace regs...

But I do have a question:
"...there is no way for us to add ports (holes) for SAT probe or additional resident thermocouples (for SAT)."

What is preventing this?

 

[Sorin]

Can't have additional holes due to how the furnace is designed. We already have specific holes for the process sensors(T/C) and these are the holes where instead of a single element T/C would like to install a dual element T/C.

 

[Ninja]

I can't think of a reason why a dual TC would introduce any error not already there (for whatever that's worth).

I would still be looking for a way to keep them separate simply to theoretically reduce lifetime costs...if one goes bad or ages, you have to replace both kind of thing.

I've never seen a furnace design that completely prohibited something as small as a thermocouple port. Even if it is simply reaming it a little wider to fit two sheaths...

 

[Sorin]

I already have replacement frequency of 6 months. Being that the resident T/C (for SAT) shall be re-calibrated/replaced every 3 months, it will double the cost for the process sensor.

 

[Ninja]

You find a compny willing to calibrate a used TC?
Please share....

 

[Sorin]

Re-calibrating T/C is more expensive than buying new ones....so...

 

[Ninja]

Yup. That is my experience.

I wasn't sure when I saw:

... shall be re-calibrated/replaced every 3 months,...

Functionally, this really means "shall be replaced every 3 months..."

I was just hoping you found the magic cheap-calibration vendor....

 

[BradM]

You find a compny willing to calibrate a used TC?
Please share....

I have run across a few of them. But you have to pay them to clean/sanitize them prior to the calibration. ($$$$)

Re-calibrating T/C is more expensive than buying new ones....so...

I agree. However, you really run across the issue of trying to do a post check on thermocouples prior to disposal.

Depending on how many thermocouples and how frequent a check, a calibrator dry block calibrator may be a decent option for you.

 

[ncwalker]

My T-couple knowledge hit list is this:

They work because of dissimilar metal bias generating a voltage. I have always tried to keep the dissimilar metals running to the transducer as much as possible. What I mean to say is this - if after exiting the furnace, you switch to say plain copper and run this to the transducer because it is far away from the furnace, expect results to vary. You would want to continue the entire run in thermocouple wire for best results. And that's it. They are pretty simple. With that said....

I can't see where physics of the situation would invalidate their close placement. What I mean is it's not like measuring the temperature with one makes the substance "cooler" or "warmer" affecting the measurement of the other.

I can see where you could mess up the installation .... It is the dissimilar metal junction where the sensing happens. So if you are cramming these things in and accidentally strip the t-couple wire creating a junction "behind" where you are trying to sense, the temperature reading you will get is at this closest (wrong) junction. Effectively because the "tip" down in the furnace is now shorted and not contributing to the voltage difference.

So as long as you are careful in the install, I'd support it.

Real easy to check - after you're done, do you get similar results? You ARE using different T-Couple styles ....