Thermocouples Hot Junction and Number of Uses Specifications

[Kapsyx]

Hello, I hope this hasn't been covered anywhere else, I did a quick search and couldn't find something related.

I'm working in the Aerospace Industry and was wondering if there are any specifications for load thermocouples hot junction? I couldn't find anything specific in AMS 2750. I know the preferred method is to weld but can the wires simply be twisted?

On a similar note, is there any specification on how to record the number of uses a TC has gone through? AMS 2750 specifies that for an expendable load TC it has to be a maximum of 30 times when used below 1200?F. Current procedure is to just write a crossmark on a sheet that contains the TC serial numbers for each time the TC went through a cycle. Right now I can, provided the information recorded is thrustworthy, know how many times a TC has been used but not at what point of time or who verified the TCs to see if they are good to be re-used after each cycle. Perhaps I'm only worried that at some point the guy crossmarking the sheet, each time a TC goes through a cycle, for some reason forgets to and the # of uses may go above 30. In any case, as I said earlier, the TCs are verified to see if they still work everytime they come out of a cycle, but then again they don't record who does this and thus the person's training traceability is lost. I'm I just being too picky?

 

[MTProcessing]

I attended PRI's pyrometry class this spring and thermocouple junctions were discussed. The instructor said that while some customer-specific specifications (e.g. Boeing) may require a particular join method, there is not an industry standard. So if you're sure you're not covered by a specific customer requirement, welding vs twisting is your choice.

For cycles, while I'm not aware of detailed requirements, I would think you'd need some way to audit the record. If an audit can pick a small time period, go through the furnace records and verify that all the uses were recorded, then you can gain confidence the system is working. I think you'll need to add at least dates and possibly times of use to the tally to get the system to a point where it can be audited.

 

[dgriffith]

With respect to the join method, the conditions of use would drive one method over another. If corrosion is a possibility then I would weld. If I just want to measure (very poorly...) the air temperature in a room with a TC for a short period of time, then I might twist them and get on with it. Remember, the point at which the two wires first touch is the junction.

 

[Kapsyx]

There's no customer requirement in this case. I guess I can just make a suggestion. How much of a difference do you think twisted vs welded junctions make in measuring the carbon fiber part temperature? In this case the TCs are inserted between plies. Internal curing specs have tolerances of ?2?F at some points of the cure, if it could be proved that using twisted junctions doesn't provide enough accuracy to be within those 2?F of tolerance, would that be good ground to start talking about modifying the requirements?


About the cycles, the traceability by dates is a good suggestion and something I was expecting to see the first time I looked at the records. Again, they are doing the minimum and at some point I feel like I'm just being too picky and looks like I'm against the keep it simple principle and my rationale is that I see the TCs malfunction too often which may or not be because of this "lack" of #uses control.

 

[dgriffith]

There's no customer requirement in this case. I guess I can just make a suggestion. How much of a difference do you think twisted vs welded junctions make in measuring the carbon fiber part temperature? In this case the TCs are inserted between plies. Internal curing specs have tolerances of ?2?F at some points of the cure, if it could be proved that using twisted junctions doesn't provide enough accuracy to be within those 2?F of tolerance, would that be good ground to start talking about modifying the requirements?

Something like this might meet your need and avoid the weld/no weld decision, especially if the lump of TC under the plies will cause the plies to take a set as they cure.
http://www.omega.com/pptst/88000_thinfoil.html
As for the ?2 ?F, some TC can meet that limit of error or better, but that doesn't include the TC readout error. You haven't said what type TC and what you're using as the temperature readout.
If I had to make a claim of how well I knew what the temperature was during curing, I would not use a twisted junction.