K Type Thermocouple

The limit is 1100C for type K.

Type B is good up to 1700C

A calibrated Spectrometer can be used for very high temperatures,but it requires a hot temperature body for calibration.

Try this.

Our use of K type thermocouple is to control a furnace of Max.attainable temperature of 1300'c for a short while and 1250'c is the max. working temp. Control is by PID programmable temperature controller working in conjunction with SSR.

Here is a chart that will hep you decide which is best for your application.

https://assets.omega.com/pdf/tables_and_graphs/thermocouple-type-b-celsius.pdf

Type "B" seems to me to be the best candidate,but I do not know your specific conditions,atmosphere,pressure,etc.

Good luck!

<...an analog input on a single loop controller...that will range up to 1316°C for Type K....>

Just because the input is suitable doesn't mean that the thermocouple itself is suitable.

Yes, one can use this thermal probe in this upper limit. But like you, Carl_E, clearly explain the accuracy of the measurement will always be in question.

Possibly, though only at the user's risk, as it is beyond the manufacturer's top range for this equipment.

How would the user establish that <...1300^oC...> has been accomplished, so as to validate the use of it, though? One would need something else that is suitable for this temperature to do that, which means that the something else could have been used instead and would therefore be a better choice.

A 1250°C operating setpoint provides some margin to the top end of a Type K.

But the hazard faced in using at Type K at this temperature is that the controller is likely to report a false temperature value, one whose value has drifted because the junction has been polluted and the thermocouple is no longer a thermocouple junction of chromel and alumel alloys, but is now a junction of unknown alloys, for which the international tables of mV vs temperature are no longer valid. The junction pollution by the movement of metallic ions is greater the higher the temperatures.

The problem is that a drifted thermocouple has for all practical purposes failed, but it has not broken or failed open which all controllers have been to detect for decades. A drifted thermcouple continues to generate a mV signal, but the mV signal cannot be correctly converted to a temperature value because the junction is no longer a junction of pure alloys, it is a junction of polluted alloys for which there is no published reference data.

Drift is a serious factor addressed by AMS2750, the US pyrometry standard for many heat treat industry groups. AMS2750 has time-at-temperature restrictions which requires monitoring how many exposures a base metal (Type K) thermcouple can have before it has to be replaced, because of the probability of drift beyond the stated number of exposures. That number of exposures has decreased in succeeding revisions of AMS2750.

There is one US vendor who offers a proprietary insulating compound, 'MI-dry' (MI is Mineral Insulation, the white powder inside of thermocouple sheath), used in place of MgO, that maintains a higher resistance over time and temperature, reducing ion migration therefore reducing the tendency to drift.

US DOE researchers published data in the mid '90's showing that a T/C junction resistance increases as the pollution factor increases. Based on that research, some PID controllers, like the Honeywell UDC's, offer a "thermocouple health" feature that reads the thermocouple resistance on a periodic basis. A new, unused junction has a very low resistance, on the order of 1-2 Ohms. As pollution occurs and the junction resistance increases, the Honeywell can alarm at one of two preset alarm points - a 'warning' and a 'failure' level.

Taking some precautions to guard against drift might provide suitable service from a base metal Type K at the operating temperature. Otherwise, it's time for noble metal T/C's.

The Thermocouple is placed within a ceramic sheath . Does that help ?

All it will do is slow down the response of the thermocouple, which could be an issue as far as the control loop and its tuning is concerned. The Ziegler-Nichols method gives a number of approaches to loop tuning and it might be found that introducing a <...ceramic sheath...> into the mix will render the previous tuning for the loop inappropriate, rendering re-tuning a necessity.

If in doubt, consult a qualified Controls Engineer.

Wouid a ceramic protection help? I can't answer that. I just realized I've only ever encountered ceramic protection tubes in use with noble metal T/C's. I suspect ceramic tubes are even more porous than the alloy protection tubes, which would expose the junction to hydrogen, Mn, or other deleterious elements.

Maybe this is why (vendor recommendation):

"such conditions usually require a noble metal thermocouple such as Pt or Pt alloys"

The best summary of K/N thermocouple issues, noting both drift and the Green Rot phenomenon is here:

https://instrumentationtools.com/thermocouples-green-rot-effect/

Thanks to everybody. This discussion opened up many aspects of K Type thermocouple worth knowing and has much academic value if not anything else .

Regards

A. Mukerji