Heat Treat Today publisher Doug Glenn has a second conversation with long-time thermocouple industry expert Ed Valykeo from Pelican Wire about T/C accuracy and classifications. Listen to learn more.
This is the second episode in a series of three on Thermocouples 101. Check out the first episode of the series here.
Below, you can either listen to the podcast by clicking on the audio play button, or you can read an edited transcript.
The following transcript has been edited for your reading enjoyment.
Doug Glenn (DG): Ed, welcome back. I'm glad you were brave enough to come back. Last time, Ed, we talked about a lot of good basic thermocouple stuff. We talked about, basically, Thermocouples101 which I mentioned last time, was one of the best and most well read articles on our website, which is great. We covered a lot of different things last time. I was just reviewing it, and it's interesting, we were talking about several different men as you gave a good history of thermocouples starting back in the early 1800's and talking about guys like Alessandro Volta, where we get the word volt, and Thomas [Johann] Seebeck and the Seebeck effect or the Seebeck coefficient, and things of that sort. We talked about all the different noble thermocouples, J, K, E, N, and T, and we talked about the N leg and the P leg on all of those which was all good. It was very interesting. If you didn't listen to the first episode, you ought to go back and listen to it. It's really a pretty good summary of thermocouples, a basic primer on thermocouples. We also did some things like vocabulary for ourselves; we learned what an EMF was, electromotive force and things of that sort. It was very good.
This time, I think we want to move on to, what we could commonly classify or in a big picture classify as, standardization and accuracy discussion. But, before we do, I've got a quick follow-up question from the last episode. We had mentioned that an EMF is produced when two dissimilar metals are joined together or placed together. There is a very, very, small electric current that's created. My question is: Can you do that with any metal? Is it possible? Or do you have to have only certain types?
Ed Valykeo (EV): Theoretically, yes, you could probably join any two different metals and produce some sort of voltage. However, the accuracy of that, and if doesn't mean anything, probably not. The thermocouple base metal thermocouples that we talked about last time, are industry known, used worldwide and, quite honestly, have been perfected over many, many years. So, yes, you could generate a volt probably from any two metals, but, really, to produce an accurate thermocouple, something you can measure temperature with, you're going to want to stick to the thermocouple types that we talked about.
And again, today, we're talking about the base metal thermocouples which are known as Type K, Type J, Type T, Type E and Type N. Those are the base metal thermocouples.
DG: Let's talk a little bit about standardization of these things, and accuracy. My understanding, Ed, is that there are one or more organizations out there in the world that deal with certifying, qualifying, or giving us standards for these thermocouples. Can you tell us a little bit about those organizations? Then, we'll jump in and talk more specifically about the classifications and accuracy.
EV: Sure. One of the bodies that we use is ASTM. In ASTM-E230, are all the thermocouple tables for the different types of thermocouples, not just the base metal, but also noble metal. It's a fairly lengthy book. All the thermocouples are based on the ITS-90 scale and that is the EMF output of each one of these thermocouples at prescribed temperatures. We could go into more detail with that if you'd like, but there are a number of ways that they have extremely accurate temperature medium to measure the thermocouple output. But, that's what the tables in ASTM-E230 are based on, the ITS-90 scale.
When we talk about ASTM, there are also a couple of other standards that we use, and we'll probably get into a little bit later in the conversation when we talk about calibrating the thermocouples themselves. So ASTM-E220 and ASTM-E207 are the two that are used in calibration of the thermocouples.
DG: But, basically, the organization that does that, I don't know if we want to call them a lab or not, but the organization that does is it ASTM.
EV: ASTM is one of the bodies that publishes the books that I call the standards for thermocouples. I think I won't be mistaken, but ITS-90 is really more an IST list. They control the ITS-90.
DG: Let's move into the accuracy standards, then. I think you mentioned the ASTM-E230. Is there anything else we need to talk about as far as the accuracy standards, or did we already hit it?
EV: Certainly, in the ASTM-E230, they spell out the different types of thermocouples, as I mentioned, the base metal thermocouples, but the accuracy of each one of those is listed in the ASTM-E230.
DG: What about classification? Let's talk about the guidelines for classifying these different thermocouples.
EV: Again, ASTM-E230, and there are other publications, but, again, we use ASTM here. The classification of the thermocouples are also spelled out in ASTM-E230 and basically, we talk about special limits of error, standard limits of error and extension grade thermocouple. Again, those can be found in E230.
DG: So, when we classify those, are we classifying them based on temperature deviations or the temperature tolerances? Is that, basically, what it is?
EV: Yes. It's based on temperature tolerance. I'd like to share a quick rule of thumb for classification of those thermocouples. So, special limits of error, basically from zero degrees Fahrenheit to 500 degrees Fahrenheit, it's + or - 2 degrees, and above 500 degrees it is + or - .4%. For example, at 1000 degrees, you're looking at + or - 4 degrees; if you have 2000 degrees Fahrenheit, the tolerance at 2000 would be + or - 8 degrees for special limits of error.
On the other side of that, you've got standard limits of error, and, basically, you could just double that. From zero to 500 degrees Fahrenheit, you're talking + or - 4 degrees; at 1000 degrees would be + or - 8 degrees and at 2000 degrees, + or -16 degrees.
Where there is some confusion, and maybe some people don't understand thermocouples, is when we talk about extension grade. There are actually two types of extension grade. There are standard limits of error and special limits of error extension grade. Extension grade is just exactly as it sounds. It carries that signal from your sensor all the way back to instrumentation rather than run maybe a little more expensive wire all the back to your instrumentation, you're going to put extension grade to continue that circuit back to the instrumentation. Extension grade is the same metals as the thermocouples. If you're using Type K sensor, then you're going to want to use Type K extension grade, and so on, for the rest of the base metal thermocouples. The difference is that the extension grade material is only guaranteed to meet the tolerances up to 400 degrees Fahrenheit. If you look at ASTM-E230, the tolerances only go, on extension grade, to 400 degree Fahrenheit. And, actually, Type T is a little bit different; Type T only goes to 200.
DG: In the heat treat industry, that's not really going to do us much good, right? I mean, most of our processes are well above 400.
EV: It is. That's why you would never use an extension grade as the actual sensor. This is some of the confusion out in the industry: “Well, I can just take my extension grade, create a junction and use it to measure temperature.” You probably could up to 400 degrees, but it's not guaranteed above that temperature, and you could get yourself in trouble.
DG: So, you run extension grade outside of the furnace because, obviously, you're not above 400, so you can use extension grade to run it. I think last time we talked about no more than 100 feet rule of thumb.
Extension grade is basically this: Here's your extension cord that you can run from your regular wire, either your standard limit of error or special limit of error, from that to the box.
EV: Exactly. And so, the key to understanding extension grade is the tolerances on that extension grade are the same – say if you have special limits extension grade – it's the same as your special limits thermocouple wire, + or - 2 degrees, in this case, up to 400. It's guaranteed to meet special limits of error and then the same thing on the standard limit side. You just double those tolerances. Again, it's really the temperature that it is guaranteed to.
DG: Very good. So those are the different classifications. We've got special limits of error, which is a tighter temperature tolerance, and then we've got standard limits of error, which is a little less tight, and they we've got our extension grade which is only classified up to 400 degrees anyhow.
I know some heat treat processes require very, very tight temperature tolerances, especially in things like aluminum brazing and things of that sort. Is it possible to get anything better than special limits of error?
EV: It is. The first thing I want to say is that they're not really recognized within ASTM, these tighter tolerances. But, in the industry, certainly in heat treating and in the pharmaceutical side where they typically use Type T, we've had many requests for tighter tolerance material. Some people call it quarter limit material or half limit material, there's a bunch of different names that it goes by. So, we go to our manufacturer's of the wire and request that and, most of the times, it's a no quote. It really comes down to more of a selection process.
For us here at Pelican Wire, we have a pretty good sized stocking program of bare conductor and sometimes what we can do is mix and match to try and meet the tighter tolerance material. There are a number of ways that some of the manufacturers, in fact, the heat treaters, will request special limits materials, that must meet + or - 2 degrees up to 1000 degrees and then .2% after that. It can be done and we do it on occasion.
DG: Let's follow up on that a little bit. How do you determine the accuracy of a lot of wire, or a spool of wire? How do you go about doing that?
EV: Let me back up just a little bit and start with the actual wire producer themselves: There are not any left in the States, so, basically, all the thermocouple wires are melted overseas, whether it be Germany, France, Sweden. When they melt, they try to meet special limits of error. Now you're talking each leg has to be melted separately; they don't melt them all at one time, right? So, each “melt” or “heat”, they are shooting to make special limits of error.
This is where some of the testing specifications come into play. ASTME-207 is a test method for single thermal element thermocouple wire. I don't want to confuse our listeners, but, again, if you think about a melter that just melted or heated a melt of wire and they process it down to wire, they only have one conductor. They want to know if that one conductor is going to potentially meet special limits of error. There is a testing specification that ASTM has (ASTME-207) that you can test a single leg thermocouple wire to see if it's going to meet special limits of error. What they do is they calibrate the single leg, they get their values (the EMF output), and they have the second other leg and they do the same thing. They, then, mathematically add the EMF of those two and go back and look at the standards to see if it's going to fall within the special limits of error.
That's how the melters, the folks that are melting the individual thermocouple legs, are doing it. We users, we are an insulator wire, we put the two legs together and now we have a thermocouple. The way we test those thermocouples is by using an ASTME-220, which is a comparison method. We're taking a known standard and we're calibrating the thermocouple wire against that standard and getting the temperature deviation from that. That's how we verify that the wire is meeting the tolerance that is requested by our customers, whether it's special limits of error, standard limits of error or even extension grade.
DG: When you say "a standard", what does that test actually look like? Are you taking a thermocouple that you know is good, sticking it in a hot furnace and your test thermocouple or are you just doing it through current testing or something like that?
EV: Good question. We actually use SPRTs (resistance thermocouples) high accuracy, that we use as our standard. They're calibrated at an outside firm, so we know what the output of that resistance thermometer is, and we calibrate our sample against that. The three things you need to do a temperature calibration is the temperature medium, the reference thermometer and the equipment to capture that output or measure the voltage that's being produced. Having those, we have our reference standard that we know the EMF or the temperature output of. Now, we put our thermocouple in the furnace and we compare the two. That's how you get your deviation.
DG: There are labs, I understand, that do these certifications and things of that sort, that certify the accuracy of the thermocouple. Now, Pelican Wire does that. You do have a lab and you do certifications, right?
EV: We do. We calibrate the thermocouples and we produce a test report showing the deviation of the thermocouple for the customer.
DG: Earlier, we were talking about standards and how there's the organization ASTM. How about for these labs? Do the labs have to meet some sort of outside third party certification?
EV: There is nothing that they have to do. I will say that there are a number of standards. We're ISO9001, but we're also seeking accreditation for 17025 so that our lab is accredited to IECISO17025, which just proves that we are a quality lab. We have our quality systems in place. We have our uncertainty budgets for all the equipment we use. A customer can feel confident that the calibration report that we provide is as accurate as possible.
DG: I think covers most of the things we wanted to cover in this episode. We talked about the standardization, the special limits of error, the standard limits of error, who are the bodies out there that do the certifications/classifications, if you will. I think we covered a good bit.
I think we were going to do one more episode, Ed, and I think we're going to talk about insulating materials. I understand that one of your colleagues is going to be there to talk about that with us, John Niggle.
EV: Yes. John Niggle will join the next podcast and talk a little bit about how now that we have the thermocouple wire, what kind of insulations do we put on that wire. It depends on the medium that it's going to be used in, the heat treater or whoever.
Doug Glenn
Publisher
Heat Treat Today
To find other Heat Treat Radio episodes, go to www.heattreattoday.com/radio and look in the list of Heat Treat Radio episodes listed.
