Thermocouples Installation And Temperature Acquisition

Are you trying to measure the temperature at specific points ON the engine, or air temperatures AROUND it? Maybe both?

If you understand the CFD software model, you should be able to estimate the number of and locations where temperature measurements would yield optimum model results/verification.

This seems to me an extremely complex modeling exercise, not only because of the complex engine geometry, but also of the multitude of different geometries of the components around it.

I have done some CFD, and will help where I can. Hopefully there is someone here that has much more experience than me.

I'm sure that there are specific CFD forums out there that would probably give you more specific help.

Anyway, good luck.

You should also register. That way, you will get an email notifying you when someone replies to your post. If you don't you won't.

Are you planning on using mineral insulated thermocouples in a stainelss steel sheath?

I recommend
- un-grounded junctions
- "special limit of error"

The response is a little slower, but the probabilities for ground loop and common mode problems diminish drastically with ungrounded junctions. Ungrounded junctions cost a buck or two more than grounded junction.

"Special limit of error" halves the error allowed by the ANSI spec for commercial T/C's. It costs a buck or two more per T/C.

The measurement point is near the tip of the sheath, where the junction is, but the thicker the sheath, the heavier gauge metal and the more mineral insulation that the heat has to transfer through to get to the junction, further slowing the response time.

As for mounting, if it's for a short term test, the mounting only needs to be maintain the location and secure enough to not fail during the test, break or melt or come undone (plastic wire ties might melt or misform if too close to an exhaust manifold). I'm thinking superglue and baling wire. I doubt the adhesive used on those adhesive backed doodads used to bundle network cabling will hold up to the heat.

The wire between the thermocouples and the terminal block for the data acquisition must be Type K thermocouple wire or Type K thermocouple extension wire, not copper wire.

Keep drafts, like a cycling A/C duct outlet away from the data acquisition device because it has a temp sensor to measure the terminal block temperature in order to do 'cold junction compensation' (CJC). Some CJC is so fast that it does not reflect the slower temp change of the actual terminal block temperature and the result is an error.

Hi,

Thanks for your interest, thanks for your answer.

I'm using Omega «thermocouple wire special limit of error» K-type (TT-K-24-SL). see this link: http://www.omega.com/Temperature/pdf/SLE_Wire.pdf

for the mounting, the best is to apply the hot junction of the thermocouple directly on the surface and i will discuss this with my project manager if we can do this or not. If not, i think to use metal plate on the surface with a metal super-glue. On the exhaust, i will use kind of «v-band clamps» to adhere the thermocouple with the exhaust surface.

For acquisition, I'm using an EDAQ-lite (http://www.somat.com/products/somat_edaq_lite.html) with a thermocouple SMART modules (http://www.somat.com/products/accessories/thermocouple_smart_module.html) (provide isolated thermocouple conditioning).

I don't know if the wires length of thermocouples and the mounting method will affect the measurement accuracy!!

Your comments !

Thanks

My impression is that many CFD studies are done with bare beaded junctions. It isn't clear whether you need the speed of response of a bare bead or not. If not, there are ring style thermocouples designed to fasten with a fastener, but obviously more mass.