High temperature pressure detection 1

[sjewenp]

i'm not sure if this is the right roeum to post this on, but it was my best guess.
i am trying to measure the pressure in a loop that has a high temperature of 950 c. i know there really aren't any pressure detectors that can handle those temps, and a line must be used to cool the fluid. i had planned on using a 't'fitting and taking a line off the third branch.
my issue is i'm working with helium at 950 c and 30 bar. leakage is a major concern and i'm on a strict budget. does anyone know a cost-effective way i can measure pressure without breaking the bank or losing my helium?

 

[iainuts]

Hi sjewenp,
The title of your thread seems to indicate you're trying to measure temperature, but then you say:

i am trying to measure the pressure in a loop that has a high temperature of 950 c. i know there really aren't any pressure detectors that can handle those temps, and a line must be used to cool the fluid. i had planned on using a 't'fitting and taking a line off the third branch.

So you're trying to measure pressure, right?

If you run a piece of tube over to your pressure measuring device, then as long as it's very narrow with respect to its length, there will be a temperature gradient in the tube from hot to some lower temperature at your pressure transducer. If the transducer is put far enough away, it will be ambient temperature and you don't have to worry about temperature. Generally a few feet of 1/4" or 1/8" tube is sufficient. Put it 10 feet away with a 1/4" line and there won't be any high temperature at the pressure transducer at all, and the pressure measurement won't be affected either.

 

[sjewenp]

i understand i can run pipe to my detector, but how do i connect to a pipe running 950 c?

 

[iainuts]

Ah. You're talking about hot work. I'm sure there are plenty of folks who can give you advice about connecting to a 30 bar, 950 C pipe here, but I'm not one of them.

 

[moltenmetal]

Is the line carrying your fluid pipe, or is it tubing?

If it's tubing, you'll need to use a cone and thread type tee. Have a look at Autoclave Engineers etc.

If it's pipe, you're talking about welding a piece of pipe or tubing onto the line, with suitable branch reinforcement per the applicable piping design code. This branch line should extend through the insulation to a point where it's cool enough that you can weld or swage on an ordinary compression-type fitting (Swagelok etc.) to connect your impulse leg to.

Once the tee is in, an tubing impulse leg to your pressure instrument will ensure that the instrument isn't fried by the heat.

How do you minimize leakage? Minimize connections other than welds, and test these welded joints per the code: the test will be significantly higher than 30 bar, taking into account the significant ratio of stress values at your extremely elevated temperature relative to room temperature. After the hydrotest, test AGAIN with helium as a pneumatic test per code- helium can sneak through cracks that water never could. The more non-welded joints, the more potential for leakage especially once you're got the system hot.

 

[iainuts]

Hi sjewenp,
Could you clarify, do you need to make this connection to the main pipe while the pipe/tube is in operation at full pressure and temperature, or can you take the system down to do work on it?

Obviously, taking it down to do work on it is the easiest thing to do. I'm under the impression you want to connect to this pipe while in operation at full pressure and temperature.

 

[sjewenp]

i guess i'm not that good at making myself clear. the assembly hasn't been built yet. we're still in the design phase. the detectors will be installed while the loop is cold.

 

[sjewenp]

i'm relatively new to the design process. is there a reference i can refer to that tells me information regarding how long my branch line should be?
(i'm using 800ht pipe, by the way)

 

[unclesyd]

I would talk with these people about your high temperature pressure transducer application.

http://www.kulite.com/news.asp

 

[btrueblood]

You can also talk to PCB, they make a specialty piezo transducer with a helium bleed coolant (for rocket engine testing).

Also, google for fiber optic pressure transducers, and/or check out this link (there are more suppliers out there); optical transducers are starting to be readily available, and have higher usage temperatures than the standard electric types...not quite 950 C, but SiC (made by Luna Innovations) comes close to that, reducing the amount of standoff you'd require.

http://www.optrand.com/

http://www.isa.org/Content/ContentGroups/InTech2/Features/20023/January4/memsfeature.pdf

 

[moltenmetal]

You don't need any kind of optical pressure transmitter or anything for rocket engines here unless your process can't tolerate the impulse leg for some reason.

Since you're using 800HT pipe, you need a suitable WELDED branch extending through the insulation. Once you're an inch or two outside the insulation with your 800HT pipe or tube branch, a suitable connection can be made to transition to stainless steel. By careful selection of what you do, you can minimize the number of joints which might leak.

An uninsulated impulse leg of even six inches will drop the temperature sufficiently for you to use a common pressure transmitter (Rosemount, Yokogawa etc.)

 

[btrueblood]

Moltenmetal, your comment is probably true, but if the OP needs any decent frequency response from the transducer, then realize that it gets affected by the impulse line.

 

[moltenmetal]

It's a gas-filled system, and this is a gas-filled impulse line, not a liquid filled capillary. The impulse line isn't a big deal in terms of frequency response.