Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

Register Log in
  • Congratulations cowski on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Designing a flow-through electrical resistivity cell 2

Status
Not open for further replies.
j.doe

j.doe

Electrical
Jun 6, 2019
32
Hi, I need to design a flow-through cell to measure the electrical resistivity of pressurized dielectric liquids when flowing across a tube. The main liquid I want to test is 134a which is thought to have a resistivity of around 10^12 ohm-m (in a static state). I want to design the cell so that the cell constant is less than or equal to 0.113/m because cells with this cell constant have been proven to be able to measure resistivities that high. There are no commercial or academic studies done on flow-through cells that measure such high resistivity values (to my knowledge).

My first question is how do I calculate the cell constant for a flow-through cell with two cylindrical electrode rings in an insulating cylinder? I understand the general equation but i'm not sure how to apply it to this specific case, as in, which surface constitutes the area and length? Another question I have is how do I attach the pieces together? I’m thinking of using PEEK for the insulating tube and have these ring electrodes on the inside of the cylinder somehow. Should I glue them somehow on the inside and drill a hole in the PEEK cylinder for the wires to connect? Thanks in advance.
 
Replies continue below

Recommended for you

Sort by date Sort by votes
You may get better accuracy with a screen, depending on the spacing between electrodes.
It may not matter that much if you are more interested in a change in resistivity than in the absolute value.
Are you trying to detect the breakdown of the oil or of the motor windings in a refrigeration system to avoid an unplanned failure?
It sounds like an interested project.
Alternately, you could use a piece of insulating tubing between two sections of conductive piping and use the piping for electrodes.
Or, practically, a section of insulating tubing to a metallic nipple and then another section of insulating tubing connecting to the main metallic piping system.
The nipple may be energized and insulated and all other metallic piping kept at ground potential for safety.
You will be measuring the resistance from the nipple in both directions.
Using rings runs the risk that much of your measuring current will be conducted and diverted through the metallic piping near each ring.
That will surely degrade your reading unacceptably.

Bill
--------------------
"Why not the best?"
Jimmy Carter
 
Upvote 0 Downvote
@waross
waross said:
You may get better accuracy with a screen, depending on the spacing between electrodes.
Click to expand...

Thanks for the reply waross. I really like your nipple idea. I was wondering what you meant by the quote above. This project is to understand the resistivity of dielectric propellants used in pharmaceutical inhalers when flowing within the valve stem of inhalers. We're not interested in the breakdown strength. I was also wondering if you have any ideas about how to calculate the cell constant in this situation? Thank you.
 
Upvote 0 Downvote
@IRstuff Does that mean the area for the cell constant is calculated from the surface area of the anode in contact with the liquid while the geometry of the cathode does not matter? Thank a lot for the link btw, that was really helpful.
 
Upvote 0 Downvote
As i understand the cell constant, it is based on the electrodes being plane surfaces that are parallel.
A screen will give you parallel electrodes.
The ring will be an annular electrode and it will be difficult to calculate the cell constant as all parts of the surfaces of the electrodes will not be the same distance apart.
Also, the current density in the liquid will not be constant in the area near the annular electrodes.
However, as the annular electrodes are moved further apart the percentage error will lessen.
You will have the same effect if you use a nipple and the existing piping for electrodes.
Best to determine the cell constant by calibrating with a liquid of known resistivity.

Bill
--------------------
"Why not the best?"
Jimmy Carter
 
Upvote 0 Downvote
waross said:
A screen will give you parallel electrodes.
Click to expand...

@waross Hey waross, Can you please explain what you mean by a 'screen' in this context and how that would give me parallel electrodes.

I think you're right about this statement, i might need to solve it numerically via FEA instead using COMSOL:
waross said:
it will be difficult to calculate the cell constant as all parts of the surfaces of the electrodes will not be the same distance apart.
Click to expand...

The problem with calibrating it is that it would be a lot easier if i could approximate the cell constant before designing and then calibrate it after construction. It would save a lot of time which i don't have. In terms of the fringe field effect, i think i might add a guard electrode on the inlet side of the anode to reduce this effect.
 
Upvote 0 Downvote
@IRstuff i understand that but that would require further adjustment of the design of the cell if the constant was found to be too high. I thought it'd be better to try to approximate it numerically and base the design on this approximation before constructing the cell, thereby saving time.
 
Upvote 0 Downvote
We used to do this with oils, but I wasn't the instrumentation guy I just built it.
The body was virgin unfilled PEEK with about 3/4" ID. We wanted slow enough flow that we wouldn't have too much turbulence, the rest of the line was 3/8".
The plates were about 5/8" square gold platted Cu.
There were a pair of longitudinal grooves in the ID of the PEEK that the plates slid into to hold them about 1/4" apart.
The leads came out though the PEEK and were heat sealed.
The fittings on the ends of the PEEK were tied together with a strap and then grounded. In some fluid we had issues with static charge build up.
Now all of that said, If I was working with fluids that had lower resistivity I would go with a pair of annular rings as the electrodes. I don't see any reason to not try using fittings separated by pieces of plastic pipe. But you might need to make the parts special so that you can have a nice smooth ID (without steps in diameter). I could see using three pieces of plastic with two Cu fittings between them used as the electrodes. This configuration should work OK with larger diameter fittings. In small diameter I would be concerned about the variations in charge density along the electrodes impacting the measurements.
Our calibrations were all empirical, using samples that we tested side by side with our standard static test unit.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy
 
Upvote 0 Downvote
@IRstuff That's actually a good idea, thank you. If i did that, would I then just add all the cell constant values for all the chunks to get the overall value?
 
Upvote 0 Downvote
@EdStainless I don't understand what you mean by plates, how does it look like exactly with a plate configuration? Also, I don't quite understand what you mean by this statement:

EdStainless said:
In small diameter I would be concerned about the variations in charge density along the electrodes impacting the measurements.
Click to expand...

Surely, a high charge density would be good to get a good current signal for dielectric liquids.
 
Upvote 0 Downvote
Screen. Visualize the flow completely blocked by a plate.
Now visualize a lot of holes in the plate.
Now replace the plate with a screen.
For fast flows, the turbulence may make this unusable.


Bill
--------------------
"Why not the best?"
Jimmy Carter
 
Upvote 0 Downvote
Actually the plates were edgewise to the flow.
You do need some electrode area, otherwise you would just use a couple of pieces of wire.
Back when I did this all of the equipment was analog, and getting signals was a real pain.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy
 
Upvote 0 Downvote
Thanks a lot guys. You've given me so many good ideas i haven't even thought of. I'll do the simulation on comsol to get an equation for the cell constant and then decide whether this plate/screen design is better or the ring one. [peace]
 
Upvote 0 Downvote
Status
Not open for further replies.

Similar threads

Cesme
  • Locked
  • Question
Soil electrical resistivity measurements for a wind turbine location
Replies
1
Views
3K
7anoter4
7anoter4
srini1785
  • Question
IP65 PROTECTION CLASS FOR NON ELECTRICAL EQUIPMENT 1
Replies
17
Views
8K
TugboatEng
TugboatEng
Hafzkk
  • Locked
  • Question
Software for electrical Simulation
Replies
1
Views
539
Overvoltage
Overvoltage
Coco_HueHueHue
  • Locked
  • Question
Electrical Contact Grease Question
Replies
5
Views
2K
Coco_HueHueHue
Coco_HueHueHue
K
  • Locked
  • Solved
Load Flow Analysis in SKM
Replies
7
Views
3K
Kanu_01
K

Part and Inventory Search

Sponsor

Back
Top Bottom