Level measurement problem 2

[bushwhacker]

I've run across an interesting continuous level measurement challenge. I need to measure the level of a mixture of Aluminum Oxide and ceramic powder, inside a closed vessel (approximately 48" high), that is kept under partial vacuum. Ultrasonic sensing is no good, because of vessel not being at atmospheric pressure (speed of sound different). If the vacuum was constant, I could apply a correction factor, but it isn't. Capacitive level sensing won't work, because the conductive powder will stick to the probe. I don't know if an RF-type level transmitter will work, never having worked with one, and the department paying for this project is reluctant to part with $$$$ for experiments. Any ideas would be appreciated. Thank you.

 

[madkungfu]

Have you considered load cells for the tank? I don't know if your application would allow for flexible couplings, to allow the tank to be freestanding..I have used Kistler-Morse load cells with pretty good results on fluid apps...10K gallon tanks.

Hope this helps..

MadKungFu

 

[ICman]

I think the load cell idea is good one. RF may have problems in the partial vacuum, as well. You could use nuclear, but then you would have the department really mad at you .

Can you use a float-type device? Or, perhaps a sensor that drops down, contacts the material, then raises? The latter isn't quite continuous, but I've used it in a similar situation (much larger tank).

I once used a "tuning fork" device as a level switch. They (Endress + Hauser, I think) may make a continuous device. Since it vibrates, I don't think the vacuum would affect it.

 

[nbucska]

Both the Al2O3 and the ceramic are very good insulators
with >2 dielectric constant -- measure capactance if the
composition of the mixture is constant.

<nbucska@pcperipherals.com>

 

[saxon]

bushwacker, load cells are good candidate, you might also want to take a look at radar level sensors. These things are reliable and highly accurate and all you need do is bolt them to a top nozzle flange. One advantage is that they are non-contact type sensors.

saxon

 

[RGCook]

A nuclear device will work but is not cheap.

I like saxon's radar suggestion. Suggest you check out and/or call Ohmart-Vega.

 

[nbucska]

Use a capacitive sensor and use a vibrator to shake the
sticking powder away.

<nbucska@pcperipherals.com>

 

[hacksaw]

vibrating probe or rotating paddle type sensors are commonly used for solids level in control and alarm functions. Cheap and reliable.

Load cells are used where more precise measurement of the quantity is needed.

cap probes are a non-flyer in the service your are describing. radar might work but you'll have to try it; there are more details involved in a proper installation and with a single point probe. Forget nuke meas.

 

[HamishMcTavish]

Forgive my ignorance but could someone tell me how the nuclear level sensing device works and what applications it is commonly used in?

Thanks, HM

No more things should be presumed to exist than are absolutely necessary - William of Occam

 

[IRstuff]

see:

http://www.omega.com/literature/transactions/volume4/T9904-14-RAD.html#rad_4

TTFN

 

[bushwhacker]

Thanks to all for suggestions - the load cell would be best way to solve this problem, and was the solution of choice. Unfortunately, the area manager decided that $600 was too much money to spend for instrumentation! He's decided that we don't need to measure that parameter after all.

&quot;You get what you pay for.&quot; (anon.)

 

[suriyab20001]

I believe rotating paddle type switches are used at most of the solid level measurments, since it is very cheap A still costlier is the mechanical rope type measurmentbut may needs a continoues maintenance . since it involves a motor and tape for auto inserstion and returning back. Load cells are static locating device which mayhelps for higher accuracy and precise measurment, sometimes may not worth for installing such expensive measurment.

i beleve for ur process paddle type level switch may be comfortable if the temperature is not too high enough.

 

[quark]

One problem with load cell based measurement is that the error in the reading gets multiplied depending upon no. of legs the vessel has, if you are measuring with only one load cell. If you use multiple cells, then the cost will be more.

RADAR type level transmitter can be a better option, if at all you want to measure the level.

Believe it or not : Had we trusted Archimedes and assigned him the work of lifting the earth(or any mass equivalent to that of earth on earth),with a lever of suitable length, it would have taken him 23 million million years to lift the earth by one centimeter, if he worked at the rate of 1 HP.

 

[JesperMP]

bushwhacker,

you should be able to make a load cell solution at approx half the stated 600$ price.

Use only one load cell.
Best is to suspend the entire vessel from one point (no calibration required).
Second best is to make a stand where the load is pivotted in one side, and placed on the load cell in the other side (as the load cell does not have 100 % of the load, some calibration and signal scaling will be required).

Shop around for load cells solutions. You should be able to get a reasonable deal if your precision requirements arent too stringent.

 

[jsummerfield]

I would start with radar. All three nuclear types are expensive and normally among the last choices. I cannot see why RF would be affected by vacuum. Only use capacitance where they are known to be successful. Consider load cells if it is a new vessel.

John

 

[SL1000]

Radar is a possibility, but may be touch and go with the quoted dielectric constant. This is on the threshold of many radar antenna systems. However if a fixing can be made on the base of the vessel then a guided radar can be used.

 

[xyzz]

SL1000,
Do you really think the guy needs an answer after half a year?