Density and Size Meaurement of rubber granules

N

Thread Starter

NB

I have vessel where 90% water level is maintained at 150 DegC and 3.5 Kg pressure. The temperature is maintained through steam. The feed is mix of steam and 1-3butane. Rubber granules of 6 to 7 mm size are produced inside the vessel. The water is kept agitated by a stirrer. Currently the operator views the size and concentration of the granule through site glasses and manually adjust the steam flow. I want to automise the process.

Please advise how to measure primarily the concentration and secondarily granules size in the vessel.

The slurry is being transferred to a stripper though 2 inch pipe. The measurement can be done in the pipe also.
 
C

Curt Wuollet

I might put a rectangular transparent section in the pipe
and set up a shadowgraph arrangement with a monochrome
camera. Grab a frame with a light pulse (if necessary) and
count blobs for density and pixels for size. The software
to do this wouldn't be difficult to write and the mechanical
dimensions of the section could be varied to obtain enough
separation for accuracy. But then I'm just an unemployed
automation bum and what do I know?

Regards
cww
 
R
In mining they sometimes use a particle size monitor but they usually require a sample. I have been out of that industry for a while so there may be some new developments. Try Outokumpu.

Roy
 
The density of the granules/crumbs are reasonable maintained. The specific density if 0.97.

Regards, NB
 
It seems to be a workable idea if no existing instrument is found. Can u please elaborate or give some further reference?

Regards
 
C

Curt Wuollet

Let's say you made a fixture with two parallel, transparent, polycarbonate plates spaced 1/2" apart. If they were say 6" by 8" that would be convenient for a commodity camera. You put frosted Mylar on one and backlight with an LED array. This gives you an evenly lighted background. The 2" pipe has an area of about 3" give or take depending on wall thickness. You adapt to cause the flow to go between the plates which are sealed top and bottom. If you were to watch, the particles are flowing by and are sharply contrasted against the background. Set up a camera to cover the area. Grab frames. If a sample has enough clear space that the particles are mostly separate, great. If not, moving the plates closer and/or making them larger will adjust.

When you can count the particles easily manually, you write software to grab frames into a buffer and do a threshold function to give you only black or white. Then you process the array to identify the particles as areas where adjacent pixels are black. You might as well count the pixels while you do this. You count the black areas to determine the density and you average the pixels per area to determine the average size. Or you can determine max/min etc. It's all just data at that point. If the flow causes blurring at frame speed, you might have to upgrade to pulsing the led array and use a camera that can sync to the flash. All of this is fairly cheap if you use the commodity cameras and hardware.

I've done this sort of thing for various physical measurements and it works well. The Linux tools are free. Once you have the numbers you can do direct control with the Linux box or send the numbers to an external controller. I've done both. There might even be commercial camera systems that can do this, but they are expensive and very difficult to stabilize as you don't have low level control being general purpose items. It can get compute intensive, but PCs today have tremendous capability.

You need

Camera.
Frame grabber card.
PC.
The fixture,
Software.

Regards
cww
 
This is where I am tempted to consider a coriolis mass flow meter either Emerson (straight tube) or E&H Promass. These measure density as well as mass flow.

You should be able also to get %mass or % volume calculations from it.

The reason I suggest mass flow is because you can get a straight through mass meter at your pipe size and because I suspect insertion type density meters may not be too clever with the particle sizes you have.

You don't care what the actual flow rate is since the only reason to use the coriolis is because it gives you a full bore density meter.
I would suggest asking the manufacturers the question. Be nice if you posted their response back here so we can learn from it.

Particle size is something I can't help you with but it might make sense to separate out the two different measurements.
 
Top