Measuring inside a pipe

[Willengos]

I am looking for sensors that will look inside a pipe. The flow is water with gel pellets and i need to know the size and number of the pellets and the flow velocity. Anyone could give me ideas of the kind of sensors that could be used?. The size of the pipe will be 7 cm and the pellets could vary from 3 to 50 mm.

 

[sreid]

If you can use a clear pipe a vision system could be programmed to identify and count pellets. Velocity could be determined by the travel dispalcement in two frame captures. Vision systems have become reasonably priced.

 

[Willengos]

Thank you sreid for your answer. I thought about clear pipe vision but I could not see how would it work for all the requirements. I understand the velocity measurement but to count the pellets and identify the size how would the vision system be applied bearing in mind that not all the pellets are the same size and the concentration is not constant. do you know?

 

[DesignerMike]

I think a Magmeter (flowmeter) can output line velocity. I'm not sure how the pellets would affect the accuracy though? I'm not an expert by any means, but there are a lot of different meters out there that can do some pretty neat stuff. You may find one that would be able to sense the pellets as the come past? Usually my customers specify what they want, but I have used Endress an Hauser and Anderson.

I agree with sreid about the clear section of pipe with different vision systems. I have sold a number of sight-glasses in tubing before...7cm is getting on the large side if you have significant pressures though.

What kind of pellet frequency are you looking at? could there be multiple pellets in the same cross-section of tubing at the same time? And what do you anticipate the line velocity to be?

 

[Willengos]

Thank you DesignerMike for your answer too,

Magmeters could be used since they provide mass flow rate which is what I need. The problem is that the pellets are of the same (or very close) density than the water and therefore i would know the total mass flow rate. With that information I wouldn't know how to get the mass of pellets. Any ideas?
In machine vision systems the mass is estimated based on the fractional fill of the screen occupied by the product. Unfortunately in this case this is not constant and it will be different for different cross-sections.

Thanks again
Will

 

[NickE]

You also might wnat to contact AutoKompu. They make various different process sensors for the mineral processing industry (ore recovery). It is vital to know size/rate etc.. info, and this company came Highly reccommended by the proffesor I had for that class... (or ask around till you find a Mineral Processing Engineer (you can call them slimers too...) )

good luck!

(maybe try

http://www.mtu.edu

and got to the materials science pages and look for Kohmar (first name) Kawatra)



Nick
I love materials science!

 

[sreid]

What I invisioned with a vision system is something that would give you an averaged, statistical number for particle number and size. Each frame could be analysed for partical size and the number of particles. The assumption would be that the partical satistics in the pipe would be represented by a 2D analysis. One could look from four sides to get a better count.