Machine Vision to test spray pattern of water

[dblk22vball]

Hello all,

I am working on some test fixtures for our manufacturing plant, and it came up in a meeting that it would be nice to measure the height and pattern of water that is coming out of a spray head. From the spout, the water needs to be a certain height, and at that certain height, if you were to place a plastic template at that height, the water pattern should fill the outlined circle (see picture), and if it does not fill the circle, it is not good.

I was thinking that LabVIEW and some sort of vision should be able to do this, however I have not been able to find any information on this type of analysis. I did download the Vision assistant from NI, but could not get it to "detect" the water droplets properly (could be I am not 100% familiar with the suite yet). I was just doing a simulated analysis, not with any live equipment (camera).

Most of what I am seeing is that a part is orientated correctly, filled with water to a certain amount, etc, not measuring the height of the water droplets and if it fills an area.

Has anyone done anything like this before?? Thanks.

[crelf]

QUOTE(dblk22vball @ Feb 6 2008, 02:12 PM)

I am working on some test fixtures for our manufacturing plant, and it came up in a meeting that it would be nice to measure the height and pattern of water that is coming out of a spray head. From the spout, the water needs to be a certain height, and at that certain height, if you were to place a plastic template at that height, the water pattern should fill the outlined circle (see picture), and if it does not fill the circle, it is not good.

Maybe I'm not understanding your diagram - is the red disk in the left side view diagram the same red disk in the top view diagrams?

[Tim_S]

QUOTE(dblk22vball @ Feb 6 2008, 02:12 PM)

Has anyone done anything like this before?? Thanks.

I've heard of spray patterns being detected before, but it was with a custom-built capacitive-based sensor array in an R&D lab environment.

What you seem to be describing is a more 3D version of web analysis (I think that's the proper term), which is detecting features of anything streaming by at high rates of speed. Such applications include detecting applied glue, thickness of paper, and bubbles in molten glass (fiberglass production). For something at lower rates of speed, there is an NI demo of detecting how full a soda bottle is on an assembly line.

[dblk22vball]

QUOTE(crelf @ Feb 6 2008, 01:42 PM)

Maybe I'm not understanding your diagram - is the red disk in the left side view diagram the same red disk in the top view diagrams?

I'll try to explain a little more.

In my diagram, I have the height referenced from the spout. On the right, I have a rectangle (labeled plastic template) with 2 red circles. The red circle is the area that the spray pattern from the spout has to fill, ie the water should fill, if not over fill the red circle. The red circle on the left is a good (pass) pattern, the red circle on the right, is an example of a bad (fail) pattern. I show 2 red circles on the template, but in practice there will just be one. I showed both to try to illustrate a good and bad pattern.

The plastic template would be held and X distance from the spout, at the height shown on the far left. I hope this helps.

Ill also look into the 3D web analysis to see what is all involved in that. thanks.

[eaolson]

QUOTE(dblk22vball @ Feb 6 2008, 01:12 PM)

I was thinking that LabVIEW and some sort of vision should be able to do this, however I have not been able to find any information on this type of analysis. I did download the Vision assistant from NI, but could not get it to "detect" the water droplets properly (could be I am not 100% familiar with the suite yet). I was just doing a simulated analysis, not with any live equipment (camera).

Could you maybe fit the smallest ellipse that fits all your water droplets inside it, and compare that with your target?