Imaging 3D Fuel Cone Spray Pattern

[jon_mcbee]

I am just starting to spec out a system for imaging the spray pattern of a fuel nozzle.  We want to be able to create a 3D image of the cone of spray.  I have been told that we want to use one camera that we can move in a circle of fixed radius around the cone in order to get the images.  This is a very different application than I have done before and am not sure where to start.  I am interested in using as much NI hardware as possible as this will be written with LabVIEW and integrated into a LabVIEW code base.  Has anyone solved a problem like this in the past, and if so I would love to hear the approach you took.

Regards

[seVEMes]

I have not done one of these applications before but it will take some thought for relating the angle of the spray pattern to the angle of the camera. Finding the angle of the spray pattern will not be too difficult using the vision VIs.

You would have to take into account the viewing angle of the camera if it is too close to figure out how to display it on a 3D graph. It sounds like a lot of geometry.

[jon_mcbee]

Thank you for replying.  After talking more with our customer I have a better understanding of the project.  The main goals are to be able to measure the angle of the spray pattern, as you mentioned, as well as be able to identify streaking in the spray pattern.  By streaking I mean that the nozzle is designed to create a spray of droplets and we need to verify that there arent any streams of water (in our test case) but a spray of droplets.  I dont know if streaking is the correct term for this as I am learning as I go, perhaps streaming is a better word to use. 

I am still trying to pickout the appropriate hardware for this application.  I don't yet know the exit velocity of the spray from the nozzle or the average droplet size so I am not sure how many frames per second I will need to capture droplets.  Truth be told I dont know how necessary it is to capture individual droplets, as long as I can somehow differentiate between droplet and stream.  Because the hardware will not be cheap I am looking for any and all advice or recommendations. 

I appreciate the help!

[seVEMes]

Well, from a software perspective this should work fine with the Vision Development Module.

From a hardware perspective you could probably use many types of hardware and still get this to work. I don't recommend USB cameras. Camera Link or FireWire are probably good choices. Choose a camera first. Makesure it can get the frame rate and resolution that you need. Then choose the card and you should be set for hardware from National Instruments. Obviously you would have to find a third party mounting system. You could also control the motor with National Instruments controlers and drives if you would like.Then comes the programming.