I’m working on a project that requires me to calculate the circular area of a column of falling grain that is being dispelled from a grain auger. The area is not constant (but the column falls fairly tight) so it is important to get rapid, accurate readings of the changing area.
I am thinking that setting up a circular ring of Infra red sensors that could be used to measure the distance from the ring to the column of grain (about 4-5 cms approx..). This could give me a plot of circular points that can be used to calculate the area.
Just seeing what would be the best way to approach this problem as I’m not going to be able to work on it until May. I’m new to Infrared sensors (and Arduino for that matter) so any feedback is greatly appreciated!
The angle of repose for the grain would be a constant I reckon (all things being equal like humidity), so if you know the amount of grain that has been deposited you could calc the volume of the cone and also the area of the base.
Dunno if that's even practical but just thinking outside the box in case.
Or are you planning to work the other way around, use the area of the base to calc the quantity. If so it's not a column but a cone so you still need to know the angle of repose.
If the height is a constant and the speed is constant, then it sounds like all you need to do is sample the diameter of the coluum fast enough and use the samples like "slices" that you "stack" mathematically on top of each other until it equals the height of the column. Each "slice" should be the "height" a single "grain" of grain (I don't know what they call it) . The question is can it be sampled fast enough and how ? How do you plan to use IR sensors , by detecting the reflection off the grain ? Would that be accurate enough ?
Optical sensing sounds like the most promising approach, but how clean is the air going to be around the falling grain? If it's dusty, it may not be easy to tell optically where the column ends.
Also, how often are you willing to go out and wipe the dust off the sensors?
I'd measure a WEIGHT of grain auger. Attach a wire to the top of this pipe, and depends how much gain inside, it will pull a wire with different strength. Knowing speed of deposition, it easy to calculate a flow in ton/sec. It would make sense to add up an accelerometer, as reported by wire (force sensor) weight 'd depends on angle.
I believe agricultural systems have been available for some time using differential GPS and augur sensors to collect data about the crop yield as it varies across a field. I'd expect the weight of crop harvested to be the data required. In some cases, this information can be used to give extra fertilizer to the under performing areas of the field.
I assume that the OP's project is doing something similar & wants to calculate the weight of grain being harvested at any moment.
Is there some way to sub-sample? Say have some minor split off that is an appropriate sub-sample of the whole, and then measure that sub-sample by weight every few seconds/minutes, then dump it? Sounds like a complicated geometry problem but any other solution seems complicated as well.
What exactly is the source of this question? That might help with better solutions.
You're still brainstorming an unknown problem.
When you make assumptions about issues, whithout sufficient information, Fate calls up a guy names 'Murphy'.
The picture is kind of a bad example of what I am interested in. The auger in question is an “8 inch” Sakundiak auger used for wheat. The grain falls straight down and has no horizontal component.
Modern harvesters do this but I am looking for auger usage like the one in that picture I will attach. The area will have porosity in it so this will have to be an experimented value that I will have to make an assumption and multiply this porosity by the area recorded.
“Optical sensing sounds like the most promising approach, but how clean is the air going to be around the falling grain?”
Dust will be an issue that is a concern with using optical sensing but how much of a concern is it? I could maybe have an attachable wiper or maybe a jet stream of air blowing from the top to blow the dust down. Also would the grain particles being round (and not flat) bounce the light source in different directions and not back into the receiver?
I think I might post this into the “sensor” section of the forums as well and see what people think.
@Graynomad
What my goal is to measure is the instantaneous area of the column of grain as it passes through the ring of sensors. What I think could work is a ring of multiple single beam sensors (the more the better) that send out a beam of light that hits the column of grain and returns it to the receiver. How fast the light goes out and returns can be used to measure distance.
The more points that we have (sensors) the more accurate our representation of the polygon will be to the actual column of grain. The shape of the column is changing constantly and will be circular-ish shaped NOT exactly perfect circle (which would be an erroneous assumption). What I want to know if this is possible to accomplish using IR sensors.
scotthetland:
Dust will be an issue that is a concern with using optical sensing but how much of a concern is it? I could maybe have an attachable wiper or maybe a jet stream of air blowing from the top to blow the dust down. Also would the grain particles being round (and not flat) bounce the light source in different directions and not back into the receiver?
The sort of optical sensor I had in mind is a simple sender/receiver pair which can detect whether the beam is being broken (i.e. by the falling grains). A few of these spaced at different distances from the center of the column of falling grain would enable you to estimate the radius and hence the area. (To deal with non-round columns would require a much larger number of beams.)
This is much simpler and cheaper than the 'laser tapemeasure' sort of measurement which seems to be implied by your original description.
This does seem like a better solution! This way you can set up a “grid” of sorts and instead of relying on the sensors to record data it will be simple binary (on/off). If the beam travels to the receiver we can say that that space is not occupied by grain, if the beam is disrupted we can say that the area is occupied by grain.
Wouldn’t this require a lot more sensors in order to make your grid accurate? And there is the concern of single grain particles that could part from the column disrupt the beam and skew the data.
What would be the price for an IR sensor transmitter/receiver pair? I guess if you are using sensors that will detect on/off instead of distance they would be much cheaper.
