Sensor Guidance for Weight Calculation (Not that easy)

Howdy,

I've got a project that will require determining the average weight of objects. This implementation will have some caveats, though.

The objects will be hay bales that are formed from 15 different individual sections that we will call "charges" in this example. A piston/ram will push a "charge" of hay against the others, which are held into a compression chamber secured on all sides -- this creates a density per "charge."

Envision the following:

Movement Path >>>

===|[15]  [14][....][02][01]

Piston      Compression

In the above, the piston is at the far left, which is pushing a "charge" numbered [15] against the previous charge [14], which is in an enclosed compression chamber (where the sides exert pressure to prevent it from moving backward). Once it has 15 "charges" the entire package will be bound together with string, and it will be pushed out of the compression chamber as another bale is formed at the front, and the process starts again.

I would like to determine a somewhat accurate weight of the bale.

Typically this is done by having a device that sits behind the compression chamber that a bale would fall into (one at a time) with a load cell at the bottom.

I'm hoping this can be done by either an in-line load cell at the back of the chamber, where we can measure just a small subset of the full bale length and perform math to get a close estimate -- or perhaps using other sensors calculate the weight based on the number and depth of "charges" against some constants such as dimensions, type of material, moisture content percentage.

Thanks in advance.

You can have a load cell for the entire compression chamber and watch the weight increase with every charge.

Otherwise you run into multiple problems. You have to know how to estimate the weight in an alternate way, based on yet unknown properties of the hay, by use of yet unknown sensors, for some yet unknown purpose. If your research in the literature or internet provides some more meaty facts you are welcome in this forum for further assistance.

DrDiettrich:
You can have a load cell for the entire compression chamber and watch the weight increase with every charge.

I don't think that would work since the only place to mount in the compression chamber would be on a surface that is actually applying considerable force against the hay to compress it -- to get any sort of number that we can use to calculate weight of the hay would almost require it all be resting on the load cell, right?

You may get better help if you post some photos or drawings of the compression chamber and surrounding structures/appurtenances.

Also, what's wrong with the "typical" approach?

DaveEvans:
You may get better help if you post some photos or drawings of the compression chamber and surrounding structures/appurtenances.

Also, what's wrong with the "typical" approach?

Let me know if this drawing helps at all.

The side plates are semi-transparent so you can see the inside. They are static. The orange plates (top and bottom) are held at a specific distance using the pressure cylinder, which causes the hay to compress.

The typical approach may be fine, but it will be a lot more expensive to create, and it might not work in all scenarios -- as there may be other items behind the compression chamber.

It may be possible to determine a weight by its oscillation in some elastic environment.

DrDiettrich:
It may be possible to determine a weight by its oscillation in some elastic environment.

Hmm, that sounds interesting. Do you have any specific material or pointers that I can research?

For further clarity -- the inside of the chamber has wedges that will keep the hay from "springing" back towards the charging ram once it retracts to fill a new "charge."

If it's possible to weigh a charge before compression, just add them up.

JCA34F:
If it's possible to weigh a charge before compression, just add them up.

Hmm, yeah there might be about 500ms or so when the hay is laying in front of the charging ram.

By design, I'm not sure I could get a sensor in there though, as the ram would then contact it when it moves.

I wonder if we put a load cell under a plate that is only about 4" wide at the very back of the chamber, and then try to take a measurement as the hay comes out of the chamber, if we could perform some math to make it fairly accurate.

There would never be a time it doesn't have hay on it since it would all come out end-to-end without any break between the bales, but with another sensor we could track the linear movement to know when a full bale has passed over the sensor -- then maybe just... idk, avg the readings accumulated since the last trigger.

Still sounds flawed.

travis88:
Hmm, that sounds interesting. Do you have any specific material or pointers that I can research?

Hooke's law explains how the bouncing frequency of a mass on a spring depends only on the mass and the stiffness of the spring (harmonic oscillator).

I also wonder what's the purpose of your intended data acquisition. The mass of a bale can be measured easily after composition.

DrDiettrich:
I also wonder what's the purpose of your intended data acquisition. The mass of a bale can be measured easily after composition.

It would be important to measure the weight alongside other sensors so that you can ensure you're always making a proper bale that isn't too heavy, or too light -- especially as field conditions change.

Sounds like a standard hay baler to me that is moving in a field, you would then need to deal with the bumps and such and vibration of the baler. Maybe some way could be found to measure the moisture in the hay, but that is beyond my pay-grade.

Hi,
Will this be on a mobile machine?
Hay Baling Machine?


Tom... :slight_smile:
PS 10sec later..... :slight_smile: :slight_smile: :slight_smile:

saildude:
Sounds like a standard hay baler to me that is moving in a field, you would then need to deal with the bumps and such and vibration of the baler. Maybe some way could be found to measure the moisture in the hay, but that is beyond my pay-grade.

You are correct! Moisture shouldn't be an issue with a conductivity moisture sensor. But, I would still need to know the weights or density of the material to then calculate.

That would be the best way from a perspective of limiting fabrication -- be able to calculate compression density from the linear movement of each "charge" stroke, and know the moisture percentage and total bale dimensions. But, I'm not sure how error prone and accurate that would be.

It's easy to just say that a 39" bale would have 15 "charges" and weigh 60 lbs at 15% moisture, then divide by "charges" and depth of a "charge" -- but, accuracy might be an issue.

TomGeorge:
Hi,
Will this be on a mobile machine?
Hay Baling Machine?


Tom... :slight_smile:

Yes, that is correct. It will be moving.

So how will you tag the individual bails with their weight if that is important? If it is a more "general" thing, then you may be generating more overhead than it is worth. I assume this is some type of school project.

zoomkat:
So how will you tag the individual bails with their weight if that is important? If it is a more "general" thing, then you may be generating more overhead than it is worth. I assume this is some type of school project.

Not a school project -- will be in production.

Tagging isn't needed. This is just to establish a baseline average to ensure the vast majority of bales produced are within a margin of tolerance. It will also be used to calculate a chemical application at a rate per x tons.

"Tagging isn't needed. This is just to establish a baseline average to ensure the vast majority of bales produced are within a margin of tolerance. It will also be used to calculate a chemical application at a rate per x tons."

Sounds even more like a school project. I suspect the baler functions to make bales of a standard size independent of the amount of square feet of ground required to form the bail. Bales per acre might be more functional as to what is happening in the field. Just seems to be too many assumptions being made for accurate data for the end result you seek. Just saying.

Dealing with the dynamics of a moving machine and moving bale would be a considerable challenge, but you may be able to get something useful from the tension in the chain (red dot) provided the cantilever portion is hinged (blue dot).

Annotation 2020-01-10 090824.jpg

Annotation 2020-01-10 090824.jpg