calculating delivery rate of running conveyor

realolman

Member
R
Join Date
Mar 2009
Location
here
Posts
584
I have a vibratory conveyor that is being fed from another conveyor. The feeding conveyor does not always deposit a consistant flow of "product" onto the vibratory conveyor... sometimes there is very little and sometimes there is a lot. Nothing can really be done about that... just the way it is.

I can weigh the product that is on the vibratory conveyor at all times, and that value is an analog input to a PLC. That is already in place

How can I determine, using the PLC... even a pretty good approximation.... the amount of product being discharged from the vibratory conveyor . The PLC has PID capability (a Micrologix). I imagine this has something to do with integration... I guess I'm asking how to have the PLC do the math,

Thanks
 
Hello realolman;
Integration is simply a sum (mathematically infinite, but not in the real world) of values captured in small, standard timeslices.
What you are trying to do is called a "totalizer", which will give you a mass value out of a mass flow.
Search for totalizers in the forum, and make sure you specify the make and model od PLC or PC program you will be using, math functions (and interrupts, and cyclic blocks...) vary wildly between brands.
Basically: Take a measure of the mass on the conveyor at standard, recurring times, that will give you a picture of the mass on that section of the coveyor. say, once every second 9you will adjust this based on your own application. Then add this value every second to a register taht will "totalize" the recorded values; then, every minute, or hour, or whatever time you need to evaluate the totalized mass, move the totalized value to another register or FIFO table so you can display, monitor, trend... the mass value on the conveyor.
Hope this helps,
Daniel Chartier
 
dchartier said:
Hello realolman;
Integration is simply a sum (mathematically infinite, but not in the real world) of values captured in small, standard timeslices.
What you are trying to do is called a "totalizer", which will give you a mass value out of a mass flow.
Search for totalizers in the forum, and make sure you specify the make and model od PLC or PC program you will be using, math functions (and interrupts, and cyclic blocks...) vary wildly between brands.
Basically: Take a measure of the mass on the conveyor at standard, recurring times, that will give you a picture of the mass on that section of the coveyor. say, once every second 9you will adjust this based on your own application. Then add this value every second to a register taht will "totalize" the recorded values; then, every minute, or hour, or whatever time you need to evaluate the totalized mass, move the totalized value to another register or FIFO table so you can display, monitor, trend... the mass value on the conveyor.
Hope this helps,
Daniel Chartier
Click to expand...

Sure does.... now that you have articulated it (extremely well), it sounds pretty simple... This is an allen bradley Micrologix PLC

Although I imagine it is easier said than done.... I'm gonna give it a whirl... thank you.

On second thought, though,I am not sure I want to totalize it... I am more interested in how much is coming off the conveyor... perhaps in similar "time slices" as it was acquired.

Although the flow onto the conveyor is heavy and light, it is evened out somewhat by the vibration of the conveyor itself.

I am then mixing it with another ingredient and I would like to be able to control the mix a little more tightly by knowing how much is leaving the vibratory conveyor.

At any rate, your very well articulated answer without a whole lot of formulas is very much appreciated(y)
 
Known values: Mass on conveyor, even distribution by the time it reaches the end of the conveyor.

Assumption: Velocity is constant

Attempting to solve this:
Output(t) (Flow rate)
Weight(t) (Volume on conveyor)

Weight/Length of conveyor = Average amount of product per unit space

I imagine you could find the velocity of the conveyor simply.

So (Weight(t)/Length) * Velocity = Output(t)

Output in a time slice is then Output(t)*Timeslice

Or (Output(t) + Output(t - Timeslice))/2
 
On second thought, though,I am not sure I want to totalize it... I am more interested in how much is coming off the conveyor... perhaps in similar "time slices" as it was acquired.
Click to expand...
I think Daniel's point is that (over some known time period, say 1 minute) you have New Total - Old Total = Amount Leaving Belt in last 1 minute.
 
Lancie1 said:
I think Daniel's point is that (over some known time period, say 1 minute) you have New Total - Old Total = Amount Leaving Belt in last 1 minute.
Click to expand...


Sorry for ressurecting this old thread, but I am revisiting this problem and I imagine others might have similar applications.

It seems to me that the problem with that is the it disregards the amount ENTERING the belt. If the same amount entered the belt as left the answer would be zero. 1000 - 1000 =0

What I am looking for is the amount the belt delivered during the time slice. in this case 0 would be wrong.
 
It seems to me that the problem with that is the it disregards the amount ENTERING the belt. If the same amount entered the belt as left the answer would be zero. 1000 - 1000 =0

What I am looking for is the amount the belt delivered during the time slice. in this case 0 would be wrong
Click to expand...
Exactly correct. You have TWO unknowns, the amount leaving the vibratory feeder, and the intermittent unknown weight ENTERING the vibratory feeder. Unless you can weight the amount entering, then the amount leaving or rate of feedout cannot be calculated.

If you could measure the amount entering the feeder, then

W0 = Amount Entering since beginning of a time period T,
W1 = Amount weighed some T time ago (Old Weight).
W2 = Amount weighed right now (New Weight),
W3 = Amount Leaving in time period T.

Flow Rate Leaving (lbs/sec.) = W3/T = (W0+W1-W2)/T seconds.

Can your PLC be told WHEN the new amount gets put into the vibrator? If so, then your Time Period could be set to be from Entering Amount to next Entering Amount. That way, the W3/T = (New Entering Amount - Old Entering Amount)/ Elapsed Time.
 
Last edited:
I concur with Lancie. When I started reading this thread just now, I thought you were being given the impossible, and I raised my eyebrows at the proposed solution, questioning maybe it is possible?

But you need a way to measure the weight at an accumulation or emptying point in order to get this approximation. Upstream would be preferred if I think I understand the intended purpose of this. You need load cells under a container that is either only receiving or discharging product at any given time.
 
I need to know the amount delivered by the conveyor in somewhat real time... It doesn't have to be perfect but I want to be as close as practical.

How about this? If I know:

weight on known length of conveyor
speed of conveyor.

Seems to me I should be able to calculate how much is being delivered in small time slices.
 
Since the product is somewhat uniform at the discharge of the conveyor because of the vibrations, the flow out should be proportional to the weight of the product on the vibrator. The more product (weight) on the vibrator, the higher the product level will be at the discharge. You can do some sampling to figure out the relationship between the the discharge rate and the weight of product on the vibrator. Put a container under the discharge to collect a sample for a set time (say 15 sec), and record the vibrator weight at the same time. Repeat as many times as necessary to get a good plot on a graph (at least 10 or more samples), then you should see a roughly straight line out of that plot. The slope of the line would then be the relationship between the discharge flow rate and the weight of the product on the conveyor. Then use whatever timeslice you need to get an accurate weight for your process.
 
maybe it's an idea to see the belt as a shift register?
like:
|w1|w2|w3|w4|w5|w6|w7|w8|

w1 is the weight entering, w8 the weight leaving.
say, wOld is the total weight before the calculation

then at a interval do this:
Code: 
w8 = value leaving
wTemp = wOld - w8
[shift values to the right]
w1 = total weight - wTemp
wOld = total weight
you have to figure out the right timebase but I would go with some feedback from the conveyor like an encoder or as simple as a cam. [edit: you said vibratory conveyor so it's probably not belt like?]
 
Last edited:
How about this? If I know:

weight on known length of conveyor
speed of conveyor.

Seems to me I should be able to calculate how much is being delivered in small time slices.
Click to expand...
Are you talking about the weight from the feeding conveyor? Yes that is the method to calculate the Weight/Speed. You can install a short section of an off-the-shelf Weigh-belt Conveyor to do that job very accurately. A weigh belt has outputs that will tell you the weight/feet or weight/second being delivered by the conveyor.
 
Do the material vary in quality or humidity?
IF TRUE THEN go for Lancie1's belt weight.
ELSE go for Mark S. method.

In a customers factory nearby, there are lots of vibrbrbrbro feeders. Mostly 5" to 8" tubes with 2 or 4 excenter weight motors. I would never in my faintest imagination have had the idea to put them on weight cells. Might work though.

Kalle
 
Lancie1 said:
Are you talking about the weight from the feeding conveyor? Yes that is the method to calculate the Weight/Speed. You can install a short section of an off-the-shelf Weigh-belt Conveyor to do that job very accurately. A weigh belt has outputs that will tell you the weight/feet or weight/second being delivered by the conveyor.
Click to expand...
This stuff is already installed for several years. Never worked very well though. Company has been back several times and failed to improve it.

It has a micologix PLC and you would not believe how it is loaded up with stuff .... loaded to the gills with expansion slots analog I/O and has I think maybe 60 some elements left to use in the program. 20 some screens in a Panelview plus.( which is a whole nuther Factory talk subject) Talks to other PLCs that control other stuff. 4-20 ma inputs to PLCs that monitor 0-10v position resistors in pneumatic cylinders with 4 solenoids to control the position of each cylinder.

Over the past years it has been in I have spent weeks trying to figure this thing out... with the help of you fine folks ... I actually enjoy the challenge and I think I'm getting close. It's hard to shut the thing down to download programs .

The PLC probably oughta be pitched and started over with something that can be edited on the fly. It's a bugger.
 
Last edited:
This stuff is already installed for several years. Never worked very well though. Company has been back several times and failed to improve it.
Click to expand...
If they tried to weigh a vibrating feeder, I imagine that is a big headache. I think the only way that would work is if the weight were averaged over some long time period. The vibrations would cause the weight reading to jump around over a wide range.

I once tried to weigh a merely-rotating mixer, not vibrating, inside a PLC program. I never could get reliable weight readings even with averaging them over 1 minute.
 
Last edited:

Similar Topics

M
Help Math reminder / calculating encoder cnts per unit
This application has a motor with encoder feedback that drives a linear actuator that moves in/out, and is at roughly 45 degs from horiz. As the...
2
Replies
19
Views
1,365
drbitboy
drbitboy
F
Calculating a running average?
I need to keep a running pass/fail yield of the previous 5,000 parts produced. I have used this formula before to calculate average: AvgValue =...
Replies
6
Views
2,156
drbitboy
drbitboy
E
Calculating Energy Savings From Removing Air Cylinders
Does anyone know how to calculate the savings from now needing to run an air compressor all day? Basically I have a design that replaced 6 * 1"...
2
Replies
26
Views
4,828
damica1
D
A
Calculating parameters of motion profile
I would like to refer to this document of which I used some data ...
Replies
1
Views
1,478
Peter Nachtwey
Peter Nachtwey
D
Calculating rate/velocity with speed and position
Seems like this should be a simple thing but I’m struggling with it. The basis is that I am trying to detect if a linear assembly is stopped or...
Replies
6
Views
3,095
kvogel
K
Back
Top Bottom