Weigh Conveyor - Example

Following my recent postings for Weigh Conveyor, using the SIWAREX MS (and S7-226) with a single SIWAREX R - SB series Loadcell, following is a sketch of what it should like.




I would like to have a discussion on the way I intend to do this, and as my share, I'm posting details of what I know.

Some basics, to remeber are,

- The Process is fairly tolerably in terms of variance of quantity.
- The Product height remains almost constant except for the START and END.



1) SPEED Calculation,
Encoder Pulse X linear distance travel

2) The HIGH level sensor stops the whole upstream equipments.

3) the LOW Level sensor stops the WEIGH CONV.

4) A VFD is being used. with several speed steps, depending on Recipes.

5) There should be a sensor (the RED circle in Center of conveyor) to start the measuring Process once the Product COVERS the PLATFORM.

Skiiping the STARTING and ENDING issues, the CONVEYOR shall be covered evenly with the PRODUCT, so the VFD does not change speed, rather it STOPS/RUNS as per instructions from the PLC.

What I need to properly understand is the measure (and control) the Product, that is in the area after the PLATFORM (distance 'C')

For instance the PRODUCT weight is 10KGS and the PLATFORM width (distance 'B') is 10CM, I get 1KG of Tobacco from delivery end, for every 1CM of displacement.

this is the simple version,

what issues I may face include,

a) Variance in product weight.
b) START/END variation.
c) Friction of the Running Belt.

Measuring is easy, I'm confused as to how I will manage the Per Linear distance (CENTIMETER in this case) of weight in TABLES?
and keep it updating ever displacement (CM)

project update

Further on the topic,

since we can only measure the WEIGHT on the platform and keep track of it as it moves, and have to ignore the PRODUCT weight on the FEED SIDE (Left). I would get this.

If the platform size is 10CM and the conveyor further onwards is another 20CM, This would mean I would need keep track of 30CM.

Now if each INTEGER stores the product weight of 1CM span, I would need 30INTEGERS? is this the way?

for instance WD01, WD02, WD03, WD04, WD05, WD06, WD07,....WD30

Now at every Pulse I get from the Encoder,
showing that the conveyor has moved 1CM, I need to move the following DATA

WD30 -> Total conveyed so far
WD29 -> WD30
WD28 -> WD29
.
.
.
.
.
WD01 -> WD02
MOVE ZERO to WD01

I've Never done this before, but as the PLCS.NET encourages ppl who put in thier share of efforts first, and ask later, I'm just following the rules.

would apprecaite any feed back.

I'm sure a few people could help with some theory on this, it is a specialist subject though so, just the belt weighing itself, let alone the code you would need to implement.

The things I would look out for as you stated, the very start and end (how often would these come along) where the weight would decrease.

The rest of the time it should be fairly even, unless you have lumping etc.

I would experiment a bit to find out the flow, the weight you read is not the weight being transferred its a proportion of what you read that is transferred, the weight is the total volume of product on the conveyor, that is length*height*width, knowing the speed you know how much length has moved.

From that you should be able to calculate the amount of product off the end.

I would prefer to do that than use some sort of indexing, it would even itself out.

Then again I would have bought a K-Tron and not had this problem

PeterW said:

I'm sure a few people could help with some theory on this, it is a specialist subject though so, just the belt weighing itself, let alone the code you would need to implement.

The things I would look out for as you stated, the very start and end (how often would these come along) where the weight would decrease.

The rest of the time it should be fairly even, unless you have lumping etc.

I would experiment a bit to find out the flow, the weight you read is not the weight being transferred its a proportion of what you read that is transferred, the weight is the total volume of product on the conveyor, that is length*height*width, knowing the speed you know how much length has moved.

From that you should be able to calculate the amount of product off the end.

I would prefer to do that than use some sort of indexing, it would even itself out.

Then again I would have bought a K-Tron and not had this problem

Click to expand...

That's exactly why we recommended getting a canned system before... Weigh belts are extremely difficult system to automate. PLC's don't integrate well - which is what you're needing.

How large is your isolated weighing range? It's not 1 cm. If you have 30 1-cm sections, you're not going to have an accurate accounting of weight. That is, part of the 1kg recorded in this current "zone" will also be part of the 1-cm zones both before and after the "current zone".

oh well.......

This is not rocket science. If you Google "poidometer" you will see that this was done with a beam scale 2 centuries ago. I've seen those antiques run and they are still pretty accurate.

We in the US forget that what is a reasonable purchase here is an unreasonable number in many other countries. Our ancestors had to invent this stuff and now we have become lazy.

PeterW said:

the weight is the total volume of product on the conveyor, that is length*height*width, knowing the speed you know how much length has moved.

Click to expand...

The weight of the material delivered is NOT the volume of the product on the conveyor. It's the volume * density (specific gravity). I'm sure you know that, so maybe a mixup in terms, but for others reading, it's important to note.

I'm with other's here. You will be much better suited with a K-Tron type canned system. The whole point of a weigh belt feeder is to compensate for variances in flow profile and specific gravity. While you may think you will have constant flow from a choke feed as you've illustrated, you have to allow for blockages, restrictions, humidity changes, temperature changes, head pressure or other reasons that you don't get perfect flow. You also have to make allowances for changes in specific gravity of your product. Again, humidity, temperature, head pressure and compaction, fines (small bits vs larger bits), etc. will all effect overall SG.

To compensate for all this, your belt will generally run at a fixed speed for a theoretical flow rate required. The load cell in coordination with the integrator will monitor the actual flow and make changes to speed if you are trying to achieve a constant mass flow, or change time of run if you are trying to achieve a totalized mass.

Robert

All above is very true and if I was doing this project in this country I would purchase a proven system and be done with it.
If you spend time in developing countries you will see a diferent mindset.
Have some sympathy for these guys, many times they have to reinvent the wheel using only what they have. Perhaps a constant volume over a period will average to a "close enough" weight.

gas said:

Robert

All above is very true and if I was doing this project in this country I would purchase a proven system and be done with it.
If you spend time in developing countries you will see a diferent mindset.
Have some sympathy for these guys, many times they have to reinvent the wheel using only what they have. Perhaps a constant volume over a period will average to a "close enough" weight.

Click to expand...

Point well taken....However, often times, in posting, not only are we addressing the original poster, but all those who come after searching for these topics. You are absolutely right, and the brute force method may certainly be good enough for what the OP needs. But for posterity sake, anyone who searches later and has the means, it's prudent I think to point out better methods.

PeterW said:

Then again I would have bought a K-Tron and not had this problem

Click to expand...

What!! and miss out all the fun!!

gas said:

We in the US forget that what is a reasonable purchase here is an unreasonable number in many other countries. Our ancestors had to invent this stuff and now we have become lazy.

Click to expand...

Ditto!

The Place I work, i.e., Mostly Asia/Middle east & African continent, and a PLC programmer gets something between US$300 - $500/ (That's per month, if he is employed, double that for a self-employed or a Free lancer, but he may or not have any project in hand at times!)

I'm not sure how good that is in your area.

gas said:

All above is very true and if I was doing this project in this country I would purchase a proven system and be done with it.

Click to expand...

Beleive me so would I, Infact that was one of my suggestion, unfortunately thats far as it goes!

gas said:

If you spend time in developing countries you will see a diferent mindset.
Have some sympathy for these guys, many times they have to reinvent the wheel using only what they have. Perhaps a constant volume over a period will average to a "close enough" weight.

Click to expand...

There's another thing you might note about world economics, Many large manufacturers are shifting thier productions from US /Europe to Asia? even thought why so?

I understand one of the reason could be, CHEAP LABOR.
Shifting a Hitech plant from thousands of Miles to obtain goal.

ADVANCE LABOR * HITECH EQUIPMENT = Price X
CHEAP LABOR * HITECH EQUIPMENT - Price X/y ('y' being the difference in Labor)

From Economic/Stats. point of view, there are Many types of customers, each having different wants, and level of desire to spend, etc

simply put, suppose u Buy 1Dozen BANANAS for a maximum price of US1.00
but if the price is reduce/increased, this could (would in most cases) impact the amount of goods (or services) acquired.

as mentioned above, the Mindset, variables, limitations, environment, etc have to be considered as well, why do you think anyone would not want a FULLY TESTED AND ALREADY DEVELOPED SYSTEM?

Now in a different scenario, I'm seeing on regular basis that Many big manufacturers do what I'm about to, amazingly they belong to Europe/US (Developed countries I suppose), just for an insight, I belong to Tobacco Industry, and major equipment suppliers (specialized for this industry) have been using lame technology (or methods), for whom? for custmers belonging to DEVELOPING COUNTRIES?
amazingly NO. the customer I'm talking about is British American Tobacco (BAT), Phillip Morris....

Now again, from MINDSET point of you, you might have noticed, that what an INTEGRATOR thinks about a project, the cost involved, the equipment that should be used, etc MAY BE TOTALY DIFFERENT TO WHAT THE COMPANY's MANAGEMENT may think, so there's a difference (STRONG) of opinion, amongst the Organisation as well.

Its not just simple. but I dont want to get involved too much into Economics, Physcology, World politics, right now.

Just consider some people opt to play the GAME in HARD MODE

robertmee said:

Point well taken....However, often times, in posting, not only are we addressing the original poster, but all those who come after searching for these topics. You are absolutely right, and the brute force method may certainly be good enough for what the OP needs. But for posterity sake, anyone who searches later and has the means, it's prudent I think to point out better methods.

Click to expand...

Yes, I somewhat Agree, this is what the site is about, to talk about different Projects, under different environment... DIFFERENT ENVIRONMENTS I MEAN...

Difference of Environment could mean, Limited Budget, Scarcity of Professional Person, Limited Equipment, meeting deadlines, etc

Pointing out a BEST SOLUTION under the variables available, to the USER, is my interest, giving him something he cannot digest would not be nice!

Regarding other posters, whenever I get to read a thread, I read it within the Original posters available resources, for learning sake, I do not take it as the Last word, so if any one gets misguided by my postings, please forgive me!