We manufacture weigh scales and have installed a number of systems till 30 Tons capacity. We have calibrated these with water to achieve desired result. I am now interested in bidding for a project where about 20 tanks of 500 Tons each are to be installted. Some would be with 4 loadcells and some with 3 loadcells + 1 Dummy cell

However i do not know how to calibrate these. Material substitution does not seem practical since it involves a very large volume, which may not be possible. Intital calibration is usually achieved after loading material atleast 2-3 times so weighbridge based transfer would also lead to problems of storage

While calibrating at lower load is an obvious solution, there is a chance of linearty error since capacities are high. I would like to know how other people have solved this problem or of the best possible solution

 

Since they are tanks, would it be possible to batch in water using an accurate meter e.g. coreolis. You woluld then batch it from the first tank to the next etc.
Regards
Roy

 

Water is obvious, and i have used it for smaller capacities. but the volumes involved are large. also piping to fill water in the first tank or transfer it to another is not in place as the tanks are intended for filling solids.

 

Hi Hemal,

I understand your situation. To the best of my knowledge there is no direct solution to your problem. I am listing a few approaches below, however I must admit I may not even be close to what you are looking for. Still these may guide you to the final solution.

1. There are a number of companies that provide load cell calibrators. However I know none which have them in the 500T range. Try the link below.
http://www.sensing-systems.com/products/calibration_kits/default.html
Honeywell is another company that develops this product. Try contacting the Honeywell design center in Pune.

2. Try Companies manufacturing artificial Loads or Hydraulic Load simulators. These are hydraulic systems based on pressure amplification principle. I remember using such a system in stress tests on large shipping containers though not in the 500T range. Perhaps you could try one of the companies that build these containers.

3. One indirect method would be to replace the 125T loadcells with say 10T loadcells having IDENTICAL characteristics (identical sensitivity, temperature and mechanical construction charecteristics) as the original loadcells from the same manufacturer. Then characterize these just as you would with water. If the loadcell manufacturer has done his part well then, you can be fairly certain your system will perform within specs when the 125T loadcells are returned to the system.

4. Characterize the loadcells individually at lower loads, such as the ones you have used in the past, Using a high precision data logger or a high precision DMM typically say a 6 & 1/2 digit kethley DMM and then plot the data to predict the nonlinearity.

regards

 

Since you have ruled out water and fixed weights would not be practical perhaps just do an electronic calibration i.e. have a load cell simulator giving you a mV output similar to a full tank. That way each tank would use the same reference point.

Another option might be to load each cell with a calibrated hydraulic jack.

BTW, if it's for solids, it's a bin, not a tank.
Is there a conveyor filling the bin, does it have a scale?

Regards
Roy

 

A system with dummy load cell will not work with solids generally. What is your accuracy expectation? What is the idea behind 3 active and 1 dummy load cell.

Patricks

 

By solids i do mean fines and other granular ingredients. The client insists on having live + dummy, expecting a great reduction in cost (24 bins )and its sometimes difficult to convince them. While there is a loading belt conveyor, it does not having weighing built in.

Besides 500 tons some of the required capacities are 1200 Tons (2 live + 2 dummy) and 6 bins are of 2000 tons (4 live + 2 dummy)

Accuracy requirement across capacities is +/-0.05%

Has anyone tried Digital load cells and is their a way out by using them ?

 

I have used digital cells (by Hardy I think). They have a chip included with the characteristics of the cell. this is read by the scale.

0.05% seems un-realistic, the only way to get that sort of accuracy would be with calibration weights.

Roy

 

If you need 0.1% or better, calibration weights is the only method. That too without any dummy load cell (all live load cells).

Many indicators have "mV/V data direct input" option. Have you considered this option ? The accuracy could be 0.1-1% depending on load cells. Again no dummy load cells and junction box to be used.
Digital or analogue (load cell) will not make much difference IMHO.

Regards
Sam

 

Seems we have reached a dead end. Weights are not possible and desired accuracy beyond reach. In two minds whether i should even quote for the inquiry. Maybe i should change the topic of discussion to "How to educate clients for specifications "

 

Yes, you should quote it but make sure you give a good explanation of why you propose to use your chosen calibration method be it electronic or hydraulic etc otherwise someone else will offer what the customer asked for then not be able to deliver.

It seems to me that your firm has the experience otherwise you would not be asking the questions. As a purchaser of many instrument systems I appreciate it when the vendors are up-front.
Good Luck
Roy

 

Mr Roy,

thanks for the confidence shown in my abilities, but no, i seriously have no clue of how to go about with this enquiry or expereince with such high capacities. My expereince is limited to 30 Ton tanks

I even tried contacting some overseas manufacturers so that i could tieup with them but people dont seem interested.

And yes some joker will quote and not deliver but its just that i don't want to be that joker, considering that the client is a premier conglomorate and value of the contract high

Hemal

 

This sounds a lot like grain bins. Measuring the level of grain bins are tricky, grain do not behave as water( water has a smooth surface). Grain piles up in the middle during filling, and the other way around during empting. So using load cells makes sense in a way.
But sometimes grain ( and the like) will stick to the walls inside and give you more load on one side.

Dummy cells will not make things easier.

Then the density of grain will vary from batch to batch, when the water content varies.

If it is grain, or the like, you never knows how much a full bin weights.

 

No, this for an iron pellet plant. 17 bins for mixed iron ore, 2 for crushed ore, 2 for pellets, 2 for bentonite, 1 each for limestone, dolomite, binder, coke and 1 for other material. I am seriously looking for someone who can partner me in the project

 

Hemal

first question you should ask your client is

How are you going to calibrate [not the correct word] - 500 Tons tank? load cell calibration would follow

 

The client somewhat rightfully states that i should answer that question, since weighing is my core competence and not his. And there are also 1200 Ton bins.

This is certainly not the first such project in the world, and there must be a method which i hoped someone would assist me with here

 

Than consider this as Pre-bid evaluation of your company or you

 

Another method sometimes used is to fill the bin then run the product into a truck which you then weigh on a road scale. 5oo tons would be quite a few truck loads.

Perhaps they would accept some method of calibrating each cell in turn like a removable hydraulic cylinder that clamps on to the cell with calibrated pressure gauge.

Moving all that product around surely at some point it is exposed to some form of weighing.

Roy

 

Also remember that errors accumulate, so if you have 50 10 Ton test weights each with an accuracy of 0.1% then your overall error will be >0.1%.

BTW. Do the tests have to be validated/audited? You may want to check out what they want.

 

Use one indicator with each load cell. So if you have 4 load cells, there will be 4 indicators. Each indicator can be calibrated by load cell manufacturer with its load cell. You need to sum the 4 weight readings to get total weight. The scale manufacturer will be "zero calibrating " the indicator at 0kg. So you need to zero the indicator when installed, with empty tank or bin weight. I would expect an accuracy of 0.1% typically. But accuracy is dependant on the installation too.


Regards
sam.sensor at gmail.com