calibration of weighing system

P

Thread Starter

pk

hello everyone
i want a small information.
we have a weight system where we are weighting the material inside a bin. The range of the system is from 0 - 300 tons. Now i want to calibrate the system. My problem is, i can't put the rated load on the bin to calibrate is for span.

we made to zero calibration and for span, what the supplier is suggesting is to calibrate it with 80T known load.

Is it right. Will it work perfectly? what about linearity?
 
R
Yes, it's difficult to do a span calibration. I assume you have several large load cells supporting the bin, or do you have the type that is added to the legs after the fact?

If you have the former:
Measure each cell mV with bin empty
Measure each cell mV with bin full
Measure the excitation voltage
Calculate the weight added to each cell, the sum of all readings is the full weight.

If you have the latter, can you weigh the product in or out with a truck scale or similar?

Load cell output is normally very linear and the nameplate data is usually reliable. Post the make/model of your system and someone will be able to help you further.

Hope this helps,
Roy
 
Hi,

We are using three loadcells in weighing system of Schenk make. My question is, if at the full load what we measure is up to 300 Tons and this scale we are calibrating has 50 or 70 Tons only for span, will it not make any problem?

If so, then why to calibrate a system which is designed for 10 Tons with 10 Tons for span? We can make span with 500 Kg also. Am I wrong?
 
R
The closer you get to full scale, the more accurate you get but how practical is it to have a 300t calibration weight, or even a 30t for that matter?

Are your load cells the type that supports the whole weight ore are they the type that is inserted into a hole drilled into the supports?
If you are filling the bin via a truck delivery can you co-relate it to a truck scale?

Do you have a belt scale feeding in/out of the bin?

Yes, I think it's reasonable to calibrate a 10t scale with a 500kG weight, it's all a matter of how accurate you need to be.
Are you using the weight as part of a batch mixing scheme or custody transfer or just simply as a percent full indication?

Regards
Roy

 
Thank you friend...

I am really concerned about the accuracy only. When I calibrate a system designed for 300 tons with 50 tons, the percentage error will be very much less. But when more load, say 280 T, is applied on it, the percentage error will increase. Won't it? Or will the percentage error remain the same?
 
R
If you were 1 tonne out at 50 that would be a 2% error. If you then measured 300 tonnes you might be 6 tonnes out. If you were 1 tonne out at 300 that's only 0.3% error. As I said before even if you had a 300 tonne calibration weight how would you place it in the bin? One possible method would be to batch in 300 m3 of water.

Please tell us more about what you are trying to do with the bin, custody transfer, batch mixing or just short term storage (surge capacity).

Regards,
Roy
 
hi roy
you are right that 300 ton can not be put in the bin. we are using it as storage. But there are methods to calibrate it. Like we can calibrated with hydraulic jack by pressing it up to equal load of 300 ton.
In fact for 60 T also, the supplier made it that way.

When I calibrated it for 60 ton, error was within +/- .15%. But when i checked it after filling 275 ton, it is showing me around .3 to.4% error, which i assume is normal as we didnt make span with rated or at least 80% of rated load.
 
T
Hello there,

It is normal to calibrate weighing systems at 10% of the full scale value. The indicator or controller extrapolates the mV values for full scale on basis of applied weight. If you are using 300T loading then you can calibrate it with 3 Tons test weight. Suppliers guarantee linearity of sensors for full range. BUT DONT EXPECT 0.1% OR SO as no system fulfills this. Just consider air loading on the bin. This will change the force on the load cells every time. In some Schenck systems they have this external disturbance compensation, but it only works only on repeatable errors and not on random disturbances.

I'm using 7 Schenck based feeders but have not seen this magical 0.1% accuracy. Best way to be accurate is to establish error in the system and compensate it in the indicator settings. You will always get compensated indication instead of measured value.

Topak_Khan
 
Which calibration method should I use?
Question
There are four types of calibration, which include:
1. Hardware Calibration Using Certified Weights
2. Hardware Calibration Using Material Substitution
3. Hardware Calibration Using Test Weights with Material Build
4. C2® Second Generation Calibration (Hardy Instruments only)

When do I use which method?
Answer
1. Hardware Calibration Using Certified Weights:
Hardware Calibration Using Certified Weights is typically used in an environment where multiple scales are in the same process and the outputs are compared. Where a standard is set and that standard must comply to allow a product release. Also when the product weighed is the same or general range. A scale is most accurate at zero and the span weight applied. To improve weighing in a particular weighing range, use that weight as the span value in the calibration procedure. Weighing 5 kg bags on a checking scale. Calibrate at zero and then span using a 1kg test weight. This creates a very accurate scale for that target weight. Scales that measure a variety of weight use the 80 to 100% of the highest expected load as the span weight.

The practical minimum weight you should use when performing a hard calibration is 10% of capacity.

The recommended minimum weight you should use when performing a hard calibration is 80-100% of your target load. Loading 500 lb. bags, the span weight is 400 lb. to 500 lb. This gives you the most accurate and repeatable weight readings.

This method uses known calibrated weights to apply a load on the load-receiving elements. Because the trace-ability of the standard is known and maintained, it is the most accurate calibration technique. Accuracies within + .05% of the applied load can be achieved with this method.

This procedure, however, may be restrictive for some industrial scales, as large capacity weights are not easily transferred to the site or are often unavailable in the field. In many other cases, the load-receiving element may not be able to accept placement of the weights, or be in a poorly or inaccessible location. An example of this would be a vessel hanging high above the factory floor or an open top vessel without any place to install test weights of a required amount.

To use the certified weights method of calibration, a set of Class "F" accuracy weights (within 0.01% of denomination) that meet the following regulations are required:
a. A total weight of 80% to 100% of the system capacity.
b. At least three weights between 10% and 100% of the system capacity to check the mid range.
c. Several low capacity weights equivalent to one or two instrument divisions.

This precaution is necessary due to the nature of a Test calibration. Test weight calibrations can hide or mask scale problems like non-linearity and binding. Without various levels of weight testing the scale health cannot be determine, just by a hardware calibration.

Most local Weights and Measures organizations have such weights available. Typical weight values are 10, 25, 50, 100, 500 and 1000 pounds. If the weighbridge is a device other than a platform, some method of loading the weights on the structure must be devised. Also insure the test weights are placed as close as possible to the center of the scale and the load is balanced. To many weights placed on one side or the other can cause a weighing moment that is not present in the normal operation of the scale. In that case material substitution would be a better calibration solution.

2. Hardware Calibration Using Material Substitution:
When certified test weights are not available or usable, accurately weighed material may be used to calibrate a system. In this method, material weighed on a secondary, calibrated scale is delivered to the site of the scale to be calibrated. The secondary calibrated scale should be of the same accuracy or greater and have a capacity approximately equal or less. A smaller scale has a finer resolution and produces a better accuracy. Material introduced, either a liquid (water) or a dry product (sand) is dependent on what the vessel is designed to weigh. The methods used for delivering the material for calibration must not be over looked since they may lose or gain weight during transit (i.e. if a truck is part of the delivered weight, it loses fuel during the transfer). Snow, rain, or other environmental characteristics might alter the precision of the weights. A liquid such as water can be measured by volume and its weight calculated. Accuracies of up to +0.15% of applied load can be achieved. A commonly used method of material substitution called "Warm Body Cal." is easy but yields a much lower accuracy. With this calibration method, multiple workers are weighted on a secondary scale, and then climb onto the scale (vessel) being calibrated. Extreme care must be taken to insure no injuries to personnel or damage to equipment occurs.

This method is used when test weights will not physically fit on the scale or are not available in large enough quantities. If it is not possible to transport sufficient amounts of weighed material a combination of method one and two may be used.

3. Hardware Calibration Using Test Weights with Material Build:
This method is used when calibrating large capacity scales for which is not feasible to obtain certified weights to equal 80% of capacity. This method employs a combination of certified test weights and material substitution. It requires class "F" certified weights equal to a minimum 5% of full capacity and a means of attaching the weights to the scale. This method also requires a means of filling, transporting and discharging the material (dry or liquid) to the scale being calibrated. In this method, the scale is calibrated with the certified test weights. The test weights are removed and material is added to the calibrated span weight. The weight is then recorded and the certified test weights are placed back on the scale and verified the material weight plus the test weight totals agree. This process is repeated until 80% to 100% of the scale capacity is reached. At that point the scale and be re-spanned to give a calibration point at the 80-100% of target load. This method can produce accuracies of up to +0.25% of the applied load.

With all these methods there is another method that removes any of the guesswork in calibration. It is a combination of electronic pre-calibration testing and test weight for verification. This is the Hardy Instruments C2 calibration method described next.

4) Hardy Instruments C2 Second Generation Calibration:
Each individual load point (load cell with C2 and mounting hardware) in a C2 system has its performance characteristics electronically stored in a memory device at the end of it's cable or in the case of Advantage Load Points internal to the load cell body. These performance characteristics decide how that load point will perform. They are measured on a NIST traceable test standard and recorded when the load cell is manufactured. The C2 system uses these parameters, the instruments' characteristics, and a reference point to mathematically calibrate the scale. If the C2 memory device is destroyed, the load point will still act as a standard load cell, and can be calibrated utilizing hardware techniques.

As with any of the described calibration methods, test weight must be used to verify any scale calibration. C2 calibration only needs 5-10% of the target weight to determine if the scale is accurate.

Through field-testing of actual installations, Hardy has found C2 to be equal to and in many cases better than calibrations using test weights. Using certified test weights with weight equal from eighty to 100 percent of the scale capacity yields the most accurate calibration. However, conditions for calibrating are often less than ideal. Many vessels lack a place to put or attach the required capacity of test weights. In addition, there may not be a way to distribute the weights equally on (or in) the scale vessel. Some vessels are mounted high above the plant floor or in areas offering limited accessibility. Still others have weight capacities far in excess of the available test weights. Multiple calibrations over time can also yield dissimilar results. This can be due to different personnel using diverse calibration methods, using different test weight quantities, or to the placement of these weights on the scale. The result of these real world issues likely would produce an error prone calibration. With C2, these considerations are no longer an issue. As soon as the scale system is installed, it can be C2 calibrated and a proper scale installation verified. The result is a calibration that is easier, quicker, safer, and typically more accurate, without the need for highly skilled labor.

Requirements for C2 Calibration:
1. C2 certified Hardy Instruments load points
2. C2 certified Hardy Instruments HI 215JB series Junction Box to handle extra wires
3. C2 certified eight-conductor cable to handle extra signals.
4. Sense jumpers removed and the sense lines installed.
5. A mechanically correct load cell and scale installation.
a. Level
b. Correct piping installation
c. Correct parameters set in the instrument.
d. Installation instructions properly followed.
6. C2 certified weighing instrument: any Hardy Instruments indicator or controller.
7. Certified test weights equal to one or two graduations and 5-10% of the target load.

Some instruments have the ability to adjust for errors in the system. These correctable errors are limited to gravity and electronic instrumentation error. Errors caused by binding, unstable mounting or outside interferences should be repaired, not corrected electronically. Refer to the troubleshooting guide in the back of the instrument's manual to help clear large errors. Typical system errors are 0.15% to 0.2% of rated capacity. Error for Gravity is around the 0.07% range and varies the closer and further your system is from the 45th parallel.

A few other instruments have a calibration programmed called "soft cal". This is typically the same as a C2 calibration. The soft cal primarily reports the results of a C2 calibration, overall sensitivity and the total range of the load cells. This program is used to correct small errors like gravity and electronic controller errors. It is only found in Hardy Instruments models HI 2151/30WC and HI 1771WS (and the WS 100 software for the HI 1746WS module).

There is now a gravity correction factor in the latest HI 3030 software release.

If you need additional information, or if you need on-site assistance call Hardy Technical Support at 800-821-5831 x1757 or 858-278-2900 x1757. We offer Local Field Service in the U.S. & Canada for system installation, startup, inspection, verification, calibration and certification, emergency troubleshooting, on-site training and routine preventative maintenance for your Process Weighing, Tension Control and Vibration Monitoring instrumentation. For the solution to your problem, Think Hardy!
 
Hi PK,

I think you did very well to get your scale that accurate, I would have been happy with 1%. If you are only using it for storage why is accuracy so important?

Your method of using a hydraulic jack sounds good, it will certainly be easier than placing weights, but how do you measure the pressure accurately? Most pressure gauges would have a worse accuracy.

An interesting post by Hardy, I have used one of their pre-calibrated systems on a cement batching plant. It worked very well and was nice not to have to mess with weights.

Regards,
Roy
 
Is the Hardware Calibration Using Test Weights with Material Build up method described here mentioned in any of the international standards?
 
Yes, it's difficult to do a span calibration. I assume you have several large load cells supporting the bin, or do you have the type that is added to the legs after the fact?

If you have the former:
Measure each cell mV with bin empty
Measure each cell mV with bin full
Measure the excitation voltage
Calculate the weight added to each cell, the sum of all readings is the full weight.

If you have the latter, can you weigh the product in or out with a truck scale or similar?

Load cell output is normally very linear and the nameplate data is usually reliable analytical balances. Post the make/model of your system and someone will be able to help you further.

Hope this helps,
Roy
I was just wondering how to calibrate electronic weighing scales ( seca and the like) in a hospital environment. We have 200 electroni weighing scales in our hospital which are mostly from Seca. We can not carry 10kgs - 150kgs of weights around the hospital to calibrate and / or measure accuracy and precision of these scales. OOPs..... it would be tedious. Is there any strain gauge dummy kind of measurement that we could simulate and measure the actual weight against the displayed value in all these electronic weighing scales? I know Wellch Allyn temperature monitors have 3 different temperature probes and / or calibrators to measure, calibarate and validate at 3 different temperatures. Similar to this do we have dummy load for electronic patient weighing scales. Electronic balances in the laboratory are easier to calibrate as weights are smaller between 1gm -5000gms but not for patient weighing scales.
 
As long as the empty weight zero is identified, the load cells are linear.
You must rely on the lab calibration of the load cell transmitters performed when purchased.
Your high bin level span is set during pre-startup check out.
 
Top