This article considers just one of the factors pertaining to the differences in load cell design and illustrates that all is not as simple as it might first appear.
The strain gauge load cell is a simple and ubiquitous force-measuring device with many useful industrial applications.
Since it was developed over 60 years ago, one of the major applications — the measurement of process vessel weight — has yielded real benefits for manufacturing companies throughout the world. Improved product quality, reduced production costs, and better inventory control are among the most important.
Essentially consisting of a few sensors glued to a piece of metal that change their resistance as the metal deflects under load, the strain gauge load cell is an easy transducer to understand. As with all apparently simple and familiar concepts, familiarity can sometimes breed contempt, or at least an indifference to the features that make a good load cell.
This article considers just one of these factors in detail, in order to tease out differences in load cell design and illustrate that all is not as simple as it might first appear.
Load cell types
The companies that make up Vishay Transducers have, in their product portfolio, in excess of 100 different load cell designs. These fall into some generic categories characterised by the type or distribution of the strains that they are designed to measure.
Compression, Tension, Shear, Single Point, Bending, and Torsion Ring load cells have been developed to meet particular needs. The experienced user is well equipped to make suitable choices from these designs. A process-manufacturing customer might have many applications and might have amassed considerable experience dealing with them.
However, a much more typical industrial user has more sporadic needs. Projects requiring weight measurement may only occur infrequently and have very different specifications when they do occur. For this group of users, there are clearly only limited opportunities to acquire direct experience or to access the current techniques that might inform the selection process.
Manufacturers of load cells will of course provide comprehensive data to describe a particular load cell within their range and will bring their own expertise to bear on the application requirements. Such information is extremely valuable, but by the very nature of the source, the information is constrained by the boundaries of each supplier’s product range.
Networks of colleagues can broaden the knowledge base, bringing their own experience to bear in order to clarify a particular issue. There is also another resource that provides such a network: The Institute of Measurement and Control. It provides many channels to facilitate information transfer.
The Weighing Panel of the Institute reporting to the Learned Society Board comprises a group of engineers from industry, academia and related organisations who have in common extensive experience in the field of weight measurement. This pool of knowledge is freely available on the internet.
Several useful documents may be freely downloaded using links from this site. They are intended to help guide a prospective industrial user of load cells through the many technical issues that arise when applying the transducers to process weight measurement.
One document that does not appear in detail is the very first project that the committee published in the early 1990’s. The document is called: ‘A Procedure for the Specification, Calibration and Testing of Strain Gauge Load Cells for Industrial Process Weighing and Force Measurement.’
This code of practice has recently been used as the basis of a new British Standard, which now supersedes it. This Standard is particularly relevant to the selection of load cells for industrial process applications.
British Standard BS8422:2003
BS8422 is intended to fill the gap left open by the various European standards for force measurement, such as OIML R60, which is applicable to load cells used for trading purposes. It is a unique standard in that it recognises that the needs of the user vary and has the flexibility to allow the user to select the calibration parameters to suit the application.
It has two main sections: Standard Calibration and Supplementary Calibration. It is expected that all load cells would be subjected to the standard calibration in which a series of forces are applied in a defined way to the load cell in the principal measuring axis for which the cell was designed.
This procedure enables the basic non-linearity and reproducibility of the transducer to be determined. The supplementary calibrations may be carried out independently and address most of the critical parameters required by industry. These calibrations are designated:
A. Repeatability
B. Hysteresis
C. Creep
D. Creep Recovery
E. Overload Effect
F. Temperature Sensitivity at Zero Force
G. Temperature Sensitivity at Maximum Calibration force
H. Concentric Inclined Application of Force
J. Eccentric Inclined Application of Force
K. Eccentric Force Effects
L. Base Loading Effects
M. Zero Force Output Stability
N. Output Stability at Maximum Calibration force
P. Damp Heat
Q. Sealing
R. Excitation Voltage Effects.
Practical example
Among this collection of possible measurements, one of the supplementary calibrations — Eccentric Inclined Application of Force – is especially interesting to the weighing engineer trying to measure the weight of a process vessel. This is because a load cell supporting a vessel is sandwiched between two surfaces that will not be completely rigid.
They may or may not be truly horizontal either unloaded or when subject to normal operating loads and they may try to move relative to one another due to thermal expansion or agitation. This poorly defined mechanical arrangement results in the load cell being subject to forces that may be inclined and eccentric to the standard loading pattern established by standard calibration tests.
The effect can manifest itself in a real installation as either an error or instability of the measurement and can be a cause for complaint in an application subject to high levels of such forces unless steps are taken in the initial design stage to select a load cell that has a lack of sensitivity to such forces.
Tests to establish this effect are seldom performed in the laboratory and even less frequently specified on the load cell data sheet. This is probably due to a variety of reasons. Prior to the emergence of the new British Standard, there was little authoritative common procedural guidance as to how such tests might be performed. To this is coupled the expense of such testing and (in the context of the number of load cells actually made) the limited numbers of applications in which such information is valuable.
To try out these new procedures Vishay Measurements Group took two of its own load cell designs to the UK National Physical Laboratory in Teddington and asked them to perform eccentric inclined loading tests on their behalf. The designs chosen were the two most common generic types used for the weighing of process vessels.
Shear beam
The shear beam cell, as its name implies, comprises a horizontally mounted beam to which strain gauges are attached. Load is applied at one end and the shear strain on a web machined into the billet is measured and output. The capacity of beam chosen was 1,000kg.
This type of load cell was designed for use in weigh scales and has a very good performance in standard calibrations. The cell can easily meet the requirements for legal weighing in trade and complies with the OIML regulations for up to 6000 divisions, which is a very high specification. The cell has also been adapted for use in vessel weighing systems using a variety of mounting hardware accessories. It is extensively used in these applications.
KIS beam
The KIS beam cell is a cylindrical shear beam load cell. It has a second concentric outer beam, which allows the measured load to be applied in line with the shear strain gauges. Again a 1000kg-capacity unit was selected.
This cell was specifically designed to have a very low sensitivity to ‘off axis’ loading and has been used successfully in ‘difficult’ process applications for many years.
Both load cells were mounted in standard-production mounting hardware. The pictures in Figure 3, taken from a computer simulation, illustrate how the strain gauges in the KIS design are located at a position of low bending strain which is the feature that is intended to give the cell a low sensitivity to adverse loads.
Tests were performed in the National Physical Laboratories 20kN dead weight testing machine in accordance with the new standard. The testing basically involves a standard calibration to establish the output of the load cell under “normal” loading conditions. This is followed by tests in which a steel wedge is placed underneath the mounting, to tilt and displace the location of the applied load to an extent laid down in the Standard.
The results of the tests showed a marked difference in the inclined eccentric loading effects as follows:
KIS Beam; e2 = 0.034% of rated load.
Shear Beam ; e2 = 0.303% of rated load.
Where e2 the difference in output at rated load between a standard calibration and the eccentric inclined calibration at the same net load.
This practical example helps to illustrate that while neither design is completely immune from the effects of misaligned loads, the double beam design would tend to provide a measurement that is ten times more accurate and stable in applications where such loads may be present and significant.
This difference in performance may or may not be of value to a prospective user, but when presented with this additional information, the user is clearly in a better position to make a judgement as to the suitability of these particular load cells for his own application.
Conclusion
The selection of load cells to fulfill a particular requirement often involves subtle judgements that are simple when the information is available and clearly presented, but complicated and potentially flawed in the absence of such information. As indicated by the title of this article, load cell selection can be either simple or complex, depending on whether one has the necessary background information.
This information is freely available; but, as with all such data, the key is knowing where to look for it. Process Weighing using load cells continues to be a powerful and useful measurement technique. Application expertise is the key to success. The author hopes that this article will help the reader gain that expertise.
[Peter Zecchin is the process weighing manager at Vishay Measurements Group.]