by Todd A. Rae
The use of load cells in product manufacturing has seen a change in emphasis from unit cost considerations alone back towards weighing performance. This recent change has been brought about by several factors including precision batch process automation requirements, material accountability/process yield reporting as well as ISO9000/FDA certification requirements.
Properly evaluating a weighing system’s performance starts with the design of system’s fundamental components, the stress measurement member, normally referred to as the ‘Load Cell’, and the appropriate mounting hardware.
All industrial load cells share some common design-principles when using strain gauges to measure the deflection or deformation of a steel or aluminum structure. By using this kind of technology, the load cell functions basically as a force-measuring device. Taking this into account means that any force which is transferred to the load cell will influence the strain gauge.
The expectation of a load cell is to get a signal that is truly proportional to the applied load (force = mass x gravity) under dynamic process conditions. The following measurable factors (errors) are influencing to all load cells and ultimately to the absolute in-process measurement output accuracy. It is important to bear in mind that a load cell measures applied force.
To accurately compare load cell performance, all of these factors must be taken into consideration. The combination of all of these possible errors can be minimized by using the right alloys, proper design of strain gauges to fit the mechanical structure, precision machining practices, special material heat treatment and hardening processes, having a highly reliable manufacturing process as well as testing and adjusting the sensor while in manufacturing. All too often, most manufacturers of shear beam and compression load cells calculate a COMBINED ERROR based on only two of these errors, HYSTERESIS and NON-LINAERITY. A really true indication about load cell accuracy is only possible by taking all of the above errors into account when calculating the overall ACCURACY CLASS.
The difficulty arises in that these terms are not standardized and only the very fine print shows what is actually used in the calculated accuracy statement. Even then the verbiage can be confusing regarding performance in less than ideal testing conditions (test lab). As a consequence, it is becoming common place for end-users to ask the load cell manufacturer to provide a written performance guarantee on the equipment proposed and the application itself. This statement should be inclusive of actual dynamic conditions such as the effects of a mixer running, tank expansion/contraction, normal operating temperature swing, and wash down operations. Performance should be stated as either± actual engineering units (lb./ kg) or± % of the actual operating live load/batch amount. For any automated batch weighing application, the performance statement should also include the measurement update time based on any averaging or delay effects of Digital Signal Processing (DSP), if required.
In relation to the design of the load cell, the appropriate mounting hardware is absolutely essential to ensure performance. Some designs are more or less sensitive to side forces, titled installations or parallelism of the mounting plates. For a load cell weighing system to accurately measure the live load or applied weight, it must be capable of eliminating forces not related to the live load while not effecting the load cell system measurement accuracy. Currently, the only recognized method of achieving this in an industrial process application is via the load cell mounting hardware.
The mounting hardware must be designed to truly allow the load cell’s top plate to float while eliminating horizontal forces and retaining the vessel’s anchoring position. All of the load cell top plates should allow for some permissible displacement in the horizontal plane from the nominal load centerline without effecting performance. Process side forces are normally developed by rotating machinery such as mixers and/or support point spread due to loading/unloading and thermal expansion/contraction of the tank itself from heating/cooling cycles, washdown and clean in place (CIP).
The mounting hardware’s top plate must also allow for up to several degrees of angular change without effecting the overall system performance. Some load cell designs allow for offset during the initial installation but when loaded, movement of the top plate is not practical without influencing the cell’s output. An easy means of determining the top plate position relative to the load cell centerline should be provided in the design, since this can be a major source of error when exceeding the manufacturer’s specifications.
Another negative influence will come from connections to the vessel, such as piping. A load cell with a small deflection or displacement is able to negate these influences. Best practice is to use flexible connections of sufficient length for all piping or transfer points.
Due to typical manufacturing tolerances associated with most load cells, field adjustment procedures are required to achieve the best possible performance. First, an adjustment of the zero output signal of the load cell is done by measuring the output signal of all participating load cells in the array and then leveling/balancing them by mechanical shimming or adjustment of leveling nuts.
Second, on high accuracy applications, load cell vendors have supplied a summing/trim junction box between the load cells and the electronics to provide compensation for the lack of standardization of the sensitivity and impedance between
load cells. Again, it is important to bear in mind that all electronic load cells start off as analog measurement devices. This box normally contains a potentiometer pot for each load cell input that allows the trimming of the excitation voltage on each load cell to make their active sensitivities similar at the time of adjustment. These trim boxes have historically been a source of failure due to water contamination and drift since they contain resistive pots. Some manufacturers have addressed these issues with additional electronic circuitry to condition and digitize the load cell output signal.
In contrast, truly matched precision load cells are manufactured to exacting specifications for both sensitivity and circuit impedance. These cells minimize balancing requirements and do not require any additional trimming or scaling (digitizing), thus allowing for calibration without test weights. Additionally, if a load cell should ever fail, it can be replaced without the need for recalibration.
This precision class of cells is found to have very good linearity and high resolution allowing for verification of the weighing system performance with a proportionally small amount of test weight as compared to its full-scale capacity. This has been beneficial to end-users by reducing their dependency on local scale calibration services associated with traditional load cells. Since each precision load cell is essentially the same as the others in the array, shifting dry bulk material load or a changing center of gravity does not effect the weight measurement. This important characteristic effectively eliminates band-aid solutions such as electronic filtering during dynamic conditions that occur in mixing tanks/vessels.
All of the analog load cells in the array are simply wired in parallel in a junction box. Some load cell manufacturers have specifically designed their load cell junction boxes such that there is minimal effect from water contamination due to condensation and/or moisture intrusion through the electrical conduit. A single ‘Home Run’ extension signal cable connects the j-box to the weigh electronics for indication, control and/ or to communicate on a particular field bus such as Profibus-DP, DeviceNet or Ethernet.
Advanced weighing electronics now available have special software routines to detect and identify failures. When used with precision matched load cells, some electronics can be changed out without recalibration. Simply enter previously saved calibration data into the new electronics in the facility instrument shop and exchange wires in the field reducing production downtime to a minimum. Industrial end-users who service their own scales have reported that tank scale maintenance costs have actually declined significantly after investing in higher quality weighing equipment.
Batch weighing is the most accurate method of dispensing, accumulating or receiving ingredients. Again, it is important to bear in mind that an inaccurate load cell system will yield inaccurate or non-repeatable batches regardless of any other issues that can be addressed separately. The latest generation of batching controllers with PLC based software code not only controls the ingredients but also can control material in-feed and/or final batch discharge. Batching systems can also bias or ratio the second, third, etc. ingredients against the actual amount of the initial (primary) ingredient charged into the weigh hopper for extremely consistent blend control.
PC based recipe and material management reporting software is also available to connect multiple batch controllers to a common data acquisition point. The remote scales normally communicate to the PC via a serial port or Ethernet connection. The batching system should allow the operator to easily create new formulas that can be downloaded to the batch controllers. All ingredient transfer amounts should be automatically sent to the PC to allow for material tracking. Any weighing or batching system considered for purchase and installation should have open system architecture. Failure to adhere to this requirement effectively leaves the end user at the mercy of the chosen equipment vendor with regard to costly code modifications/rewrites.
Process automation, whether a weighing system or a complete batch controls system, is a capital investment with identifiable areas of return on investment (ROI) based on its performance. Because the load cells are an integral part of the ROI equation, it is vitally important to take the time necessary to choose a qualified vendor and product. Multi-million dollar projects have been subject to costly delays due to the inadequate performance of the weighing system when subjected to dynamic process conditions. Load cells, once installed, become an integral part of your process and should be expected to provide many years of reliable and trouble free performance.
The load cell manufacturer should have a reference list with names and numbers to contact regarding past performance history regarding accuracy and reliable usage of the load cell technology being offered. Load cells are offered with warranties varying from one to three years plus. Highly reliable/dependable precision load cells are hermetically sealed having IP68/NEMA 6 watertight environmental rating and are provided with Limited Lifetime Warranties guaranteeing a maintenance free installation.
The author, Todd A. Rae, is president of Instruments, Controls and Equipment (ICE) Corporation, a Maryland-based manufacturers’ representative firm. Mr. Rae has over 20 years experience in the field of process control and weighing systems. He can be reached at 410-452-0070 / 443-928-9500.
* * *
Previous Feature: Elevating Educational Prerequisites For Future Engineers | Next Feature: Duplicating Nature’s Nanoscale Construction Techniques