*As Featured on NI.com
Original Authors: Paul Riley, Computer Controlled Solutions Limited
Edited by Cyth Systems
The Challenge
Overland and maritime transport of shipping containers using railways and cargo vessels is the most fuel-efficient shipping method relative we have today. To load railcars and cargo vessels requires colossal drive trains capable of safely moving shipping containers several tons in weight. To ensure the success of these drive trains, we needed to create an equally colossal test rig capable of safely applying hundreds of kilonewtons of torque and move a 25-ton positioning platform with submillimeter accuracy.
The Solution
Creating a data acquisition and control system using LabVIEW and CompactRIO for deterministic control so that one can safely handle and test a 30-ton drive train with high accuracy. We used the CompactRIO platform, with its user-reconfigurable FPGA and real-time deterministic operating system, to develop a solution that would not have been possible using existing programmable logic controllers (PLC).
Left: Hydraulic load and torque testing rig, Right: Container cranes (also known as gantry cranes) contain drive trains tested using CompactRIO.
Finding the Best Electronic Solution
Maritime transport is crucial to the world’s economy. With 90 percent of the world’s trade shipped by sea, it is the most cost-effective way to move bulk cargo across the globe. Ocean freight services also have a smaller carbon footprint than air freight. Efficiency continues to improve with the newer, larger generation of container ships exceeding 400 m in length and weighing up to 200,000 tons. Moving these giants requires a power plant the size of a three-story office building to drive gearing and propellers of huge proportions. To ensure this drive train succeeds, we needed a colossal test rig.
With a rig of these proportions and vast power capabilities, we needed an approachable development platform that keeps all aspects of control, acquisition, and safety in one system. Normally, when a rig like this is designed, there is a logger from one supplier, multiple PID control systems from another, and logic handling from yet another.
The downside with this approach is that it can end up separating responsibilities and skills across suppliers with each expecting to work on their own specialty as there is no commonality between the tools. Integrating the individual components can be complicated and costly, in both time and budget. Identifying the root cause of problems can take time as everyone oversees only their specific area of specialty. Many different vendors are used, which limits scalability due to the time spent on integration when any single component changes. This makes future proofing a design challenging.
We chose LabVIEW and CompactRIO as they have simple integration and expandability. Because they are part of the same platform, everyone can be familiar with every part of the system. This makes fault finding significantly easier. Scalability becomes straightforward too as CompactRIO features an expandable chassis for new I/O and sensible programming in LabVIEW makes adding new features a quick process.
Left: Electronics Control Cabinet, Right: NI-9038 & NI-9063 cRIO Controllers
Why CompactRIO?
We use CompactRIO and its I/O because the platform is:
Modular—Off-the-shelf hardware is available quickly and worldwide.
Parallel—All logic is efficiently coded in FPGA firmware with a 25ns response. FPGA use is a key decision; it means all the critical control and acquisition is handled at the same time instead of using microprocessors that must execute all logic in a series where one process can hang all the others.
A Single Development Environment—The whole project is self-contained in a LabVIEW project file, so no third-party add-ons can upset maintenance in future years.
Expandable—The modular nature and rack mounting of the CompactRIO product means we can easily expand the system in the future. Here, we used three CompactRIO systems all synchronized and handling hundreds of I/O with a high-bandwidth acquisition of data.
Moving 25 Tons and Applying Large Load
The test rig has a 25-ton platform that supports the 30-ton drive train. We raise and lower this with submillimeter accuracy to engage splines and then apply up to 400 kN of tensile or compressive load.
We also accurately apply 300 kNm of torque to the unit whilst it is rotating at constant speed. Considering the average sports car can produce 0.5 kNm of force, we were dealing with considerable torque. We applied the load using two large actuators precisely synchronized and able to swap between a displacement and load control mode. We based our solution on custom-written closed-loop control code on the FPGA. Based on standard PID control loops, the CompactRIO platform helped us design more complex algorithms to account for precise dual control of load and displacement. For accurate and noise-free measurement, we used digital devices wherever possible wired directly to the FPGA. We measured the torque frequency signal and absolute encoder data, which results in total calibration accuracy, 100 percent linearity, and high-speed measurement for critical feedback channels.
Flexible Software Design Criteria: How DIAdem and TDMS Are Essential
As this was a new machine for testing units at the end of production, we still faced questions about how to apply the high loads and torques, rather than just simply testing, analyzing, and reporting. Our engineers needed ultimate flexibility in using this rig for research, quality testing, and production test processes.
The solution was to design the software with a clear status ribbon along the main screen. This ribbon clearly indicated the full range of loading, installing, and testing the unit in such a way that the operator could step forward and backward at any time along the process. If required, the operator could then go once through a whole test process or skip parts or perform retests at will. The issue with this approach is: how do you acquire all this data in a tidy format, in one file, and analyze it with any popular package when you don’t know what data you will collect and in what order until runtime?
This is where saving in a Technical Data Management Streaming (TDMS) format helps. We open a new file when the unit under test is loaded and can then save separate data blocks at will, with full calibration information, varying channel count, and frequency as required. Data is grouped by type so that if an engineer performs a retest, the new data can be logically stored next to the data of the first test. This data is in a compact, single file that can easily be loaded into Excel, DIAdem, The MathWorks, Inc. MATLAB® software, and more with very clear metadata for analysis and reporting.
Fast Fault Detection
Detecting a transducer fault quickly is critical with a test rig of this power and size. Writing algorithms in LabVIEW on the FPGA allowed us to constantly monitor all critical transducers. Any failure instantly puts the rig into a controlled and safe shutdown procedure and clearly indicates the nature of the fault and its location to the operator for quick repair.
Original Authors: Paul Riley, Computer Controlled Solutions Limited
Edited by Cyth Systems
Comments