More and more companies are engaging in the creation of Digital Twins. What exactly is a Digital Twin? A Digital Twin is an up-to-date digital model of a real world asset or structure that incorporates relevant historical and current information regarding the characteristics of the asset or structure. For example, a Digital Twin of a tank would likely include dimensions, age, construction materials, coating history, information on the contents of the tank (historical as well as current), test data, and other information on the physical structure and operating conditions. The test data would likely include temperature vs. time, acoustic emission vs. time, corrosion damage, mechanical damage, and metallurgical properties. This collection of information, our Digital Twin, enables the performance of the tank to be modeled using a variety of software tools (e.g. Abaqus from Solidworks). Operators can also use the Digital Twin to migrate from scheduled maintenance to predictive maintenance. Effective implementation of Digital Twins has resulted in savings from 10% to 50%.
Creating a Digital Twin requires data…lots of data. Because all assets have a three dimensional shape, a digital model of the three dimensional shape is a critical component of the digital twin. The 3D model might be available in CAD. However, for many older assets, 3D CAD models are not available. Often, even when 3D CAD models are available, the as-operated condition of the asset is not the same as the as-designed CAD model. For this reason, low-resolution (10mm point spacing) 3D models are generally acquired when creating a Digital Twin. Below is a low resolution model of a boiler room.
Key sections of the asset are tagged with information markers that can be used to access additional data (e.g. temperature data, damage data, fluid flow data). The 3D digital model of the above boiler room is shown below as a collection of points.
The collection of points is 110MB of data. The collection of points is difficult to navigate, hence the need for software to enable a 3D walk-through. We can, however, slice through the points to locate specific sections of the boiler room that need further investigation.
In software, we have focused our attention on a small area of the boiler room. This area represents less than 1% of the total 3D data that comprise the boiler room.
In the above image, we have focused further to a single section of piping. We will now look at the damage associated with the piping. To accurately view the damage, we need a much high resolution image. The native resolution of the 3D digital model of the boiler room is 10mm. The resolution of the 3D image of the damaged area, shown below is white, is 0.1mm – 100 times more detail. This detail is necessary to identify the damage, in this case the damage is corrosion, and to determine the impact of the damage. The high resolution 3D image is 50MB of data and represents approximately 1/10,000th of the total surface area of the boiler room. Building accurate Digital Twins requires acquisition of large amounts of data.
Assessment of the damage can be done using either traditional API 579 methods or finite element analysis (FEA). In this case, we have calculated the impact of the damage use an FEA model. The image below shows a plot of the Von Mises stress associated with the corroded area.
Based on the results of the FEA modeling, the corrosion on the pipe has reduced the overall burst pressure of the pipe by 5%. When assessing the overall performance of this asset, incorporation of the impact of damage is a necessary part of the Digital Twin.
This case study has focused on one area of damage of an asset. A complete Digital Twin would incorporate not only damage from one section but all of the damage to the asset. In addition to the damage information acquired using NDT methods, the Digital Twin would need to include dimensions, age, construction materials, coating history, information on the contents of the tank (historical as well as current), and other information on the physical structure and operating conditions.
Contact us for more information on how we can help you build a Digital Twin.