Phased Array Availability Modeling and Simulation: Techniques for Efficient and Effective Performance Modeling
Introduction and Availability Overview
Raytheon has a legacy of delivering high-availability phased array radars on ground, sea and airborne platforms. A sample of our 30+ year history of phased array systems is illustrated in Table 1. This broad range of experience has allowed Raytheon to develop sophisticated availability modeling and simulation techniques, which are continually refined to assess performance for evolving and varied mission and support environments.
While we typically think of system “reliability,” or the probability of failure-free mission performance, as the common measure of system dependability, it is the system “operational availability” that has gained recent visibility in terms of mission readiness and system support costs. Availability is the probability of being able to accept a mission with full capability given that the mission may be requested at any random time. The joint probability of being able to accept a mission (availability) and then complete the mission (reliability) is a common measure used in system effectiveness models.
Due to the thousands of elements in a typical phased array, the rate of single item failure is relatively high (daily in certain very large arrays). Therefore, performing immediate maintenance on a single failed item is often not practical or cost effective. Phased arrays typically operate in accordance with periodic scheduled maintenance, whereby a system will deploy for weeks or months with no array face maintenance. To accommodate this support approach, Raytheon designs arrays such that when failures inevitably occur, performance degrades gracefully while maintaining specified levels of performance until scheduled maintenance can be performed. The challenge for reliability and availability engineers is to ensure and verify that the array design has sufficient margin to successfully complete “maintenance-free” missions, and that sufficient support is provided (e.g., spares and personnel are made available) to restore the array to full health prior to the next deployment.
High Availability Phased Array Radar Antenna Architectures
The high availability of active phased array radars is made possible by the antenna architecture’s ability to degrade gracefully. The architecture is designed to efficiently distribute waveforms, power and control to an array of hundreds or thousands of individual transmit (Tx) and receive (Rx) elements. Because transmit and receive elements are distributed on an array in large numbers, individual elements make a relatively small contribution to system performance; thus, failures are very tolerable. Transitioning upstream from the transmit and receive elements, the antenna’s power, radio frequency (RF) and control line replaceable units (LRUs) support dedicated groups of Tx/Rx elements and therefore make a larger contribution to system performance and decreasing tolerability of failures. The degree of performance degradation from any individual LRU failure increases as the failures occur farther upstream within the distribution chain. Figure 1 summarizes the reliability architecture of a gracefully degrading phased array antenna and the associated failure effect and maintenance concepts linked to primary antenna functions.
In addition to high availability, a well distributed, fault tolerant, phased array radar architecture offers attractive and flexible maintenance and support options. For example, if projected array failures during a deployment are tolerable, there is no need for costly on-board spares; nor is there a need for the associated on-board maintenance manpower. The resulting ability to defer maintenance and centralize spares significantly reduces the logistics footprint. It is interesting to note that deferred maintenance does not improve costs by reducing the number of overall repairs. Rather, it improves costs by way of efficient batching of repairs without impact to mission operations.
While the performance of our phased array antennas degrades gracefully as failures gradually accrue, Raytheon’s continuous health assessment function provides status on all failures, regardless of criticality. This health monitoring of arrays provides the status and data to allow commanders to monitor degradation, and it enables fix-or-fight decisions prior to mission.
Phased Array Radar Availability Performance Simulation & Modeling
Raytheon employs a variety of modeling and simulation techniques to accurately evaluate array availability performance. Tools that focus on inherent availability are implemented during architecture and preliminary design trades. Inherent availability assumes an ideal support (maintenance and spares) environment. The Raytheon-developed Phased Array Combinatorial Reliability Analysis Tool (PACRAT) is a Monte Carlo simulation tool that allows rapid assessment of all possible projected failure combinations during a specified deployment or mission period. PACRAT computes the probability of mission success and availability where success is defined as performance above specified thresholds. Typical array performance thresholds include antenna sensitivity (which is expressed in the availability simulation as the maximum number of failed Tx/Rx modules), and antenna pattern degradation (which is expressed in the simulation as both the maximum number of failed Tx/Rx modules and the maximum number of Tx/Rx module groupings). These groupings are specific to the architecture and relate to failures of upstream assemblies. Before tools like PACRAT were available, phased array availability calculations required time consuming individual calculations for the thousands of possible failure combinations that could occur during a mission. Using a 16 element array example, Figure 2 illustrates how PACRAT significantly reduces the analysis and computational effort. PACRAT also provides an effective approach for conducting design trades involving array module density (e.g., elements per field replaceable assembly, assemblies per upstream assembly, etc.).
In addition to inherent availability, an operational availability assessment is required to optimize radar support concepts and to verify array compliance to reliability and availability specifications in an operational environment. The additional maintenance and support inputs required to compute operational availability include spares provisioning, system corrective and preventive maintenance, depot repair factors, transportation, and maintenance manpower. Raytheon uses a variety of powerful operational availability simulation tools to support these types of analyses and to develop a cost-effective system support structure.
While tools like PACRAT provide rapid inherent availability calculations for “maintenance free” portions of operation, operational availability requires more complex calculations and tools. This complexity is driven by:
- Multiple maintenance strategies (e.g., replace some array items more frequently while deployed and defer others until end of deployment).
- Multiphase mission profiles with phase-dependent allowances for maintenance and varying operational tempo.
- Complex failure dependencies.
- Limited support resources (spares, personnel, funding).
- Time-dependent failure rates (e.g., “wear out”).
- Need for optimization analysis (e.g., spares cost, maintenance personnel).
To assess operational availability, more complex Monte Carlo Reliability Block Diagram (RBD) simulation techniques are used. First, a RBD model must be constructed to accurately capture the array’s fault tolerance. Next, all the blocks must be populated with the required attributes (e.g., failure rate and distribution, repair rate, etc.), and the spares and maintenance resource pools must be defined and/or assigned. Finally, a model comprising the mission phases that define the RBD operational environment is employed. Once the models and resource pools are established, a Monte Carlo simulation of system operation is executed for the defined missions and the entire system service life. Figure 3 summarizes the failure and support flows of this modeling and simulation approach.
Raytheon relies on commercially available simulation capabilities such as Windchill®Quality Solutions™ Reliability Block Diagram to develop the specific RBD models and to simulate system availability performance. When required, Raytheon also employs customer-specified simulation tools to support modeling and simulation (M&S) integration with a user’s higher-level “system of systems” model. Regardless of the M&S tool used, results are only valid if populated with valid data inputs. Raytheon’s decades of field reliability analysis, operations and sustainment (O&S) and depot repair operations provide a wealth of real-world operational data required for accurate Ao forecasts.
Raytheon also uses its M&S capability to help customers determine the additional support required to “dial up” availability. For instance, an additional user investment in on-board spares can significantly reduce the probability of system down-time due to insufficient provisions, and thus provide much higher operational availability for extremely critical missions.
Operational availability (Ao) modeling and simulation is not just a technique to support Ao requirements verification. Raytheon routinely extends Ao M&S into the O&S phase. Raytheon’s models can easily transition from “predicted” inputs to “field-observed” inputs that are obtained from the mission environment. The integration of models with the Failure Reporting Analysis and Corrective Action System (FRACAS) and Reliability Centered Maintenance (RCM) activities provides a powerful feedback loop, which includes field failure rates, repair cycle times and wash-out rates. This living availability model enables accurate system availability performance assessments, helps identify candidates for technology refresh, optimizes scheduled maintenance, and serves to measure and prioritize reliability growth initiatives throughout operations and sustainment.
Share This Story