(703e) Built-in-Tests for Thermal Fluid Systems of Aerospace Applications

The reliability, cost, safety and environmental impact of
complex and uncertain engineering systems that are prone to faults, drives
fault detection and isolation (FDI) methodologies to the forefront. More
specifically, the efficiency and accuracy of these FDI methods are increasingly
important among competitive industries, such as aerospace, where system
downtime can have a significant impact on profit. Advancement of FDI methods
can increase system reliability, availability, and safety through more accurate
fault diagnosis. Model-based FDI methods entail advantages over hardware
redundant or data-driven approaches in their absence of need for additional
hardware, as well as the increased accuracy over wide operating regions, due to
their use of first principles and/or empirical correlations.^1,2 Recent advances in model-based FDI methods are
applicable to nonlinear systems of thermal fluid dynamic processes, such as
aircraft environmental control systems (ECS). For instance, looking at the
example of an aircraft ECS, significant fouling can occur over time within its
primary heat exchanger that can lead to performance degradation and abrupt
downstream faults. Current monitoring techniques, like Kalman
filtering or observer-based methods, which utilize statistical inferences, such
as cumulative sum tests, are incapable of accurately detecting the incipient phenomena
of fouling over short periods of time due to the small deviations in the system
performance during that time span.³ As a result, the aerospace
industry typically uses a shotgun approach in which the heat exchanger is
removed for maintenance at arbitrarily chosen time intervals.

In this presentation we propose a method for initiated
built-in-tests (iBIT) for aircraft ECS heat exchanger
fouling detection, in which the extractable test information is maximized on
the basis of a system model. Heat exchanger fouling detection, in terms of
quantification of its severity, is critical for aircraft maintenance scheduling
and safe operation. We focus on methods for offline fouling detection during
aircraft ground handling, where the allowable variability range of admissible
inputs is wider. We explore methods of optimal experimental design to estimate
heat exchanger inputs and input trajectories that maximize the identifiability of fouling. Fouling metrics, such as thermal
fouling resistance, are treated as parameters and analyzed along with a
combination of uncertain system inputs, such as inlet temperatures, moisture
content, and mass flows. The iBIT design vector, i.e. the system admissible inputs at
which fouling is to be estimated, is manipulated to optimize experimental
information and expressed via the Fisher Information matrix. Nominal and
optimal iBITs are compared with respect to their
capability to identify fouling, through parameter estimation, using the
corresponding heat exchanger measurements. It is shown that the proposed iBIT methodology for heat exchanger fouling identification
allows for accurate and precise estimation of the heat exchanger fouling, when
this would have been infeasible with current conventional methods and without
the addition of extra measurement devices or other BIT equipment. We close by
generalizing the iBIT formulation for continuous
non-linear systems with uncertainty.

Acknowledgment

This work was sponsored by the UTC Institute for Advanced
Systems Engineering (UTC-IASE) of the University of Connecticut and the United
Technologies Corporation. Any opinions expressed herein are those of the
authors and do not represent those of the sponsor. Help and guidance by Modelon and Modelon-AB are
gratefully acknowledged.

References

1.  Hwang, I., Kim,
S., Kim, Y. & Seah, C. E. A survey of fault
detection, isolation, and reconfiguration methods. IEEE Trans. Control Syst.
Technol. 18, 636?653 (2010).

2.  Simani, S. Model-based fault diagnosis in dynamic systems
using identification techniques. (Universita degli Studi di MODENA e REGGIO
EMILIA, 2000). doi:10.1007/978-1-4471-3829-7

3.  Isermann, R. Model-based fault-detection and diagnosis - Status
and applications. Annu. Rev. Control 29,
71?85 (2005).