Professional Documents
Culture Documents
ABSTRACT
Structural Health Management (SHM) and its related technologies offer the
combined benefits of reducing maintenance costs while preserving or improving
structural design performance. Advanced SHM technology implementation for in-
service aerospace platforms requires a disciplined approach to quantify how each
application can provide benefit in the context of design practices and design and
maintenance criteria. The focus of this paper will be to describe SHM from a broader
integrated design perspective and vision for the future. This will be accomplished by
first providing a brief historical illustration of how design practices have evolved
along with the ability to collect in service knowledge. An overview of SHM from a
design and integration perspective will then be provided, followed by some
examples that demonstrate SHM implementation in the context of a broader design
and maintenance perspective. The paper will conclude with a list of challenges for
the SHM community in the context of a broader integrated design vision.
INTRODUCTION
Jerry Young, Eric Haugse and Chris Davis, Boeing Co., P.O. Box 3707, Seattle, WA 98124, U.S.A.
3
Figure 1. Design practices have evolved with increased service knowledge.
factors such as humidity and salt exposure were incorporated into supplemental
inspection programs and other preventative maintenance standards.
Finally, the failure of the very flexible Helios vehicle as a result of coupling
between the control system and an extreme gust environment taught us about the
importance of load monitoring to aide control-structure interactions [3]. Full scale
testing often uncovers anomalies which after investigation and with more analysis
and test become explainable. We evolve and improve by incorporating this
knowledge into our design process and identify technological advancements for use
on future vehicles.
At the same time that we are testing, monitoring, inspecting our products, the
performance of our systems are increasing significantly. Missions now might last
days where 50 years ago hours were exceptional. The life of our industry’s products
is now decades where years would have been acceptable in the past. Payloads and
efficiency also have increased significantly. Our ability to analyze has taken us
from static sizing to design for safe life/fail safety. Design cases or load cases have
gone from a handful to tens of thousands. And finally the data associated with each
of these designs and their certification has grown from notebooks to terabytes. It is
truly a daunting task to create knowledge from of all of this data and information.
So we are more informed with each design cycle as we learn about the last. Our
assurance of safety has evolved with our experience and knowledge. In Ulf
Goranson’s 1997 keynote speech at the Stanford workshop for Structural Health
Monitoring, he characterized structural safety as an evolutionary accomplishment
[4]. He described the evolutionary approach based on attention to design features,
acquisition and review of fleet service data, use of technology standards to ensure
analysis commonality across the fleet, and finally teardowns on older aircraft. He
detailed how designing and monitoring the health of an airframe during its life was
an integral process. Experience through test and in service operations was evaluated
and incorporated through “standard technology” back into the design and process.
Ulf showed how velocity, gravitational acceleration, and altitude (VGH) sensors
were used to alter operations of the fleet and improve life. Similarly, design
technology standards reduced the number of fatigue findings in subsequent
generations of airframe structure and the use of teardown and inspection improved
these standards for the future. The approach was institutionalized at Boeing through
the use of the Boeing Design Manuals as shown in Figure 2. Ulf noted in
conclusion that “monitoring” the health of the fleet and institutionalizing
disciplined technology standards resulted in an order of magnitude reduction in
fatigue and corrosion findings in the second generation fleet.
4
Figure 2. Disciplined technology standards have enabled an evolution in design.
5
Figure 3. SHM technology must be integral with design and design criteria
6
Figure 5. New capabilities help all phases of a design’s life cycle to enable gathering data
that develops and optimizes in service knowledge and its use.
and appetite for analysis has kept pace with Moore’s law challenging software,
hardware and engineering capability. Today we are able to look at our products
from atoms to airplanes. Computational materials tools are used as well as
computational fluid dynamics (CFD) to understand the details of every aspect of
our design. We put sensors on the vehicle to measure what actually happens to the
vehicle in service. This diagnostic tool set is tightly coupled together and
interconnected across scales in an effort to optimize knowledge about a product
throughout its life cycle.
7
Figure 6. Tomorrow’s vision must include a much more integrated solution.
The use of advanced SHM technologies like flight parameter based load
monitoring can be used to trigger inspections autonomously. In one scenario,
operational loads can be monitored and used to set manual inspection intervals as
opposed to the current method that uses an analytically estimated load spectrum. In
a more rigorous scenario, load monitoring can again be used to set inspection
intervals, but the inspections would be carried out autonomously using sensors,
material modeling, and damage sizing algorithms as opposed to manual inspections.
This type of change would of course need to be considered in the context of overall
aircraft system inspection practices.
In addition to reducing maintenance costs, it is suggested that this approach
could also be used early in the product development phase to reduce design
conservatism with the understanding that there would be more knowledge about the
8
usage and integrity of the structure during its operation. Advanced SHM
technologies then become design elements, providing additional data, information,
and knowledge to better optimize at the system level, across the life cycle of the
platform.
To assess the potential impact that load monitoring could have on inspection
interval, Boeing utilized real time operational load data recorded by the University
of Dayton Research Institute (UDRI). The FAA’s Airborne Data Monitoring
Systems Research Program funded this data collection. The data was used by
Boeing to perform a durability and damage tolerance analysis and compare the
results to the same analysis using design load spectrum. As shown in Figure 8, for
many of the airplane data sets, preliminary results indicate that inspection intervals
could be increased based on load monitoring and thereby demonstrates potential
benefit. It should, however, be pointed out that this analysis is still in progress and
that inspection intervals might not be defined based on fatigue loading alone.
The example shown in Figure 9 illustrates the use of flight parameter data and
design modeling tools to evaluate the impact of a hard landing event. The pilot-
initiated reporting of such events results in maintenance actions that require in-
depth inspections which can cause significant down-time and labor costs. The
complexity of the dynamic event coupled with the random “event severity
estimation ability” of the pilot can result in false report rates (i.e., no damage found
during the required inspection) in excess of 85% of the time. The maintenance
process could be significantly improved if knowledge of the loads generated during
the landing event were available to guide the inspection or, in some cases, perhaps
even negate the need for one. Direct measurement of load-related quantities (e.g.,
strain and/or acceleration) at every significant structural item that is sized or
affected by landing loads can be costly. The goal of this study was to show how
flight-parameter-based models could be constructed to estimate landing loads that
would, in turn, aid the maintenance process. The challenge was to develop the right
model with the right set of inputs that was reliable and economically and
technically feasible.
Fortunately, high fidelity physics-based dynamic models of the landing event
exist and are validated. These time-domain models couple rigid and flexible
9
Figure 9. Hard landing detection system demonstrates load monitoring capability.
structural dynamic models with aerodynamic models. The approach state (e.g.,
attitude, velocity, acceleration) of the aircraft along with aircraft configuration (e.g.,
weight, c.g. location) and runway condition (e.g., dry, wet, icy) define the initial
conditions and configuration of the model. The model outputs time varying loads at
predetermined significant structural items. In order to provide such high fidelity
results, however, the models must also be quite detailed and thus require significant
computational resources to run. As it would not be feasible to field a practical
solution with such a complicated model, the technology challenge was to develop a
lower fidelity model that could approximate the response of the detailed model
using readily available data and information. Readily available data and
information included flight parameter data (e.g., airspeed, sink rate, roll and pitch
attitude, etc.) that would typically be available on one of the many aircraft bus
systems and potentially additional structural sensors (e.g., strain gage,
accelerometer, temperature, etc.).
An in-depth study was performed that traded model inputs and topologies.
Model performance was ranked by the ability to correctly predict the loads with
limited false positives and no false negatives. The “tolerance” for false positives
was traded against adding dedicated structural sensors. Results of the study showed
that a flight parameter only solution could reduce the percentage of false positives
from 85% to approximately 5%.
The example illustrated in Figure 10 is focused on structural criteria to address
damage caused impact events during operational usage. Once an event is identified
by the ground or flight crew, it must be measured to determine the level of damage.
If damage is within predefined range, a rapid field repair process could be used to
restore strength to ultimate load carrying capability. Advanced SHM technology
offers the potential to significantly reduce the time to determine damage size and
minimize costly down time for the aircraft.
The SHM damage evaluation method will have some uncertainty that must be
considered as indicated by the damage detection error bar in Figure 10. This
10
example offers the possibility of trading SHM sizing capability and reliability
against structural design criteria and repair capability.
The field repair example was an eye opener with respect to SHM design
requirements and objectives. Sizing damage in three dimensions at barely visible
impact damage (BVID) levels to meet Design Requirements and Objectives
(DR&Os) was a significant challenge. This application demonstrated the need to
develop the methods and tools to integrate SHM hardware design and data
processing algorithms with structural design criteria. Models for sensor systems and
processing algorithms needed to be coupled with physics based material models to
develop and trade virtual design options prior to validation testing.
This example illustrated in Figure 11 is a condition based maintenance (CBM)
approach. It combines the concepts of operational load monitoring, with in service
damage monitoring, in the context of reliability based criteria. The calculation of an
Figure 10. Damage sizing information is considered an element of repair design and
criteria to help optimize the repair design process.
Figure 11. A change in analysis and design criteria enables new technology integration [5].
11
Equivalent Initial Flaw Size (EIFS) along with operational loads collected by an
individual aircraft usage monitoring system are key pieces of information needed to
enable a reliability based approach. Sufficient crack data is collected during fatigue
testing to develop initial EIFS probability distributions. The EFIS distributions are
then used during operations along with load monitoring data to predict damage
growth information. It is suggested that an additional and continued improvement to
this approach could be accomplished by collecting critical crack length detection
data during operational usage. A probability distribution for crack length detection
in time could then be used to further update EIFS distributions and thereby further
improve reliability based estimates for damage growth.
This is a very illustrative example as it brings together advanced SHM
technologies in the context of reliability based criteria needed to provide the kind of
knowledge that enables a CBM approach. It also offers the ability to update
knowledge as we learn from operational data collected over a product’s life.
As illustrated in Figure 12, each of the examples in the last section described
applications where advanced SHM technologies could be integrated, at varying
levels, as part of an aerospace vehicle’s design criteria and maintenance practice.
In Ulf Goranson’s 1997 keynote speech, he characterized structural safety as an
evolutionary accomplishment. He described an evolving approach based on
attention to design features, acquisition and review of fleet service data, and the use
of technology standards to ensure commonality across the fleet. He detailed how
designing and monitoring the health of an airframe during its life was an integral
process and how life cycle experience was evaluated and incorporated through
“standard technology” back into the design process.
In the past twelve years we have seen significant improvements in software and
computing capability, diagnostic hardware, engineering methodologies, and new or
enhanced design criteria. However, implementation of advanced SHM technologies
into design practices and operational platforms has been limited.
To increase implementation of new SHM technology, tomorrow’s vision must
continue to focus on an integrated approach, where designs and design criteria are
developed with advanced SHM technologies in mind. SHM technologies must be
developed by evaluating them in the context of a new CONOPS paradigm, new
criteria, and advancements in other design tools and practices.
Figure 13 lists some suggested areas for development in the context of an SHM
process flow from data to decision making. As discussed in the first part of this
paper and demonstrated in the examples, it has been found that the most effective
development of new technologies should be done in the context of this overall
process that starts with understanding the CONOPS for which decisions will be
made about the integrity of the structural design. Next we need to understand the
data and first hand information needed to develop knowledge about the state of the
material and structure used in the design in the context of the CONOPS. The
challenges shown in Figure 12 are based on this understanding and are focused on
gathering high quality data, processing it with analytic methodologies that best fit
the targeted design to glean the most knowledge of state, and allowing for changes
in criteria and CONOPS that optimizes the safety and performance of our products.
12
Figure 12. Application examples demonstrate strong tie between design practices and
criteria and the integration of advancements of SHM technologies.
Figure 13. SHM challenges to meet a disciplined approach for integration with design and
maintenance practices.
REFERENCES
1. Babish, IV, C.A. 2008. “USAF ASIP: Protecting Safety for 50 Years”, presented at the 2008
Aircraft Structural Integrity Program Conference, 2-4 December 2008.
2. Redmond, Gerard. 2001. “From ‘Save Life’ to Fracture Mechanics – F111 Aircraft Cold
Temperature Proof Testing at RAAF Amberley,” presented at the 10th Asia-Pacific Conference
on Non-Destructive Testing Conference, 17-21 September 2001.
3 Noll, T.E., S.D. Ishmael, B. Henwood, M.E., Perez-Davis, G.C. Tiffany, J. Madura, M. Gaier,
J.M. Brown, and T. Wierzbanowski. 2007. Technical Findings, Lessons Learned, and
Recommendations Resulting from the Helios Prototype Vehicle Mishap.. UAV Design
Processes/Design Criteria for Structures, pp. 3.4-1 – 3.4-18.
4. Goranson, U.G. 2008. “Jet Transport Structures Performance Monitoring”, presented at the
1997 International Workshop on Structural Health Monitoring, 18-20 September 1997.
5. Gallagher, J.P., C.A. Babish, and J.C. Malas. 2005. Damage tolerant risk analysis techniques for
evaluating the structural integrity of aircraft structures. Proceedings of the Structural Integrity
in Transportation Systems Symposium, 11th International Conference on Fracture, March, in
Turin, Italy.
6. Torng, T., K. Chan, and K. Liu. “Development of an Integrated Durability and Damage
Tolerance Risk Assessment Strategy for ASIP”, ASIP Conference (2009).
13