_edited.png)
SMART Risk Assessment
Under sponsorship from the Federal Aviation Administration (grants 12-G-012 and 16-G-005), the University of Texas at San Antonio (UTSA) and partners St. Mary’s University, Textron Aviation, Nuss Sustainment Solutions, and Fieldstone Software, have developed a software suite to assist engineers in performing risk assessments of aircraft structures, as well as other aerospace structures, such as transport, rotorcraft and military. Two software products are available: SMART|LD (probabilistic S-N), and Smart|DT (probabilistic damage tolerance). SMART|LD considers classic fatigue failure with probabilistic loading, S-N curves, Stress Severity Factor (SSF), and an option for user-defined spectra and SN Curves. SMART|DT considers explicit crack growth with probabilistic loading, fracture toughness, da/dN material data, initial crack sizes, geometry, and inspection capabilities. Interfaces to Afgrow, Nasgro, and Fastran are available. Training material is provided at this web site and at relevant conferences. The software programs are to be considered beta and regularly updated by UTSA.
Developed By:
Harry Millwater (PhD, PI - UTSA)
Juan Ocampo (PhD, StMU)
Beth Gamble (TA)
Chris Hurst (TA)
Marv Nuss (NSS)
Nathan Crosby (PhD, AeroMatter, LLC)
Sohrob Mattighi (Program Manager FAA)
Mike Reyer (FAA Kansas City Office)
Sara Restrepo-Velasquez (PhD Student, UTSA)
Products:
Click on the buttons to see the capabilities of each software
Technology Summary
White - Version 1; Red - New in Version 1.1; Blue - Future Version
-
Loading Generation
–Computed from exceedance curves (Internal library and user exceedance option) – Weighted usage available.
​
–Flight Duration/Velocity and weight matrices, Design load limit factors, one-g stress, and ground stress as user input.
​
-
Damage accumulated using Miner’s rule
–Safe-Life calculations (in # of flights and # of hours) using MC sampling
​
–Accumulated damage calculation based on the user number of flight hours.
​
–Probability of failure computed using MC sampling
​
-
Multiple random variables
–Library of exceedance curves (weighted mix ok) – Option for user input exceedance.
​
–Flight duration, a/c velocity, one-g stress, and ground stress
​
–PSN curve constructed from constant amplitude tests – Option for user input PSN
​
–Sink Rate
​
–Random damage coefficient.
​
–Stress Severity Factor (SSF) option
​
-
Text output files showing Monte Carlo results
​
-
Sensitivities computed using correlation and scatter plots
​
-
Life distribution and hazard rate calculation
​
-
Standard Fortran 95/03, Unix and Windows
​
-
GUI
​
-
Fleet Management
-
Probability calculations
–Single flight probability of failure (Lincoln and Freudenthal)
​
–Precent cracks detected at inspections
​
–Remaining Useful Life Calculations
​
–Crack size distributions
​
-
Probabilistic methods
–Standard Monte Carlo (multi-threaded)
​
–Numerical integration
​
–Weighted Branch Integration Method
​
–Adaptive Multiple Importance Sampling
​
-
Random variables
–ai, Kc, Evd, yield strength, ultimate strength, da/dN, hole diameter, hole offset, crack aspect ratio
​
-
Crack growth
–Master curve (ai, Kc and yield strength) interpolation (user input)
​
–Through, Corner, Surface cracks
​
–Hypergrow (efficient crack growth analysis)
​
–Failure due to fracture instability or net section yield
​
–Direct link to Nasgro
​
-
Extreme Value Distribution
–User input, e.g., Gumbel, Frechet , and Weibull-Max.
​
–Ultimate/Limit load (deterministic)
​
–Computed from exceedance curves, weight matrix, etc. (Gumbel, Frechet , and Weibull-Max)​
​
-
Inspection capabilities
–Any number of inspections
​
–Arbitrary repair crack size distribution (lognormal, Weibull, deterministic, tabular)
​
–Arbitrary POD (lognormal, deterministie, tabular)
​
–User defined probability of inspection
​
–Different repair scenarios between inspections
​
-
Probabilistic Database
–Material Properties (Kc, dadN, yield, ultimate) Aluminum, steel, titanium
​
–Equivalent initial flaw size
​
–Probability of Detection Curve
​
–Hole diameter
​
–Edge distance
​
–Exceedance curves
​
-
Loading Generation
–Computed from exceedance curves (Internal library and user exceedance option) – Weighted usage available.
​
–Flight Duration and weight matrices, design load limit factors, one-g stress, and ground stress
​
–Stresses and/or flights randomizations
​
–Spectrum editing option (Rainflow, rise/fall, Dead band)
​
–User-defined spectra (Afgrow format)
​​​​​​
-
Optimized Inspections
–Risk based: time of inspection determined due to risk constraint
​
–Cost based: time and type of inspection determined due to risk constraint
​
-
Fleet Management
​
-
Computational implementation
–Standard Fortran 95/03/08, Windows, Mac OS and Linux
​
–HPC Implementation (parallel and vectorized)
​​