Unconfirmed Minutes SAE Fatigue Design and Evaluation Committee Fatigue Life Prediction Division

April 16 and 17, 1996 Raleigh, North Carolina

Divisional Progress Meeting (April 16):

  1. The meeting was called to order and the minutes from the October meeting in Toronto were approved as written.
  2. Phil Dindinger (for Tom Cordes) described to the committee a weld fatigue life prediction study that they have performed. In this study, specimens with two weld geometries were tested using the SAE Log Skidder Bending History and resulting fatigue lives were determined. Committee members are urged to make fatigue life predictions for the specimens and compare methods and results. Material properties, geometry, and service history data will be provided to those interested in participating. See the minutes for the planning session below for more detail.

  3. Yung Lee of Chrysler Corporation gave a presentation on "Load Acquisition and Damage Assessment of a Vehicle Bracket Component". Mr. Lee described a five step design process: prototype, road load data acquisition, damage analysis, determination of life test bogey, and product validation test. Central to this process is the determination of road loads, from which a damage assessment can be made to develop a laboratory test that is correlated to the proving ground. Considering the transducer requirements, the effort was made to use matrix techniques with strain gages to convert the component (bracket) into a force transducer. Papers by Masroor and Zachary at Iowa State University, and Wickham and Galliart at Deere and Co. were cited. After discussing some sources of error, the process was outlined. Strain gage locations are first determined using brittle coatings. Load-strain relationships are measured and the gage locations are optimized using design of experiments techniques. Proving ground strain histories are then gathered and the resulting loads are calculated using the load-strain matrix. Fatigue damage calculations are performed and a life test bogey determined.
  4. Yung Lee of Chrysler gave a second presentation entitled "Reliability Based Damage Assessment". He discussed the issue of determining the Weibull B10 life with 90% confidence considering test driver variability. Working with load-life curves, they shift the mean curve (50% reliability, 50% confidence) to 95% reliability, 95% confidence. To understand driver variability, he described a program where vehicles were instrumented with strain gaged axle shafts. Three drivers were selected to drive two different proving grounds road schedules and resulting usage data were recorded and analyzed. The adjusted load life curves and driver variability data are then incorporated into the fatigue damage assessment.

Divisional Planning Meeting (April 17):

Discussions were focused on the following five areas.

Weld Fatigue Study

Plans were made to put together a "problem package" for the weld fatigue study discussed in the progress meeting of April 16. Available data include local strain material properties data, dA/dN and delta K data for crack propagation, geometry details including the weld toe radius, service history (Log Skidder Bending), and normalized fatigue test life data. The objective is to make absolute fatigue life predictions and discuss results and techniques at the next meeting. It was suggested that welding codes such as BS5400, ASME, and techniques such as the hot spot stress method be used for comparison. Current plans (at the time of this writing) are to try to make the data available for distribution at the October meeting.

Notched Shaft

It was suggested that a summary table of all test results should be added to the Web Site.
It was proposed that the work done at VPI (presented by Patrick Le Moal at the main committee meeting) be correlated to the previous induction hardened test results and reported at the next meeting.
It was mentioned that shot peening may be a possible extension for the notched shaft program.
A question arose regarding the status of the variable amplitude work on the hardened notched shaft. The results of the committee work on that shaft are to be documented in the successor to AE14.

Composites

As noted in the Materials Properties Division Meeting, additional SRIM specimens have been tested at R= 0.1 and R = -1. The suggestion was made to use A. Kallmeyer's damage parameter to analyze these data. See SAE paper 930405, "Low Cycle Fatigue Behavior and Variable Amplitude Fatigue Life Calculations for an SRIM Polymer Matrix Composite".

Powder Metals

Current efforts on the new material were discussed. Ralph Stephens gave a summary of work being done at the University of Iowa. Suggestions were made to evaluate the applicability of local strain and/or crack propagation approaches, and to investigate mean stress behavior of the "new" material.

Thermal Fatigue

Thermal fatigue issues were briefly discussed. It was suggested that thermal fatigue be included for presentation topics in the future.