Prestressed bolt joints under dynamical load Kerekes Ervin
Prestressed bolt joints under dynamical load Kerekes Ervin*, Dr. Szabó Bertalan**, Bernhard Unger* * Steyr-Daimler-Puch AG, Technologie Zentrum Steyr ** Department of Steel Structures, Technical University of Budapest
Contents • Introduction (TZS BME) • Tests (test bench, load spectras, failures) • FE model (whole structure, simplified bolt model, FE loads, stress distributions) • FEMFAT concept (theoretical background) • Fatigue estimation (fatigue load, damage distributions) • Comparison between the tests and the calculations • Conclusions
We can find everywhere bolt joints Notch effect L Detachable J
Test bench
Three different end plate versions
The load spectras and the results of the tests A 1/1 and A 1/2 A 3
The half FE model of the test bench Contact area
Simplified FE model of the bolt joint Advantages: • few number of elements • linear • very fast • enough precision
Load Case 1, Prestress of the bolts (minimal load), • On the I-beam • On the flange plate
Load Case 2, The total load (maximal load), • On the I-beam • On the flange plate
FEMFAT technical contents Load spectra Technological influence factors Material data FE stress data (linear stress tensors) Endurance safety factor Damage factor Brake safety factor
FEMFAT calculation concept Load spectra s ai s mi FE model s i s mi s ai N Neuber hypothesis S/N curve s k s ai s. D Relative stress gradient D 1 ND e s mi SD D 2 i c* s. D k c* ND k s. D c* sm . . . Dn i
Load case definition for Fatigue Analysis Amplitude stress: Mean stress:
Damage distribution on A 1/1 without sequence effect • On the I-beam • On the flange plate
Damage distribution on A 1/2 without sequence effect • On the I-beam • On the flange plate
Damage distribution on the A 1/1 and the A 1/2 cases with sequence effect • In the A 1/1 case on the flange plate • In the A 1/2 case on the flange plate
The photos of the breaks in the A 1/1 and the A 1/2 cases • In the case of A 1/1 • In the case of A 1/2 Damage !
Comparison the tests and the FEMFAT results Test break FEMFAT calculation without sequence effect FEMFAT calculation with sequence effect
Why have we got difference? Calculation side • • Test side • • Linear and simplified hypothesis Material properties from literatures Simplified bolt model Lifetime ~ sk ð small amplitude stress changing means high lifetime changing ! D=1. 77 D=1. 07 D=0. 67 Test failures Too few tests
Conclusions • From the tests and fatigue estimation we can say, the dangerous places for this kind of load conditions are the middle screw bores on the flange plate. • The too high prestress is the reason of this weak point • The prestresses on the bolts mean a constant stress, and this is dominant for this type of load.
- Slides: 20