|
|
Residual stress measured in a hardened (carburized and quenched) steel
ring
Paper on these measurements and some nice FEM predictions:
M. B. Prime, V. C. Prantil, P. Rangaswamy and F. P. Garcia,
"Residual Stress Measurement and Prediction in a Hardened Steel Ring,"
Materials Science Forum, Vols. 347-349, pp. 223-228, 2000.
preprint (pdf) (LA-UR-99-5103)
The finte element model described in this paper includes
phase transormation kinetics and transformation induced plasticity (TRIP).
The Specimen:
- The specimen was a carburized low alloy 0.2% carbon steel ring
- Approximate Dimensions: 2.25" Outer Diameter (57mm)
1.4" Inner Diameter (35mm)
0.8" Height (20mm)
- At the left of the specimen in the photo you can just make out
some of the wire used for the EDM cut
|
Ring specimen after EDM cut. |
The test
- Two MM CEA-06-032UW-120 strain gauges were used (these have
gauge length of 0.032" (0.81mm))
- A slot was made using a Wire EDM (Electric
Discharge Machining) at the MST-7 Precision Manufacturing Shop
and a 0.002" (0.05mm) diameter tungsten wire
- Cuts were made in 0.004" (0.1 mm) increments to a final
depth of 0.054"
- The EDM machine is an anti-electrolysis Mitsubishi SX-10
- During cutting the specimen was submerged in temperature controlled
de-ionized water. This is necessary for cutting and also reduces
noise in the strain data
|
|
The raw data - measured strains
Great Sensitivity! - Look how big the strain readings are
The results - residual stress
- This is Hoop residual stress in the first mm or so of depth
- See paper (pdf) for information on finite
element model and surface x-ray measurements
- The residual stress is given as a 4th order polynomial from the least
squares fit used to reduce the data
- How well does this fit the strain data? See below
Fit to the strain data
- The solid lines are the strains predicted analytically by the residual
stress distribution shown above
- The symbols are the actual data
- The ± 1 microstrain fit error corresponds well with how well
you can measure strain
- The compliances were calculated using the solution for a slot of finite
width, reference # 20
NOTE: The raw data from this test and any other data I have can be made
available for research purposes. Just email
me. |
|