Title: Block Shear : A Review
1Block Shear A Review
- Review of LRFD Rules
- Test Results
- LRFD rules v. test results
- Suggestions for improvement
2Block Shear Issues
- Gusset plates
- Coped beams
- Angles, tees
3An aside this is what were used to seeing
4But this is what actually happens (in gusset
plates)
5Failure mode in gusset plates
Failure is always by tensile fracture, regardless
of geometric proportions Shear yield along
vertical planes (through gross area, not
net) Failure is controlled by ductility not
strength.
6Another example of a gusset type test
7So, for gusset plates
- Use a model that combines
- tension fracture (on net tension area)
- shear yield (on gross shear area)
8G. Y. Grondin I did add the hole on the tension
face. If you dont want it, click on it and
delete. Your original line is behind.
Failure mode in coped beams
?y
?ult
9Coped Beam(holes in two lines)
10Coped Beam(single bolt line, slotted
holes)
11Some observations
- Mode of failure different for gusset plates v.
coped beams (as would be expected) - Does amount of rotation permitted in beam tests
affect results?
12AISC LRFD Predictions
- Shear yield (Ag) tensile fracture (An) or
- Tensile yield (Ag) shear fracture (An)
13AISC rules contd
with an upper limit, which is tension rupture
(on net tension area) shear rupture (on net
shear area)
14Difficulties.
- The when statements are illogical
- The Commentary is in conflict with the equations
(...use the larger) - The equations predict a failure mode that does
not exist in the tests - The predictions are not in good agreement with
the tests
15Gusset plate tests
- 117 tests from 6 different sources
- Good range of geometries represented
- Steel grades from 36 ksi to 100 ksi yield
16Gusset plate results
Source No. Tests Test / LRFD(1999) Comments
1 28 1.22 (0.08) Std. dev. thus (0.07) Results always conservative Predicted failure mode usually wrong
2 1 1.18 (0.00) Std. dev. thus (0.07) Results always conservative Predicted failure mode usually wrong
3 4 1.22 (0.07) Std. dev. thus (0.07) Results always conservative Predicted failure mode usually wrong
4 3 1.35 (0.02) Std. dev. thus (0.07) Results always conservative Predicted failure mode usually wrong
5 8 1.21 (0.04) Std. dev. thus (0.07) Results always conservative Predicted failure mode usually wrong
6 73 1.18 (0.05) Std. dev. thus (0.07) Results always conservative Predicted failure mode usually wrong
17Coped beam tests
- 38 tests from 5 different sources
- Limited range of geometries represented (but,
improving) - Steel grades from 36 ksi to 115 ksi yield
- Amount of rotation a non-controlled variable
(except in one test series)
18Coped beam results
Source No. Tests Test / LRFD(1999) Comments
1 7 0.70 (0.09) Std.dev. thus (0.19) Results often non-conservative
2 5 0.99 (0.19) Std.dev. thus (0.19) Results often non-conservative
3 1 0.95 (0.00) Std.dev. thus (0.19) Results often non-conservative
4 8 1.13 (0.08) Std.dev. thus (0.19) Results often non-conservative
5 17 1.06 (0.12) Std.dev. thus (0.19) Results often non-conservative
19What about angles and tees?
- Angles, tees
- 41 tests, 4 different sources
- The parameters involved now include shear lag
- Behave more like gussets than coped beams
20Angles and tees
- Researchers include Epstein, Gross, Orbison,
Barthels (and their colleagues) - Many more variables than for gusset plates and
coped beams - Variables include shear lag, use of staggered
holes, and so on.
21A better model
- Gusset Plates tensile ultimate on net shear
yield on gross - Coped Beams ½ tensile ultimate on net shear
yield on gross
22Results using new model
G. Y. Grondin For gusset plates, the mean is
1.19 and standard deviation 0.06 if we use what
has been adopted by S16-2001, i.e. we take the
least of yield on the gross shear area, Agv, or
ultimate on the net shear area, Anv. If you go
with what was originally proposed (yield on gross
shear), the numbers that you have are correct.
- Test / Predicted
- Gusset Plates 1.11 (0.10)
- Coped Beams 1.14 (0.13)
23Apply the suggested gusset model to tees and
angles
- Overall result (all four sources) is good
test-to-predicted 0.93 (0.05) - Results for Epstein stand out as inconsistent
with the others - Existing rules also give good agreement
24Concluding comments
- LRFD gives conservative results for gusset
plates, non-conservative results for coped beams - Degree of conservatism is inconsistent
- Predicted failure mode is often incorrect
25Satisfactory results using
- Gusset plates, angles
- Coped beam webs
26..with an upper limit
e.g., for gusset plates
27(No Transcript)
28(No Transcript)
29(No Transcript)
30(No Transcript)
31(No Transcript)