Hello Dan:

Your report, "Cyclic Comparison Testing of Light Wood Framed Shear Walls,"
was brought to our attention. From a cursory review, the following comments
are offered for your consideration.

1. The 4240 lb test load (unit shear load of 760 lb/ft) for the plywood/OSB
walls represents 2.9x the allowable design load of 1451 lb (= 260x5.58) for
the tested construction. The test load was in the upper range of where we
would expect failure; e.g., strength limit state (SLS). Based on other
tests now in progress (see below), this test load was about 11% more than
the maximum shear capacity of similarly-constructed shear walls. This high
loading would affect the hysteretic energy absorbed by these walls, when
compared to the reduced loading for gwb walls (see 2. below).

2. Likewise, the 4240 lb test load (unit shear load 324 lb/ft, or 162 lb/ft
per side) for the gypsum wallboard (gwb) walls was about 2.2x the allowable
design load of 1962 lb (=150x0.5x13.08x2) for the tested construction.
Based on other tests now in progress (see below), the test load for GB1-a
was about 42% of the maximum shear capacity of unblocked gwb walls with
similar construction and reduced nailing (7" oc). The unblocked gwb shear
walls would be expected to have lower stiffness and strength than the tested
blocked gwb wall construction.

The test load subsequently was doubled (Test GB1-b) to 324 lb/ft per side
(0.8g), at which point the wall failed quickly. This test load was about
84% of the maximum shear capacity of unblocked gwb walls. The results
probably are influenced by the prior loading (GB1-a).

Test GB2 was conducted at 243 lb/ft per side (0.6g) and did not fail in
eleven loading cycles. Based on other tests now in progress (see below),
the test load for GB2 was about 63% of the maximum shear capacity of
unblocked gwb walls.

In comparison to unblocked gwb walls, blocked gwb walls should have greater
stiffness and maximum shear capacity. Thus the test loads for the blocked
gwb walls are even more conservative in comparison to expected maximum shear
capacity. These factors should be considered when comparing the performance
of gwb walls vs. walls with wood structural panels.

3. OSB has demonstrated similar stiffness and ductility to plywood wall
sheathing in cyclic load shear wall testing by APA and City of Los Angeles
(see below). Preliminary testing indicates that maximum shear strength of
OSB walls might be improved if fastener edge distance is increased along top
and bottom plates from 3/8" to 3/4". Gypsum wallboard is shown in tests to
have low displacement (compared to wood structural panels) at yield limit
state and strength limit state. Your report concludes with a very negative
connotation about the performance of wood structural panels and OSB in
particular, compared to gwb. These conclusions are not supported in cyclic
load testing by others. Load capacities of walls at specified wall
displacements needs to be considered when combinations of wall sheathing
materials are used in building construction. E.g., walls sheathed with gwb
appear to reach yield and possibly strength limit state conditions
substantially before these conditions occur with wood structural panel shear
walls.

4. Editorial corrections:
a. Plywood and OSB test wall design load: Z should be 0.40 (instead of
0.04). Calculated demand shear V = 1458 lb.
b. Gypsum wallboard test wall design load: Z = 0.40, V = 1943 lb.
c. Description of test panels (PW1): 3/8" plywood has three plies (instead
of four).
d. Description of test panels (OSB1 and OSB2): The standard is PS 2-92
(instead of PS 1-95, which covers only plywood).

Additional Information

APA sponsored shear wall tests at Univ. of CA - Irvine (UCI), which followed
the SEAOSC cyclic loading protocol. We gradually approached SLS by
displacement-based cyclic loading (0.2" increments to 1.6" displacement,
then 0.4" increments to 3.2" displacement). We found SLS for wood
structural panel shear walls to occur at about 1.5" displacement (1.5% story
drift), with load factors (max. unit shear/allow. design unit shear)
averaging about 2.4. (Ref: APA Report 158). In your tests with plywood and
OSB, the test loading caused initial displacements of 1% to 1.8% drift,
which indicates that the assumed test loading was too high for the tested
construction. No load-displacement curves for gwb walls were transmitted
with your report draft.

The City of Los Angeles is expanding the data base on the preliminary APA
tests with a series of over 100 shear wall tests at UCI, to evaluate the
1997 UBC shear wall design values for cyclic (reversed) loading under
seismic conditions. These tests are nearing completion (about 80 tests have
been completed to date).

Please contact me if you have questions on these observations.

John D. Rose
Senior Engineer
Technical Services Division

APA - The Engineered Wood Assn.
PO Box 11700
Tacoma, WA 98411-0700
Tel: 253-565-6600 Ext. 482
Fax: 253-565-7265
E-Mail: john.rose@apawood.org

Hi John,

Thank you for your interest in the shear wall report. It is good to finally
get someone to read it and make intelligent comments. I have already made
the editorial changes that you suggested with the exception that PW1 really
was a 4-ply assembly. The two plywood samples were purchased by the
students working on the tests at local lumber yards. I have left the
calculation results at 3 significant figures.

This series of tests was first conceived back when the 1988 code reduced the
capacity of gyp shear walls by 50%. Although I agreed that a reduction was
called for, 50% seemed like it might be a somewhat arbitrary factor
(although I think most of what we know about earthquakes is 2 significant
figures at best). My real interest was the gyp design levels and we were
going to test plywood only for baseline values to compare with the gyp. OSB
was added to the test series when funding was offered for the two OSB tests.
I would have preferred to test APA sheathing but the students did not buy
any.

The problem, as I saw it, was to find a test method to determine if the code
capacity for gyp was appropriate . Almost any wall sheathing material has
some capacity to resist lateral forces, the real question is not what
material, if any, is "best" but what is the appropriate design capacity for
the material so that the resulting building is safe. With the correct
design capacities and a good design method, a safe building can be built
with plywood, gyp or OSB sheathing.

I had planned to use plywood as a baseline because it is commonly believed
to perform well in earthquakes (meaning that the established capacity and
design method are good). Although I expected variation in results for wood
product sheathing materials, I was surprised by the results I got. Since
the two OSB samples were from the same batch, it may simply have been an
unfortunate pick. I know that APA has consistently got good results with
OSB. I would like to test other assemblies, especially unblocked gyp.

The report is posted and may be viewed at:

http://www.engr.sjsu.edu/dmerrick/shearwalls/

All the diagrams are available there. I would like to post your comments if
you will permit me.

Thank you again,
Dan

Hi Dan:

Thanks for your response. The cyclic load testing by UC-Irvine with gwb
walls provides a basis for re-evaluating the UBC values for shear walls with
this material. They tested two conditions - 1/2" unblocked and 5/8" blocked
(3 replications each). I was surprised at the yield limit state (YLS) and
strength limit state (SLS) values for the gwb walls under cyclic loading,
but the displacement at these conditions was low - about 1/8" or less at
YLS, and about 3/8" to 5/8" at SLS, indicating the brittle nature of the
resulting system. (Stucco attached directly to wood framing had even lower
values).

1/2" gwb unblocked, 5d nails @ 7/7" oc - YLS 240 plf, SLS 388 plf
5/8" gwb blocked, 6d nails @ 4/4" oc - YLS 380 plf, SLS 694 plf

Current thinking is that LRFD design values should be based on YLS, with
some reduction or resistance factor (phi) to account for variability, test
vs. field construction etc. For ASD, a further reduction factor of 0.71
(1/1.4) is proposed based on 1997 UBC seismic design provisions.

Prof. Gerard Pardoen/UC-Irvine Dept. of Civil & Environmental Engineering is
project leader for the City of Los Angeles/UC-Irvine shear wall test
project. His e-mail address is gpardoen@uci.edu. They have summarized the
results of the first 27 groups of tests (3 replications each, totalling 81
tests). A report will be forthcoming from two UC-Irvine graduate (MS)
students. There are plans to conduct 24 or so additional tests in the
project, but these have not been defined.

Incidentally, you might want to recheck the records of the plywood for PW1.
If it was truly 4-ply panel construction, it is likely that it was 15/32"
plywood (Span Rating 32/16). There are a few (very few) plywood mills that
make 4-ply 7/16" plywood (Span Rating 24/16), but I've never heard of 4-ply
3/8" thick plywood (Span Rating 24/0).

You are welcome to circulate my comments if you think that would be helpful.

John Rose/APA

-----Original Message-----
From: Dan Merrick [mailto:danmerr@ix.netcom.com]
Sent: Wednesday, December 01, 1999 8:53 AM
To: John Rose
Subject: Re: Cyclic Comparison Testing of Shear Walls