Bending Strength of Panelized Decking from Black Hills Ponderosa Pine Lumber

Donald C. Markstrom and Edwin H. Oshier

Introduction

Panelizing laminated decking showsconsiderable promise for utilizing low-grade lumber from pon- derosa pine (Pinus ponderosa Laws.) and other species, and for recovering some of the lost sheath- ing market. This product can be designed for speci- fic stiffness and strength requirements, can be manu- factured into panel form for reduced installation costs, and could utilize large amounts of low-grade lumber. Lower grades of lumber, with lower moduli of elas- ticity and rupture, can be used as core material, with higher grades of lumber resawn for the outer plies (fig. 1 ).

The purpose of this study was to provide design information for laminated decking of Black Hills ponderosa pine lumber. The primary objective was to determine and compare both stiffness and ultimate load-carrying capacity of decking fabricated with different lumber grades.

It has been established that moisture content, density, knots, cross grain, shakes, checks, borer holes, wave, and decay affect the bending strength of wooden members.-^ The most frequent and important strength-reducing defects in the lower common grades are knots. The modulus of rupture and modulus of elasticity are reduced by the lower compressive and tensile strength of the knot, associ- ated cross grain, and by stress concentrations bor- dering the knot.

Methods

Fabricating the Panels

Rough air-dried 1- by 6-inch ponderosa pine boards, with defect types, sizes, and distributions representative of the lumber grades to be used in the test panels, were selected by a lumber grader at Black Hills sawmills. Forty test panels, 10 for each lumber face and core combination, were fabri- cated in the following combinations:

Faces

3 Common 2 Common 1 & 2 Clear 1 & 2 Clear

Cores

4 Common 4 Common 4 Common 1 & 2 Clear

TlguA2. 1.--Tz6t paml Mlth Kej,m(id 4/4 lamboji iacu glae.d p^AptncLLculoAlij to thz mmlnaZ 1-indt coh.t booAcLs.

^Markwardt, L. J., and Wood, L. W. Simvli- fied principles for structural grading of timber. USDA Forest Serv. , Forest Products Lab. Rep. 2112, 19 p., illus. 1958.

-1-

No decay or open defects greater than 1/16 inch in width were allowed in any member except those permitted on grade 4 Common. The moisture con- tent of the rough 4/4 lumber, measured with an electric moisture meter on 10 randomly selected boards within each grade, averaged 11.5 percent and ranged from 10 to 13 percent. Stress distri- bution was sampled with three 1-inch sections sawed from three randomly selected boards within each lumber grade. The prong-type stress specimens indicated very little or no detectable stress within the sample boards.

The rough lumber for face boards was resawn with a vertical line bar resaw, jointed on one edge and ripped on the other edge to SYa-inch widths, and crosscut to 50-inch lengths (fig. 2). The rough lumber for core boards was planed on both faces to 25/32-inch thicknesses, jointed on one edge and ripped on the other edge to 5'/2-inch widths, and crosscut to 27y2-inch lengths. Nine core boards and 10 face boards were randomly selected and arranged for each panel within each lumber face and core combination; upper and lower face boards were oriented at right angles to core boards. The five face boards were stapled together at the ends to facilitate handling and assure tight joints when pressed. The staples were removed when the lami- nated panel was trimmed to a 48-inch length.

F-cguAe 2. --Pe^awinq Kough 4/4 ZixmbeA into (jace booAcU. Thz 6an) wcus ieX to pfioduc^ {\act boa/icU with 7/16-inch thicknUi. Tka> w<u mceji^cutii to p^ovid^ {'^on. vta.ni.nq a. mooth gluincf 6aA()ace., and having a {sinal uni/)0m tkicknui, o{) 3/S inch {)0n. all f,ac^ booAdk.

Phenol-resorcinol adhesive was spread on both surfaces of the core boards at the rate of 65 to 70 pounds per thousand square feet. Glue was not spread on either the face boards or the edges of the core boards. The panels were pressed at 150 pounds per square inch (p.s.i.) for 6 hours at 80° to 90°F. (fig. 3). The cured panels were planed on both faces to a uniform thickness, and trimmed to a 24-inch width and 48-inch length. The panels were stickered and kept at a moisture content of 1 1 to 1 2 percent until tested for strength.

=^ n -\ -\

TiguJiz S.--Tivz panels wqaz a6imbtQ.d and pAeMcd at om tarn. Uotn squeeze oiut o{) giuz.

Testing the Panels

The panels were statically bent over a 44-inch span with equal loads applied at quarter points (fig. 4). The movable head of the Tinius Olsen 400,000 pound Super L Testing Machine^ was driven at the rate of 0.34 inch per minute. A Tinius Olsen Model D-2 Deflectometer and Model 51 Recorder measured and recorded loads and cor- responding midspan deflections between the sup- ports. A tripoint deflectometer fabricated with ply- wood, three cap screws, and an Ames gage meas- ured midspan deflection of the constant moment portion of the panel between the load points (fig. 5). A Bolex 16 mm. camera synchronized with a

Trade and company names are used for the benefit of the reader, and do not imply endorsement or preferential treatment by the U. S. Department of Agriculture .

-2-

Sanborn 150 Recorder photographed the Ames gage every seco nd.

The length, width, and thickness of the panels were measured within the ±0.3 percent accuracy stipulated by ASTM for tests of veneer, plywood, and other glued construction.-^ Deflections were measured and recorded to the nearest 0.001 inch.

A 1 -inch-wide section sawed across the width of the panel after testing was weighed to the near- est 0.01 gram, ovendried at 103°C., and reweighed to determine the moisture content.

FlquA^ 4. --Tut pamJti wz>it loaded at qaahtzn. points, and mZdipan dz{)t^cJu.on beXweew f^appofvU) wa^ mecLiuAed ^^)AJih a dtf^ltcXormtoA. Two 4- bu 25-lnch itainlej>6 6te.^l vlatej> weAe placiid btti^idm thz panel, and each load- ing edqe of) the static bmdilnq tool. Gfiapkite 6pfLead beJioem the platen to alZow ecUiU movement.

Results

Strength of Panels

Perhapsthe most significant results werethatthe panels with grades 2 Common faces and a 4 Com- mon core were neither significantly stronger nor stiffer than those with grades 3 Common faces and a 4 Common core (table 1). Also panels with grades 1 and 2 Clear faces and with a 1 and 2 Clear core, while stronger than those with a 4

American Society for Testing and Mate- rials. Standard methods of testing veneer, ply- wood,and other glued veneer constructions , ASTM Designation: D 805-63, p. 218-221. In_ 1966 Book of ASTM Standards, Part 16, Structural Sandwich Constructions ; Wood; Adhesives. Philadelphia, Pa. : Amer. Soc . Test. Mater. 1966.

FtguAe 5 .--TnlpoLnt dellzdtomeXeA locaXtd beJx^een load potnti to me,aj>uAe de{^lecJu.oni, (^on caZculxvting modalwi eZa^tZcAXy ,

zx.c£udtng isheoA..

Common core, were not significantly stiffer. As might be expected, the panels with grades 1 and 2 Clear faces and with either a 1 and 2 Clear or a 4 Common core were stronger and stiffer than those with either 2 or 3 Common faces and a 4 Common core. The between-panel variation of both strength and stiffness was greatest for the 3 Common face and 4 Common core combination.

Table 1. --Strength and stiffness of panels with different lumber grades in the faces and coresi'

Lumber grade

Modul us of

rupture^/

Modulus of elasticityi/

Faces

Cores

Incl uding shear

Excluding shear

p.s. i . X 10^ p. s . i .

1&2 Clear 1&2 Clear 9150 ± 500 1.45 ± 0.06a 1.61 ± 0.07a

1&2 Clear 4 Common 8100 ± 380 1.40 ± 0.04a 1.55 ± 0.06a

2 Common 4 Common 5790 ± 620a 1.25 ± 0.05b 1.38 ± 0.08b

3 Common 4 Common 5310 ± 810a 1.17 + 0.07b 1.33 ± 0.10b

i'&rade combinations with same letter are same-- (5 per- cent) Duncan's multiple range test.

95 percent confidence limit.

The average and range of certain test variables are shown in table 2. The height and width values ore each averaged from six measurements within the constant moment portion of each beam. The average specific gravity based on total panel weight and volume for ail panels was 0.46, greater than

-3-

Table 2. --Average and range of measured or calculated variables for four tests of panel construction

	Measured or calculated variables
Construction of			Moi sture		Stiffness
test panels	Thickness	Width	content (ovendry weight)	Speci fic gravity	Including Excluding shear strain shear strain (EI) (EI) 1 2	Ultimate load
	Inches	Inches	Percent	(Lb.-inch^/inches of width)xlO	' Pounds

Faces--1&2 Clear Core — 1&2 Clear

Average

Range

Faces 1&2 Clear Core --4 Common

Average

Range

Faces --2 Common Core — 4 Common

Average

Range

Faces--3 Common Core --4 Common

Average

Range

1.441

1.442

1.439

1.440

24.02

10.4

0.47

307

340

11,700

1.433 to 1.445 24.00 to 24.03 9.5 to 11.2 0.45 to 0.48 287 to 336 312 to 371 10,000 to 13,100

24.02

10.9

0.46

297

329

10,400

1.400 to 1.443 24.00 to 24.03 10.2 to 11.4 0.44 to 0.47 278 to 317 299 to 357 9,350 to 11,800

24.01

11.3

0.44

262

289

7,370

1.436 to 1.441 23.98 to 24.02 10.9 to 11.6 0.44 to 0.45 237 to 291 241 to 305 6,300 to 10,000

24.01

10.9

0.46

247

280

6,800

1.437 to 1.441 23.98 to 24.03 10.4 to 11.4 0.44 to 0.48 218 to 287 241 to 323 5,130 to 9,800

the 0.41 value for ponderosa pine at 12 percent moisture content in the Wood Handbook.-^

Strength Formulas

The formulas to calculate stiffness and the moduli of elasticity and rupture are:

STIFFNESS:

Including shear strains—

{EI)^=^(3L;-4A^) 1

where

(El), = stiffness including shear strains, pounds-inch^ P| = load within proportional limitof beam, pounds A = distance from support to load point, inches L, = span length between supports, inches A, = deflection produced at midspan relative to supports by P, , inches

^U. S. Forest Products Laboratory. Wood Handbook. U. S. Dep. Agr . , Agr . Handb. 72, 528 p. 1955.

Excluding shear strains—

P AL 2

where ^

(Eljj = stiffness excluding shear strains, pounds-inch' = load within proportional limitof beam, pounds A = distance from support to load point, inches L 2 = span length within constant moment section

over which deflection was measured, inches ^2 = deflection produced at midspan of Lj by

load, Pj, inches

MODULUS OF ELASTICITY: Including shear strains—

where

Ef^ = modulus of elasticity of faces (parallel to grain), p.s.i.

I J = moment of inertia of face plies, inches'* Ec = modulus of elasticity of core (perpendicular to grain), assumed to be 0.05 Ej , p.s.i. I c = moment of inertia of core, inches'* (El) = value from test data

-4-

Excluding shear stirains—

where

E, = modulus of elasticity of faces (parallel to

2 . > .

gram), p.s.i.

Ij = moment of inertia of face plies, inches^ E^ = modulus of elasticity of core (perpendicular to grain) assumed to be 0.05Ej , p.s.i. I J. = moment of inertia of core, inches^ (El)2 = value from test data

MODULUS OF RUPTURE;

" - 0.4541"

where

MOR = modulus of rupture

Mj^ = maximum bending moment, inch-pounds

C = distance from neutral axis to extreme fiber, inches

T = thickness of panel, inches W = width of panel, inches

Most of the panels failed suddenly with no visible compression failure.

The load-deflection curves were linear to failure, which indicated sudden tension failure and little or no compression failure. Compression failure was apparent on only three of the panels.

Conclusions

An important conclusion from this study is that stiffness of panels fabricated with 3 Common faces and 4 Common cores is as high or higher than con- ventional panelized decking designed for 4-foot spans. The measured stiffness (El) of the 10 test panels averaged 247,000 and ranged from 218,000 to 287,000 pounds-inch^per inch of width. The maximum bending moment of the 10 test panels averaged 1560 and ranged from 1170 to 2240 inch-pounds per inch of width.

A second conclusion is that only panels fabri- cated with grade 1 and 2 Clear faces can be thinner or span a greater joist spacing than those with 3 Common faces and 4 Common cores. Theoretically, the lumber grade of the faces, not that of the core, affects both stiffness and the maximum bending moment the most. The ratio of the moment of inertia of the face section to that of the core of a 1 .441 -inch-thick panel with a 0.781-inch core is cal- culated at 5.3: 1 .

This strength data will help a design engineer to determine working stresses when using this type of panel.

Agricaltare — CSU, Ft Collins

-5-

in T3 03 C

I . +-> TD

I > IB

■ S- OJ !- I OJ i- O

OO O ■ —

U_ O

1 +-> C_)

■ m c

. QJ -r- - i- 03 l/J T O +-> C

: u- c •.-

+-> u s- ■ o o

E o

J I 0) 8

w OJ

o ^

0 o

CU . — rtj —

O O -r- OO

Q. ' -a ^ • •!->

C +-> <TJ Q- C

c O Ln E

" Ol S- -r-

Q_ VI O)

O u O D

1— CD o-s: I

O TJ f— .

r— Q. c un

•r- fO (O O

D. or CO

0 E -Q O E U

•!-> O S- 1-1

0 ^ D

O U> a

0 ?

-D — .E

O) 0 D

C "D Q- ^ O

^ ^ O

"D ^= —

Q- i_i

E E o

0 U

CD ^

(/) "O

<u c:

: o ) • -a

) > 00

- s- ai s- 1 <u s- o

O 1— : Li_ O

. O) •.- -

03 CO

n o +-I c

: Li. c •.-

3 1 —

: =C o .—

> Q s: o

> OO CJ

I =)

^ +->

I u s-

) • o o

1 s- q; u- ■ OJ

o E E o u

J3 D

o ^

t3 O

O -

O u

o o

o ^

0 0) o _

c +J re o- c

fO CD CO CD

C O LD E

CU ^ -r-

. S- CU " S-

<_)+-> "O CM CU

CO C Q-

c ■■- o -o

> 00

r— • C71CSJ

I — Q. c en

•p- 03 CO O

m a. q: CO

E U

O) 0 o ^

c "D Q- V) CU

:g 2 2 g

I O ^; O -

^ 1§ § ^

4-> O i. i-H

o o 0 u

l/l XJ CU c

. 4J XJ

> CO ID S- 0 S- 0 S- O C/1 O I —

U- o

, +J c_)

CO c:

0) -r- - S- CO CO

1 o +-> c

Ll_ c -1- 3 r— CC O r—

I Q s: O

I C/1 C_3

• o o

s_ ci: u-

0 1 — OJ

C -I- +->

•r- i/1

c -t-> CO a. c

CO CD CO 0

c o un E

03 c ta;

C •'— CO 0 t— 4

o "o 1— • cncM

O C Q- C LD

0 •!- OJ OJ O

" CO =n Q. q; CO

D ■-

t o

J I

>- o

Q. 0 C C c CO

o o

0 CL U

_Q r: ^ ^

o ^

0 0 o j;: ~

O D

Q_ 01

o) £

c o

0 E -Q O E U

+J o

CO 1 —

S. <-< as

0 ^

"D —

O) 0

^ O

0 >

M >

	«
	to
	O



	c
	o
	Cb
	to
	s
	s
	■ti
	ft.
	r
	CO
	+->
	o
	3
	"a
	o
	S-
	a.
	+-)
	CO
	0
	s_
	o
	H-

	c
	0

	E


	CO

	0
0
	OJ
o
u

M- Q. cj

E E o 0

^ o

u • +-> -a

OJ > l/l 03

I— S_ 0 s_

CD 0 S- O

en O I—

E LL. O

o +-> tj

S- CO C

0 -I—

S- 03 CO

CD O +-> C

C Ll. C -r-

E o J I

CO c

O -

0 . J3

S- C 3 c/1 O

0 1— 03 C -r- +-> ■■- in

C +-> OJ CL C

' 0 s- •,-

f— CD 03 C C •■-

o -a ■

-Q ^ J; ^

r. CO

■1- 03 03 O

n= Q. q: CO

JO D

0 0

c >-

D D

C2_ CO

o) 2 c o

— u c o E E o E U

-a

CD 0 o ^

S- .

4- > O </>

^ CD

5- i-H

0 =

E J o o

CD ^ ^