Missouri University of Science and Technology

Scholars' Mine

International Specialty Conference on Cold- (1988) - 9th International Specialty Conference

Formed Steel Structures on Cold-Formed Steel Structures

Nov 8th, 12:00 AM

Behavior of Continuous Span Purlin Systems

Roger A. LaBoube
Missouri University of Science and Technology, [email protected]

M. Golovin

D. J. Montague

D. C. Perry

L. L. Wilson

Follow this and additional works at: https://scholarsmine.mst.edu/isccss

Part of the Structural Engineering Commons

Recommended Citation
LaBoube, Roger A.; Golovin, M.; Montague, D. J.; Perry, D. C.; and Wilson, L. L., "Behavior of Continuous
Span Purlin Systems" (1988). International Specialty Conference on Cold-Formed Steel Structures. 1.
https://scholarsmine.mst.edu/isccss/9iccfss-session1/9iccfss-session2/1

This Article - Conference proceedings is brought to you for free and open access by Scholars' Mine. It has been
accepted for inclusion in International Specialty Conference on Cold-Formed Steel Structures by an authorized
administrator of Scholars' Mine. This work is protected by U. S. Copyright Law. Unauthorized use including
reproduction for redistribution requires the permission of the copyright holder. For more information, please
contact [email protected].
 Ninth International Specialty Conference on Cold-Formed Steel Structures
St. Louis, Missouri, U.S.A., November 8-9,1988

BEHAVIOR OF CONTINUOUS SPAN PURLIN SYSTEMS

R. A. LaBoube l , M. Golovin2, D. J. Montague 3 ,
D. C. Perry4, L. L. WilsonS,

ABSTRACT
Cold-formed steel structural shapes are the mainstay of metal building roof
systems in the United States. Because of the measurable economic gains that
can be derived from an optimum design of such a roof system, it is imperative
that the synergism of the system, i.e., the combined structural resistance
provided by the structural members, roof panels and their attachments, be
considered in the structural design methodology.
This paper will discuss a test program conducted to provide insight into the
behavior of a typical metal building roof system subjected to wind uplift
loading. Both C and Z shaped structural members, continuous over three spans,
were investigated. The tests were representative of a conventional through
fastened roof assemblage. Results of the test program, as well as an
easily applied design approach, is presented.

1Butler Manufacturing Company, Grandview, MO

2CECO Building Division, Columbus, MS
3Mesco Metal Buildings Corp., Grapevine, TX
4Southern Building Code Congress International, Birmingham, AL
5American Building Company, Eufaula, AL

191
 192

INTRODUCTION
Cold-formed steel structural shapes, C and Z sections, are the mainstay of the
metal building roof system in the United States. These shapes are generally
designed as beams, continuous over the main frames, having the top flanges
attached to metal roof sheets. However, when subjected to wind uplift, the
top flange is the tension flange and the bottom flange functions as the
compression flange in the positive moment region of the span. The tendency,
in conventional steel design, is to assume that because the compression flange
is not fully braced, the design of the member should assume a laterally
unbraced condition. This is unduly conservative, and measurable economic
gains can be derived from recognizing the synergism of the roof system, i.e.,
the combined structural resistance of the structural members, roof panels
and their attachments.
To gain insight into the behavior of a typical metal building industry roof
system, a series of full scale tests were conducted. The test program focused
on the uplift capacity of both C and Z section roof systems as used by members
of the Metal Building Manafucturers Association. This paper describes the
test program, discusses the test results and presents a simplified design
procedure for assessing the uplift capacity of a roof system.
TEST PROGRAM
The test program consisted of a total of 19 full scale tests. Fourteen test
specimens used Z-section beams and five used C-section beams. The Z and C
sections were selected to provide a test program that would envelope most of
the section geometries commonly used in the metal building industry. The
following summarizes the range of variation in the key section geometric
parameters:
Flange width - 2 to 3 in. (50.8 to 76.2 mm)
Web depth - 6.5 to 10 in. (165 to 254 mm)
Thickness - .056 to .101 in. (1.42 to 2.57 mm)
Edge Stiffener - 42 to 90 degrees.
Lap length - 30 in. to 72 in. (76.2 to 183 mm)
Span length - 20 ft. to 30 ft. (6.1 to 9.1 m)
Table 1 provides a summary of each test specimen and it's geometry.
All test specimens used an industry standard galvanized roof panel, formed
from 24 or 26 gao thick sheet steel. The panels were 36 in. (91.4 cm) wide
with 1.25" (31.8 mm) deep corrugations, 12 -in. (30.5 cm) on center. The
attachment of the roof panel to the beam section was accomplished by using
commonly used self-drilling screws (No. 12 SOS). The screws were located at
12 in. (30.5 cm) centers along the length of the beam secion. Because lateral
stability of a beam section is enhanced by the rotational restaint that is
provided by the panel and it's attachment, the panel and fastener type are
significant parameters.
 193

TEST SETUP
All specimens were tested as continuous be'ams subjected to a uniformly
distributed load. Each test specimen consisted of two beam sections
(pur1ins), continuous over three spans and affixed to roof panels. Figure 1
is a schematic of the test setup, and Figures 2 and 3 are photographs of a
typical test setup. A detailed discussion of the test setup is given in
Reference 1.
As depicted in Figures 2 and 3, the test specimen was constructed in a
pressure test chamber. A simulated wind uplift load was applied by
pressurizing the test chamber. The applied load was recorded by use of
manometers located at each end of the test chamber. The vertical deflection
of each pur1in was recorded in the end bays, and at the location of the
maximum deflection based on linear elastic beam theory. Because of the
unsymmetrical geometry of the pur1in cross section, the compression flange
will also displace horizontally under load; this deflection was also measured.
Both the vertical and horizontal displacements were measured by utilizing a
surveyor's level and targets (Ref. 1).
TEST RESULTS
All test specimens were loaded until failure of one of the purl ins was
obtained. Table 2 provides a summary of the failure load for each specimen.
The failure load is given by the pressure at failure times the tributary width
of roof panel that is supported by each purl in. Failure typically manifested
itself as a local buckling of the web and flange at the location of maximum
applied moment in one of the end spans. Figure 4 shows typical failure
conditions.
Also listed in Table 2 is the tested yield strength for each purlin specimen
evaluated in accordance with the procedures of ASTM A370.
EVALUATION OF TEST RESULTS
The continuity of the Z-sections is achieved by nesting of the pur1in sections
at the frame line, or intermediate support. The literature contains little
information regarding the behavior of continuous span pur1ins, and the ability
of the nested purlins to develop continuity. Reference 3 indicates that to
achieve full continuity, the lap length should be at least 1.5 times the depth
of the section. For simi1iar purlin sections in this test program, the
assumption of 1.5 times the depth is a reasonable length to develop full
continuity. This is based on the comparison of the failure load for tests No.
1 and 3 and tests No. 2 and 4. As given in Table 3, the failure load for
these test specimens is virtually independent of the lap length, which
suggests that linearly elastic beam theory can be used to evaluate the applied
internal forces with sufficient accuracy.
 194

Previous research (Ref. 2) presented an analytical technique for calculating

the actual stresses that would be experienced in a purlin section, having an
un braced compression flange, subjected to wind uplift. However, to estimate
the ultimate failure mode required an iterative calculation procedure that was
not well suited for routine design. Therefore, a simple, straightforward
empirical procedure was developed based on the test results for both
continuous span purlins, as described previously, and simple span purlins as
given in Reference 2 and unpublished tests conducted at Butler Research.
The following simplified design approach for evaluating the capacity of a C or
Z beam section, attached to roof panels, and subjected to a wind uplift load,
is based upon applying a reduction factor to the calculated fully braced
moment capacity. The nominal moment capacity is determined based on the
provisions given in the 1986 edition of the AISI Specification (4), assuming
full lateral support along the length of the member. Recognizing the
relationships between applied load and moment based on elastic beam theory,
the nominal load was calculated and is listed in Table 4.
Also given in Table 4 is the tested failure load. Based on the given test
results, the load capacity, of a C or Z beam section braced on its tension
flange and having its compression flange unbraced. is a constant relationship
to the fully braced member. This relationship is evident by the ratio of the
tested failure load to the nominal load (Table 4).
For purpose of design, it is recommended that the following expressions be
used to calculate the nominal capacity of C and Z beams:
Mn = R Se Fy (1)
Where
R = 0.70 for continuous span Z sections
= 0.60 for continuous span C sections
Se = the elastic section modulus of the effective section at Fy
Fy = yield stress.
Because the reduction factor, R, was experimentally determined, the use of Eq.
1 should be limited to through fastened roof assemblies and for members having
the following geometric limits:
- Purl in depth less than 10 in. (25.5 cm)
- The free flange is a stiffened compression element
- 60 < web depth / thickness < 170
- 2 < web depth / flange width < 5
- 16 < flange flat width / thickness < 43
 195

CONCLUSIONS
The rigourous, analytical evaluation of the strength of a Z or C beam roof
system subjected to wind uplift requires an iterative calculation procedure,
that does not lend itself to routine design. Therefore, an simple, easily
applied empirical procedure has been developed.
REFERENCES
1. Perry, D., et.al., "Continuous Span Purlin Uplift Tests," Final Report,
Metal Building Manufacturers Association, October 1987.
2. Pekoz, T., and Soroushian, D., "Behavior of C and Z Purlins Under Uplift,"
Proceedings of the Sixth International Specialty Conference on
Cold-Formed Steel Structures, November 1982, University of Missouri-Rolla,
Ro 11 a, MO.
3. Robertson, G.W., and Kurt, C.E., "Behavior of Nested Z-Shaped Purlins,"
Proceedings of the Eighth International Specialty Conference on
Cold-Formed Steel Structures, November 1986, University of Missouri-Rolla,
Rolla, Mo.
4. Cold-Formed Steel Design Manual, American Iron and Steel Institute,
Washington, D.C., 1986 Ed.
 196

TABLE 1
GENERAL SECTION GEOMETRY

Test Section Section Flange Material Span Lap

No. Type Depth Width Thickness Length Length
(In.) (In.) (In.) (Ft.) (Ft.)
1 Z 9.5 3.00 0.0701 30 6.0
2 Z 9.5 3.00 0.1013 30 6.0
3 Z 9.5 3.00 0.0696 30 2.5
4 Z 9.5 3.00 0.1002 30 2.5
5 Z 9.5 3.00 0.0695 20 2.5
6 Z 6.5 2.00 0.0587 20 4.0
7 Z 6.5 2.00 0.0613 20 4.0
8 Z 6.5 2.00 0.0600 20 4.0
9 Z 6.5 2.00 0.0590 20 4.0
10 Z 8.5 2.00 0.0847 24 5.0
11 Z 8.5 2.00 0.0845 24 5.0
12 Z 10.0 2.75 0.0833 24 5.0
13 Z 8.5 2.50 0.0613 24 4.0
14 Z 9.5 3.00 0.0660 20 2.5
15 C 9.0 2.38 0.0607 24 4.5
16 C 7.0 2.50 0.0598 24 4.5
17 C 8.5 2.50 0.0897 24 5.0
18 C 8.5 2.50 0.0561 20 4.0
19 C 8.5 2.50 0.0895 30 6.0
Notes: 1. For definition of lap length see Fig. 5.
2. 1 in. = 25.4 mm.
3. 1 ft. = 0.3048 m.
 197

TABLE 2
TEST RESULTS
Test Tested Failure
No. Fy Load
(ksi) (l b/ft}
1 63.1 138
2 58.1 234
3 61.6 136
4 58.2 218
5 61.3 280
6 55.2 109
7 57.4 124
8 56.1 121
9 55.8 119
10 60.2 181
11 57.8 193
12 61.5 286
13 58.5 134
14 63.9 241
15 56.5 128
16 54.6 98
17 58.3 217
18 65.3 156
19 58.3 135

TABLE 3
LAP LENGTH COMPARISON
Test Required Test Failure
No. Lap Length Lap Length Load
(ft.) {ft.} (l b/ft.}

1 2.375 6.0 138

3 2.375 2.5 136
2 2.375 6.0 234
4 2.375 2.5 218
Notes: 1. Required lap length based upon Ref. 3, and shown on Fig. 5 as
greater than 1.5 D.
2. 1 ft = 0.3048 m
3. 1 lb = 4.448 N.
 198

TABLE 4
EVALUATION OF DATA
Test Nominal Test Test
No. Calculated Failure Calculated
Load Load
{lb/ftl {lb/ftl
1 193 138 0.72
2 283 234 0.83
3 189 136 0.72
4 274 218 0.80
5 421 280 0.67
6 170 109 0.64
7 190 124 0.65
8 177 121 0.68
9 171 119 0.70
10 294 181 0.62
11 281 193 0.69
12 400 286 0.72
13 186 134 0.72
14 406 241 0.59
Mean D:7O
Standard Dev. 0.06
15 213 128 0.60
16 145 98 0.68
17 337 217 0.64
18 288 156 0.54
19 222 135 0.61
Mean ---o.6l
Standard Dev. 0.05
Notes: 1. 1 ft = 0.3048 m
2. 1 lb = 4.448 N.
 @ TARGET BULKHEAD DROPPED
ROTARY BLOWER WITH
INTO TEST CHA~BER
~ ANTI-ROLL "Un TUBE ~ANOMETER
DEVICE FLEXIBLE HOSE

,-
n.=
:~
1
=c=~--
SECTION "A"

'@II
: /:
=r- I I I
I I DIAL GAUGE TO
~ 0,
I

o I MEASURE DE-
.tt\ I
FLECTION OF
END SUPPORT

SPAN SPAN SPAN

TEST CHAMBER 151'-0" LONG

continuous top panel ....
1~" 9" 1~" <J:)

10'-B"
2 I" 'I 2
<J:)

~il:I, ~
• u -u

:2

-
'" It
'11~
~
(FACE TO FACE OF CLAMPING ANGLES)
2'-9~" 5'-0"

6 MIL IpOLYETHY7ENE FILM ;;

...--PANELS g' -7" LONG
1,---(2)
!
1/2"
A307 BOLTS
2'-9!"

,s -----..u
. [
~

\.
-
,,,
:2

ANTI-ROLL .1
. 12" B

DEVICE
=c::JI 1~" 9" 1~"
SUPPORT BEAM
2 II II 2

jjf
E3

,
I
I - - - - SUPPORT BEAM COLUMNS - - -
'" ;--TIE-OOWN BEAM /16" PL
r- -+ (TYP.) wi
, 9/16" ,s
J HOLES
3/411112~111 3/4"
t SIDE OF
10'-7"
DETAIL OF
TEST BOX SECTION "A" "Discreet" (not continuous support element)
ANTI-ROLL DEVICE

FIG. 1 TEST SET-UP FOR PURLIN UPLIFT TESIS

 200

FIG, 2 PHOTOGRAPH OF TEST SETUP

 201

FIG, 3 - PHOTOGRAPH OF PURLINS AND SUPPORTS

 202

FIG. 4 TYPICAL PURLIN FAILURE MODE

 203

0...
0 0 «
-l
::;;:
0 «
w
I.I"l
co
I
N
1\ :r: -l
I-
(9 «
u
J1 z
W
-l 0...
u >-
I-
0...
«
-l I.I"l

(9

u..

0 0
1--- - -