Design of Metal Building Roof Purlins

Including System Reliability Effects

S. Torabian, B.W. Schafer

June 2020

COLD-FORMED STEEL RESEARCH CONSORTIUM

REPORT SERIES
CFSRC R-2020-05
CFSRC Information
The Cold-Formed Steel Research Consortium (CFSRC) is a multi-institute consortium of university
researchers dedicated to providing world-leading research that enables structural engineers and manufacturers
to realize the full potential of structures utilizing cold-formed steel. More information can be found at
www.cfsrc.org. All CFSRC reports are hosted permanently by the Johns Hopkins University library in the
DSpace collection: https://jscholarship.library.jhu.edu/handle/1774.2/40427.

Funding Acknowledgment and Disclaimer

This report was prepared with partial support from the Metal Building Manufacturers Association (MBMA). A
portion of the work was conducted through contract with Simpson Gumpertz & Heger (SGH). Any opinions,
findings, and conclusions or recommendations expressed in this publication are those of the authors and do
not necessarily reflect the views of MBMA or SGH.
Design of Metal Building Roof Purlins Including System Reliability Effects

S. Torabian, B.W. Schafer

1. ABSTRACT

This report provides a framework to incorporate structural system reliability effects in the design
of roof purlins in a typical metal building. Today every roof purlin is considered as a separate
component and the effect of spatial variation in the demand loads and potential redistribution and
load sharing in the roof system capacity are ignored in design. Component reliability is established
by first-order reliability methods implemented through load and resistance factor design. Based
on recent work in loading bearing cold-formed steel framing systems the load and resistance
factor design framework is extended from components to systems through an additional
resistance factor to account for system influence. An archetypical metal building is designed and
selected for this study. Monte Carlo simulations of a segment of the metal building roof are
performed with consideration of both randomness in the demands and capacity and employing
geometric and material nonlinearity in the response model of the roof. The simulations indicate
that the system effect in metal building roofs is beneficial, and increases in the design capacity
when evaluated against demands may be appropriate. Sensitivity to the target reliability (allowed
probability of failure), deflection limits, and modeling assumptions are observed and discussed.
Preliminary factors to account for roof system reliability are provided.

2. INTRODUCTION

In metal building design today every component in the building sees a separate design check. If
any one component fails we predict the entire building fails. That is, we use component reliability
to try to design the system. In the future one can readily imagine a more involved framework.
Engineers establish the desired system reliability (i.e. desired system failure probability) and the
reliability of a component matters only insofar as it influences the system. One approach to realize
this vision, that allows current workflows to largely continue, would be to embed the influence of
the system within the engineer’s individual component checks – this approach is explored herein.

In today’s design framework (e.g., ASCE7-16 for demand, AISI S100-16 for capacity) component
reliability, 𝛽, is established through first-order second-moment reliability methods implemented
through load and resistance factor design generally expressed as

𝜙𝑅! ≥ 𝑐Σ𝛾𝑄!∗ (1)

where 𝜙 is the resistance factor, 𝑅! is the nominal strength of the component, 𝑐 is a factor that
converts load demand to load effect (axial force, moment, etc.) and can be understood as the
result of conducting a linear structural analysis, 𝛾 is the load combination effect (e.g. 1.2 in 1.2D),
and 𝑄!∗ is the nominal demand (e.g., D). The component reliability consistent, 𝛽, consistent with
Figure 1 is established as
!
#$% &
" #
𝛽= '
(2a)
!
$%&"'

1
 #$()# /+# )
≅ (2b)
-.!( /."(

Figure 1: Basic component reliability concepts (a) definition of the randomness 𝑄 and 𝑅;
(b) definition of the reliability index b (AISI 2016)

where 𝑅 and 𝑄 are random variables, subscript 𝑚 denotes mean, and 𝜎 standard deviation – 𝑅0
is the mean resistance, 𝑄0 is the mean demand (load) effect, 𝑉) is the coefficient of variation for
the resistance, and 𝑉+ is the coefficient of variation for the load effect. The mean factors are
connected to the nominal values in Eq. (1) by:

𝑅0 = 𝑀0 𝐹0 𝑃0 𝑅! (3)

𝑄0 = 𝑐ΣB𝑄!∗ (4)

where 𝑀, 𝐹, and 𝑃 are the material, fabrication, and professional factors where subscript 𝑚 refers
to their mean values and 𝐵 is the bias factors between the nominal loads and the mean load. For
the case where demand = capacity, substituting Eq. (1) into Eq. (3) we can see that 𝑅0 may also
be expressed as:

𝑅0 = 𝑀0 𝐹0 𝑃0 (1/𝜙)𝑐Σ𝛾𝑄!∗ (5)

Eq. (5) shows that the resistance factor, 𝜙, may be used to tune 𝑅0 and as a result, 𝛽. Substitute
Eq. (4) and (5) into (2b) for the direct expression.

For a system a direct approach to calculating 𝛽 becomes more complex. While Eq. (2a) still holds
the resistance is now a function of a nonlinear analysis and the conversion from demand (D, L)
to demand effect (force, moment) is also no longer linear. For an existing design we seek a
modification to Eq. (1) to account for this effect:

𝜙𝑅121 𝑅! ≥ 𝑐Σ𝛾𝑄!∗ (6)

Where 𝜙 and other variables are unchanged from before, but 𝑅121 accounts for the difference
between component and (sub)system reliability. If failure of the component equates to failure of
the system then 𝑅121 = 1. If the system is “brittle” and system failure is generally weaker than the
individual component (such as the classic linked chain that fails when the weakest link fails) then
𝑅121 < 1, while if there is beneficial system effect, for example through load redistribution, then
𝑅121 > 1. The simplest approach is to select an 𝑅121 and then use simulation to assess the
statistics of the demand/capacity (𝑅/𝑄) ratio for use in Eq. (2a).
This basic approach has been studied for floors framed with repetitively placed cold-formed steel
joists (Chatterjee 2016; Chatterjee et al. 2017; Smith et al. 2016 and 2018). For this system 𝑅121
was found to be significantly greater than 1.0. A value of 1.25 was recommended from the analysis
in Smith et al. (2018) where it was assumed that a the floor target reliability would be the same
as the member component reliability (𝛽=2.5). The source of the beneficial system effect included
(a) excess component reliability through the use of the same member across the floor, (b)
beneficial load re-distribution from joist-to-joist under overload conditions, (c) beneficial decreases
in the variation of the strength due to a large number of inter-connections between joists and
sheathing which braces the joists, and (d) benefits from spatial variation in the demands. Note,
wood design also uses a repetitive member factor of 1.15, which corresponds to a 15% increase
in capacity (ANSI and AWC 2012).

For metal buildings it is possible to imagine a complete system analysis-based approach for the
entire building system and abandoning traditional component checks. This approach has been
explored for steel building frames and racks (e.g., Buonopane and Schafer 2006, Zhang et al.
2018). While this is a promising long term approach, in the near term it is considered more likely
that (a) only well-defined subsystems will leverage system reliability, and (b) component checks
will remain. Thus, the notion of the component level system reliability effect of 𝑅121 in Eq. (6) is
pursued here. For metal buildings, the roof purlin system is closest to the previously studied cold-
formed steel framing floor system, and is thus selected for the small studies herein.

It is worth noting that expectations for the benefit of system reliability are lower for metal building
systems than typical cold-formed steel framed building systems. Of the 4 system benefits
observed in cold-formed steel floor systems, only two are likely for metal building roofs: beneficial
load re-distribution, and spatial variations in the demands across the subsystem. Compared with
cold-formed steel framed building roofs, metal building roof purlins have larger spacing, less and
lighter inter-connects to sheathing, and are optimized member-by-member, bay-by-bay, across
load cases in a manner not typical in cold-formed steel framing.

3. ARCHETYPE METAL BUILDING

An archetype building was specifically designed for this study by an MBMA member (MBMA
2019). This archetype building is focused on the roof and diaphragm, frame members and lateral
load resisting systems are not explicitly studied unless they have some effects, such as the strut
forces in the roof.

The specifications of the archetype metal building are summarized below:

• The building is 100’ (clear span) x 125’ (length) x 20’ (Eave) with 0.49:12 roof slope.
• Design loading is based on 2018 IBC, ASCE7-16, Risk Category II for the Washington
DC geographical location.
• The frame spacing is 25 ft, and purlins are continuous (lapped connections) over 5 bays
of 25 ft.
• The seismic force in the Washington DC area is SDC B, and it is all “R=3.0” systems for
the moment and braced frames.
• The LRFD method is used in the design.
• The roofing is a standard Through-Fastened Roof (TFR., PanelRib), with self-drilling
screws, over purlin spacing of 5 ft. Additional sub-purlins are placed in the corner zones
for high wind uplift.
 • Non-proprietary connections or clips are used.
• The longitudinal, lateral force resisting system is braced frames in two side-walls, and
roof-truss diaphragm. All are designed as tension-only X-diagonals.
• Secondary framing, purlins, and girts are 8.5” deep Zee-sections, but with different
thickness, as needed.
• Roof live, and snow load patterns are always applied as stand-alone loads in addition to
dead loads (only). Pattern loads are not combined with wind, and seismic loads and the
live load are applied as the uniform live load. The same assumption is true for the snow
loads.
• End post spacing was increased slightly, to force wider strut space in the roof plane. As
a result, there are 5 identical purlin lines between strut/collector lines.

Figure 2: Archetype building (annotated)

Figure 2 shows the overall details of the archetype along with the designated area under study
with five purlins. Although the loading is not precisely the same for the five selected purlins, as a
common industry practice, the beam cross sections are assumed to be similar. The struts are not
included in the roof sub-model to avoid complexities due to wind or seismic axial forces in these
members.

Figure 3 summarizes all loading assumptions made by the designer and reflected in the
calculation package. For more information on the archetype model, please see the Metal building
archetype design (MBMA 2019).
 Figure 3: Loading assumption from the calculation package

4. ROOF MODEL

The roof model in this study is a sub-model consisting of five purlins, along with cross-elements
representing the decks and intermediate supports at the location of the frames. The numerical
modeling has been performed in MASTAN 2, and the pre-processing and post-processing has
been done via a MASTAN batch file in MATLAB to enable Monte Carlo (MC) simulations.

Figure 4 shows the geometry and boundary conditions of the purlin sub-assembly. In addition to
the purlins, the steel decks passing over the purlins are also modeled to enable load redistribution
between the purlins. As shown in Figure 5, the roof distributed loads are applied to the decks, and
the decks are modeled as beam elements connected to the purlins to transfer the loads. The deck
elements are assumed to behave elastically, and only the purlins have plastic hinges to enable
plastic behavior and load redistribution in the analyses.

The “2nd-order Nonlinear” analysis module in MASTAN provides a load factor at which the
displacements become very large, and the structure is thus unstable. The load factor is the
reserve capacity of the system for the loads applied in the model. For an applied load, MASTAN’s
load factor can be understood as a resistance-to-demand factor.
 Figure 4: Geometry and boundary conditions of the purlin sub-assembly model

Figure 5: Beam loading via deck members

4.1 Purlin nominal capacity in the roof model

Figure 6 shows the purlin nominal flexural capacity along the length in positive and negative
bending moments, extracted from the original beam design. The variable strength of the joist
along the length is due to different unbraced lengths and different thicknesses of the purlin along
the length as well as lapping of the individual purlins at supports.
 Figure 6: Purlin nominal flexural capacity along the length

4.2 load combinations

The controlling load combinations in the purlin design are summarized in Table 1.

Table 1: Design load combinations

Positive (gravity) pressure design load combination Load distribution
• Snow dominated: 1.2D+1.6PF2 (pattern2) Partial load Full on 2 spans (LC9)
• Snow dominated: 1.2D+1.6PF2 (pattern3) Partial load Full on 2 spans (LC10)
• Gravity dominated: 1.2D+1.6L+0.5W Wind load, Pressure (LC19)
Negative (uplift) pressure design load combination Load distribution
• Wind dominated: 0.9D-1.0W Wind load, Suction (LC22)

For the simulations we desire the mean demand, i.e. Eq. (4), not the nominal. The bias factors
(𝐵) are summarized in Meimand and Schafer (2014) and provided in Table 2(a). Accordingly, the
load combinations in the analyses, including bias factors, are provided in Table 2(b).
V.M. Meimand, B.W. Schafer / Structural Safety 48 (2014) 25–32

Table 2: Mean load combination tables

Table 1
coefficient of variation of the resistance as
(a) bias Coefficient
and variability from
of variations and bias Meimand
factors. and Schafer (2014)
+ V P2 (4) Dead Live Wind Snow Earthquake

Chapter F with sample size effect included: 7-05 7-10

VQi 0.1 0.25 0.37 0.37 0.26 1.38

+ C P V P2 (5) Bi 1.05 1.0 0.92 0.575 0.82 α

(b) resulting mean load combinations

Positive
mand is connected to the nominal loads (gravity) pressure
as follows:
2. unfactored load combination Load distribution
Load combinations
"
c Bi Q i • Snow dominated: (6) 1.05D+0.82PF2 (pattern2) Partial load Full on 2 spans (LCU9)
ms across all loads (e.g., D, L, W), • Snowloads
dominated: 1.05D+0.82PF2 (pattern3)
Based on ASCE7-05 [4] the following load combinations should be
c converts (e.g. Partial load Full on 2 spans (LCU10)
considered when designing structural members:
) to load effects (e.g., compression• force Gravity
in a dominated:
stud), 1.05D+1.05L+0.575W Wind load, Pressure, (LCU19)
ASCE7-05
Negative (uplift) pressure unfactored
factor between specified loads ( Q i ) and mean loads
(1) 1.4D
load combination Load distribution
• Wind
t note that the coefficient of variation of VQdominated:
is load 1.05D-0.575W
(2) 1.2D + 1.6L Wind load, Suction, (LCU22)
endent, which may be expressed as follows: (3) 1.2D + 1.6L + 0.5S
! $ (4) 1.2D + 1.0L + 1.6S
%2 # %2
Qi Bi Q i VQi (5) 1.2D + 1.0L + 1.6W + 0.5S
= # (7) (6) 1.2D + L + 0.2S + E
Bi Q i
(7) 0.9D − 1.6W
(8) 0.9D − 1.0E
t maximum load) the design capacity is equated to
where Dead (D), Live (L), Snow (S), Wind (W), and Earthquake (E)
and (to reach the desired target reliability):
loads are defined in ASCE7. Note, the effects of these loads on the
(8) demands of a component, are the focus of this work. When the live
load is less than 100 psf (common in cold-formed steel structures),
4.3 Resistance Variables

As shown in Figure 6, for positive and negative pressure design load combinations, the strength
pattern along the length is different, due to different moment distributions and different unbraced
lengths of the purlins in positive and negative pressures. The nominal flexural capacity along the
length are extracted from the original beam design. While the distribution of the capacity is
assumed to be deterministic, the purlin capacity in each bay and the connection region is assumed
to be independent random variables with the capacity equal to the nominal capacity multiplied by
material (𝑀0 ), fabrication (𝐹0 ), and professional factors (𝑃0 ), equal to 1.1, 1.0, and 1.0,
respectively, per AISI S100 chapter K, for CFS joists. Accordingly, the mean capacity is 1.1 times
the nominal capacity as follows:

𝑅0 = 𝑀0 𝐹0 𝑃0 𝑅! or 𝑅0 = 1.1𝑅! (7)

The coefficient of variation (COV) is also calculated to be 0.15 (𝑉) = 0.15) per AISI-S100 chapter
K, for CFS joists. 𝑅0 and 𝑉) are used to generate random strength variables of each purlin along
the length. The strength of adjacent purlins are assumed to be independent, and also the strength
along the length of each purlin is independent, including the splice region.

4.4 Load Variables

The load pattern for all loads is taken as a deterministic input, but the load magnitudes for each
bay and each joist are taken as an independent random variable. Load random variables for each
load type are produced using the bias factor and a coefficient of variations in the literature, as
described in the following sections.

4.4.1 Dead and live loads

Dead and live loads are uniformly distributed loads, but to include the load variability, the load
magnitude in the tributary area of each joist is assumed to be an independent random variable.
For each analysis, the distributed gravity loads are taken from a pool of random variables
generated with normal distribution. For dead loads, a bias factor and a coefficient of variations of
1.05 and 0.10, respectively, are considered per Meimand and Schafer (2014). For live loads, a
bias factor and a coefficient of variations of 1.00 and 0.25, respectively, are used to generate
Gaussian random variables.

4.4.2 Snow load

Per design load combinations, the 2nd and 3rd bay of each joist gets 100% of the mean snow load,
and the rest of the bays gets 50% of the mean snow load, as shown in Figure 7. For each analysis,
there are a total of 5 independent random loads for each beam, for a total of 25 independent
random loading variables. A bias factor and a coefficient of variations of 0.82 and 0.26,
respectively, per Meimand and Schafer (2014), is used to generate random variables with a
normal distribution.
 Figure 7: Snow load pattern

4.4.3 Wind load

The wind load pattern is taken per ASCE7-16 load patterns, as shown in Figure 8 (ASCE and SEI
2010). The wind load pattern is taken as a deterministic input, but the wind pressure magnitude
in each bay is a random variable. The wind bias factor for ASCE7-10 is 0.575, and the coefficient
of variation is 0.37 (Meimand and Schafer, 2014; ASCE and SEI 2010, 2016). It is assumed that
ASCE7-16 has the same bias and standard deviation. For simplicity, the wind random variables
are generated with a normal distribution.

Figure 8: Wind load pattern and nominal magnitudes

4.4.4 Load summary

Figure 9 summarizes all the deterministic load patterns and magnitudes over the roof region under
study. Figure 10 shows a single realization of the simulation for the same load patterns. All
variables are generated with magnitudes following a normal distribution. As shown, the loading is
variable for adjacent joists and along each joist. Each model in the MC analysis is subjected to a
different realization of the random loads (similar to Figure 10) and averages across simulations
converge to the original deterministic distributions (Figure 9).

Figure 9: Deterministic load pattern and magnitudes (lengths are in inches)

 Figure 10: One realization of random load variables (lengths are in inches)

5. RELIABILITY ANALYSIS APPROACH

As shown in Figure 1(a), the load effect, 𝑄, and the resistance, 𝑅, are random variables and the
probability of exceeding a limit state assuming a lognormal distribution shown in Figure 1(b) can
be calculated as Eq. (2a), i.e.:
!
#$% &
" #
𝛽= '
(8)
!
$%&"'

where, 𝑚 refers to the mean and 𝜎 the standard deviation of the natural log of the resistance-to-
load (𝑅/𝑄) ratio. This study considers a direct approach in addressing the system reliability of the
roof purlins. This approach can be summarized in the following main steps:

Step 1: Generate independent random variables for resistance (𝑅) of all purlins, including initial
guess for 𝑅121
Step 2: Generate independent random variables for all applied loads (𝑄)
Step 3: Perform a series of 2nd-order nonlinear analyses and determine the 𝑅/𝑄 ratios (or load
factor in MASTAN) at which collapse or other performance criteria is met
)
Step 4: Calculate ln ?+ @ and 𝜎#$%!&
0 "
Step 5: Calculate the reliability factor 𝛽 from Eq. (8)
Step 6: If 𝛽 > 𝛽3 , (where 𝛽3 = target reliability) increase 𝑅121 and go back to Step 1
To incorporate the system reliability effect on the member design, the initial 𝑅121 factor is assumed
to be 1.0 and if the reliability index (𝛽) is larger than the target reliability index (typically 𝛽3 =2.5
for members and systems), then the whole loop can be repeated with slightly higher 𝑅121
(increments of 0.05) until the time 𝛽 is not smaller than the target reliability index, 𝛽3 . The number
of random variable samples needs to be large enough to make sure the calculated logarithmic
mean and logarithmic standard deviation of the results is stable with less than 1% error.

6. SIMPLIFIED EXAMPLE

Before performing MC simulations, the merits of using the system reliability approach instead of
the element reliability approach are studied using a simplified example as follows. Consider the
following estimation of the reliability index for the roof under the uplift load combination. The
applied load to the beam is an entirely uniform wind dominated unfactored load of 1.05D-0.575W.
The load pattern is defined in Figure 9. The resistance of the purlins along the length is defined
as 1.10 times the nominal capacity of the purlin as shown in Figure 5.

MASTAN 2nd-order nonlinear analysis has been performed, and the resistance-to-load ratio is
calculated to be 2.561 at collapse, as shown in Figure 11. The figure shows the location of the
plastic hinges and the load factor at which the plastic hinge formed. Accordingly, the reliability
index can be approximated from Eq. (2b) and calculated using 𝑉+ = 0.31 (Meimand and Schafer
)#
2014), = 2.561, and 𝑉) = 0.15:
+#

!
#$% # & #$ (5.789)
"#
𝛽= =
√;.97( /;.<9(
-.!( /."(

𝛽 = 2.73 > 𝛽3 = 2.5

The calculated system reliability index of 2.73 is larger than the target reliability index of 2.5, and
therefore, it is expected that it may be possible to achieve a system reliability factor, 𝑅121 , larger
than 1.0. The 𝑅121 factor is determined in the following section using a large number of models in
MC simulations.

Figure 11: Applied load ratio for the uplift load case
7. MONTE CARLO SIMULATIONS

Monte Carlo simulations have been performed following the steps explained in Section 5. To
determine the minimum number of models required to get a stable standard deviation (STD) or
Coefficient of Variation (COV) of the 𝑅/𝑄 or load factors, a large number of simulations have been
performed for one load combination and the results are shown in Figure 12. As shown, after about
100 simulations, COV and STD are stable and not changing with a higher number of simulations.
To make sure that an adequate number of models are used in the simulations, a total of 200
simulations are considered for each load combination in the following.

Figure 12: Standard deviation (STD) on and coefficient of variation (COV) convergence

Figure 13 shows the load factor (𝑅/𝑄) distribution at collapse for different load cases and 𝑅121
factors, changing between 1.0 and 1.35, with no deflection limit state. As expected, the average
𝑅/𝑄 factors decrease as the 𝑅121 factor in increased. The histograms of the 𝑅/𝑄 distribution show
less spread for the gravity dominated load case (1.2D+1.6L+0.5W) and more spread for snow
and wind dominated load cases.

By assuming an equal likelihood of occurrence for each load combination, a combined 𝑅/𝑄
distribution is provided for each 𝑅121 factor, including all load combinations, as shown in Figure
14. The system reliability index, 𝛽, is calculated using Eq. (8). The logarithmic mean of the 𝑅/𝑄
ratio and the logarithmic standard deviation per each 𝑅121 factor is calculated based on the results
of 800 simulations for a total of four load combinations.

The results in Figure 14 show that for 𝑅121 =1.0, the system reliability index of 3.24 is significantly
higher than the target reliability index of 2.5. This shows the potential merits of using the system
reliability approach rather than the member reliability approach. By increasing 𝑅121 to 1.15, the
reliability index is close to the target index of 2.5. It can be concluded that the strength of the
purlins can be increased by 15% to account for the system reliability effect. As shown, design with
a 𝑅121 >1.15 results in failing to meet the target reliability index.

In the 2nd-order nonlinear simulations in MASTAN, there is a possibility that the model converges
at extensive/unrealistic deformations. In these cases, it is required to limit the deformations to
avoid unrealistic results. Accordingly, a deflection limit of L/40, which corresponds to about 1.25%
total plastic hinge rotation, is considered in Figures 14 and 15. The 1.25% total plastic hinge
rotation is generally achievable for cold-formed steel members (Ayhan and Schafer 2017) but is
not the focus of this study. As shown in Figure 15, the 15% increase in the purlin strength still
provides a system reliability index of 2.51, which is more than the target reliability index.

Figure 13: “Load factor at collapse” distribution for different load cases and Rsys values
(no deflection limit)
Figure 14: Load factor at collapse distribution and system reliability index for different Rsys
values (no deflection limit)
Figure 15: “Load factor at collapse” distribution for different load cases and Rsys values
(including deflection limit state of L/40)
Figure 16: Load factor at collapse distribution and system reliability index for different Rsys
values (including deflection limit state of L/40)
8. DISCUSSION

Adoption of an 𝑅121 factor for use in design would require careful definition of the “system” and
ensuring that the system and its possibilities were faithfully represented in the analysis justifying
the selected 𝑅121 . The preliminary analysis provided herein suggests that even though metal
building system purlins are highly optimized to existing load cases, and relatively sparsely spaced
in the roof, it may be possible to benefit from system reliability.

The roof model employs plastic hinges for the purlins that assume elastic perfectly-plastic (EPP)
response. Although total plastic rotations allowed are small, the EPP assumption is not strictly
true. Further, the model does not capture out-of-plane demands that result as the purlins collapse
and attempt to twist under load. Deck-to-purlin connections, which may limit the benefits of the
deck under large deformations, are not considered in the developed model.

The use of independent random variables (even though mean-centered about the deterministic
values) for the load magnitude may artificially increase the spatial variation. Correlated random
variables or other more sophisticated treatments of the spatial load variation may be warranted.

The controlling load cases were down-selected from a large number of considered load cases
through the use of conventional design. The impact of load cases not included in the final MC
simulations remains unknown. Additional examination of the best methods to limit the number of
considered load variations, but still capture the large number of load cases commonly considered
in design, is needed.

The single system analysis approach of Section 6 provided estimated response similar to the
more involved MC simulations with a single nominal collapse analysis. Suggesting that simpler
means to utilize load redistribution may be possible. If a single nonlinear collapse analysis was
completed for every load case this may be as useful as the MC simulations performed herein on
a small number of load cases – further work in this regard is needed.

9. SUMMARY AND CONCLUSIONS

This study has provided an analytical approach to incorporate system reliability effects in the
design of roof purlins in metal buildings. It is intended to reflect the system effects in a component
design factor, namely 𝑅121 , that can be used to increase the design capacity of the purlins. Based
on an archetype metal building design Monte Carlo simulations have been performed to study the
reliability of a group of roof purlins connected via profiled steel decks. The geometric and material
nonlinear collapse analyses of the roof segment have been performed in MASTAN to find the
ultimate resistance-to-capacity ratios, and the results have been used to calculate the reliability
index for a group of five purlins designed with different 𝑅121 factors. The results showed the
beneficial effect of including system reliability effects in the design of the roof purlins. The 𝑅121
factor is estimated to be about 1.15, which means a 15% increase in the purlin capacity in design.
A deflection limit of L/40 was considered in the simulations, which corresponds to about 1.25%
total plastic hinge deformation at failure. More studies are required to establish a path to evaluate
the rotational capacity of the purlins and incorporate the nonlinear behavior of the connecting
steel decks into the simulations. The results provided here can be interpreted as a proof of
concept for the effect of system reliability on the design of steel purlins, but more research is
required to provide an 𝑅121 factor that reflects all characteristics of the actual roofing system.
10. ACKNOWLEDGMENTS

The advice and consultation of Prof. Sanjay Arwade at the University of Massachusetts-Amherst
was beneficial during the development of this work. This work was seeded by a small gift from
the Metal Building Manufacturers Association (MBMA). A portion of the work was conducted
through contract with Simpson Gumpertz & Heger (SGH). Any opinions, findings, and conclusions
or recommendations expressed in this publication are those of the authors and do not necessarily
reflect the views of MBMA or SGH.

References

AISI-S100-16 (2016) North American Specification for the Design of Cold-Formed Steel Structural
Members. American Iron and Steel Institute, Washington, DC ANSI/AISI-S100-16.

ANSI and AWC (American National Standards Institute and American Wood Council). (2012).
“National design specifications for wood construction.” ANSI/AWC NDS 2012, Leesburg, VA.

ASCE and SEI (Structural Engineering Institute). (2010). “Minimum design loads for buildings and
other structures.” ASCE/SEI 7-10, Reston, VA.

ASCE and SEI (Structural Engineering Institute). (2016). “Minimum design loads for buildings and
other structures.” ASCE/SEI 7-16, Reston, VA.

Ayhan, D., and Schafer, B. W. (2017). “Characterization of in-plane backbone response of cold-
formed steel beams.” J. Const. Steel Res., 132, 141–150.

Buonopane, S.G., Schafer, B.W. (2006). “Reliability Implications of Advanced Analysis in the
Design of Steel Frames.” ASCE, Journal of Structural Engineering. 132 (2) 267-276.
(doi:10.1061/(ASCE)0733-9445(2006)132:2(267))

Chatterjee, A. (2016). “Structural system reliability with application to light steel-framed buildings.”
Ph.D. dissertation, Virginia Polytechnic Institute, Blacksburg, VA.

Chatterjee, A., Arwade, S. R., Schafer, B. W., and Moen, C. D. (2017). “System reliability of floor
diaphragms framed from cold-formed steel with wood sheathing.” J. Struct. Eng.,
10.1061/(ASCE)ST.1943-541X.0001958, 04017208.

Meimand, V. Z., & Schafer, B. W. (2014). Impact of load combinations on structural reliability
determined from testing cold-formed steel components. Structural safety, 48, 25-32.

MBMA (2019) Metal Building Archetype for Roof System Reliability Studies. (private
communication)

Smith, B. H., Arwade, S. R., Schafer, B. W., and Moen, C. D. (2016). “Design component and
system reliability in a low-rise cold-formed steel framed commercial building.” Eng. Struct., 127,
434–446.
Smith, B. H., Chatterjee, A., Arwade, S. R., Moen, C. D., & Schafer, B. W. (2018). System
Reliability Benefits of Repetitive Framing in Cold-Formed Steel Floor Systems. Journal of
Structural Engineering, 144 (6) 04018061.

Zhang, H., Liu, H., Ellingwood, B.R., Rasmussen, K.J.R. (2018). “System Reliabilities of Planar
Gravity Steel Frames Designed by the Inelastic Method in AISC 360-10.” ASCE, J. of Struct. Eng.
144 (3) 04018011.