First Order Reliability Method (FORM) and

its applications in Civil / Structural

Engineering

Lessandro Estelito O. GARCIANO

Department of Civil Engineering
De La Salle University

Engr. Maynard Oliver C. GONZALES

Graduate Student, DLSU
 Outline of the presentation
o Introduction

o Basic Theory of First Order Reliability Method (FORM)

Civil / Structural Applications

o A hypothetical example
o Partial safety factor calculation for bending strength parallel to
fiber

o Summary
 Introduction
o Engineering design is the assurance of structural performance
with economic constraint

o Our primary responsibility (role) as Structural Engineers is to

assure performance (including safety) of structures

o The achievement of this objective, is not a simple problem,

especially for large and complex structures

o Occasionally structures fail to perform their intended function,

including (hopefully) rare cases of collapse
Effects of extreme winds
Effects of extreme earthquakes
 Structural Failure
o Structural failure occurs when a component or
components cannot perform its intended function
❖ failure occurs when Vu exceeds Vcritical
❖ steel beam may fail when a plastic hinge is developed
❖ exceedance of critical buckling strength

o The concept of a “limit state” is used to help define failure

in the context of structural reliability analyses
 Limit State
o A “limit state” is a boundary between
𝑿
desired and undesired performance 2
of a structure
Unsafe region 𝑔 𝑿 < 0

o This boundary is often represented Limit state function 𝑔 𝑿 = 0

mathematically by a limit state
function or a performance function

Safe region 𝑔 𝑿 > 0

o This undesired performance can 𝑿1

occur in many modes of failure, e.g., The limit state concept
cracks, corrosion, excessive
deformation, local buckling etc.
 Limit State (Performance) Function
o The limit state, corresponding to the boundary between desired and
undesired performance, is when 𝐺 𝑅, 𝑆 = 0; (𝑅 = 𝑆)
o The structure is safe (desired performance) if 𝐺 𝑅, 𝑆 ≥ 0; (𝑅 > 𝑆)
o The structure is not safe (undesired performance) if 𝐺 𝑅, 𝑆 < 0; (𝑅 < 𝑆)

The limit state concept

𝑥2′
y’
Failure State M > 0 g(X1, X2) < 0

Limit State M = 0 g(X1, X2) = 0

d g(X1, X2) > 0

Safe State M > 0

x’ 𝑥1′
 Examples of Limit State Functions
o moment capacity of a compact steel beam
Fy = yield stress
𝐺 𝐹𝑦 , 𝑍, 𝑆 = 𝐹𝑦 𝑍 − 𝐷 + 𝐿 + 𝑊 + 𝐸 Z = plastic section modulus

brittle fracture due to crack instability 

( )  S 
N (t )
o 2
a d = a o exp C Y  i
m

𝐾  i =1 
𝐺 𝑅, 𝑆 = −𝑆
( )  m
N (t )
𝑌 𝜋𝑎𝑑 
m
S i
m
= N (t ) 2 2  SH 1 +
m
 
i =1  2 
K = fracture toughness
o hypothetical model of a deteriorating structure
Y = geometry correction factor
𝐺 ∙ = 𝒙1 + 𝒘𝑘 − 𝒙2 𝑡 − 𝒙3 ad = accumulated damage
C, m = crack propagation parameters
Sim = cumulative dynamic stress
range during time t
N = no. of load cycles up to time t
 Some types of limit states
o Ultimate limit states (ULS) – related to the loss of load-
carrying capacity.

o Serviceability limit states (SLS) – are related to gradual

deterioration, user’s comfort or maintenance costs. May
or may not be directly related to structural integrity

o Fatigue limit states (FLS) – related to loss of strength

under repeated loads. Related to accumulation of
damage and eventual failure under repeated loads.
 ULS modes of failure
o exceeding the moment carrying capacity
o formation of a plastic hinge
o crushing of concrete in compression
o shear failure of the web in a steel beam
o loss of the overall stability
o buckling of flange
o buckling of web
o weld rupture
 SLS modes of failure
o excessive deflection

o excessive vibration

o permanent deformations

o cracking
 FLS modes of failure
o crack formation and propagation until rupture

o failure of pre-stressing strands of concrete structures

 Basic Theory
Let us define the following random variables: The reliability maybe formulated also by convolution with
respect to r, as shown below.
𝑅 = the supply capacity ∞
𝑆 = the demand requirement 𝑃𝐹 = න 1 − 𝐹𝑠 (𝑟) 𝑓𝑟 (𝑟)𝑑𝑟
0

𝑃𝐹 = 𝑃 𝑅 < 𝑆

= ෍ 𝑃 𝑅 < 𝑆|𝑆 = 𝑠 𝑃(𝑆 = 𝑠)

𝑎𝑙𝑙 𝑦

If R and S are statistically independent; that is

𝑃 𝑅 < 𝑆|𝑆 = 𝑠 = 𝑃 𝑅 < 𝑆

For continuous R and S the above equation for the

probability of failure becomes

∞
𝑃𝐹 = න 𝐹𝑟 (𝑠)𝑓𝑠 (𝑠)𝑑𝑠
0
 Basic Theory

The overlap region depends on the relative The overlap region depends on the degree
positions of 𝑓𝑥 𝑥 and 𝑓𝑦 𝑦 . of dispersions in 𝑓𝑥 𝑥 and 𝑓𝑦 𝑦 .
These dispersions can be expressed in terms
of the coefficient of variations 𝛿𝑥 and 𝛿𝑦 .

Any measure of safety or reliability therefore should be a function of the relative positions of 𝑓𝑥 𝑥 and 𝑓𝑦 𝑦 as well as the
degree of dispersions.
 Probability of Failure
Consider the performance function
𝐺 𝑅, 𝑆 = 𝑅 − 𝑆

The probability of failure

𝑃𝑓 = ෍ 𝑃 𝑅 = 𝑟𝑖 ∩ 𝑆 > 𝑟𝑖

= ෍ 𝑃 𝑆 > 𝑅|𝑅 = 𝑟𝑖 𝑃 𝑅 = 𝑟𝑖

for the continuous case

+∞
𝑃𝑓 = න 1 − 𝐹𝑆 𝑟𝑖 𝑓𝑅 𝑟𝑖 𝑑𝑟𝑖
−∞
+∞
=1−න 𝐹𝑆 𝑟𝑖 𝑓𝑅 𝑟𝑖 𝑑𝑟𝑖
−∞
 Probability of Failure
Alternative formulation
𝐺 𝑅, 𝑆 = 𝑅 − 𝑆

The probability of failure

𝑃𝑓 = ෍ 𝑃 𝑆 = 𝑠𝑖 ∩ 𝑅 < 𝑠𝑖

= ෍ 𝑃 𝑅 < 𝑆|𝑆 = 𝑟𝑖 𝑃 𝑆 = 𝑠𝑖

for the continuous case

+∞
𝑃𝑓 = න 𝐹𝑅 𝑠𝑖 𝑓𝑠 𝑠𝑖 𝑑𝑠𝑖
−∞
 An example
o The convolution integral X Type a b
Load, S Uniform 38.4 141.9
Resistance, R Uniform 115.4 184.6
 An example
1
𝑓𝑠 𝑥 = = 9.63 × 10−3
141.9 − 38.4

𝑥 − 115.4
𝐹𝑅 𝑥 =
184.6 − 115.4
+∞
𝑃𝑓 = න 𝑓𝑠 𝑥 𝐹𝑅 𝑥 𝑑𝑥
−∞

+∞
𝑥
=න 9.63 × 10−3 − 1.67 𝑑𝑥
−∞ 184.6 − 115.4

= 0.0483
= 4.83%
 Basler / Cornell Notation
o The first two moments of M are
𝑀 = 𝑅 - 𝑆

𝑀 = 2 𝑅 +2 𝑆

o The safety index is

𝑀
=
𝑀

o The failure probability

𝑃𝑓 = (−)
 Basler / Cornell Notation
o the first two moments of M are
X Type xi xi
𝑀 = 150 − 90 = 60 Load, S Normal 90 30
Resistance, R Normal 150 20
𝑀 = 202 +302 = 36.1

o the safety index is

𝑀 60
= = = 1.7
𝑀 36.1

o The failure probability is

𝑃𝑓 = (−1.7) = 0.049 = 4.9%

Representation as a joint pdf
o The joint pdf is X Type a b
𝑓𝑋,𝑌 × 141.9 − 38.4 184.6 − 115.4 = 1 Load, S Uniform 38.4 141.9
Resistance, R Uniform 115.4 184.6
1
𝑓𝑋,𝑌 = = 1.39 × 10−4
7162.2

o The failure probability is

𝑃𝑓 = 𝑉 = 0.5 × Δ𝑆 × Δ𝑅 × 𝑓𝑋,𝑌
1 2
= 141.9 − 38.4 1.39 × 10−4
2

𝑃𝑓 = 0.0489
 First Order Reliability Method (FORM)

Probability Integration Probability Integration in X-space

FORM comes from the fact that the performance function g(x) is approximated by the first order Taylor expansion (linearization)
First Order Reliability Method (FORM)

Direct evaluation of 𝑃𝑓 = ‫𝑔׬‬ 𝑓

𝒙 <0 𝑥
𝒙 𝑑𝒙 is
extremely difficult for the following reasons:

o a number of RVs are involved, so the probability

integration is multi-dimensional

o the integrand 𝑓𝑥 𝒙 is the joint pdf of X and is

generally a non-linear and multi-dimensional
function

o the integration boundary 𝑔 𝑿 = 0 is also multi-

dimensional and usually a non-linear function

In this regard, FORM and SORM have been

developed in the area of structural reliability
 Steps involved in FORM
STEP 1
Simplify the integrand 𝑓𝑥 𝒙 so that the contours become
more regular and symmetric

STEP 2
Approximate the integration boundary 𝑔 𝑿 = 0.

After these steps, an analytical solution to the probability

integration can be found.
 Step One
Simplify the integrand
Transform the random variables from their original random space into a standard
normal space

𝑿 = 𝑋1 , 𝑋2 , ⋯ , 𝑋𝑛 X-space U-space
𝑼 = 𝑈1 , 𝑈2 , ⋯ , 𝑈𝑛

This transformation from X to U is based on the condition that the cdfs of the RVs remain the same before and after the transformation.
 Step One
Simplify the integrand
After the transformation, the performance function
becomes
𝑌 = 𝑔(𝑼)

After the transformation, the probability integration

becomes

𝑃𝑓 = 𝑃{𝑔 𝑼 < 0} = න 𝜙𝑈 𝑼 𝑑𝑼
𝑔 𝑼 <0

where 𝜙𝑈 𝑼 is the joint pdf of U. Since all RVs are Probability integration after the transformation
independent, the joint pdf is the product of the
individual pdfs of the standard normal distribution
and is then given by
𝑛
1 1 2
𝜙𝑈 𝑼 = ෑ 𝑒 −2𝑢𝑖
𝑖=1
2𝜋

Therefore, the probability integration becomes

𝑛
1 1
−2𝑢2𝑖
𝑃𝑓 = න ⋯න ෑ 𝑒 𝑑𝑢1 𝑑𝑢2 ⋯ 𝑑𝑢𝑛
𝑔 𝑢1 ,𝑢2 ,⋯𝑢𝑛 <0 𝑖=1
2𝜋
Probability Integration in U-space
 Step Two
Approximate the integration boundary
FORM uses a linear approximation (the first order The mathematical model for locating the MPP is
Taylor expansion) given by

𝑔 𝑼 ≈ 𝐿 𝑼 = 𝑔 𝒖∗ + ∇𝑔 𝒖∗ (𝐔 − 𝒖∗ )𝑇
𝑛
where 𝐿(𝑼) is the linearized function, 𝒖 = ∗ 1 1 2
max ෑ 𝑒 −2𝑢𝑖
𝑢1∗ , 𝑢2∗ … . 𝑢𝑛∗ is the expansion point, T stands for 𝑢 2𝜋
𝑖=1
transpose, and ∇𝑔 𝒖∗ is the gradient of 𝑔 𝑼 at subject to 𝑔 𝑼 = 0
𝒖∗ .

𝜕𝑔 𝑼 𝜕𝑔 𝑼 𝜕𝑔 𝑼
∇𝑔 𝒖∗ = , ,⋯, , อ since
𝜕𝑈1 𝜕𝑈2 𝜕𝑈𝑛
𝒖∗
𝑛
1 1
− 𝑢𝑖2 1 1
− σ𝑛 2

Expand the performance function 𝑔 𝑼 at a point ෑ 𝑒 2 = 𝑒 2 𝑖=1 𝑢𝑖

𝑖=1
2𝜋 2𝜋
that has the highest contribution to the probability
integration. 1
1 − 𝑢𝑖2
maximizing ς𝑛𝑖=1 𝑒 2 is equivalent to minimizing
The point that has the highest probability density on 2𝜋
the performance 𝑔 𝑼 = 0 is termed as the Most σ𝑛𝑖=1 𝑢𝑖2 .
Probable Point (MPP).
 Step Two
Approximate the integration boundary
The model for the MPP search can be re-written as

min 𝒖
൝ 𝑢
subject to 𝑔 𝑼 = 0

where 𝒖 stands for the norm (length or

magnitude) of a vector, namely,
𝑛

𝒖 = 𝑢12 + 𝑢22 + ⋯ + 𝑢𝑛2 = ෍ 𝑢𝑖2

𝑖=1

o As shown above, the MPP is the shortest distance point from

the limit state 𝑔 𝑼 = 0 to the origin O in the U-space.
o The minimum distance 𝛽 = 𝒖∗ is called the reliability
index.
 Step Two
Approximate the integration boundary
Recall zero at the failure surface The equation below indicates that 𝐿 𝑼 is a linear
function of standard normal variables. Therefore,
𝑔 𝑼 ≈ 𝐿 𝑼 = 𝑔 𝒖∗ + ∇𝑔 𝒖∗ (𝐔 − 𝒖∗ )𝑇 𝐿 𝑼 is also normally distributed.
Its mean is given by 𝑛 the mean of 𝑈𝑖 = 0
𝜕𝑔 𝑼 𝜕𝑔 𝑼 𝜕𝑔 𝑼
∇𝑔 𝒖∗ = , ,⋯, , อ 𝐸𝐿 𝑼 = 𝐸 𝑎0 + 𝐸 ෍ 𝑎𝑖 𝑈𝑖
𝜕𝑈1 𝜕𝑈2 𝜕𝑈𝑛
𝒖∗ 𝑖=1
𝑛 𝑛
𝜕𝑔 𝑼 𝜕𝑔 𝑼
𝐿 𝑼 =෍ ቤ 𝑼𝒊 − 𝒖∗𝒊 𝜇𝐿 = 𝑎0 = − ෍ ቤ 𝒖∗𝒊
𝜕𝑈𝑖 𝒖∗ 𝜕𝑈𝑖
𝑖=1 𝒖∗
𝑖=1
𝑛 𝑛
𝜕𝑔 𝑼𝜕𝑔 𝑼 and its standard deviation is given by
=෍ ቤ 𝑼𝒊 − ෍ ቤ 𝒖∗𝒊
𝜕𝑈𝑖 𝒖∗ 𝜕𝑈𝑖 ∗ 𝑛 𝑛 2
𝑖=1 𝑖=1 𝒖 𝜕𝑔 𝑼
𝜎𝐿 = ෍ 𝑎𝑖2 = ෍ ቤ
𝑛 𝜕𝑈𝑖
𝑖=1 𝑖=1 𝒖∗
𝐿 𝑼 = 𝑎0 + ෍ 𝑎𝑖 𝑈𝑖
𝑖=1 Therefore, the probability of failure is calculated by
𝜕𝑔 𝑼
σ𝑛𝑖=1 ฬ 𝒖∗
−𝜇𝐿 𝜕𝑈𝑖 𝒖∗ 𝒊
𝑛 𝑝𝑓 ≈ 𝑃 𝐿 𝑼 < 0 = Φ =Φ
𝜕𝑔 𝑼 𝜕𝑔 𝑼 𝜎𝐿 2
𝑎0 = − ෍ ቤ 𝒖∗𝒊 𝑎𝑖 = ቤ 𝜕𝑔 𝑼
𝜕𝑈𝑖 𝜕𝑈𝑖 σ𝑛𝑖=1 ฬ
𝑖=1 𝒖∗ 𝒖∗ 𝜕𝑈𝑖 𝒖∗
 Summary
The procedure for FORM is as
follows:

o Transform the original random

variables from X-space to U-space
by Rosenblatt transformation.

o Search the MPP in U-space and

calculate the reliability index β.

o Calculate the probability of failure

Flowchart of the MPP search

Reliability Index  and
Probability of failure Pf

Pf 
10-1 1.28
10-2 2.33
10-3 3.09
10-4 3.71
10-5 4.26
10-6 4.75
10-7 5.19
10-8 5.62
10-9 5.99
 Hypothetical Problem
The strength (X1) and maximum The performance function becomes
stress (X2) of a structural component 𝑔 𝑼 = 𝑈1 𝜎𝑋1 + 𝜇𝑋1 − 𝑈2 𝜎𝑋2 − 𝜇𝑋2
are normally distributed. Use FORM
= 20𝑈1 + 200 − 10𝑈2 − 150
to compute the probability of failure
of the component. = 20𝑈1 − 10𝑈2 + 50
RV Type  
X1 Normal 200 20
X2 Normal 150 10

Performance Function:
𝑔 𝑿 = 𝑋1 − 𝑋2

𝑋1 − 𝜇𝑋1 𝑋2 − 𝜇𝑋2 𝑔 𝑼 = 20𝑈1 − 10𝑈2 + 50

𝑈1 = 𝑈2 =
𝜎𝑋1 𝜎𝑋2

𝑋1 = 𝑈1 𝜎𝑋1 + 𝜇𝑋1 𝑋2 = 𝑈2 𝜎𝑋2 + 𝜇𝑋2

 Hypothetical Problem
∇𝑔 𝒖0 = 202 + −10 2 = 22.3607

∇𝑔 𝒖0
𝑎0 = = 0.8944, −0.4472
∇𝑔 𝒖0

𝑔 𝑼 = 20𝑈1 − 10𝑈2 + 50
𝛽0 = 𝒖0 = 0

∗ 0 0
𝑔 𝒖0
𝒖 = −𝒂 𝛽 +
𝑔 𝑼 = 20𝑈1 − 10𝑈2 + 50 𝛻𝑔 𝒖0
50
Start from the initial point 𝐮0 = 0,0 = − 0.8944, −0.4472 0+
22.3607
𝜕𝑔 𝜕𝑔 The reliability index is
= 20 = −10
𝜕𝑈1 𝜕𝑈2
𝛽 = 𝒖∗ = 2.2361
𝜕𝑔 𝜕𝑔
∇𝑔 𝒖0 = , = The probability of failure is
𝜕𝑈1 𝜕𝑈2
20, −10 𝑝𝑓 = Φ −𝛽 = Φ −2.2361
𝑔 𝒖0 = 50
= 0.0127
 Hypothetical Problem
An analytical solution to this problem exists since Using Monte Carlo Simulation (MCS)
𝑔 𝑿 is a linear combination of normally distributed
RVs 𝑋1 and 𝑋2 .

𝜇𝑔 = 𝜇1 − 𝜇2 = 50

𝜎𝑔 = 𝜎12 + 𝜎22 = 22.36

The reliability index is

50
𝛽= = 2.2361
22.36

The probability of failure is

𝑝𝑓 = Φ −𝛽 = Φ −2.2361

= 0.0127
Using 100,000 sample realizations

𝑝𝑓 = Φ −𝛽 = Φ −2.2360

= 0.0127
 Investigation of an alternative testing protocol to determine the
shear strength of bamboo parallel to the grain
Gabrielle Luisa Cantos, Luis Lopez, Richard De Jesus, Corinna Salzer and Lessandro Estelito Garciano
Maderas. Ciencia y Tecnologia, 21(4), pp. 559 - 564, 2019

ABSTRACT: The common testing protocol to

determine the shear strength of bamboo parallel
to grain is ISO 22157 (2004). However widely
documented limitations of this test procedure calls
for an alternative testing protocol. This study
compared ISO 22157 (2004) (TP1) and 2 alternative
testing protocols proposed by Base Bahay (TP2a
and 2b) in determining shear strength of bamboo
parallel to grain. The study used a local bamboo
known as “kawayang tinik” (Bambusa blumeana).
Results showed that TP1 produced more pure
shear failures at 65% compared to TP2a and TP2b
at 55% each. Shear strength values between the
three methods were not significantly different,
although, TP2b resulted in the lowest coefficient of
variation at 18%. Also, TP2b was found to be the
most effective in terms of specimen preparation,
ease of assembly of testing equipment and low
variability data.
 Assessment of testing protocols for bamboo
for tension parallel to the fibers
Martin Dela Cruz, Luis Felipe Lopez, Lessandro Estelito Garciano and Richard De Jesus
International Journal of GEOMATE, 19(74), 31–36, 2020

ABSTRACT: To determine tension parallel to fiber properties of Bamboo, one

can employ ISO 22157. However, several studies highlighted some challenges
in using the method such as premature failure of testing due to force at
clamping and slipping failure at the grips. Hence, this study aimed to look for
a solution to address these challenges. This study investigated, both
qualitatively and quantitatively, three testing protocols in determining
bamboo’s tensile strength parallel to fiber – (1) ISO 22157, (2) a modification
of ASTM D143, and (3) Pittsburgh Method. The success rate, test duration,
load rate and transmission, specimen preparation, equipment fabrication,
and execution of the three protocols were compared. The physical properties
of bamboos, such as moisture content, density, and shrinkage, were also
measured in the study. The results of the study showed that the modified
ASTM D143 test is the most viable method to use in testing the tensile
strength parallel to fiber of bamboos. Modified ASTM D143 produced the
highest tensile strength with value equal to 100.36 MPa, compared to
specimens tested under ISO 22157 with tensile strength only reaching 94.11
MPa, while Pittsburgh Test produced the lowest tensile strength of 76.78
MPa. Modified ASTM D143 also yielded the lowest confidence interval which
implied good consistency. Modified ASTM D143 is the recommended test
protocol based on the results of this study as it gained the highest success
rate during testing, lowest testing duration and the easiest to execute.
 Establishing the strength parameters parallel to fiber of
Dendrocalamus Asper (Giant Bamboo)
Anika Paula De Jesus, Lessandro Estelito Garciano, Luis Lopez, Diane Megan Ong, Ma. Chrissel Paula Roxas,
Mikhaela Andrea Tan and Richard De Jesus
International Journal of GEOMATE, 19(74), 31–36, 2020

ABSTRACT: The supply of timber is becoming scarce in

the Philippines. Dendocalamus asper (or giant bamboo)
is a candidate for safe, sustainable, and low-cost
alternative housing to timber. However, its mechanical
properties are yet to be established. In this paper, the
compressive strength, shear strength, and tensile
strength parallel to fiber of Dendrocalamus asper are
determined using ISO 22157 and ASTM D143 tests. For
the latter, a slight modification of ASTM D143 was
employed. The result yielded an improved tensile
strength parallel to fiber. The tensile strength was also
tested using a modified version of ASTM D143, wherein
the length of the test piece was changed to ensure
failure within the gauge length. Two hundred (200)
samples of 2m-long Dendrocalamus asper (giant
bamboo) poles were prepared and used for testing.
Equipment was fabricated to the dimensions of the test
sample. The obtained tensile strength parallel to fiber
using the ISO 22157 method had an average strength of
312.78 MPa for specimens with attached hardwood tabs
and 424.43 MPa for specimens with attached softwood
tabs while with the Modified ASTM D143-94 method,
the average strength was 269.86 MPa. The shear
strength parallel to the fibers had an average strength of
10.64 MPa at the internode and 11.87 MPa at the node.
Lastly, the compressive strength parallel to the fibers
had an average strength of 63.42 MPa at the internode
and 55.55 MPa at the node
 Comparative Analysis of Shear Strength Parallel to Fiber
of Different Local Bamboo Species in the Philippines
Bautista, Lessandro Estelito Garciano and Luis Lopez
Sustainability 13, 8164, 2021

ABSTRACT: There are limited published studies related to the mechanical properties of bamboo species in the
Philippines. In this study, the shear strength properties of some economically viable bamboo species in the
Philippines were properly characterized based on 220 shear test results. The rationales of selecting this
mechanical property are the following: (1) Shear strength, parallel to the fiber, has the highest variability among
the mechanical properties; and (2) Shear is one of the governing forces on joint connections, and such
connections are the points of failure on bamboo structures when subjected to extreme loading conditions. ISO
22157-1 (2017) test protocol for shear was used for all tests. The results showed that Bambusa blumeana has
the highest average shear strength, followed by Gigantochloa apus, Dendrocalamus asper, Bambusa
philippinensis, and Bambusa vulgaris. However, comparative analysis, using One-way ANOVA, showed that shear
strength values among these bamboo species have significant differences statistically. A linear regression model
is also established to estimate the shear strength of bamboo from the physical properties. Characteristic shear
strength is also determined using ISO 12122-1 (2014) for future design.

Species All specimens

n fv, mean fv,c
G. apus 30 10.77 5.11
B. philippinensis 30 9.68 7.26
B. vulgaris 30 9.26 6.46
D. asper 30 10.31 6.98 ISO 22157:2019 Test Loading until Failure Typical Shear Failure
Set-up
B. blumeanaτ 100 11.44 5.15
B. blumeanaα 15 12.18 9.62
 Determining the characteristic bending strength and modulus of
elasticity of Philippine Bambusa blumeana (Kawayan tinik) and the
relationship between its physical and flexural properties
Marion Ryan A. Vicencio, Lessandro Estelito Garciano and Luis Lopez

ABSTRACT:Bambusa blumeana (locally known as Kawayan tinik) is the

most used and one of the most economically important bamboo species
in the Philippines. The study examined 70 samples for its physical
properties and conducted four-point bending test to determine its
flexural properties. The flexural properties investigated in this study were
ultimate bending moment (Mult), bending strength (fm,0), bending
stiffness (EI), and the modulus of elasticity (E). The characteristic values
were also evaluated to summarize the strength of the material. The
characteristic bending strength and modulus of elasticity were computed
and some of the secondary properties such as compression and tension
strength parallel to fibers, and shear strength were also derived. The
characteristic properties of B. blumeana were found to be: bending
strength (fm,0,k) - 55.90 MPa; mean modulus of elasticity (Emean) –
18.67 GPa; and 5th percentile modulus of elasticity (E0.05) – 12.896 GPa.
The secondary properties derived were: compression strength parallel to
fibers (fc,0,k) – 27.95 MPa; tension strength parallel to fibers (fc,0,k) –
55.90 MPa; and shear strength (fv,k) – 3.73 MPa. The physical properties
of the samples were measured such as the outside diameter (D), wall
thickness (t), linear mass (q), density (ρ), external taper (αe), internal
taper (αi), ovality (do), eccentricity (ec), and moisture content (ω). The
correlation between the physical properties were investigated and found
out that that there is strong correlation (R2 > 0.7) with ec and αe;
moderate correlation (0.5 > R2 > 0.7) between αi and αe, and αi and ec;
weak correlation (0.3 > R2 > 0.7) between q to D and t; and negatively
weak correlation between t compared to αe and ec. Examining the
relationship between the physical and flexural properties show that ρ
have moderate correlation (R2 = 0.5612) with the fm,0 and weak
correlation (R2 = 0.4372) with E; q have strong correlation (R2 = 0.8335)
with Mult and moderate correlation (R2 = 0.6826) with EI; D have weak
correlation (R2 = 0.3161) with Mult and moderate correlation (0.5689)
with EI. Other studies show that D could have better and stronger
correlation with M and EI but the difference could be attributed to the
difference in the species being studied. The results of the study could
possibly be used in the estimation of the flexural properties
 Proposed bamboo connection method for cement
bamboo frame
Lessandro Estelito Garciano , Richard De Jesus, Jonathan Sebastian Isleta, Jeslie Lim, Paul Daniel Mabborang, Dennis
Christian Santos and Luis Lopez

ABSTRACT: Bamboo is recognized as an

alternative material in construction. Exploring
connection types using bamboo may help
promote the use of bamboo as well as improve
structural reliability. This study investigated
three types of connection methods: a J-bolt
attached to a threaded bolt connected to a
bamboo culm with nodes at both ends, filled
with cement mortar to reinforce the J-bolt and
threaded by a bolt (type I); the second type is
similar to the first except for the absence of a
node at the end of the bamboo culm, and the
third type is also similar to the first except for
the absence of the cement mortar. The strength
and cost of each connection were compared and
evaluated. ISO 22156 and ISO 6891 were used as
reference standards for the testing. Test results
indicated that the second connection type is
16% stronger in terms of allowable tensile load
and 27% cheaper compared to the first
connection type. The third connection method
attained the lowest tensile capacity but the least
costly among all the connection methods.
 comparative study evaluating testing methods for the
tensile strength of bamboo perpendicular to the fiber
Carlo Dela Cruz, Luis Felipe Lopez, Lessandro Estelito Garciano and Richard De Jesus

ABSTRACT: In the Philippines, there is a consistently

high interest in the development of alternative
construction materials. Bamboo, aside from it being
abundant in the country, has the potential to
become a fully-fledged structural material once
local standardization is established. The tensile 3
strength of bamboo perpendicular to fiber is one of 2.5
mechanical properties and needs an established

Distribution
2
testing method. It is one of the dominant modes of
1.5
failure in structural applications, specifically
connections, and it is frequently the limit-state 1

governing its use. This study tested and evaluated 0.5

two currently existing alternative methods by other 0
researchers, as well as an additional, and proposed 0.00 0.50 1.00 1.50 2.00

testing method - Modified Colombia Test, in

determining the perpendicular tensile strength of
bamboo using a local bamboo specie – Bambusa
Blumeana (“Kawayang Tinik”). The MC Test is the
most conservative and consistent with different
loading mechanism and with a simple design and
execution. The Pittsburgh method however,
performed better than ASTM D143 More robust
setup, less demanding on precision esp. in specimen
fabrication
 Experimental Study on the Dowel-Bearing Strength
of Bambusa blumeana
Bangoy, Cres Dan Jr. O.; Zeng, Steven Royce A.; Falcon, Jedelle Y.; Lorenzo, Hannah Amyrose F.; Garciano, Lessandro
Estelito; Lopez, Luis; Cacanando, Carlo Joseph

ABSTRACT: Bamboo is a valuable and sustainable

construction material in the Philippines, especially in rural
areas, where it is locally sourced and abundant. However, the
use of bamboo in construction pre-sents significant
challenges due to its longitudinal fibers, thin walls, hollow
nature, and geometric irregularities. To address these
challenges, bamboo connections are necessary to join two or
more bamboo and distribute the load to other elements.
Although ISO 22157 contains test methods to evaluate
bamboo's physical and strength properties, testing methods
for dowel-bearing strength, a property crucial in bamboo
connections, still need to be improved. This study aimed to
evaluate the dowel-bearing strength of treated Bambusa
blumeana culms, a widely used bamboo species in the
country. Results showed that nodal placement significantly
affects the ultimate dowel-bearing strength of bamboo
connections. The top nodal placement yielded the highest
ultimate dowel-bearing strengths (6,450.97 kN to 22,901.63
kN) while the none and bottom nodal placements yielded
the lowest ultimate dowel-bearing strengths, ranging
(2,736.85 kN to 15,796.55 kN). Furthermore, samples with
no nodes exhibited brittle failure, while those with nodes
seemingly demonstrated strain hardening, delaying failure.
Specimens with nodes along the middle had vary-ing failure
modes, dependent on physical properties. The study's
findings demonstrate the im-portance of nodal placement in
improving the strength properties of bamboo connections
for sus-tainable construction. The results could inform the
design and construction of bamboo structures, particularly in
rural areas where bamboo is widely available and used as a
building material.
 Experimental Studies on The Ductility of a Proposed Bamboo
Connection with Embedded Rebar subjected to a Shear Force
Lorenzo Rio H. Cadiz, Louis Angelo C. Cabingan, Anjealyn Joyce A. Dy, Lenard Jan C. Guillermo, Vince Gem Sacdalan,
Lessandro Estelito Garciano, Lopez, Luis and Carlo Joseph Cacanando

ABSTRACT: Bambusa blumeana, a predominant bamboo

species in the Philippines, has been successfully used as a
structural element in the local construction industry.
Although the mechanical properties of different bamboo
species have been the subject of research in recent years,
studies on the strength of bamboo connections are limited. In
this study, bamboo connections involving embedded rebars
and steel clamps are proposed. A total of 120 bamboo
samples, categorized based on variations in rebar dimensions
and steel clamps, were tested. A point load with a load rate
of 10 mm/s was applied to each bamboo specimen, which
resulted in a shear load and minimal bending moment. The
results show that the average maximum shear force observed
for the unclamped samples was 6.48 kN and 14.0 kN for the
clamped samples. The initial findings of the clamped samples
suggest a notable increase in the maximum shear force of
approximately 115.75% greater than the unclamped samples.
For unclamped samples, it was observed that the dominant
mode of failure was the splitting of bamboo. Whereas for
clamped samples, three modes of failure were observed, e.g.
loosening of the steel clamp, snapping of the steel clamp at
the fastening point, and fracture of the steel clamp at the
midpoint. The results suggest that the inclusion of a steel
clamp in the bamboo connection with embedded rebar
improves the ductility of the bamboo connection.
Furthermore, the proposed connection increases the
ultimate failure load of the bamboo connection.
 Establishing the characteristic compression strength
parallel to the fiber of five local Philippine bamboo species
Cañete, Christy S., Navarra, Althea R., Panti, Christine Abegail T., Rubinas, Kerby D., Lessandro Estelito Garciano and ,
Lopez, Luis

ABSTRACT: Bamboo is considered a sustainable

construction material due to its ability to grow quickly
and its mechanical properties that are comparable to
timber. Contributing to the current effort to establish
the structural bamboo standards in the National
Structural Code of the Philippines (NSCP), this study
establishes the characteristic compressive strength of
four bamboo species: Bambusa vulgaris (36 samples),
Dendrocalamus asper (36 samples), Bambusa blumeana
(94 samples), and Guadua angustifolia Kunth (30
samples). The samples were subjected to compressive
loading following ISO 22157-1 (2017). The characteristic
compressive strength values obtained, according to ISO
12122-1 (2014), were 40.35 MPa for B. vulgaris, 40.21
MPa, for D. asper, 46.63 MPa for B. blumeana, and
36.99 MPa for G. angustifolia Kunth. Simple linear
analysis, One-way ANOVA, and Welch’s t-Test were used
to analyze the correlation models and establish a
comparative analysis of the effect of nodes, and
geometric and physical properties on the compressive
strength of bamboos. Comparing the characteristic
compressive strength obtained from this study to the
strength of unseasoned structural timber of Philippine
woods, all bamboo species showed higher strength
values, and thus have great potential as an alternative
construction material to timber
Design Codes
 Design Codes
o Design codes play a central role in the building process because they specify the requirements that the
designer must satisfy so that the minimum acceptable safety level is provided

o Reliability-based codes have been developed in the US, e.g.

✓ Structural Steel buildings – ANSI/AISC 360-16
✓ Loads on buildings – ASCE/SEI 7-22
✓ Highway bridges – AASHTO, LRFD 2020
✓ Wood building structures - ANSI / AF & PA NDS – 2018

o The common feature of current codes is that they provide guidance on determining design loads and design
load carrying-capacities of structural members

o Safety reserve is implemented through conservative load and resistance factors applied to the design loads
and design load-carrying capacities
 Role of a code in the building design process
o The building process includes o The major parties can be put into four categories
❖ Planning ✓ Owner / investor

❖ Design ✓ Designer

❖ Manufacturing of materials ✓ Contractor

✓ User / operator
❖ Transportation
❖ Construction
❖ Operation / use
❖ demolition
 Code Levels
o Level 1 Code
❖ These codes use deterministic design formulas

❖ The safety margin is introduced through central safety factors (ratio of design
resistance to design load) or partial safety factors (load and resistance factors)

❖ A type of equation used in a Level I code is shown below.

factored nominal resistance ≥ total factored nominal load

𝜙𝑅𝑛 ≥ ෍ 𝜆𝑖 𝑄𝑛𝑖
 Code Levels

𝜙𝑅𝑛 ≥ ෍ 𝜆𝑖 𝑄𝑛𝑖

𝑓𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦 𝑓𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦

𝑄 𝑅
𝜇𝑄 𝑄𝑖 𝜆𝑄𝑖 𝜙𝑅𝑛 𝑅𝑛 𝜇𝑛
mean load Factored Reduced mean load
load Resistanc
Design e Design
load Resistanc
e
 Calibration of partial safety factors
for a level I Code
o Level 1 Code in the LRFD
𝜙𝑅 ≥ ෍ 𝛾𝑖 𝑄𝑖
o Design equations following this format contain partial safety factors f and g
that must be calibrated based on the target reliability index adopted by the
code

o To determine the partial safety factors, a target  is specified and then the
required mean values of the resistance and loads to achieve the target is
calculated.
o This means that we need to find the design point {z*} corresponding to the
target .
Calibration of partial safety factors
for a level I Code
Calibration of partial safety factors
for a level I Code
Calibration of partial safety factors
for a level I Code
Data and Results
Bending strength parallel to fibers
 Experimental Set-up

𝐹𝑢𝑙𝑡 × 𝑎
𝑀𝑢𝑙𝑡 =
2
a = shear span
Load δ
𝑀𝑢𝑙𝑡 × 𝐷
Bambo 𝑓𝑚 = D
2𝐼𝐵
o
𝜋
𝐼𝐵 = 𝐷4 − 𝐷 − 2𝑡 4
D = Diameter
a a 64
δ = Culm wall thickness
Resistance Data
Resistance Data

Parameter Symbol Distribution  

Bending fb Normal 84.3 22.7

strength 2 1
Diameter D Normal 91.5 8.10
2
Culm thickness d 3-parameter 7.34 0.61
Weibull
 Goodness-of-fit tests
bending strength diameter Thickness
Distribution AD p-value Distribution AD p-value Distribution AD p-value
Normal 0.391 0.375 Normal 0.21 0.856 Normal 2.322 < 0.005
Lognormal 0.44 0.286 Lognormal 0.219 0.834 Lognormal 1.069 0.008
3-parameter Lognormal 0.291 * 3-parameter Lognormal 0.209 * 3-parameter Lognormal 0.251 *

Exponential 25.084 < 0.003 Exponential 39.017 < 0.003 Exponential 32.983 < 0.003

2-parameter 9.818 < 0.010 2-parameter 11.242 < 0.010 2-parameter 5.189 < 0.010
exponential exponential exponential
Weibull 0.441 > 0.250 Weibull 0.815 > 0.034 Weibull 3.865 < 0.010
3-parameter Weibull 0.212 > 0.500 3-parameter Weibull 0.177 > 0.500 3-parameter Weibull 0.378 0.43
Smallest extreme value 1.666 < 0.010 Smallest extreme value 1.278 < 0.010 Smallest extreme value 6.144 < 0.010

Largest extreme value 0.556 0.164 Largest extreme value 0.878 0.023 Largest extreme value 0.391 0.250
Gamma 0.294 > 0.250 Gamma 0.211 > 0.250 Gamma 1.426 < 0.005
3-parameter Gamma 0.28 * 3-parameter Gamma 0.21 * 3-parameter Gamma 0.296 *
Logistic 0.551 0.112 Logistic 0.329 > 0.250 Logistic 1.432 < 0.005
Loglogistic 0.533 0.132 Loglogistic 0.334 > 0.250 Loglogistic 0.779 0.024
Target Reliability
ASCE 7-22 Minimum Design Loads and Associated Criteria for Buildings
Load Distribution and Parameters

ASCE 7-22 Minimum Design Loads and Associated Criteria for Buildings
Limit State Equation

𝑓𝑏 𝐼
𝐺= − 0.125 𝜌𝐷 + 𝜌𝐿 𝑠ℓ2
𝑐
where
I = moment of inertia
D = diameter
c = D/2
rD = dead load
rL = live load
s = spacing
Load distribution and parameters

𝑤ℓ2
𝑀𝑄 =
8
where:
𝑤 = 𝜌 × 𝑠𝑝𝑎𝑐𝑖𝑛𝑔 𝑜𝑓 𝑓𝑙𝑒𝑥𝑢𝑟𝑎𝑙 𝑚𝑒𝑚𝑏𝑒𝑟𝑠
𝑤 = uniformly distributed load (kN/m)
𝐿 = length of individual span
𝜌 =uniform design load (kPa)

Harries K. A., Trujillo, D., Kaminski, S., & Lopez, L. F. (2022). Development of load tables for design of full-culm bamboo. European Journal of
Wood and Wood Products, 80(3), 621–634.
Partial safety factors for Bending
Sanchez (2019) Pfailure β γ
0.0026 2.79 1.20 Proposed γ
β γ 0.00233 2.83 1.25
0.00195 2.89 1.30 MCS 1.5 - 2.0
1.0 1.0752
0.0018 2.91 1.35 Sanchez, 2019 1.69 - 2.13
2.0 1.5151 0.00172 2.92 1.40 ISO 22156 2.00
2.33 1.6949 0.00143 2.98 1.45
3.0 2.1277 0.00139 2.99 1.50
3.5 2.5641 0.00115 3.05 1.55
4.0 3.030 0.00109 3.06 1.60
0.00103 3.08 1.65
0.00095 3.10 1.70
0.00094 3.11 1.75
0.0008 3.16 1.80
0.00074 3.18 1.85
0.00069 3.20 1.90
0.00065 3.22 1.95
0.00059 3.24 2.00
ISO 22156 (ASD)
𝟏
𝑓𝑚 = 𝑓𝑖𝑘 𝐶𝑅 𝐶𝐷𝐹 𝐶𝑇
𝑭𝑺𝒎

𝑓𝑖𝑘 = Characteristic Strength as per ISO 12122-1

𝐶𝑅 = 0.9 Non-redundant elements
𝐶𝐷𝐹 = 0.65 Transient Loads Class 2
𝐶𝑇 = 1.00 Class 2 Temp. ≤ 38 °C
𝐹𝑆𝑚 = 2.0 𝐹𝑎𝑐𝑡𝑜𝑟 𝑜𝑓 𝑆𝑎𝑓𝑒𝑡𝑦 𝑓𝑜𝑟 𝐵𝑒𝑛𝑑𝑖𝑛𝑔
ISO 12122-1 (5th percentile)
Total factor of safety = 3.42 𝑋𝑖𝑘 = 𝑋0.05,0.75 = 𝑋0.05 1 −
𝑘0.05,0.75𝑉
𝑛
 Calculations
Formulate the limit state function:
𝑔=𝑅−𝑆
𝑔(𝑓𝑀 , 𝐷, 𝐿) = 𝑓𝑀 − 𝐷 − 𝐿
𝛾𝑓𝑀 𝑓𝑀 ≥ 𝛾𝐷 𝐷 + 𝛾𝐿 𝐿

where 𝛾𝑓𝑀 , 𝛾𝐷 𝑎𝑛𝑑 𝛾𝐿 are Load and Resistance Factors

Calculate the equivalent normal parameters:

𝜇𝑥𝑒 = 𝑥 − 𝜎𝑥𝑒 Φ−1 𝐹𝑥 𝑥

𝑒
1
𝜎𝑥 = 𝜙 Φ−1 𝐹𝑥 𝑥
𝑓𝑥 𝑥
Calculations
In the study conducted:
𝛾𝑓𝑀 𝑓𝑀 ≥ 𝛾𝐷 𝐷 + 𝛾𝐿 𝐿
𝛾𝑓𝑀 = 0.8479 𝛾𝐷 = 1.255 𝛾𝐿 = 1.6028
1
= 𝐹𝑀 = 1.1794
𝛾𝑓𝑀
𝐹𝑀 ≈ 1.2

ISO 22156 for LRFD:

𝐹𝑀 = 1.5
Summary
 Summary

o The basic theory of FORM was discussed with a hypothetical example

o The use of reliability theory to determine partial safety factors using the
bending strength test results of bamboo was also presented

o The study can be improved if more tests are conducted

o Other strength test results such as compression and shear parallel to grain can
also be used for future studies
 References
o Probability Concepts in Engineering: Emphasis on Applications to Civil and
Environmental Engineering, 2nd Ed. (2007), Ang, A. H-S and Tang, W. H.

o Computer Intensive Methods, Coles, S. (2002)

o Reliability of Structures (2000), Nowak, A. and Collins, K., McGraw-Hill Book

Co., Singapore

o Exploring Monte Carlo Methods (2012), Dunn, W. and Shultis, J.

Thank you for your attention!
End of Presentation!