ENP-Constantine 3rd year Mechanical engineering Mechanical engineering Department

Chapter II: Bending deformation of isostatic and hyperstatic beams

II.1 Introduction:
The search for the elastic line or deformed line of a primarily straight beam with constant or
variable cross-section, primarily subjected to bending, constitutes one of the fundamental problems
of strength of materials. The centroidal axis of the beam, the locus of gravity centers, situated in
the plane of the neutral fibers, bends and undergoes a deformation without any change in length.
Shear deformations are typically very small and negligible compared to bending deformations.
II.2 Hyperstatic beams:
Hyperstatic beam refers to beams whose reactions at the supports cannot be determined solely by
the equations of statics. The degree of hyperstability of the beam is equal to the number of
redundant unknowns compared to the three equilibrium equations of statics. Examples of
hyperstatic systems are numerous, with many engineering structures being hyperstatic, such as self-
supporting frames, continuous beams supported at multiple points, etc.

Figure 1: Hyperstatic beams

II.3 Deformation by analytical integration
After deformation, the centroidal line of the beam becomes a plane curve in the case of a problem
within a plane, or a spatial curve in the case of any deflected bending.
II.3.1 Slope and displacement by integration
The equation of the elastic curve will be defined by the coordinates y and x. And so, to find the
𝟏
deflection 𝒚 = 𝒇(𝒙) we must be able to represent the curvature (𝝆) in terms of y and x. The
curvature is shown that this relationship is
𝒅𝟐 𝒚⁄ 𝒅𝟐 𝒚⁄
𝟏 𝒅𝒙𝟐 𝑴 𝒅𝒙𝟐
= 𝟑 we have = 𝟑
𝝆 𝒅𝒚 𝟐 𝑬𝑰 𝒅𝒚 𝟐
[𝟏+( ) ]𝟐 [𝟏+( ) ]𝟐
𝒅𝒙 𝒅𝒙

Apart from a few cases of simple beam geometry and loading, this equation is difficult to solve,
because it represents a nonlinear second-order differential equation. Fortunately, it can be modified,
because most engineering design codes will restrict the maximum deflection of a beam or shaft.
1
 ENP-Constantine 3rd year Mechanical engineering Mechanical engineering Department

𝒅𝒚
Consequently, the slope of the elastic curve, which is determined from , will be very small, and
𝒅𝒙
its square will be negligible compared with unity. Therefore, the curvature can be approximated by
𝟏 𝒅𝟐 𝒚 𝑴 𝒅𝟐 𝒚
= . With this simplification, can now be written as 𝑬𝑰 =
𝝆 𝒅𝒙𝟐 𝒅𝒙𝟐

It is also possible to write this equation in two alternative forms. If we differentiate each side with
𝒅𝟐 𝒚
𝒅𝑴 𝒅(𝑬𝑰 𝟐 )
respect to x and substitute 𝑽 = , we get: 𝑽 (𝒙) = 𝒅𝒙
𝒅𝒙 𝒅𝒙

𝒅𝟐 𝒚
𝒅𝑽 𝒅𝟐 (𝑬𝑰 )
𝒅𝒙𝟐
Differentiating again, using 𝒘 = , yields: 𝒘 (𝒙) =
𝒅𝒙 𝒅𝒙𝟐

For most problems the flexural rigidity (EI) will be constant along the length of the beam.
Assuming this to be the case, the above results may be reordered into the following set of three
equations:
𝑑4 𝑦
𝐸𝐼 4 = 𝑤(𝑥)
𝑑𝑥
𝑑3𝑦
𝐸𝐼 3 = 𝑉(𝑥)
𝑑𝑥
𝑑2𝑦
{𝐸𝐼 𝑑𝑥 2 = 𝑀(𝑥)
II.3.2 Boundary Conditions:
Solution of any of these equations requires successive integrations to obtain y. For each integration,
it is necessary to introduce a “constant of integration” and then solve for all the constants to obtain
a unique solution for a particular problem. For example, if the distributed load w is expressed as
𝒅𝟒 𝒚
a function of x and 𝑬𝑰 𝒅𝒙𝟒 = 𝒘(𝒙) is used, then four constants of integration must be evaluated;
however, it is generally easier to determine the internal moment M as a function of x and use
𝒅𝟐 𝒚
𝑬𝑰 𝒅𝒙𝟐 = 𝑴(𝒙), so that only two constants of integration must be found.

Most often, the integration constants are determined from boundary conditions for the beam, Table
1. As noted, if the beam is supported by a roller or pin, then it is required that the displacement be
zero at these points. At the fixed support, the slope and displacement are both zero.

2
 ENP-Constantine 3rd year Mechanical engineering Mechanical engineering Department

Table 1: Supports in beam and their boundary conditions

Supports in beam Boundary conditions

*y = 0
Roller

*y = 0
Pin

*y = 0
Roller

*y = 0
Pin

*y = 0
*θ=0
Fixed end

II.3.3 Continuity Conditions:

Since beams can support portions of a distributed load and concentrated forces and couple
moments, the internal shear and moment functions of x will be discontinuous, or their slopes
will be discontinuous, at points where the loads are applied. Because of this, these functions
must be determined for each region of the beam between any two discontinuities of loading. It
consists of a series of several distributed and concentrated loads, figure 2 (a), then several functions
must be written for the internal moment, each valid within the region between two discontinuities.
For example, the internal moment in regions AB, BC, and CD can be written in terms of the x1, x2,
and x3 coordinates selected as shown in figure 2 (b).
When each of these functions is integrated twice, it will produce two constants of integration, and
since not all of these constants can be determined from the boundary conditions, some must be
determined using continuity conditions. For example, consider the beam in figure 3. Here two x
coordinates are chosen with origin at A. Once the functions for the slope and deflection are
obtained, they must give the same values for the slope and deflection at point B so the elastic curve
is physically continuous. Expressed mathematically, these continuity conditions are u1(a) = u2(a)
and y1(a) = y2(a). They are used to evaluate the two constants of integration. Once these
functions and the constants of integration are determined, they will then give the slope and
deflection (elastic curve) for each region of the beam for which they are valid.

3
 ENP-Constantine 3rd year Mechanical engineering Mechanical engineering Department

Figure 2 Figure 3
• Example:
The beam shown in figure 4 a supports the triangular distributed loading. Determine its maximum
deflection. EI is constant.

figure 4
II.4 Method of superposition
𝒅𝟒 𝒚
The differential equation 𝑬𝑰 𝒅𝒙𝟒 = 𝒘(𝒙) satisfies the two necessary requirements for applying the
principle of superposition; i.e., the load w(x) is linearly related to the deflection y(x), and the load
is assumed not to significantly change the original geometry of the beam or shaft. As a result, the
deflections for a series of separate loadings acting on a beam may be superimposed. For example,
if y1 is the deflection for one load and y2 is the deflection for another load, the total deflection for
both loads acting together is the algebraic sum y1 + y2. Using tabulated results for determinate the
deflection.

4
ENP-Constantine 3rd year Mechanical engineering Mechanical engineering Department

5
 ENP-Constantine 3rd year Mechanical engineering Mechanical engineering Department

• Example:
Determine the displacement at point C and the slope at the support A of the beam shown in Figure
5. EI is constant.

Figure 5
II.5 Conjugate beam method
This method allows for the simple calculation, in the case of concentrated or uniformly distributed
loads, or couples, of the linear deformation and rotation of the beam at particular points without
using integral sums. The principle of this calculation was developed by Mohr and is called the
conjugate beam method.
II.5.1 Method exposition
This method is based on the analogy between the differential relationships of the bending moment
and linear deformation:

2. Deformation of a beam subjected to a

1. Bending moment of a beam stressed by a
bending moment
continuous load w

𝑑2 𝑀 𝑑 2 𝑦 𝑀(𝑥)
= 𝑤(𝑥) =
𝑑𝑥 2 𝑑𝑥 2 𝐸𝐼

These two second-order differential equations are of the same type; hence their solutions are
similar.
II.5.2 Principle:
The calculation of the deflection 𝒚(𝒙) can be replaced by that of a virtual moment Mv of a so-called
conjugate beam subjected to:
1. For straight beams with constant flexural rigidity, 𝑬. 𝑰𝒛 = 𝒄𝒐𝒏𝒔𝒕𝒂𝒏𝒕, by the actual
bending moment Mf.

6
 ENP-Constantine 3rd year Mechanical engineering Mechanical engineering Department

2. For a beam with variable flexural rigidity, E = constant and Iz variable, by the ratio of the
actual bending moment to the sectional quadratic moment 𝑴𝒇 /𝑰𝒛 .

Real beam Conjugate beam Surface and area

Figure 6: Common real and conjugate beams, position of the centers of gravity
The conjugate beam method is primarily used to calculate the deformations of beams with constant
cross-sections. The virtual beam is loaded by 𝒘𝒗 = 𝑴𝒇 . The efforts in the beam will be divided
by the product 𝑬𝑰𝒛 to find the linear and angular deformations.
II.5.3 Supports of the conjugate beam
The conjugate beam must have joints or fixed supports corresponding to the deformation conditions
imposed by the actual beam. The rules to follow to obtain a compliant conjugate beam are as
follows:
1. A hinged support at the end of the actual beam remains a hinged support on the conjugate
beam.
2. A hinged support with a cantilever transforms into a hinge.
3. A fixed end on the actual beam becomes a free end on the conjugate beam.
4. A free end on the actual beam becomes a fixed support on the conjugate beam.
5. A hinge, without support, on the actual beam transforms into a support on the conjugate
beam.

7
 ENP-Constantine 3rd year Mechanical engineering Mechanical engineering Department

II.5.4 Center of gravity of areas

When the straight beam is subjected to concentrated forces, the bending moment diagram consists
of a polygonal line. If the load is uniformly distributed, either in sections or interspersed with
concentrated forces, the bending moment diagram is composed of arcs of second-degree parabolas.
Finally, if the load is distributed linearly, the bending moment diagram exhibits one or more arcs
of third-degree parabolas.
The positions of the centers of gravity for the fundamental cases are provided in Figure 6.
Decomposing the area of the bending moment diagram into fundamental areas makes it easy to
find the forces on the lines of action of the forces or at the end of the beam, on the actual supports,
or potentially the position of maximum deformation.
II.5.5 Constant section beams
The examples provided are simple and fundamental applications of the method in cases of loads
on straight beams.
II.5.5.1 Concentrated force F at an arbitrary point between the two supports.
𝑭𝒄 𝑭𝒃
The support reactions are 𝑹𝑩 = and 𝑹𝑪 = . The maximum bending moment occurs on the
𝒍 𝒍
𝑭𝒃𝒄
line of action of force F and is equal to: 𝑴𝒇𝒎𝒂𝒙 = 𝒍

Figure 7: Beam on two supports without overhang and corresponding virtual load.
1 𝑙+𝑐 𝑙 𝐹𝑏𝑐 𝐹𝑏𝑐(𝑙+𝑐)
• Virtual support reactions: 𝑅𝐵𝑣 = ( 𝑙 ) ( 3 ) (2) ( 𝑙 ) = ( 6𝑙 )
1 𝑙 + 𝑏 𝑙 𝐹𝑏𝑐 𝐹𝑏𝑐(𝑙 + 𝑏)
𝑅𝐶𝑣 = ( ) ( )( )( )=( )
𝑙 3 2 𝑙 6𝑙

𝐹𝑏𝑐(𝑙+𝑐) 𝐹𝑏𝑐(𝑙+𝑏)
• Beam angles: 𝜃𝐵 = and 𝜃𝐶 =
6𝐸𝐼𝑧 𝑙 6𝐸𝐼𝑧 𝑙
𝑏𝐹𝑏𝑐(𝑙+𝑐) 𝑏 𝑏 𝐹𝑏𝑐 𝐹𝑏 2 𝑐 2
• Virtual bending moment: 𝑀𝑓𝑣𝐹 = ( ) − (3 ) (2 ) ( )=
6𝑙 𝑙 3𝑙
𝐹𝑏 2 𝑐 2
• Maximum deformation: 𝑦𝐹 = 𝑓 = 3𝐸𝐼𝑧 𝑙

8
 ENP-Constantine 3rd year Mechanical engineering Mechanical engineering Department

II.5.6 Variable cross-section members

If the straight beam has a variable cross-section, the virtual load to be applied on the conjugate
beam is given by 𝒒𝒗 = 𝑴𝒇 /𝑰𝒛 , which is the ratio of the actual bending moment divided by the
sectional quadratic moment. The principle of the solution remains the same as before. Mohr had
proposed a graphical method to find the value of angular deformations at the supports and
maximum or minimum linear deformations. This method uses the funicular polygon. One of the
major difficulties in applying this method is to correctly determine the magnitude of deformations
considering the scales introduced for the representation of the member, the loads, and the polar
distances.
II.5.6.1 Numerical Mohr method
We propose to adapt Mohr's fundamental method to the current conditions of numerical calculation.
The member with a variable cross-section is represented at a suitable scale, and the equilibrium of
the member allows finding the magnitude of the bending moment at each section. For a statically
determinate member on two hinged supports B and C, subjected to concentrated forces applied
between the supports, the numerical method consists of successively calculating the following
quantities:
1. Virtual linear load: it is found by dividing the actual bending moment by the sectional
quadratic moment of the considered section: 𝒘𝒗𝒊 = 𝑴𝒇𝒊 /𝑰𝒛𝒊.
2. Decomposition of the member into sections with constant quadratic moment, and linearly
variable bending moment. The initial point on the left is designated as 0, and the end of
each section is numbered in ascending order, see Figure 8.
3. Representation of the virtual beam with the virtual linear load: on each section, the virtual
linear load is: on the left 𝒘𝒗𝒊𝒈 , on the right 𝒘𝒗𝒊𝒅 .
4. Calculation of the virtual load on each section: it is found by the area of each trapezoid:
(𝒙𝒊 − 𝒙𝒊−𝟏 )(𝒘𝒗𝒊𝒈 + 𝒘𝒗𝒊𝒅 )
𝑭𝒗𝒊 = 𝑨𝒗𝒊 = 𝟐
. The total virtual load is the algebraic sum of the
component virtual loads 𝜮 𝑭𝒗𝒊.
5. Determination of the virtual support reactions: using the relation for calculating the abscissa
of the center of gravity of the trapezoid, as shown in Figure 8 on the right, the virtual support
reaction RCv is equal to the sum of the static moments of surface with respect to point B
divided by the distance to the support l:
𝑛
1 𝒘𝒗𝒊𝒈 + 𝟐 𝒘𝒗𝒊𝒅 𝒘𝒗𝒊𝒈 + 𝒘𝒗𝒊𝒅
𝑅𝐶𝑣 = . ∑ [𝑥𝑖−1 + (𝑥𝑖 − 𝑥𝑖−1 ). ] . (𝑥𝑖 − 𝑥𝑖−1 ).
𝑙 3(𝒘𝒗𝒊𝒈 + 𝒘𝒗𝒊𝒅 ) 2
𝑖=1

9
 ENP-Constantine 3rd year Mechanical engineering Mechanical engineering Department

The virtual support reaction RBv will be calculated using the same relation, adopting
positive abscissas to the left from point C. Checking the sum 𝜮 𝑭𝒗 = 𝟎 will verify the
accuracy of the calculation.
6. The virtual shear force in the conjugate beam is proportional to the angular deformation of
the beam. At a section located at abscissa xi, the angular deformation is found by:
𝑘
1
𝜃𝑥𝑖 = . (𝑅𝐵𝑣 − ∑ 𝐴𝑣𝑗𝑔 )
𝐸
𝑗=1
7. The virtual bending moment in the conjugate beam is proportional to the linear deformation
at that point. At a section located at abscissa xi, the linear deformation can be found by the
expression:
𝑘
1
𝑦𝑥𝑖 = . (𝑥𝑖 𝑅𝐵𝑣 − ∑(𝑥𝑖 − 𝑥𝑗 + 𝑥𝑐𝑗 ). 𝐴𝑣𝑗𝑔 )
𝐸
𝑗=1

with xcj the distance to the center of gravity of the partial surface, measured to the left.

Figure 8: Beam with variable section

II.6 Continuous hyperstatic beams
If all the support reactions of a straight beam primarily subjected to bending are found from the
equations of static equilibrium, the problem is said to be statically determined and the beam is said
to be statically determinate. When the equations of statics do not allow finding the equilibrium of
10
 ENP-Constantine 3rd year Mechanical engineering Mechanical engineering Department

the member, the system is said to be hyperstatic, and the problem is statically indeterminate. The
degree or order of hyperstaticity is determined by the number of unknowns that cannot be found
from the conditions of static equilibrium.
Let's recall the following properties: a hinged support imposes only a concentrated force as a
support reaction, while a fixed support involves both a couple and a force. To find the equilibrium
of a body subjected only to parallel forces, two equilibrium equations are required, either one or
two moment sums, or a force sum and a moment sum.
II.6.1 Straight beam on two or three supports
To solve a statically indeterminate continuous beam problem in bending, several methods are
available to the engineer. In this subsection, we will discuss the principle of superposition of
deformations and forces.
II.6.1.1 Fixed-end beam on two supports
The straight beam, with a constant section, is fixed at one end, supported at the other by a
deformable hinged support. The load on the beam can be arbitrary: concentrated forces and/or
distributed loads. The solution method involves the following operations:
1. Search for static equilibrium.
2. Removal of the hinged support and calculation of the linear deformation at this support
under the effect of the load applied to the beam.
3. Application of the support reaction to eliminate the deformation at the hinged support.
4. Superposition of forces and deformations to find the resultant quantities.

11
 ENP-Constantine 3rd year Mechanical engineering Mechanical engineering Department

II.6.1.2 Beam fixed at both ends

The straight beam with a constant section is fixed at both ends. However, these two fixations serve
only to maintain the direction of the deformation along the original axis at these points. Therefore,
the beam can move axially. The load on the beam can be concentrated and/or distributed. The
solution method involves:
1. Searching for static equilibrium and determining the degree of hyperstaticity.
2. Removing the two fixations and finding the angular deformation at the two assumed hinged
supports under the effect of the load.
3. Loading the beam by the two fixing moments and calculating the corresponding angular
deformations to eliminate those produced on the hinged beam.
4. Superposition of forces and deformations to find the resultant quantities.

II.6.1.3 Beam on three hinged supports

The beam with a constant section on three hinged supports, the first of which can be fixed, and the
other two on rollers. The load on this continuous beam can be either concentrated and/or
distributed. The solution method involves:
1. Finding the static equilibrium of the hyperstatic beam.
2. Removing the central support and determining the deformation of this member on two
hinged supports, on the line of action of the central support, while keeping the original load.
3. Calculating the deformation of the beam under the effect of the central support reaction to
straighten the member and impose its original shape.

12
 ENP-Constantine 3rd year Mechanical engineering Mechanical engineering Department

4. Superimposing the forces and deformations to find the resultant quantities.

II.7 Equation of the three moments

The equation of the three moments, demonstrated by Clapeyron, is applicable to continuous beams
with constant flexural rigidity in each span. It is a very elegant method for calculating the bending
moments at the supports and finding the successive support reactions.
II.7.1 Conventions and substitutions
A straight beam on two hinged supports is statically determinate. A continuous straight beam on n
hinged supports is hyperstatic of order n-2. To solve this kind of hyperstatic problem, it is necessary
to establish n-2 deformation relations in order to determine the initially indeterminate values. To
correctly apply the equation of the three moments, the angular deformations and bending moments
at the supports are assumed to occur as follows:
1. The rotations of the sections at the supports occur inward toward the spans.
2. Bending moments are counted as positive if they are directed inward toward the spans.
3. The equation of the three moments is still applicable when the continuous beam has one or
two cantilever supports or when one or both ends are fixed.

Figure 9 Sign conventions for rotations and moments at supports.

13
 ENP-Constantine 3rd year Mechanical engineering Mechanical engineering Department

• Cantilever section: This section is not included in the equation of the three moments. The
cantilever section is replaced by the moment of the load and by the load at the cantilever.
This operation amounts to reducing, at the support, the efforts applied on the cantilever
section.
• End section fixed: Since the rotation of the beam at the fixed end is zero, this particular
condition can be simulated by replacing the fixed end with a fictitious span of zero length.
If a couple is applied at a support, this effort can be transferred to the left span or the right span, or
shared between the two spans simultaneously by arbitrarily decomposing it.
II.7.2 Formulation
The equation of the three moments is based on the fact that at the supports, the bending moment
does not change, unless there is an external couple, and that the beam is continuous. The inclination
of the deformed line is identical on the left and right of each hinged support. The continuous beam
is decomposed into simple beams hinged at each support and subjected at these points to moments
of unknown forces.
Let's consider a continuous beam on n supports supporting concentrated loads, distributed loads,
or couples. Let's try to find the angular deformation of this member at a support. Let the two
successive spans be of lengths li-1 and li, and let i-1, i, i+1 be the three hinged supports. The bending
moments at the supports are respectively Mi-1, Mi, and Mi+1. The inclination of the deformation at
point i must be the same for the two sections. This rotation is equal to the difference between the
deformation due to the actual applied load and that caused by the action of moments Mi-1 and Mi
on the section (i-1) → i. Taking into account the sign adopted for the moments, inward toward the
considered section, the angle to the left of the beam is given by:
𝑅𝑣𝑔 𝑀𝑖−1 𝑙𝑖−1 𝑀𝑖 𝑙𝑖−1
+ + = 𝑅𝑜𝑡𝑎𝑡𝑖𝑜𝑛 𝑡𝑜 𝑡ℎ𝑒 𝑙𝑒𝑓𝑡 𝑜𝑓 𝑠𝑢𝑝𝑝𝑜𝑟𝑡 𝑖.
𝐸𝐼𝑧𝑔 6𝐸𝐼𝑧𝑔 3𝐸𝐼𝑧𝑔

Let's express the same characteristics for the rotation to the right of support i:
𝑅𝑣𝑑 𝑀𝑖 𝑙𝑖 𝑀𝑖+1 𝑙𝑖
+ + = 𝑅𝑜𝑡𝑎𝑡𝑖𝑜𝑛 𝑡𝑜 𝑡ℎ𝑒 𝑟𝑖𝑔𝑡ℎ 𝑜𝑓 𝑠𝑢𝑝𝑝𝑜𝑟𝑡 𝑖.
𝐸𝐼𝑧𝑑 3𝐸𝐼𝑧𝑑 6𝐸𝐼𝑧𝑑
In these two expressions, Rvg and Rvd represent the virtual support reactions of the conjugate beam
corresponding to the left segment or the right segment. Since the beam must remain continuous,
we can set them equal to each other:
𝑅𝑣𝑔 𝑀𝑖−1 𝑙𝑖−1 𝑀𝑖 𝑙𝑖−1 𝑅𝑣𝑑 𝑀𝑖 𝑙𝑖 𝑀𝑖+1 𝑙𝑖
+ + = −( + + )
𝐸𝐼𝑧𝑔 6𝐸𝐼𝑧𝑔 3𝐸𝐼𝑧𝑔 𝐸𝐼𝑧𝑑 3𝐸𝐼𝑧𝑑 6𝐸𝐼𝑧𝑑

14
 ENP-Constantine 3rd year Mechanical engineering Mechanical engineering Department

By writing the terms containing the moments at the supports on the left side, and the terms
containing the virtual shear forces on the conjugate beams on the right side, the general relationship
is expressed as:
𝑙𝑖−1 𝑙 𝑙 𝑙 𝑅 𝑣𝑔 𝑅𝑣𝑑
𝑀𝑖−1 . + 2 𝑀𝑖 . ( 𝐼𝑖−1 + 𝐼 𝑖 ) + 𝑀𝑖+1 . 𝐼 𝑖 = −6 ( 𝐼 + )
𝐼𝑧𝑔 𝑧𝑔 𝑧𝑑 𝑧𝑑 𝑧𝑔 𝐼𝑧𝑑

This general relationship of the equation of three moments simplifies if the quadratic moments
are the same in both segments:
𝑀𝑖−1 . 𝑙𝑖−1 + 2. 𝑀𝑖 . (𝑙𝑖−1 + 𝑙𝑖 ) + 𝑀𝑖+1 . 𝑙𝑖 = −6(𝑅𝑣𝑔 + 𝑅𝑣𝑑 )

This is the equation of three moments. The simplification by 𝑰𝒛𝒈 = 𝑰𝒛𝒅 is obviously possible only
if the two quadratic moments are constant and equal in both segments or spans.
II.7.3 Equilibrium and forces
If n is the number of articulated supports, it is necessary to establish n-2 equations of the three
moments by choosing each time a pair of spans. Solving the system of linear equations allows
finding the bending moments at all supports. The static equilibrium of each span is calculated by
introducing the applied load on the span and the bending moments at the ends of the isolated
section. The total support reaction RBi on the continuous beam at point i is equal to the sum of the
partial support reactions RBig and RBid from the two adjacent spans. In general, the support reaction
is found by the expression:
𝑴𝒊−𝟏 − 𝑴𝒊 𝑴𝒊+𝟏 − 𝑴𝒊
𝑹𝑰 = 𝑹𝑰𝒈 + 𝑹𝑰𝒅 = 𝑹′𝑰𝒈 + 𝑹′𝑰𝒅 + +
𝒍𝒊−𝟏 𝒍𝒊

with:
• R'Ig: reaction on support i due to the load on the left span.
• R'Id: reaction on support i due to the load on the right span.
• (Mi-1 – Mi)/li-1: support reaction due to the moments Mi-1 and Mi on the left span.
• (Mi+1 – Mi)/li: support reaction due to the moments Mi+1 and Mi on the right span.
To construct the bending moment diagram, it is sufficient to build the moment diagrams in each
span assuming the continuous beam articulated at each support. At the supports, plot the moments
Mi, positive upwards, negative downwards, and connect these values to the supports with straight
line segments. The area between the two lines represents the resulting bending moment diagram.
II.8 Matrix transfer method
The matrix transfer method was developed by Falk. It allows for the simultaneous determination
of forces Vi and bending moments Mfi, as well as deformations θi and yi in a straight member with

15
 ENP-Constantine 3rd year Mechanical engineering Mechanical engineering Department

variable section subjected to concentrated forces, distributed loads, or couples. It can determine
these quantities for both statically determinate and statically indeterminate structures.
II.8.1 Equation formulation
The principle of the method involves dividing the straight member into various elements:
1. Span: a portion of the member between two consecutive supports.
2. Segment: a portion of the span subjected only to a uniform distributed load w, or no-load
w, having a constant flexural rigidity EIz, along which the shear force V, bending moment
Mf, rotation of the section θ, and linear deformation y do not undergo any discontinuity.
3. Boundary: a particular cross-section, always located between two segments. This
additional concept allows for the introduction of discontinuities in load, rigidity,
concentrated forces, or couples.
To determine the deformation of the member and the values of unknown forces, the method relies
on the equilibrium of each segment and the continuity of the member. In this chapter, we will first
address statically determinate members with variable sections in a single span, subjected to forces
located in one plane.
We can express the expression in general form between points 0 and 1:

𝑙2 𝑙3 𝑤𝑙 4
1 𝑙 − −
(2𝐸𝐼z ) (6𝐸𝐼z ) (24𝐸𝐼z )
𝑦 𝑦
𝜃 𝑙 𝑙2 𝑤𝑙 3 𝜃
0 1 − −
𝑀f = (𝐸𝐼z ) (2𝐸𝐼z ) (6𝐸𝐼z ) ⋅ 𝑀f .
𝑉 𝑤𝑙 2 V
0 0 1 𝑙 −
{ 1 }1 2 { 1}
0 0 0 1 −𝑤𝑙
(0 0 0 0 1 )
{𝐕𝟏 } [𝑀1 ] {𝐕𝟎 }

Let V0 and V1 denote the vectors on the left and right of the segment, and let M1 be the square
matrix, called the segment matrix, the final vector on the right is given by:
𝑽𝟏 = 𝑴𝟏 . 𝑽𝟎
This matrix expression is valid for any segment of known dimensions, with determined initial load
and deformation.

16
 ENP-Constantine 3rd year Mechanical engineering Mechanical engineering Department

II.8.2 Final vector

For an elastic system composed of n segments, the matrix equations defining the whole are written
successively:
𝑽𝟏 = 𝑴𝟏 . 𝑽𝟎
𝑽𝟐 = 𝑴𝟐 . 𝑽𝟏 = 𝑴𝟐 . 𝑴𝟏 . 𝑽𝟎
𝑽𝟑 = 𝑴𝟑 . 𝑽𝟐 = 𝑴𝟑 . 𝑴𝟐 . 𝑴𝟏 . 𝑽𝟎
𝑽𝒏 = 𝑴𝒏 . 𝑽𝒏−𝟏 = 𝑽𝟏 = 𝑴𝒏 . 𝑴𝒏−𝟏 … … 𝑴𝟐 . 𝑴𝟏 . 𝑽𝟎
To find the vector of forces and deformations to the right of a segment, the vector of forces and
deformations to the left of the segment is pre-multiplied by the segment matrix.

Figure 10 Sign conventions and division of a piece into segments.

17