2-CYCLE MOMENT DISTRIBUTION FOR THE ANALYSIS

OF CONTINUOUS BEAMS AND MULTI-STOREY FRAMED

STRUCTURES

Chan Kam Leong

B.Sc(Eng.), MSc, P.Eng, C.Eng, MICE, MIStruct E, MIEM, MACEM K.L. Chan & Associates, Consulting Engineers
E-mail: klchanassociates@yahoo.com

ABSTRACT
In the 2-Cycle Moment Distribution, moments are distributed twice regardless of the number of spans. Moments are carried over
first and are included with fixed-end moments before the distribution is made. Both dead load and total load are distributed
simultaneously to obtain critical moments at supports as well as spans. Besides, maximum moments in the columns can be
obtained fairly quickly.

Keywords: Adjustments, Balancing Moments, Carry-Over, Distribution Factor, Fixed-End Moment, Maximum Moment,
Midspan Moment, Minimum Moment, Support Moment

INTRODUCTION THE CONCEPT OF FIXED-END MOMENT

Rigid frame and continuous beam analysis is a very In Figure 1(a), beam AB deflects under a load P and the
important subject for the structural engineer. tangents at the ends will rotate through angles θA and θB.
The method of moment distribution was first introduced by In Figure 1(b), the end at A is restrained by a moment MAB
Professor Hardy Cross to his students at the University of and the angle change at A is smaller than θA.
Illinois (USA) in 1924 and published in 1929 [1]. It has been In Figure 1(c), when the angle changes are zero at both
considered one of the greatest contributions ever made to supports A and B, the beam AB is said to have fixed ends, and
structural theory. The method has provided a means to analyse the restraining moments are called fixed-end moments, MFAB
many types of continuous frames, which were formerly and MFBA.
designed by empirical rules or approximate methods, can be The fixed-end moment is very useful in beam design since
analysed with accuracy and comparative ease. it is independent of other members in the frame and also is a
Thirty years later, the Portland Cement Association of the major part of the actual end moment in the beam. One
USA published the 2-Cycle Moment Distribution Method in objective in frame analysis is to determine the minor correction
1959 [1]. This is not a new method and it has been tested over to the fixed-end moment to give the actual moment.
a period of years in the analysis of numerous building frames.
Its speed and accuracy certainly are of great assistance to
designers.
The author came to know about this method when he was
working for a consulting firm in London in 1972. He wrote an
article on the 2-Cycle Method which was published in the IEM
Bulletin in August 1978 [2] and Dr. Leong Tuck Wah of the
University of Singapore also wrote and published in the IEM
Bulletin in December 1978 [3].
It is considered necessary to revise the article. However, the
readers are assumed that they are familiar with the Hardy Cross
moment distribution method.
As the name implies, the 2-Cycle Moment Distribution
distributes moments twice regardless of the number of spans in
a continuous frame. Moments are carried over first and are
included with fixed-end moments before the distribution is
made. And both dead load (D.L.) and total load (T.L.) are
Figure 1: Beam with various degrees of restraint
distributed simultaneously to obtain critical moments at supports
as well as spans. Besides, maximum moments in the columns
can be obtained fairly quickly. Whereas in the Cross Method the STIFFNESS AND CARRY OVER FACTOR
process of distribution is repeated for many cycles in order to In Figure 2, beam AB of constant cross section is simply
bring balancing or carry-over moments to very small supported at B and fixed at A. A counter-clockwise rotation of
magnitudes. It can only obtain moments at supports but not at θB may be effected by applying a counter-clockwise moment of
spans. Besides, dead load and total load are distributed MBA at B, and this in turn induces a resisting moment MAB on
separately. The Cross Method is comparatively time consuming. the member at A.

Journal - The Institution of Engineers, Malaysia (Vol. 68, No. 1, March 2007) 27
 CHAN KAM LEONG

In Figure 3(a), the fixed-ended beam has central portion

sags (plus) and the outer portions hump (minus).
In Figures 3(b) and 3(c), the clockwise and counter-
clockwise rotation about a central support B, of a continuous,
fix-ended frame show that the beam sags on one side and
humps on the other side. It is also clear that the beam sags at
Figure 2: Moments in beam with one end fixed, other end being one end of a member because of joint rotation, it will hump at
rotated the opposite end. Similarly, under the action of a moment, a
column can be treated as a beam when viewed from the right
It can be proved that MBA= KθB and MAB= 1/2 MBA, where K and the same sign convention for beams can be applied.
is the stiffness of the member. For the members with constant
MOMENT DISTRIBUTION AT ONE JOINT
section, K equals 4EI , which is referred to as the absolute In Figure 4(a), the frame consists of four members fixed at
L their far ends. The moment U will rotate at joint B until the sum
I
value. A relative value of K = is preferred when E is of the resisting moments induced in the four members is equal
L to U. Since all members are rigidly connected at B, each
constant throughout a frame, where E is the modulus of member will rotate through the same angle at this joint.
elasticity, I is the second moment of inertia and L is the In Figures 4(b) and 4(c), joint B is being rotated clockwise by
member length. The two equations show that: an external moment U = 69kNm. Stiffnesses, distribution factors,
distributed moments and carry-over moments are as shown.
1. The stiffness K at B equals the moment at B required to give Figure 4(a) indicates that the clockwise rotation of joint B
B a unit rotation when A is fixed. creates a hump to the left, but a sag to the right. Therefore 19
2. The moment required to rotate B through a given angle is is negative but 18 is positive. There is also a sag at A and a
proportional to the stiffness K. hump at D; therefore the carried-over moments are +10 at A
3. Applying a moment MBA at B will induce at A a moment and -9 at D. In moment distribution, U is called the
MAB = 1/2 MBA. The factor of 1/2 is called the “carry-over “unbalanced moment” and is computed as the numerical
factor”. difference between adjacent fixed-end moments.
Viewing from the right, the vertical members (which
The concepts of stiffness and carry-over factor together become horizontal), Figure 4(a) indicates that the clockwise
with the concept of fixed-end moment are used in moment rotation of joint B creates a hump to the left, but a sag to the
distribution. right of B. Therefore 22 is negative and 10 is positive. There is
also a sag at E and a hump at C; therefore the carried-over
SIGN CONVENTIONS moments are +11 at E and -5 at C.
Fixed-end moments are considered negative on both sides Figure 4(d) shows a BMD drawn on the tension face of
of a joint. Hence, moments are negative in “humps” (tension in a member.
top) and positive in “sags” (tension in bottom).
MAXIMUM SUPPORT MOMENT
As the name implies, the 2-Cycle Moment Distribution only
‘distributes’ twice regardless of the number of spans in a
continuous beam or frame. It also distributes total load and dead
load simultaneously. This can be best illustrated by an example.
Let us consider the example of a 4-span continuous beam
as shown in Figure 5. The problem is to determine maximum
support moments. Figure 6 contains five groups of calculations
for moments at the supports.
To assume maximum moment at support A, place total load on
span AB and dead load on span BC as shown in Figure 6. As long
as support B is considered fixed, the end moments at B are 67.5 to
the left and 20 to the right. The difference is 47.5. When B is
released, the moment distributed to the left = 47.5 x 2/5, and the
moment carried over to A while A remains fixed is 47.5 x 2/5 x 1/2
= 9.5. The counter-clockwise rotation of joint B creates a hump in
the beam at A. Therefore the moment of 9.5 carried over to A is
negative. Joint B is then relocked in its new position. Now turn to
A which so far has been considered locked. The original fixed-end
moment is -67.5, but the release and rotation of B transfers an
additional moment to A, and at this stage the modified total fixed-
end moment is -67.5 – 9.5 = -77. Since there is no fixed-end
moment to the left of A, the unbalanced moment equals 77.
Releasing A and permitting it to rotate induces a distributed
Figure 3: Signs illustrated by means of curvature and deflection of
beams and columns
moment at A equals to 77 x 1 = 77. When joint A rotates clockwise

28 Journal - The Institution of Engineers, Malaysia (Vol. 68, No. 1, March 2007)
 2-CYCLE MOMENT DISTRIBUTION FOR THE ANALYSIS OF CONTINUOUS BEAMS AND MULTI-STOREY FRAMED STRUCTURES

0.35 carried over to B. When A and C are released they rotate so as

to create a hump on one side and a sag on the other side of the fixed
joint B, therefore 33.75 is negative and 0.35 is positive. While B is
still considered fixed, the modified total fixed-end moments at B
are -101.25 to the left and -29.65 to the right and out of balance at
B is -101.25 + 29.65 = -71.6. Release joint B, multiply 71.6 by the
distribution factors 2/5 and 3/5, and record the results 28.64 and
42.96 as shown. The counter clockwise rotation at joint B creates a
sag to the left (+28.64) and a hump to the right (-42.96). Distributed
or balancing moments at joints C, D and E are determined by the
same procedure as shown in Figure 6.
The operations illustrated in Figure 6 show that moments are
distributed twice. Moments are carried over first and are
included with fixed-end moments before the distribution is
made. In the previous calculations, the five groups have five
different arrangements of loads i.e. total loads on spans adjacent
to the particular joint at which maximum moments are computed
but dead loads only on the next outside spans. Yet the
calculations as shown in Figure 6 can be consolidated into one
single group without any interference as shown in Figure 5. It is
clear that the results as shown in Figure 6 between the 2-Cycle
Moment Distribution and computer analysis are quite close.

THE METHOD OF CALCULATING

MAXIMUM SUPPORT MOMENT:

Step 1 Write down D.F., MD.L & MT.L. ,

where M = - WL for a uniformly distributed load, W.

2

12
Step 2 Calculate and write down C/O
C/O = 1/2 x D.F x (MT.L - MD.L)

Step 3 Obtain total unbalanced moment (∑M) by adding

MT.L to C/O

Step 4 Balancing moments are of opposite sign to reduce

larger moment.

Step 5 Add ∑M to balancing moment for max. support

moment.

MAXIMUM MOMENT AT MIDSPAN

To determine maximum moments at midspan the positive
midspan moments shown as +33.75, +15.0, + 23.44 and +33.75
are taken from Figure 5 for beams with fixed ends. Certain
corrections are to be added to these moments in order to obtain
the final maximum moments at midspan.

The procedure is illustrated for span BC in Figure 7.

Distribution = D.F (MT.L - MD.L).

The distributed moments induced at B and C give 9 and 0.69

respectively.
Distributed moment: [K/∑K] x U, carry-over moment:1/2 x [K/∑K] x U The counter-clockwise rotation of moment at B creates a hump to
Figure 4: Frame consisting of four members with far ends fixed the right of B and a sag to the left of C. Therefore, 9 is negative and 4.5
is positive. The clockwise rotation of moment at C creates a hump to
it tends to create a sag in the beam at A. Therefore 77 is positive the left of C and a sag to the right of B. Therefore 0.69 is negative and
and the final maximum moment at A is -77+ 77 = 0. 0.35 is positive. The counter-clockwise rotation of 0.35 creates a hump
To determine moments at B, begin with releasing joints A and to the right of B. Therefore 0.21 is negative. The clockwise rotation of
C, Figure 6 shows how to compute the two moments 33.75 and 4.5 creates a hump to the left of C. Therefore 2.5 is negative.

Journal - The Institution of Engineers, Malaysia (Vol. 68, No. 1, March 2007) 29
 CHAN KAM LEONG

Figure 5: Max. M (Support & Midspan) for a continuous beam over supports providing no restraint to rotation at A, B, C and D

30 Journal - The Institution of Engineers, Malaysia (Vol. 68, No. 1, March 2007)
 2-CYCLE MOMENT DISTRIBUTION FOR THE ANALYSIS OF CONTINUOUS BEAMS AND MULTI-STOREY FRAMED STRUCTURES

Figure 6: Moment distribution illustrated in its various elements

Journal - The Institution of Engineers, Malaysia (Vol. 68, No. 1, March 2007) 31
 CHAN KAM LEONG

= 9 x 0.5 (- 1 + 1 - 5 ) + 0.69 x 5 (- 1 + 1 - 3 )
2 0.5 9 2 0.5 5

= 4.5 (- 2 + 1 - 5 ) + 0.35 (- 2 + 1 - 3 )
2 9 2 5

= 4.5 (- 1 - 5 ) + 0.35 (-1 - 3 )

2 9 2 5

= [ - 4.5 (1 + 5 ) ] + [ - 0.35 (1 + 3 ) ]
2 9 2 5

= [ - 1 (1 + D.F.) x C / O ] + [ - 1 (1 + D.F.) x C/O ]

2 2

5 3
As and are the distribution factors at the right hand
9 5
support and left hand support respectively; 4.5 and 0.35 are the
C/O at the right hand support and left hand support respectively.
Therefore, they are the adjustment or correction due to right
hand support and the adjustment or correction due to the left
hand support. These adjustments or corrections added to
midspan moment to obtain maximum midspan moment.
Figure 7: Two corrections for Midspan M
THE METHOD OF CALCULATING
The corrections or adjustments at supports B and C are MAXIMUM MIDSPAN MOMENT IS AS
(-9+0.35-0.21) and (-0.69+4.5-2.5) respectively. Therefore the FOLLOWS:
corrections or adjustments at the midspan are (-9 + 0.35 - 0.21)
2 Step 1 Write down mid-span moment due to total load as if
and (-0.69 + 4.5 - 2.5). It can be shown that the corrections or beam is fully fixed at each end.
2
WL2
MT.L. = for a uniformly distributed load, W.
adjustments are - 1 (1 + D.F) x C/O due to left hand support 24
2
Step 2 Calculate & write down ‘adjustment’ due to left hand
1
at B and - (1 + D.F) x C/O due to right hand support at C support.
2
i.e. Adj. Ml = -1/2 (1 + D.F) x C/O
as follows:
Step 3 Calculate & write down ‘adjustment’ due to right hand
-9 + 0.35 - 0.21) + -0.69 + 4.5 - 2.5
support.
2 2 i.e. Adj. Mr = -1/2 (1 + D.F) x C/O

= -9 + 4.5 - 2.5 + -0.69 + 0.35 - 0.21

Step 4
Add both adjustments to midspan moment to obtain
2 2 max. midspan moment.
5 3
= -9 + (9 x 0.5) - (9 x 0.5 x 9 ) + -0.69 + (0.69 x 0.5) - (0.69 x 0.5 x 5 )

2 2

32 Journal - The Institution of Engineers, Malaysia (Vol. 68, No. 1, March 2007)
 2-CYCLE MOMENT DISTRIBUTION FOR THE ANALYSIS OF CONTINUOUS BEAMS AND MULTI-STOREY FRAMED STRUCTURES

MINIMUM MOMENT providing no restraint to rotation in Figure 5. A check on span

Span BC is shorter than the adjacent spans. It is possible that BC shows that the minimum moment at midspan is +2.85.
negative moments may extend across the shortest span. It is
therefore necessary to calculate the minimum moment at midspan DETERMINATION OF COLUMN MOMENTS
of span BC. For multi-storey buildings, it is considered satisfactory to
The loading in Figure 8(a) has dead load only on span BC and compute column moments under the same assumption used for
total load on the adjacent spans. The procedure is the same as that beam moments i.e. far ends of columns are fixed above and
described in previous sections. The results show that a minimum below the floor at which moments are to be determined. Column
moment of -7.06 at midspan. This shows that the entire span BC moments are computed for unbalanced floor loading, that is live
is hogging throughout and reinforcement must be provided load on one side only.
accordingly for concrete beam. In Figure 9, live load is placed on the alternate spans as shown
on load pattern A and load pattern B.

THE METHOD OF
CALCULATING THE
MAXIMUM COLUMN
MOMENTS :

Step 1 Write down the distribution factors

for the columns and for the beams
and the MD.L and MT.L according to
the load pattern.

(y) C/O = (-20 + 46.88) x 5/9 x 1/2 = +7.47 Step 2 Calculate and write down the carry-over moment
(z) C/O = (-20 + 67.5) x 3/5 x 1/2 = +14.25 (C/O) and C/O = 1/2 x D.F x (MT.L - MD.L)
(a1) Adj. Ml = - 1/2 (1+3/5) (7.47) = -5.98
(b1) Adj. Mr = - 1/2 (1+5/9) (14.25) = -11.08 Step 3 Obtain total unbalanced moment (∑M) by adding
MT.L to C/O

Step 4 Maximum column moments are obtained by

multiplying the difference of the beam moments at the
joint by the distribution factors of the columns.

The results of the column moments are shown in Figure

9(a). A summary of the maximum moments obtained using the
2- Cycle Moment Distribution is given as in Figure 9(b)
whereas the maximum column moments determined by
computer analysis is as shown in Figure 9(c). It can be seen that
the differences between the maximum column moments
obtained by the two methods are small in values. It is important
to note that the deflections of columns follow the direction of
the beam unbalanced moment at the joint but the column
moments rotate in the opposite direction and bending moment
diagram should be drawn on the tension face of a member in
normal practice.
It can be seen that the maximum column moments in
columns A, B and C are obtained from load pattern A,
whereas load pattern B gives maximum column moments in
column D.

ACCURACY OF THE 2-CYCLE METHOD

Figure 8: Minimum moment at midspan AGAINST COMPUTER PROGRAMME
For a continuous beam over supports providing no restraint to
Hence, minimum moment at midspan is obtained by adding rotation as shown in Figure 5 the difference between the 2-Cycle
the adjustments due to right hand support and due to left hand Method and a computer programme is 11% maximum for
support to midspan moment. maximum support moments, whereas for maximum midspan
Figure 9 is a sub-frame of a multi-storey structure. The moments the difference is 4% maximum for end spans , and is 17%
calculations and results of the maximum moments at supports maximum for the internal spans. However, the author’s experience
and spans in beams of the sub-frame is as shown in Fig. 10 as confirms that the difference gets smaller when the live load is
against the result for a continuous beam over supports increased.

Journal - The Institution of Engineers, Malaysia (Vol. 68, No. 1, March 2007) 33
 CHAN KAM LEONG

The results for maximum moments (support and midspan) in In addition to continuous beams and multi-storey
beams of frame in Figure 10 show that the difference between the building frames subject to vertical loads, the 2-Cycle
2-Cycle Method and a computer programme is negligible. Method can be used to analyse other types of structures
For maximum column moments in Figure 9, the difference such as continuous beams subject to settlement at
between the 2-Cycle Moment Distribution and a computer support, portal frames and other rigid frames subject to
programme is 12% maximum, whereas, the difference is as large horizontal and vertical loads. And even frames with side
as 150% in column C under load pattern B. However, it can be sway can be solved by this method. After going through
ignored as the moment is very small. the method and its application properly, one will realise

Figure 9(a): Column moments by the 2-cycle moment distribution

34 Journal - The Institution of Engineers, Malaysia (Vol. 68, No. 1, March 2007)
 2-CYCLE MOMENT DISTRIBUTION FOR THE ANALYSIS OF CONTINUOUS BEAMS AND MULTI-STOREY FRAMED STRUCTURES

Figure 9(b,c): Maximum column moments

Figure 10: Max. M in beams of frame shown in Figure 9

Journal - The Institution of Engineers, Malaysia (Vol. 68, No. 1, March 2007) 35
 CHAN KAM LEONG

only support moment but also span moment and column

moment can be obtained fairly quickly by using this
method. A structure basically consists of beams and
columns, through the 2- Cycle Method, one can feel,
understand and appreciate structures better and thus gain
confidence in structural design. The 2-Cycle Method can
be used to counter check computer software as no
structural engineer should ever use unfamiliar software
without applying some verification. Hence, it helps an
engineer to be a computer- aided designer rather than just
a computer operator. ■

REFERENCES
Figure 11: Moment envelop diagram in beams of frame in Figure 9
[1] Portland Cement Association, Continuity in
Concrete Building Frames Practical Analysis for
Vertical Load and Wind Pressure, 1959.

[2] Chan Kam Leong, “Two Cycle Moment

Distribution”.
Bulletin of the Institution of Engineers, Malaysia,
August, 1978.

[3] Leong Tuck Wah, “Analysis of Multi-Storey Frame

Structures Using The Two Cycle Moment
Distribution Method.” Bulletin of the Institution of
Engineers, Malaysia, December 1978.

PROFILE
Engr. K. L. Chan obtained his B.Sc (Eng)
from the National Cheng Kung University,
Taiwan and M.Sc in Construction
Management from the University of Bath,
England. He is Chartered Civil &
Structural Engineer, C.Eng and P.Eng. He
is a MICE, MIStruct E, MIEM and
MACEM.

He has more than 30 years of experience in

C&S works and has worked in Malaysia,
how useful it is. Its importance just cannot be over- Singapore and the U.K. He is the founder
of K.L. Chan & Associates, consulting
emphasised. Structural engineering is an art of science; it
engineers and a practising consulting C&S
is more an art than a science. Hence, structural engineers engineer.
are strongly recommended to learn and use this useful
technique. In order to have a good understanding of the
method, readers are advised to read this article
t h o r o u g h l y.

CONCLUSIONS
Moment coefficients may be used only if loads and
spans meet the code requirement. The Cross Method is
too time-consuming. It can give support moment only on
one load case at a time and the process of moment
distribution is repeated until the moments to be carried
over are small enough to be neglected. Whereas the 2-
Cycle Method is comparatively simpler and faster; not

36 Journal - The Institution of Engineers, Malaysia (Vol. 68, No. 1, March 2007)