Design of Tudor Arches

with
Structural Glued Laminated Timber

Copyright © 2011
American Institute of Timber Construction

American Institute of Timber Construction

7012 South Revere Parkway Suite 140, Centennial, CO 80112
Phone: 303/792-9559 Fax: 303/792-0669
www.aitc-glulam.org
 CONTENTS

1. Structural Glued Laminated Timber Arches .......................................................................................... 1

2. Preliminary Tudor Arch Design ............................................................................................................. 9

Nomenclature .................................................................................................................................. 9
Procedure ...................................................................................................................................... 11
Example 2-1 .................................................................................................................................. 19

3. Arch Design Using Structural Analysis Software ................................................................................ 27

Nomenclature ................................................................................................................................ 27
Procedure ...................................................................................................................................... 30

4. Arch Design Using A Spreadsheet....................................................................................................... 37

Nomenclature ................................................................................................................................ 37
Procedure ...................................................................................................................................... 43
Example 4-1 .................................................................................................................................. 73

5. Arch-Diaphragm Interaction ................................................................................................................ 85

Nomenclature ................................................................................................................................ 85
Procedure ...................................................................................................................................... 86
Example 5-1 .................................................................................................................................. 89

6. Seismic Design ..................................................................................................................................... 93

Nomenclature ................................................................................................................................ 93
Seismic Coefficients...................................................................................................................... 94
Detailing Requirements................................................................................................................. 95
Drift Limits ................................................................................................................................... 98
Seismic Load Distribution ............................................................................................................ 98
Procedure ...................................................................................................................................... 98
Example 6-1 .................................................................................................................................. 99

7. Specifying Arch Geometry ................................................................................................................ 105

A. Arch Layup........................................................................................................................................ 107

Grade Requirements by Zone...................................................................................................... 107
Zone Thicknesses ........................................................................................................................ 108
Effect of Taper on Reference Design Values .............................................................................. 108
Alternate Arch Layups ................................................................................................................ 109

B. Arch Lay-out ..................................................................................................................................... 111

 Chapter 1
Structural Glued Laminated Timber Arches

Structural glued laminated timber (glulam) has many advantages over traditional sawn members
or other engineered wood products including its ability to be manufactured in a variety of shapes
from straight beams and columns to graceful curved members. Glulam can be manufactured
with constant cross section along the length or with taper to meet architectural requirements.
The glulam arch fully takes advantage of the unique properties of laminated timber construction.

Since their U.S. debut in 1934, in a school gymnasium in Peshtigo, Wisconsin, glulam arches
have been successfully used in a variety of structures including churches, schools, warehouses,
barns, aircraft hangars, and others. Several of the early U.S. glulam arch structures, including
the original gymnasium are still in use today (A.J. Rhude 1996).

Glulam arches continue to be popular for use in large open structures such as churches and
gymnasiums because of their excellent structural performance, inherent fire resistance, and
aesthetic appeal. Laminated timber arches are also used for vehicle and pedestrian bridges.

Figure 1.1 illustrates a number of common arch configurations. Arches may be of either two- or
three-hinged design. The AITC Timber Construction Manual (TCM) indicates that Tudor arches
can generally be used economically for spans of up to 120 feet and that parabolic and radial
arches can be used economically for spans of up to 250 feet (AITC 2004, p. 6).

Radial Tudor

Gothic Reverse Curve

A-frame Parabolic
Figure 1.1. Common arch configurations.
1
The most common arch configuration in use today is the three-hinged Tudor arch. It provides a
vertical wall frame and sloping roof that are commonly used for modern structures. Its
appearance is also pleasing to most people. Because of its popularity, the three-hinged Tudor
arch is the focus of this technical note, however, the same principles can be used to design other
types of glulam arches.

Arch Design

The procedures presented in this document can be used to design tudor arches, however, other
suitable methods may exist. Alternate methods, which have been validated through tests,
extensive experience, or analyses based on recognized theory, can also be used. Manufacturers
of structural glued laminated timber arches and engineering firms specializing in laminated
timber design may have proprietary methods for design. Finite element modeling or other
numerical methods may be used to compute member forces and deflections under load.

Aesthetic Considerations

Because structural glued laminated timber arches are typically exposed to view, the aesthetic
appeal of the arch is an important design consideration in addition to its structural performance.
The building designer and owner should verify that the final geometry meets the aesthetic
requirements and any other architectural requirements for the design. Because of the numerous
material and geometric parameters involved in arch design, multiple geometries can be utilized
to meet the same structural requirements.

Dimensional Considerations

The length of arch segments is generally limited by transportation and erection constraints.
Properly designed moment splices may be used in long span arches to facilitate shipping (AITC
2004, pps. 280-288).

Arches can pose significant challenges in shipping, particularly with tall wall heights and roofs
with low pitch. Critical shipping dimensions are shown in Figure 1.2. To facilitate shipping, it
may be necessary to design and manufacture the arch halves with moment splices occurring in
the arms (Figure 1.3). These are typically placed at inflection points, or other locations with low
bending moments. The arch can also be designed and manufactured with a detached haunch to
reduce its overall width for shipping (Figure 1.4). Shipping widths of up to 12 ft can typically be
accommodated. Wider loads may be possible depending on the distance from the manufacturing
plant to the jobsite and the route the shipment will follow. Glulam manufacturers can often
provide guidance regarding maximum shipping dimensions for various localities.

For arches with deep haunches, manufacturing constraints may also dictate the use of a detached
haunch (Figure 1.4) to facilitate passing the arch through a surface planer. Glulam
manufacturers should be contacted early in the design process to determine their maximum
recommended haunch depths.

2
 Wshipping

Lshipping

Figure 1.2 Critical Shipping Dimensions

3
Figure 1.3 Arch with Moment Splice in Arm (Schematic, some details omitted)

Figure 1.4 Arch with Detached Haunch

4
Structural Design

The design of arches is an iterative process. Experience will lead to less design iterations.
Design checks must be made at several sections along the arch. Because the number of
calculations required is large, the use of a spreadsheet or structural analysis software is
recommended for arch design and analysis.

The basic design procedure includes the following steps:

1. Determine all applicable loads and load combinations. Roof and wall loads can typically
be assumed to be uniform if decking is applied directly to the arch (or if purlins are
closely spaced) or they may be concentrated at purlin points. Appropriate loads and load
combinations can be obtained from the applicable building code or from the building
official having jurisdiction.

2. Determine a trial geometry. The outside geometry (wall heights, peak height, outside
span) of the arch is typically dictated by architectural constraints of the building. To
complete a trial geometry, the designer must choose a width, end depths, a radius, and
angles of taper for the wall leg and roof arm. Critical geometric features of a Tudor arch
are shown in Figure 1.5.

The base and crown sections of the arch must be adequately sized to resist the reaction
loads at those locations and to accommodate the required fasteners. Typical connection
details are shown in AITC 104 (AITC 2003). All other sections must be sized to resist
bending moment, axial forces, and shear forces and the combined effect of these as
appropriate.

3. Divide the arch into segments, locate sections of interest, and calculate their section
properties. Because of the complex geometry of the arch and the variability of loading
conditions, deflections are typically estimated using the principle of virtual work
(programmed into a spreadsheet) or using structural analysis software. Additionally, the
critical cross section for stress analysis will not be obvious. Therefore, several sections
along the length of the arch, taken perpendicular to the laminations, should be chosen for
evaluation. To minimize calculations, the same sections chosen for analysis of stresses
can be used for virtual work calculations.

4. Determine forces and moments with corresponding stresses on each section. The bending
moment, shear force, and axial force must be determined for each section of interest.
These forces and moments can be determined using free-body diagrams at the sections of
interest or structural analysis software. The resulting stresses can be calculated for each
section.

5
 dc

dc'

φt,L φb,L

Hc αarm,L

αleg,L

H w,L
RL

d'b,L
LL

db,L
Figure 1.5. Arch geometry. (Only left half shown for clarity.)

5. Determine the adjusted design values (allowable stresses) for each section. Allowable
stresses in compression, bending, and shear, must be determined for each of the sections
chosen for evaluation. Reference design values must be adjusted by all appropriate
adjustment factors specified in the National Design Specification® (NDS®) for Wood
Construction.

In addition to the factors described in the NDS®, the stress interaction factor, CI,
described in the AITC Timber Construction Manual (TCM) (5th ed. p. 108) should be
applied for sections taken through the straight-tapered segments of the arch. For arches
with taper sawn on the outer faces, it is also appropriate to reduce the flexural design
value to account for the loss of high grade laminations due to the tapered cut (Appendix
A).

6
 Out-of-plane buckling of arches should be considered between points of lateral support
through application of the column stability factor, CP. In-plane buckling of the straight
segments of the arch should also be considered through application of the column
stability factor, CP. These segments can be analyzed as tapered columns modeling the
large end as fixed and the small end as pinned for the arch legs and modeling both ends
as pinned for the arch arms (assumes inflection point is near upper tangent point). The
column length for the arm is approximated as the distance from the upper tangent point to
the peak, and the column length for the leg is approximated as the distance from the
lower tangent to the base.

It is not necessary to evaluate in-plane buckling in the curved segment of Tudor arches,
because secondary moments due to the eccentricity caused by buckling will generally be
small compared to the bending moment caused by the curved shape. Therefore, for in-
plane buckling in the curved segment, the column stability factor can be taken equal to
unity (CP = 1.0).

6. Evaluate combined stresses on each section. Arches simultaneously resist both axial and
flexural stresses, so stress interactions must be considered. The combined effect of axial
and flexural stresses at each section must satisfy the NDS® interaction inequalities.
Combined flexure and compression are usually critical, however, for some loadings,
flexure and axial tension may be developed.

Furthermore, the abrupt change in section at the haunch of a Tudor arch causes stress
concentrations that can be accounted for by multiplying the stresses calculated by the
flexure formula for that segment by the following factor:
⎛ 180D − θ ⎞
Kθ ,haunch = 1 + 2.7 tan ⎜ ⎟
⎝ 2 ⎠
where: θ = the included angle between the outer faces of the wall leg and roof arm

The factor, Kθ,haunch, should be applied to the stresses calculated by the flexure formula
prior to checking the section for combined stress interaction. However, when using
structural analysis software, it may be more convenient to reduce the allowable stress by
dividing by Kθ,haunch, rather than increasing the flexure formula stress by multiplying by
the factor.

7. Evaluate radial stresses in the curved segment. Loads that result in an increase in the
radius of curvature cause radial tension stresses in the curved segment of the arch. Loads
causing a decreased radius of curvature result in radial compression stresses. These
stresses may be calculated using the following equation.
6M
f r = Kr 2
bd haunch

2
⎛d ⎞ ⎛ 180D − θ ⎞
where: K r = 0.29 ⎜ c ⎟ + 0.32 tan1.2 ⎜ ⎟
⎝ Rm ⎠ ⎝ 2 ⎠

7
 9. Use the principal of virtual work (or structural analysis software) to calculate arch
deflections for each load combination. Because of the complex and varied geometry of
Tudor arches, deflections are generally determined using approximate analysis methods,
such as by using the principle of virtual work. By using the same sections chosen for
stress calculations, the number of calculations can be reduced. Deflections should be
checked against applicable limits. In the absence of specified deflection limits, it is
recommended that the limits of Table 1.1 be applied.

Table 1.1. Recommended Deflection Limits for Tudor Arches.

Haunch (Horizontal) Arch Arm Crown (Vertical)
Hw l arm LL + LR
With Brittle Finishes
120 360 360
Hw l arm LL + LR
With Flexible Finishes
60 180 180

10. Design and detail connections. Shear plates are typically used to transfer the vertical
reaction forces at the peak connection. Bolts with steel side plates are used to transfer
tension forces between members at the peak. The base connection typically consists of a
bearing seat to resist outward thrust, with a bolt to resist inward thrust and uplift.

11. Iterate until an acceptable design is obtained. Because it is unlikely that an optimum
design will be obtained on the first try, iteration will typically be necessary.

Conclusion

This chapter has provided an overview of the arch design process. Chapter 2 describes a
simplified method for preliminary arch design. Chapter 3 presents a simplified, approximate
method for modeling Tudor arches for use in structural analysis software. Chapter 4 discusses a
rigorous analysis method that can be programmed into a spreadsheet, and should be considered
when three-hinged Tudor arches will be designed regularly. Chapter 5 addresses the analysis of
system effects of using a structural diaphragm and shear walls with Tudor arches for simple
rectangular structures. Chapter 6 covers the design of arches to resist seismic loads. Chapter 7
discusses the requirements for specifying a Tudor arch once the design is complete.

Works Cited

AITC. 2004. Timber Construction Manual. John Wiley & Sons, Inc. Hoboken, New Jersey.

AITC. 2003. Typical Construction Details. AITC 104-2003. American Institute of Timber
Construction. Centennial, Colorado.

Rhude, A. J. 1996. Structural Glued Laminated Timber: History of Its Origins and Early
Development. Forest Prod. J. 46(1):15-22.

8
 Chapter 2
Preliminary Tudor Arch Design

Because arch design can be quite complex and involved, many designers may choose to have the
arches engineered by the structural glued laminated timber manufacturer or an engineer
specializing in custom glulam structures. However, a method of approximating the geometry of
an arch is useful to both the architect and engineer for preliminary design purposes. This allows
the designer to visualize the appearance of the arches in the structure and illustrate this for his
client. It also gives the final designer a starting point for his structural analysis.

This chapter presents one method of developing a symmetric arch geometry. This method
involves multiple approximations and should not be used for the final design, but should
give a reasonable approximation of an efficient arch geometry for preliminary purposes. The
final arch geometry must be evaluated by a qualified engineer to ensure that it is adequate for all
anticipated loads.

Nomenclature

A distance from the haunch to the approximate tangent points

b arch width
CD load duration factor
CI stress interaction factor for tapered members
CM wet-service factor
Ct temperature factor
CV volume factor
Cx horizontal reaction force at arch peak connection
Cy vertical reaction force at arch peak connection
CΔ geometry factor
CΔ ,u geometry factor for unloaded edge distance
D dead load
db depth at arch base measured perpendicular to laminations
db′ depth at arch base measured horizontally
dc depth at arch peak measured perpendicular to laminations
d c′ depth at arch peak measured vertically
d LT depth at section through lower tangent point
dUT depth at section through upper tangent point
dx depth at section through point x
EC minimum shear plate end distance for parallel-to-grain loads for a geometry factor
of CΔ = 1.0 or 0.83

9
ED minimum shear plate loaded edge distance for perpendicular-to-grain loads for a
geometry factor of CΔ = 1.0 or 0.83
Eα loaded edge distance for shear plate connection
Eα ,CΔ shear plate loaded edge distance corresponding to a geometry factor, CΔ
Eα ,1.0 required shear plate loaded edge distance for a geometry factor, CΔ = 1.0
Eα ,0.83 required shear plate loaded edge distance for a geometry factor, CΔ = 0.83
Fbx reference design stress in bending
Fvx reference design stress in shear
Hw height of wall
k1 factor for proportioning unbalanced snow load
k2 factor for proportioning unbalanced snow load
LL span of left arch half
LR span of right arch half
N reference capacity for single shear plate in sloped end grain
nmin minimum number of shear plates required to transfer vertical reaction force at
peak
nx number of rows of shear plates
ny number of shear plates per row
P parallel-to-grain reference capacity for single shear plate
Q90 perpendicular-to-grain reference capacity for single shear plate in end grain
R radius of curvature
Rx , L horizontal reaction at base of left arch half
Rx , R horizontal reaction at base of right arch half
Ry , L vertical reaction at base of left arch half
Ry , R horizontal reaction at base of right arch half
S snow load
SC minimum shear plate spacing parallel to grain for parallel-to-grain loads for a
geometry factor of CΔ = 1.0
SD minimum shear plate spacing perpendicular to grain for perpendicular-to-grain
loads for a geometry factor of CΔ = 1.0
Sα spacing for shear plate connection
Sα ,CΔ shear plate spacing corresponding to a geometry factor, CΔ
Sα ,1.0 required shear plate spacing for a geometry factor, CΔ = 1.0
Sα ,0.5 required shear plate spacing for a geometry factor, CΔ = 0.5
x horizontal distance to critical section for flexure in arch arm (estimated)
x2 horizontally projected distance from haunch to approximate upper tangent point

10
 y1 vertical distance from base to approximate lower tangent point
y2 vertically projected distance from base to approximate upper tangent point
α arm taper angle in arch arm
α leg taper angle in arch leg
ωD uniformly distributed dead load
ωS uniformly distributed snow load

φ roof slope

Procedure

The arch geometry will be established based on the D + S loading. (For areas without snow load,
roof live load can be substituted). The required depths in the arch leg are determined based on
the balanced snow loading or the unbalanced loading, whichever results in larger base reaction
forces. The depths in the arch arm are controlled by the unbalanced snow loading. The
following steps are used to establish the geometry:

1. Choose angles of taper, radius of curvature, wall height, peak height, span, and arch
width (Figure 1.2). The outside arch geometry (wall height, peak height, and span) will
generally be dictated by the building dimensions. The angles of taper should generally be
less than 3° for the arch arm and less than 5° for the arch leg (steeper tapers can be used
if they are properly accounted for in design). The radius of curvature of the inside face is
commonly chosen as 7 ft 0 in. for Southern Pine and 9 ft 4 in. for other softwood species.
The arch width is often chosen so that it exceeds the minimum requirements for Heavy
Timber Construction in addition to structural requirements.

2. Choose material properties (species, design stresses: Fbx, Fvx). For Douglas fir arches,
common properties are: Fbx = 2400 psi and Fvx = 190 psi. For Southern Pine arches,
common properties are: Fbx = 2400 psi and Fvx = 215 psi.

3. Locate approximate tangent points (Figure 2.1).

⎛ 90 − φ ⎞
A = R tan ⎜ ⎟ [2-1]
⎝ 2 ⎠
x2 = A cos (φ ) [2-2]
y2 = H w + A sin (φ ) [2-3]
y1 = H w − A [2-4]

11
 φ

A
y2
Hw

y1

x2
Figure 2.1. Location of approximate tangent points.

4. Calculate the reactions and pin forces for balanced (Figure 2.2) and unbalanced (Figure
2.3) snow loads. These forces are calculated based on the outside geometry of the arch,
because the actual centerline of the members is not known. This produces a reasonably
accurate estimate of the forces.

Balanced Snow Load (Figure 2.2)

Ry , R = Ry , L = (ωD + ωS ) L2L [2-5]

Rx , R = Rx , L =
1 ⎡ ( ωD + ωS ) L2L ⎤
⎢ R y ,L LL − ⎥ [2-6]
Hc ⎣ 2 ⎦
ωD + ωS ωD + ωS

Cx Cx

Cy Cy

R x,L R x,R

R y,L R y,R

Figure 2.2. Dead load plus balanced snow load.

12
 Unbalanced Snow Load (Figure 2.3)

Ry , L =
( k1ωS + ωD )( 3LL ) + ( k2ωS + ωD )( LL ) [2-7]
4

C y = ( k1ωD + ωS ) LL − Ry , L [2-8]

L2L
Ry , L LL + ( k1ωS + ωD )
Rx , L = 2 [2-9]
Hc

k2ωS
k1ωS

ωD ωD

Cx Cx

Cy Cy

R x,L R x,R

R y,L R y,R

Figure 2.3. Dead load plus unbalanced snow load (simplified).

5. Estimate the required depth at the lower tangent point, d LT . The minimum depth at the
lower tangent point is estimated based on the bending moment created by the horizontal
reaction (from the balanced snow load or unbalanced snow load, whichever gives larger
reaction) applied at a distance, y1 , from the approximate tangent point. The depth should
be increased approximately 5% over that calculated based on moment alone to account
for compression-flexure interaction.

13
 6 y1 Rx , L
d LT > 1.05 [2-10]
bFbx CD CI CV CM Ct

CD is 1.15 for snow loading (1.25 for roof live load). C I is calculated based on the
chosen material properties and leg taper angle. A rough estimate of the volume factor,
CV , at the tangent point can be obtained from one of the following expressions: CV ≈
0.95-0.0015*(LL) for southern pine or CV ≈ 0.90-0.0025*(LL) for other species, where LL
is expressed in units of feet.

The depth at the lower tangent point should not be greater than 6 times the width for
arches braced against lateral buckling by decking or closely spaced girts or purlins and
should not be greater than 5 times the width for arches without such lateral bracing. If
the required depth calculated by Equation [2-10] is not within these limitations, the width
must be increased.

6. Estimate the required depth at the base. The minimum depth at the base is estimated
based on the horizontal reaction force and is also determined based on the estimated
tangent point depth and the amount of taper chosen for the leg. The greater depth from
these two calculations should be chosen. Generally the base depth is also chosen to be a
minimum of 1.5 times the arch width. The base depth may also be chosen based on the
minimum size required for Heavy Timber Construction.

Required base depth for shear:

3Rx
db′ > [2-11]
2bFvx CD CM Ct

Required base depth to get the required lower tangent depth with chosen leg taper angle:

d b′ > d LT − y1 tan α leg [2-12]

Occasionally, the required depth at the base for shear may be greater than the depth
required at the tangent point. In such cases, the base depth required for shear should be
chosen and the leg taper angle should be set to zero to avoid excessive depth at the
tangent point.

The base depth should generally not be chosen as less than 1.5 times the arch width. If
this criterion governs the selection of the depth, the leg taper angle can be reduced to
avoid excessive depth at the lower tangent point.

7. Estimate the required depth at the crown. The minimum depth at the crown is estimated
based on the shear plate connector requirements to resist the vertical shear force at the
crown due to unbalanced snow loading and based on flexure in the loaded arm due to
unbalanced snow load.

14
Required crown depth based on shear plate capacity:

• Calculate reference capacity for a single shear plate in sloped end grain using
Hankinson’s formula for the chosen roof slope and arm taper.

PQ90
N= [2-13]
( ) (
P sin 90 − φ − α arm + Q90 cos 2 90D − φ − α arm
2 D
)
• Adjust the capacity for load duration and apply the geometry factor for reduced
unloaded edge distance. Shear plate design values are tabulated for unloaded
edge distances of 1.75 in. and 2.75 in. for 2-5/8 in. and 4 in. shear plates,
respectively. For narrower edge distances typical of standard glulam arches, a
reduction in capacity is required. Geometry factors for reduced unloaded edge
distance are shown in Table 2.1 for standard arch widths. The effect of group
action will generally be small and can be ignored at this point.

Table 2.1. Geometry Factors, CΔ,u, for shear plate capacity based on arch width.
Arch width Geometry Factors, CΔ,u
b (in.) 2-5/8 in. Connectors 4 in. Connectors
3 0.88 --
3-1/8 0.91 --
5 1.0 0.93
5-1/8 1.0 0.95
5-1/2 or wider 1.0 1.0

N * = NCD CM Ct CΔ ,u [2-14]

• Determine the minimum number of connectors, nmin, to transfer the load by

dividing the vertical crown reaction force by the partially adjusted capacity of an
individual connector (round up to nearest whole number).

Cy
nmin = [2-15]
N*

• Determine the number of rows of connectors, nx,. Table 2.2 gives the maximum
number of rows of connectors based on arch width. (Rows are oriented vertically)

15
 Table 2.2. Maximum number of rows of connectors based on arch width
Arch width Maximum rows of connectors, nx
b (in.) 2-5/8 in. Connectors 4 in. Connectors
3, 3-1/8, 3-1/2 1 --
5, 5-1/8, 5-1/2 1 1
6-3/4 1 1
8-1/2, 8-3/4 2 1
10-1/2, 10-3/4 3 2

• Calculate the required number of connectors per row, ny.

nmin
ny = [2-16]
nx

Except for very shallow members, a minimum of two shear plates is generally
required per row to prevent significant reductions in member shear capacity due
to notch effects (NDS 3.4.3.3).

• Determine the geometry factor for optimal spacing and edge distance:

Cy
0.5 ≤ CΔ = ≤ CΔ , u [2-17]
nN *

• Determine the required spacing for the cases of CΔ = 0.5 and CΔ = 1.0.

The required spacing for CΔ = 1.0 is determined based on the angle of the
connector with respect to the grain of the member using the following
equation (adapted from Equation 5.14 from the 5th edition TCM):

SC S D
Sα = [2-18]
S sin ( 90 − φ − α arm ) + S D2 cos 2 ( 90 − φ − α arm )
2
C
2

where: Sα = Required spacing of connectors

SC = 6.75 in. for 2-5/8 in. shear plates
= 9 in. for 4 in. shear plates
SD = 4.25 in. for 2-5/8 in. shear plates
= 6 in. for 4 in. shear plates

The required minimum spacing for CΔ = 0.5 is 3.5 in. for 2-5/8 in. shear
plates and is 5.0 in. for 4 in. shear plates.

16
 • Calculate the required loaded edge distance for CΔ = 0.83 and CΔ = 1.0.

EC ED
Eα = [2-19]
E sin ( 90 − φ − α arm ) + ED2 cos 2 ( 90 − φ − α arm )
2
C
2

Eα = Required loaded edge distance

EC = 5.5 in. for 2-5/8 in. shear plates and CΔ = 1.0
= 4.25 in. for 2-5/8 in. shear plates and CΔ = 0.83
= 7 in. for 4 in. shear plates and CΔ = 1.0
= 5.4 in. for 4 in. shear plates and CΔ = 0.83
ED = 2.75 in. for 2-5/8 in. shear plates and CΔ = 1.0
= 1.5 in. for 2-5/8 in. shear plates and CΔ = 0.83
= 3.75 in. for 4 in. shear plates and CΔ = 1.0
= 2.5 in. for 4 in. shear plates and CΔ = 0.83

• Determine the required spacing and loaded edge distance corresponding to the
optimal geometry factor using linear interpolation. If the optimal geometry factor
is less than 0.83, use the minimum edge distance determined for CΔ = 0.83.

⎛ C − 0.5 ⎞
Sα ,CΔ = ⎜ Δ ⎟ ( Sα ,1.0 − Sα ,0.5 ) + Sα ,0.5 [2-20]
⎝ 1 − 0.5 ⎠

⎛ C − 0.83 ⎞
Eα ,CΔ = ⎜ Δ ⎟ ( Eα ,1.0 − Eα ,0.83 ) + Eα ,0.83 [2-21]
⎝ 1 − 0.83 ⎠

Choose spacing, S, and end distance, E, equal to or greater than those calculated.
The chosen spacing and end distance should be rounded up to the nearest 1/16 in.
or other practical unit of measure to facilitate fabrication.

• Calculate the required depth based on connector spacing and edge distance:

d c′ ≥ d c′ shear plates = ( n y − 1) Sα + 2 Eα [2-22]

Required crown depth based on flexure in the arm and chosen arm taper angle:

Estimating that the point of inflection occurs near the upper tangent point of the
loaded arch for the unbalanced snow loading and that the crown depth is
approximated by the depth required for the shear plates, the location of the critical
section in the upper arm of the loaded arch (measured from the crown) can be
determined from the following relationship, which was derived for this case from
17
 equation 4.32 of the AITC Timber Construction Manual (5th edition) (x is
measured horizontally from the crown):

⎛ d c′ ⎞
cos (φ )( LR − x2 ) ⎜⎜ ⎟
⎝ cos (α arm ) ⎟⎠
x≈ [2-23]
⎛ cos (φ ) ⎞ ( LR − x2 )
2 ⎜⎜ d c′ ⎟⎟ + tan α arm
⎝ cos (α arm ) ⎠ cos (φ )

The required depth at that section can be approximated as:

⎛ x ⎞
3 (ωS + ωD )( LR − x2 ) x ⎜1 − ⎟
⎝ LR − x2 ⎠
dx > [2-24]
bFbx CD CI CM Ct

For tapered arms, the reference flexural stress, Fbx , used in [2-24] should be
reduced to account for the loss of high strength material at the surface due to
tapering, unless the arch is laid up with a uniform-grade lay-up or the high grade
material is maintained throughout the length of the arch (Appendix A). Either of
these options may increase the cost of the arch. For standard arches with tapered
arms, a 10% reduction is suitable for preliminary design purposes.

The corresponding depth at the crown can be estimated as:

⎛ x ⎞
dc > d x − ⎜ ⎟ tan α arm [2-25]
⎝ cos φ ⎠

d c cos (α arm )
d c′ = [2-26]
cos φt

The greater depth from connector requirements or from arm flexure is chosen as the
minimum depth of the arch at the crown.

With the preceding steps completed, a trial arch geometry is defined. The trial geometry should
be further adjusted to ensure that the upper and lower tangent point depths are within
approximately 10% of each other. The trial arch should be drawn to scale and the geometry
should be adjusted if necessary for aesthetic considerations.

At this point, the preliminary design of the arch is completed. The final design must include a
thorough analysis of the arch subject to all applicable loads and load combinations by the
procedures discussed in Chapters 3-5. An example including the preliminary design procedure
described in this chapter follows.

18
Example 2-1: Design of Tudor Arch

Given: A building with outside dimension of 50 ft by 60 ft is to be constructed using Southern

Pine Tudor arches spaced at 15 ft.

The arches will be subject to normal temperatures and dry-use conditions.

The arches will be symmetric with a maximum roof height of 24 ft. and a wall height of
12 ft as shown in Figure E4.1.

24 ft

12 ft 12 ft

25 ft 25 ft

Figure E2.1. Outside geometry of arch.

The curved portion of the arches will have a radius of 10.5 ft at the inside face and a
lamination thickness of ¾ in.

The arch has continuous lateral bracing along the wall and roof.

The building has no shear walls at the ends, so the arches will be assumed to support all
vertical and lateral loads acting parallel to the planes of the arches.

Material Properties: Fbx+ = 2000 psi

Fvx = 215 psi

Loads: the loads have been determined as shown in Figures E2.2 through E2.4. The
wall dead loads are assumed to act at the outer face of the arch.

19
 225 plf

150 plf 150 plf

Figure E2.2. Dead loads.

375 plf

Figure E2.3. Balanced snow load.

585 plf
113 plf

Figure E2.4. Unbalanced snow load.

20
Wanted: Perform preliminary design of a Tudor arch with a width of 6.75 in. based on the dead
plus snow load combinations:

(1) D + S (balanced snow)

(2) D + S (unbalanced snow)

For final design, additional load combinations will need to be considered.

Solution:

The required depth at the lower tangent is commonly governed by the balanced snow
loading. The depths in the arch arm are commonly controlled by the unbalanced snow
loading.

Approximate reaction forces (using the outside arch geometry) for each considered snow
load combination are included in Table E2.1.

Table E2.1. Approximate reactions for D + S load combinations.

Load R y ,L Rx , L Cy Cx R y ,R Rx , R
D+S (Balanced) 16800 7810 0 -7810 16800 7810
D+ S (Unbalanced) 13200 7470 -2950 -7470 19100 7470

Estimating the required depth at the crown based on connection requirements

The minimum depth at the crown will be estimated based on the vertical reaction force at
the peak connection due to unbalanced snow loading and the number of shear plates
required to transfer the load.

For 4 inch shear plates in Southern Pine:

P = 4360 lb
Q = 3040 lb
Q90 = 0.6Q = 0.6 ( 3040 lb ) = 1824 lb

The shear plates will be loaded at an angle to the grain of 90° − φ − α arm .

⎛ 12 ft ⎞
φ = arctan ⎜ ⎟ = 25.64°
⎝ 25 ft ⎠
α arm = 2.5°
90° − φ − α arm = 90° − 25.64° − 2.5° = 61.86°

21
Hankinson’s formula can be used to determine the reference capacity for a single shear
plate loaded at an angle to the grain of 90° − φ − α arm .

PQ90
N=
P sin ( 90° − φ − α arm ) + Q90 cos 2 ( 90° − φ − α arm )
2

N=
( 4360 lb )(1824 lb )
( 4360 lb ) sin 2
( 61.86° ) + (1824 lb ) cos 2 ( 61.86° )
N = 2095 lb

The partially adjusted capacity (dry-use, normal temperatures, snow load) for a single
shear plate is:

N * = NCD CΔ ,u = ( 2095 lb )(1.15 )(1.0 ) = 2410 lb

Based on the vertical reaction at the crown of C y = −2950 lb , a minimum of two shear
plates will be required:

Cy 2950 lb
nmin = *
= = 1.22 ∴ Use two shear plates
N 2410 lb

The geometry factor for optimal spacing and edge distance is:

Cy
0.83 ≤ CΔ = ≤ CΔ ,unloaded edge distance
nN *
Cy 2950 lb
CΔ = = = 0.612 ≥ 0.5 ∴ CΔ = 0.61
nN 2 ( 2410 lb )

The required spacing between shear plates for CΔ = 1.0 is calculated as:

SC S D
Sα ,1.0 =
( ) (
SC2 sin 2 90D − φ − α arm + S D2 cos 2 90D − φ − α arm )
Sα ,1.0 =
( 9 in )( 6 in )
( 9 in ) sin 2 ( 61.86D ) + ( 6 in ) cos2 ( 61.86D )
2 2

Sα ,1.0 = 6.41 in

The required spacing between shear plates for CΔ = 0.5 for 4 in. shear plates is:

Sα ,0.5 = 5.0 in

22
The required spacing between shear plates for CΔ = 0.61 is determined by linear
interpolation:

⎛ C − 0.5 ⎞
Sα ,0.61 = ⎜ Δ ⎟ ( Sα ,1.0 − Sα ,0.5 ) + Sα ,0.5
⎝ 1 − 0.5 ⎠
⎛ 0.61 − 0.5 ⎞
Sα ,0.61 = ⎜ ⎟ ( 6.41 in − 5.0 in ) + 5.0 in
⎝ 1 − 0.5 ⎠
Sα ,0.61 = 5.3 in ∴ Choose Sα = 5.5 in

The actual geometry factor based on the spacing of 5.5 in. is:

⎛ S − Sα ,0.5 ⎞ ⎛ 5.5 − 5.0 ⎞

CΔ = ⎜ α
⎜S ⎟⎟ (1.0 − 0.5 ) + 0.5 = ⎜ ⎟ (1.0 − 0.5 ) + 0.5 = 0.68
⎝ α ,1.0 − Sα ,0.5 ⎠ ⎝ 6.41 − 5.0 ⎠

The minimum loaded edge distance for CΔ = 0.83 is calculated as:

EC ED
Eα =
( ) (
EC2 sin 2 90D − φt − α arm + ED2 cos 2 90D − φt − α arm )
Eα ,0.83 =
( 5.4 in )( 2.5 in )
( 5.4 in ) sin 2 ( 61.86D ) + ( 2.5 in ) cos 2 ( 61.86D )
2 2

Eα ,0.83 = 2.75 in ∴ Choose Eα = 2.75 in

The minimum vertical depth at the crown to accommodate the two 4 in. shear plates is:

d c′ = Sα + 2 Eα = 5.5 in + 2 ( 2.75 in ) = 11.0 in

The depth measured perpendicular to the laminations at the crown is:

d c′ cos φ (11.0 in ) cos ( 25.6° )

dc = = = 9.9 in
cos α arm cos ( 2.5° )

A depth, d c′ , of 11.0 inches is sufficient to accommodate the use of two 4 in. shear plates.
The geometry factor will be the lesser of that determined for spacing and edge distance,
CΔ = 0.68. The adjusted capacity (dry-use, normal temperatures, snow load) of the
connection is:
N ′ = 2 ( P90°−φ−αCD CΔ Cg ) = 2 ( 2095 lb )(1.15 )( 0.68 )(1.0 )
N ′ = 3280 lb ≥ C y = 2950 lb ∴ OK

23
Estimating the required depth at the crown based on flexure in the arm

The location of the upper tangent point of the loaded arch half can be approximated by:

(
x2 = A cos (φ ) = ( 79.3 in ) cos 25.64D = 71.5 in )
Estimating that the point of inflection will occur near the upper tangent point of the
loaded arch for the unbalanced snow loading and that the crown depth is approximately
equal to the depth required to accommodate the shear plates at the crown, the location of
the critical section in the upper arm of the loaded arch (measured from the crown) can be
determined from the following relationship, which was derived for this case from
equation 4.32 of the AITC Timber Construction Manual (5th edition):

⎛ d c′ ⎞
cos (φ )( LR − x2 ) ⎜
⎜ cos (α ) ⎟⎟
x≈ ⎝ arm ⎠

⎛ cos (φ ) ⎞ ( LR − x2 )
2 ⎜⎜ d c′ ⎟⎟ + tan α arm
⎝ cos (α arm ) ⎠ cos (φ )
⎛ 12.5 in ⎞
( )
cos 25.64D ( 300 in − 71.5 in ) ⎜ ⎟
⎜ cos 2.5D ⎟
⎝ ⎠( )
x≈
⎛ (
cos 25.64 D
⎞ )
( 300 in − 71.5 in ) tan 2.5D
2 ⎜12.5 in ⎟+ ( )
⎜
⎝ cos 2.5 ⎟⎠
D
( )
cos 25.64D ( )
x ≈ 76.7 in

The required depth at that section can be approximated as:

⎛ x ⎞
3ω y , R ( LR − x2 ) x ⎜ 1 − ⎟
⎝ LR − x2 ⎠
dx >
bFbx CD CI
⎛ 76.7 in ⎞
3 ( 67.5 lb/in )( 300 in − 71.5 in )( 76.7 in ) ⎜ 1 − ⎟
dx > ⎝ 300 in − 71.5 in ⎠ = 13.5 in
( 6.75 in )( 2000 psi )( 0.9 )(1.15)( 0.926 )

24
 The corresponding depth at the crown can be estimated as:

⎛ x ⎞ ⎛ 76.7 in ⎞
dc ≈ d x − ⎜ ⎟ tan α = 13.5 in − ⎜ ( )
⎟ tan 2.5D = 9.8 in
⎝ cos φ ⎠
arm
⎜ cos 25.6D
⎝ ( ) ⎟
⎠

d c′ =
d c cos (α arm )
=
(
9.8cos 2.5D ) = 10.9 in
cos φ (
cos 25.6D )
A trial depth, d c′ , of 11.0 inches will be chosen (controlled by connector design). The
depth of the crown (measured perpendicular to the laminations) is dc = 9.9 inches as
previously calculated.

Estimating the depth at the upper tangent based on chosen crown depth and taper angle

The depth of the upper tangent can be estimated as:

⎛L −x ⎞ ⎛ 300 in − 71.5 in ⎞
dUT ≈ dc + ⎜ R 2 ⎟ tan α arm = 9.9 in + ⎜ ⎟ tan 2.5D = 21.0 in ( )
⎝ cos φ ⎠ ⎜
⎝ cos 25.6
D
⎟
⎠ ( )
Estimating the required depth of the lower tangent point

The minimum depth at the lower tangent point will be estimated based on the moment
created by the horizontal reaction applied at a distance, y1,L , from the approximate
tangent point. The leg taper is chosen as 4° for the determination of C I . The volume
factor will be estimated as 0.95.

⎛ 90D − φ ⎞ ⎛ ⎛ 90D − 25.64D ⎞ ⎞

A = R tan ⎜ ⎟ = (126 in ) ⎜⎜ tan ⎜ ⎟ ⎟⎟ = 79.28 in
⎝ 2 ⎠ ⎝ ⎝ 2 ⎠⎠
y1 = H w − A = 144 in − 79.28 in = 64.72 in
6 y1 Rx , L 6 ( 64.72 in )( 7812.5 lb )
d LT > =
bFbx CD CI CV ( 6.75 in )( 2000 psi )(1.15)( 0.83)( 0.95)
d LT > 15.74 in

The trial depth should be a minimum of 5% larger than calculated based on flexure alone,
to accommodate the interaction of compression and bending stresses. Additionally, it is
recommended that the upper and lower tangent point depths be within approximately
10% of each other. To satisfy both objectives, a trial depth of 19 in. will be chosen.

25
Estimating the required depth at the base

The minimum depth at the base will be estimated based on the horizontal shear reaction
and also determined based on the estimated tangent point depth and the amount of taper
chosen.

3Rx , L 3 ( 7812.5 lb )
db′ > = = 7.022 in
2bFvx CD 2 ( 6.75 in )( 215 psi )(1.15 )

( )
db′ > d LT − y1 tan α leg = 19 in − ( 64.7 in ) tan 4D = 14.5 in

A trial depth of 14.5 inches will be chosen for the base.

Results: b = 6.75 in
H w, L = H w, R = 144 in = 12 ft
H c = 288 in = 24 ft
db′, L = db′, R = 14.5 in
d c′ = 11.0 in
RL = RR = 126 in = 10.5 ft
α leg ,L = α leg ,R = 4D
α arm ,L = α arm ,R = 2.5D

A drawing of the trial arch is shown in Figure E2.5. The appearance is determined to be
acceptable. The preliminary design is complete.

11.0 in

24 ft

12 ft 10.5 ft 12 ft

14.5 in
25 ft 25 ft
Figure E2.5. Trial arch with αarm = 2.5° and αleg = 4°.

26
 Chapter 3
Arch Design Using Structural Analysis Software

The complexity of Tudor arch geometry and loading makes analysis and design cumbersome
without the use of a spreadsheet or structural analysis software. Chapter 4 presents the formulas
necessary to create a spreadsheet for the analysis of Tudor arches. This chapter presents
procedures for designing a Tudor arch using structural design software, including assumptions
that can be made to reasonably approximate the shape of arches as a series of linear elements.

Nomenclature

a coefficient for calculation of effective column depth

AL distance from the haunch to the approximate tangent points on the left arch half
AR distance from the haunch to the approximate tangent points on the right arch half
b arch width
Cc curvature factor
CD load duration factor
CI stress interaction factor for tapered members
CL beam stability factor
CM wet-service factor
CP column stability factor
Ct temperature factor
CV volume factor
d section depth
d LD depth of section through lower discontinuity
dUD depth of section through upper discontinuity
d min depth of smaller end of tapered column
d max depth of larger end of tapered column
dP effective depth for tapered column buckling calculations
Ex reference modulus of elasticity
E x′ adjusted modulus of elasticity
f bx bending stress due to applied loads
Fbx reference design stress in bending
Fbx′ adjusted design stress in bending
Fbx* reference design stress in bending adjusted by all applicable factors except CL
Fbx** reference design stress in bending adjusted by all applicable factors except CV
fc compression stress parallel-to-grain due to applied loads
Fc reference design stress in compression parallel-to-grain
Fc′ adjusted design stress in compression parallel-to-grain

27
FcE Euler buckling stress
fr radial stress due to applied loads
Fr′ adjusted radial design stress, either tension or compression
Frc reference design stress in radial compression
Frc′ adjusted design stress in radial compression
Frt reference design stress in radial tension
Frt′ adjusted design stress in radial tension
ft tension stress parallel-to-grain due to applied loads
Ft reference design stress in tension parallel-to-grain
Ft′ adjusted design stress in tension parallel-to-grain
Fvx reference design stress in shear
Fvx′ adjusted design stress in shear
Hc height of arch at peak
H w,L height of wall for left arch half
H w,R height of wall for right arch half
Kr radial stress shape factor
K r ,h radial stress shape factor at the haunch
K r , LD radial stress shape factor at the lower discontinuity
K r ,UD radial stress shape factor at the upper discontinuity
Kθ bending stress shape factor
Kθ ,h bending stress shape factor at the haunch
Kθ , LD bending stress shape factor at the lower discontinuity
Kθ ,UD bending stress shape factor at the upper discontinuity
l e ,arm ,L effective length for column buckling calculations for left arch arm
l e ,arm ,R effective length for column buckling calculations for right arch arm
l e,leg ,L effective length for column buckling calculations for left arch leg
l e ,leg ,R effective length for column buckling calculations for right arch leg
LL span of left arch half
LR span of right arch half
M bending moment on a section
Pcx horizontal point load applied at crown of arch
Pcy
vertical point load applied at crown of arch
RL radius of curvature for left arch half
Rm radius of curvature measured to the mid-depth of arch
Rm , LD radius of curvature measured to the mid-depth of arch at lower discontinuity

28
Rm ,UD radius of curvature measured to the mid-depth of arch at the upper discontinuity
RR radius of curvature for right arch half
Rx , L
horizontal reaction force at left base
Rx , R
horizontal reaction force at right base
Ry ,L
vertical reaction force at left base
Ry ,R
vertical reaction force at right base
V shear force on section
x2,L horizontally projected distance from haunch to approximate upper tangent point
for left arch half
x2,R horizontally projected distance from haunch to approximate upper tangent point
for right arch half
y1,L vertical distance from base to approximate lower tangent point for left arch half
y1,R vertical distance from base to approximate lower tangent point for right arch half
y2,L vertically projected distance from base to approximate upper tangent point for left
arch half
y2,R vertically projected distance from base to approximate upper tangent point for
right arch half
β LD angle between horizontal line and line from center of curvature through lower
discontinuity
βUD angle between horizontal line and line from center of curvature through lower
discontinuity
φ roof slope
φL roof slope on left arch half
φR roof slope on right arch half
θ included angle between arch leg and arm
θL included angle between leg and arm of left arch half
θR included angle between leg and arm of right arch half
ωrx ,L
horizontal distributed load on roof arm of left arch half
ωrx ,R
horizontal distributed load on roof arm of right arch half
ωry ,L
vertical distributed load on roof arm of left arch half
ωry ,R
vertical distributed load on roof arm of right arch half
ωwx ,L
horizontal distributed load on wall leg of left arch half
ωwx ,R
horizontal distributed load on wall leg of right arch half
ωwy ,L
vertical distributed load on wall leg of left arch half
ωwy ,R
vertical distributed load on wall leg of right arch half

29
Procedure

1. Determine applicable loads and load combinations. Roof and wall loads can typically be
assumed to be uniform if decking is applied directly to the arch (or if purlins are closely
spaced) or they may be concentrated at purlin points. Appropriate loads and load
combinations can be obtained from the applicable building code or from the building
official having jurisdiction. Figure 3.1 illustrates a general loading for consideration.
ωry,L
ωry,R

Pcx

ωrx,R
ωrx,L Pcy

Hc θR
θL
ωwx,R H w,R
H w,L ωwx,L ωwy,L ωwy,R

R x,L R x,R

R y,L R y,R

LL LR

Figure 3.1. Arch loading diagram.

2. Model the arch as linear elements by connecting the outside faces of each arch half with
a circular arc having a radius equal to the radius used for the inside face of the arch.
Equations 3-1 through 3-8 can be used to locate the tangent points as shown in Figure
3.2.

φL φR
AR

AL

AR

AL y 2,R
y 2,L H w,R
H w,L
y 1,R
y 1,L

x2,L x2,R
Figure 3.2. Simplified arch model with locations of tangent points.

30
 ⎛ 90 − φL ⎞
AL = RL tan ⎜ ⎟ [3-1]
⎝ 2 ⎠
x2,L = AL cos (φL ) [3-2]
y2,L = H w,L + AL sin (φL ) [3-3]
y1,L = H w,L − AL [3-4]

⎛ 90 − φR ⎞
AR = RR tan ⎜ ⎟ [3-5]
⎝ 2 ⎠
x2,R = AR cos (φR ) [3-6]
y2,R = H w,R + AR sin (φR ) [3-7]
y1,R = H w,R − AR [3-8]

3. Divide the arch into segments for analysis. Each arch leg, arm, and curved portion
should be broken into multiple segments.

4. Determine the forces and moments on each segment. Each applicable load combination
should be used in the calculation of forces and moments.

5. Determine adjusted design values for each segment. Estimate geometry-dependent

adjustment factors, such as CP, CL, CV, and CI. For tapered sections outside of the
tangent points, where high-grade material on the face is removed, exposing lower grades
toward the core, a reduction in the reference flexural design value is appropriate
(Appendix A).

Fbx′ = FbxCDCM Ct ( CL or CV ) CI Cc [3-9]

Fvx′ = FvxCDCM Ct [3-10]
Fc′ = FcCDCM Ct CP [3-11]
Ft′ = Ft CDCM Ct [3-12]
Frt′ = Frt CDCM Ct [3-13]
Frc′ = FrcCDCM Ct [3-14]
E x′ = E xCM Ct [3-15]

6. Establish a trial geometry. Calculate the required depths at the arch bases and crown
using shear stresses and any connection requirements. Calculate the required depths at
the tangent points based on flexural requirements. Establish a trial geometry that satisfies
these depth requirements, keeping in mind that the interaction equations for bending and
axial loads must also be satisfied. The angle of taper on the arch legs should generally
not exceed 5° and the angle of taper on arch arms should generally not exceed 3° (steeper
tapers can be used if they are properly accounted for in design).

31
7. Calculate geometry dependent adjustment factors. The calculated factors should be
compared to the estimated factors, and the geometry should be adjusted as necessary.

The straight arm and leg portions should be evaluated for in plane buckling as tapered
columns, modeling the large end as fixed and the small end as pinned for the arch legs,
and modeling both ends as pinned for the arch arms (assumes that the inflection point is
near the upper tangent point).

Representative depths, dP, for the column stability calculations can be calculated using
the following equation (NDS 3.7.2).

⎡ ⎛ d ⎞⎤
d P = d min + ( d max − d min ) ⎢ a − 0.15 ⎜1 − min ⎟⎥ [3-16]
⎢⎣ ⎝ d max ⎠ ⎥⎦

where: a = 0.70 for the arch leg

= 0.50 for the arch arm

Effective lengths for the column stability calculations for in-plane buckling can be
estimated as:

l e,leg ,L = 0.7 y1,L [3-17]

LL − x2,L
l e,arm ,L = [3-18]
cos φL
l e,leg ,R = 0.7 y1,R [3-19]
LR − x2,R
l e ,arm ,R = [3-20]
cos φR

8. Evaluate combined stresses on each section. Arches act as combined axial and flexural
members, so stress interactions must be considered. The combined effect of axial
compressive stresses and flexural stresses at each section must satisfy the following
inequality:

2
⎡ f ⎤ ⎡ f bx ⎤
⎢ c ⎥ +⎢ ⎥ ≤ 1.0 [3-21]
⎢⎣ Fc′ ⎥⎦ ⎢⎣ Fbx′ ⎡⎣1 − ( f c FcE )⎤⎦ ⎥⎦

The following inequalities must be satisfied for sections stressed in combined bending
and tension:

ft f
+ bx* ≤ 1.0 [3-22]
Ft ′ Fbx

32
 f bx − f t
≤ 1.0 [3-23]
Fbx**

The abrupt change in section at the haunch causes a redistribution of flexure stresses that
is accounted for by multiplying the bending stress, f bx , calculated using the flexure
formula by an empirical bending stress shape factor, Kθ. (To facilitate the use of
structural analysis software, it may be more convenient to reduce the allowable stress by
dividing by Kθ, rather than adjusting the applied flexure stress.) For the section through
the haunch, this factor can be calculated using the following equation:

⎛ 180D − θ ⎞
Kθ ,h = 1 + 2.7 tan ⎜ ⎟ [3-24]
⎝ 2 ⎠

Linear interpolation between a value of Kθ ,h at the haunch and a value of 1/CI at the
tangent points can be used to estimate factors for the sections located between the section
through the haunch and sections through the tangent points.

If a detached haunch is used, two points of discontinuity will be created on the back face
of the arch. At the lower point of discontinuity, the bending stress shape factor, Kθ, can
be calculated using Equation [3-25]. At the upper point of discontinuity Kθ can be
calculated using Equation [3-26].

⎛β ⎞
Kθ , LD = 1 + 2.7 tan ⎜ LD ⎟ [3-25]
⎝ 2 ⎠
⎛ 90D − φ − βUD ⎞
Kθ ,UD = 1 + 2.7 tan ⎜ ⎟ [3-26]
⎝ 2 ⎠

For sections between the points of discontinuity and the tangent points, Kθ can be
determined by interpolation. Between the two points of discontinuity, interpolation can
be used to calculate Kθ for intermediate sections or the larger value can be used.

Shear stresses on each section must satisfy the following inequality.

3V
Fvx′ ≥ [3-27]
2bd

33
 Evaluate radial stresses in the curved segment. Loads that result in an increase in the
radius of curvature cause radial tension stresses in the curved segment(s). Loads causing
a decreased radius of curvature result in radial compression stresses. These stresses must
satisfy the following inequality:

6M
Fr′ ≥ f r = K r [3-28]
bd 2

The radial stress shape factor at the haunch can be calculated as:
2 2
⎛d ⎞ ⎛ 180D − θ ⎞ ⎛ dc ⎞ 1.2 ⎛ 90 − φ ⎞
D
K r ,h = 0.29 ⎜ c ⎟ + 0.32 tan1.2 ⎜ ⎟ = 0.29 ⎜ ⎟ + 0.32 tan ⎜ ⎟ [3-29]
⎝ Rm ⎠ ⎝ 2 ⎠ ⎝ Rm ⎠ ⎝ 2 ⎠

Linear interpolation between a value of K r ,h at the haunch and a value of 0 at the tangent
points can be used to estimate factors for the sections between the section through the
haunch and sections through the tangent points.

For arches with detached haunches, Kr can be calculated for the section through the lower
point of discontinuity using Equation [3-30] and can be calculated for the section through
the upper point of discontinuity using Equation [3-31]. Interpolation can be used to
estimate factors for sections between the tangent points.
2
⎛ d ⎞ 1.2 ⎛ β ⎞
K r , LD = 0.29 ⎜ LD ⎟⎟ + 0.32 tan ⎜ LD ⎟ [3-30]
⎜R ⎝ 2 ⎠
⎝ m , LD ⎠
2
⎛ d ⎞ 1.2 ⎛ 90 − φt − βUD ⎞
D
K r ,UD = 0.29 ⎜ UD ⎟⎟ + 0.32 tan ⎜ ⎟ [3-31]
⎜R 2
⎝ m ,UD ⎠ ⎝ ⎠

9. Check deflections against any applicable limits. Typically, horizontal deflections should
be checked at the haunch and vertical deflections checked at the peak. Deflection in the
arch arms should also be checked. In addition to deflections from applied loads,
shrinkage of the laminations as the arch seasons in service will cause changes to the
shape of the arch that should be considered. A calculation method to estimate
displacements caused by changes in moisture content is presented in the next chapter.

In lieu of specified deflection limits, the limits of Table 1.1 are recommended for arches.

10. Design and detail connections. Shear plates are typically used to transfer the vertical
reaction forces at the peak connection. Bolts with steel side plates are used to transfer
tension forces between members at the peak. The base connection typically consists of a
bearing seat to resist outward thrust, with a bolt to resist inward thrust and uplift.

11. Iterate as necessary until a final design is established.

34
Conclusion

This chapter presented a simplified arch geometry for use with structural analysis software. The
results of the analysis using the simplified geometry are approximate, but will be generally close
enough for design applications. Modeling with structural analysis software is particularly useful
for the design of arches in three-dimensional structural systems. For simple three-hinged tudor
arches, development of a spreadsheet to automate the procedures of the next chapter is
recommended.

35
36
 Chapter 4
Arch Design Using a Spreadsheet

The procedure presented in this chapter precisely defines the arch geometry. Sections through
the arch are taken perpendicular to the laminations. The approximations and simplifying
assumptions used in the previous chapter have not been applied. A designer who expects to
design arches on a regular or repeated basis should consider developing a spreadsheet to apply
this method.

Nomenclature

aL length of a line from left base to crown

aR length of a line from right base to crown
Cc curvature factor
CD load duration factor
CI stress interaction factor for tapered members
CL beam stability factor
CM wet-service factor
CP column stability factor
Ct temperature factor
CV volume factor
Cx horizontal component of pin force at arch crown
Cy vertical component of pin force at arch crown
D dead load
b arch width
d depth of section of interest, measured perpendicular to the laminations
db depth at arch base, measured perpendicular to the laminations
db′ depth at arch base, measured horizontally
d b,L depth at left arch base, measured perpendicular to the laminations
d b′,L depth at left arch base, measured horizontally
d b,R depth at right arch base, measured perpendicular to the laminations
d b′,R depth at right arch base, measured horizontally
dc depth of arch half at crown, measured perpendicular to the laminations
d c′ depth at arch crown, measured vertically
dc e effective depth for notch effect calculations, measured perpendicular to the
laminations
d c′ e depth of member at peak after subtraction of effective notch, measured
vertically
d c ,L depth of left arch half at crown, measured perpendicular to the laminations

37
d c ,R depth of right arch half at crown, measured perpendicular to the
laminations
d haunch ,L depth of section through haunch in left arch half
d haunch , R depth of section through haunch in right arch half
dL depth of section of interest on left arch half
d LD depth of section through the lower discontinuity
d LT ,L depth of section through lower tangent point of left arch half
d LT ,R depth of section through lower tangent point of right arch half
d max, L maximum depth of left arch half when haunch is detached
d max, R maximum depth of right arch half when haunch is detached
dR depth of section of interest on right arch half
dUD depth of section through the upper discontinuity
dUT ,L depth of section through upper tangent point of left arch half
dUT ,R depth of section through upper tangent point of right arch half
Ex reference modulus of elasticity
E x′ adjusted modulus of elasticity
Ex min modulus of elasticity for buckling about the x-axis
E y min modulus of elasticity for buckling about the y-axis
f bx bending stress due to applied loads
Fbx reference design stress in bending
Fbx+ reference design stress for positive bending (inside face of arch in tension)
Fbx− reference design stress for negative bending (outside face of arch in
tension)
Fbx′ adjusted design stress in bending
Fbx* reference design stress in bending adjusted by all applicable factors except
CL
Fbx** reference design stress in bending adjusted by all applicable factors except
CV
fc compression stress parallel-to-grain due to applied loads
Fc reference design stress in compression parallel-to-grain
Fc′ adjusted design stress in compression parallel-to-grain
FcE Euler buckling stress
fr radial stress due to applied loads
Fr′ adjusted radial design stress, either tension or compression
Frc reference design stress in radial compression

38
Frc′ adjusted design stress in radial compression
Frt reference design stress in radial tension
Frt′ adjusted design stress in radial tension
ft tension stress parallel-to-grain due to applied loads
Ft reference design stress in tension parallel-to-grain
Ft′ adjusted design stress in tension parallel-to-grain
Fvx reference design stress in shear
Fvx′ adjusted design stress in shear
Hc overall height of arch
H w, L wall height on left arch half
H w,R wall height on right arch half
I moment of inertia of section of interest
Kr radial stress shape factor
K r ,haunch radial stress shape factor at haunch
K r , LD radial stress shape factor at the lower discontinuity
K r ,UD radial stress shape factor at the upper discontinuity
Kθ bending stress shape factor
Kθ ,arm bending stress shape factor for curved section above haunch
Kθ ,haunch bending stress shape factor at haunch
Kθ ,LD bending stress shape factor at the lower discontinuity
Kθ ,leg bending stress shape factor for curved section below haunch
Kθ ,UD bending stress shape factor at the upper discontinuity
LL outside span of left arch half
LR outside span of right arch half
m1 moment on section of interest due to virtual load P1
m2 moment on section of interest due to virtual load P2
m3 moment on section of interest due to virtual load P3
m4 moment on section of interest due to virtual load P4
m5 moment on section of interest due to virtual load P5
m6 moment on section of interest due to virtual load P6
M moment on section of interest caused by applied loads
MCin − service expected equilibrium moisture content of the arch
N reference capacity for single shear plate in sloped end grain
n number of shear plates connector units at peak connection
P normal force on section of interest caused by applied loads
P parallel-to-grain reference capacity for single shear plate

39
Pcx horizontal point load applied at crown of arch
Pcy vertical point load applied at crown of arch
P1 virtual point load applied vertically at crown
P2 virtual point load applied horizontally at crown
P3 virtual point load applied horizontally at left haunch
P4 virtual load applied horizontally at right haunch
P5 virtual load applied perpendicular to right arm
P6 virtual load applied perpendicular to left arm
Q90 perpendicular-to-grain reference capacity for single shear plate in end
grain
RL radius of curvature of left arch half
Rm radius of curvature measured to the mid-depth of arch
Rm, LD radius of curvature measured to the mid-depth of arch at lower
discontinuity
Rm ,UD radius of curvature measured to the mid-depth of arch at the upper
discontinuity
RR radius of curvature of right arch half
Rx , L horizontal reaction force at left base
Rx , R horizontal reaction force at right base
Ry ,L vertical reaction force at left base
Ry ,R vertical reaction force at right base
s length of segment for virtual work calculation
S snow load
V shear force on section of interest caused by applied loads
Vr′ member shear capacity accounting for notch effect
W wind load
x x-coordinate chosen for analysis
xc x-coordinate of crown
xcenterline ,L x-coordinate of center of section through ( xinner ,L , yinner ,L )
xcenterline,R x-coordinate of center of section through ( xinner , R , yinner ,R )
xinner ,haunch ,L x-coordinate of inner edge of a section through the left haunch
xinner ,haunch , R x-coordinate of inner edge of a section through the right haunch
xinner ,L x-coordinate at any point along inner face of left arch half
xinner ,R x-coordinate at any point along inner face of right arch half
xLT ,L x-coordinate of lower tangent point on left arch half
xLT ,R x-coordinate of lower tangent point on right arch half

40
xouter ,L x-coordinate of outer edge of section through ( xinner ,L , yinner ,L )
xouter ,R x-coordinate of outer edge of section through ( xinner , R , yinner ,R )
xP5 ,centerline x-coordinate of load P5 applied at the centerline of the member
xP6 ,centerline x-coordinate of load P6 applied at the centerline of the member
xUD , L x-coordinate of upper discontinuity on left arch half
xUD , R x-coordinate of upper discontinuity on right arch half
xUT , L x-coordinate of upper tangent point on left arch half
xUT , R x-coordinate of upper tangent point on right arch half
xwall ,L x-coordinate of vertical load due to left wall
xwall ,R x-coordinate of vertical load due to right wall
x0,L x-coordinate of the center of curvature of left arch half
x0,R x-coordinate of the center of curvature of right arch half
yc y-coordinate of crown
ycenterline ,L y-coordinate of center of section through ( xinner ,L , yinner ,L )
ycenterline,R y-coordinate of center of section through ( xinner , R , yinner ,R )
yinner ,haunch , L y-coordinate of inner edge of a section through the left haunch
yinner ,haunch , R y-coordinate of inner edge of a section through the right haunch
yinner ,L y-coordinate at any point along inner face of left arch half
yinner ,R y-coordinate at any point along inner face of right arch half
y LT ,L y-coordinate of lower tangent point on left arch half
y LT ,R x-coordinate of lower tangent point on right arch half
youter ,L y-coordinate of outer edge of section through ( xinner ,L , yinner ,L )
youter ,R y-coordinate of outer edge of section through ( xinner , R , yinner ,R )
yP5 ,centerline y-coordinate of load P5 applied at the centerline of the member
yP6 ,centerline y-coordinate of load P6 applied at the centerline of the member
yUT , L y-coordinate of upper tangent point on left arch half
yUT , R y-coordinate of upper tangent point on right arch half
y0,L y-coordinate of the center of curvature of left arch half
y0,R y-coordinate of the center of curvature of right arch half
α arm angle of taper of arch arm
α arm ,L angle of taper of arm of left arch half
α arm ,R angle of taper of arm of right arch half
α leg angle of taper of arch leg

41
α leg ,L angle of taper of leg of left arch half
α leg ,R angle of taper of leg of right arch half
β haunch ,L angle between horizontal line and line from center of curvature through
haunch for left arch half
β haunch ,R angle between horizontal line and line from center of curvature through
haunch for right arch half
βL angle between horizontal line and section of interest on left arch half
β LD , L angle between horizontal line and lower discontinuity on left arch half
β LD , R angle between horizontal line and lower discontinuity on right arch half
β LT ,L angle between horizontal line and line from center of curvature through
lower tangent point for left arch half
β LT ,R angle between horizontal line and line from center of curvature through
lower tangent point for right arch half
βR angle between horizontal line and section of interest on right arch half
βUD , L angle between horizontal line and upper discontinuity on left arch half
βUD , R angle between horizontal line and upper discontinuity on right arch half
βUT ,L angle between horizontal line and line from center of curvature through
upper tangent point for left arch half
βUT ,R angle between horizontal line and line from center of curvature through
upper tangent point for right arch half
Δ1 , Δ cy vertical displacement at peak
Δ 2 , Δ cx horizontal displacement at peak
Δ3 , Δ x ,L horizontal displacement at left haunch
Δ 4 , Δ x ,R horizontal displacement at right haunch
Δ 5 , Δ arm, L displacement of left arch arm
Δ 6 , Δ arm, R displacement of right arch arm
Δη change in angle η due to moisture content changes
ΔηL change in angle ηL due to moisture content changes
ΔηR change in angle ηR due to moisture content changes
Δxc horizontal displacement of crown due to moisture content changes
Δxhaunch ,L horizontal displacement of haunch due to moisture content changes
Δxhaunch ,R horizontal displacement of haunch due to moisture content changes
Δyc vertical displacement of crown due to moisture content changes
φt angle of roof slope
φt ,L angle of roof slope of left arch half
φt ,R angle of roof slope of right arch half

42
 φb angle of soffit slope
φb,L angle of soffit slope on left arch half
φb,R angle of soffit slope on right arch half
γ angle between horizontal line and line from base to crown
γL angle between horizontal line and line from left base to crown
γR angle between horizontal line and line from right base to crown
η angle between vertical line and arch roof
ηL angle between vertical line and roof of left arch half
ηR angle between vertical line and roof of right arch half
θ included angle between the outer faces of the arch leg and arm
θL included angle between the outer faces of the left arch leg and arm
θR included angle between the outer faces of the right arch leg and arm
ωrx ,L horizontal distributed load on roof arm of left arch half
ωrx ,R horizontal distributed load on roof arm of right arch half
ωry ,L vertical distributed load on roof arm of left arch half
ωry ,R vertical distributed load on roof arm of right arch half
ωwx ,L horizontal distributed load on wall leg of left arch half
ωwx ,R horizontal distributed load on wall leg of right arch half
ωwy ,L vertical distributed load on wall leg of left arch half
ωwy ,R vertical distributed load on wall leg of right arch half
ξL angle used in calculation of ψ L
ξR angle used in calculation of ψ R
ψ angle of rotation of arch leg due to moisture content changes
ψL angle of rotation of left arch leg due to moisture content changes
ψR angle of rotation of right arch leg due to moisture content changes

Procedure

The analysis procedure includes the following steps:

1. Determine all applicable loads and load combinations. Roof and wall loads can typically
be assumed to be uniform if decking is applied directly to the arch (or if purlins are
closely spaced) or they may be concentrated at purlin points. Appropriate loads and load
combinations can be obtained from the applicable building code or from the building
official having jurisdiction.

2. Define the arch geometry (Figure 4.1). The outside geometry (wall heights, peak height,
outside span) of the arch is typically dictated by architectural constraints of the building.

43
 To complete a trial geometry, the designer must choose a width, end depths, a radius, and
angles of taper for the wall leg and roof arm.

dc

dc'

φt,L φb,L

Hc αarm,L

αleg,L

H w,L
RL

d'b,L
LL

db,L
Figure 4.1. Arch geometry. (Only left half shown for clarity.)

2.1. Choose the angles of taper for the wall legs, αleg,L and , αleg,R, and the angles of
taper for the roof arms, αarm,L and αarm,R. (Taper in the wall legs should generally not
exceed 5° and taper in the roof arms should generally not exceed 3°, however, steeper
tapers can be used if they are properly accounted for in design.)

2.2. Calculate the angles of the roof slopes, φt,L and φt,R, and the angles of the soffit
slopes, φb,L and φb,R.

44
 For the left arch half:

⎛ H c − H w,L ⎞
φt ,L = arctan ⎜ ⎟ [4-1]
⎝ LL ⎠
φb,L = φt ,L + α arm ,L [4-2]

For the right arch half:

⎛ H c − H w,R ⎞
φt ,R = arctan ⎜ ⎟ [4-3]
⎝ LR ⎠
φb,R = φt ,R + α arm ,R [4-4]

2.3. Determine the end dimensions for each arch half.

It is convenient to choose the end depths based on the depth of the cuts at the ends.
However stresses should be checked based on the depths of sections taken
perpendicular to the laminations. The following equations relate the depths measured
perpendicular to the laminations to the depths of the end cuts.

The depth at the base, d b , measured perpendicular to the laminations, is calculated as:

d b′
db = [4-5]
cos α leg

The depth at the crown, d c , measured perpendicular to the laminations, is calculated

as:

d c′ cos φt
dc = [4-6]
cos (α arm )

2.4. Choose the radius of curvature for each arch half. The inside radius for the curved
portion of the arch half should be a minimum of 100 times the lamination thickness
for Southern Pine and hardwood species and 125 times the lamination thickness for
other softwood species to enable bending of the laminations to the required curvature.
Using standard lamination thicknesses, minimum radii are:

7 ft 0 in. for Southern Pine and hardwood species

9 ft 4 in. for other softwood species

Tighter curves require the laminations to be planed to a thinner dimension and are
significantly more expensive.

45
2.5. Locate the centers of curvature. Cartesian coordinates with the origin at a point on
the floor under the arch crown are used to locate the centers of curvature (Figure
4.2). These coordinates are used in subsequent calculations. These values can be
determined using the following equations.

(x0,L,y0,L)

(0,0)
Figure 4.2. Coordinate system and center of curvature for left arch half.

For the left arch half:

⎛ 1 1 ⎞
H c − d c′ + RL ⎜
⎜ sin α
− ⎟ − ( LL − d b′,L ) cot α leg ,L
⎝ leg , L cos φb,L ⎟⎠
x0,L = for: α leg ,L > 0 [4-7]
cot α leg ,L − tan φb,L
− LL + d b′,L + RL for: α leg ,L = 0

y0, L = cot α leg , L ( x0, L + LL − db′, L ) −

RL
sin α leg , L
or [4-8]
RL
y0, L = x0, L tan φb , L + H c − d c′ −
cos φb , L

46
 For the right arch half:

⎛ 1 1 ⎞
H c − d c′ + RR ⎜
⎜ sin α
− ⎟⎟ − ( LR − d b′,R ) cot α leg ,R
⎝ , cos φ , ⎠
= for: α leg ,R > 0 [4-9]
leg R b R
x0,R
tan φb,R − cot α leg ,R
LR − d b′,R − RR for: α leg ,R = 0

y0,R = cot α leg ,R ( LR − x0,R − d b′,R ) −

RR
sin α leg ,R
or [4-10]
RR
y0,R = − x0,R tan φb,R + H c − d c′ −
cos φb,R

3. Draw the arch to scale and verify that its appearance is satisfactory. Adjust the trial
geometry until a satisfactory shape is obtained.

4. Divide the arch into segments, locate sections of interest, and calculate their section
properties. Because of the complex geometry of the arch and the variability of loading
conditions, deflections are typically estimated using the principle of virtual work.
Additionally, the critical cross section for stress analysis will not be obvious. Therefore,
several sections, taken perpendicular to the laminations, should be chosen for evaluation
along the length of the arch. To reduce calculations, the same sections chosen for
analysis of stresses can be used for virtual work calculations.

4.1. Locate the sections through the haunches and determine their depths. The sections
through the haunches are located at angles of βhaunch,L and βhaunch,R on radial lines from
the respective centers of curvature (Figure 4.3).

For the left arch half:

⎛H −y ⎞
β haunch ,L = arctan ⎜⎜ w,L 0,L ⎟⎟ [4-11]
⎝ LL + x0,L ⎠
xinner , haunch ,L = x0,L − RL cos β haunch ,L [4-12]
yinner , haunch ,L = y0,L + RL sin β haunch ,L [4-13]
LL + x0,L
d haunch ,L = − RL [4-14]
cos β haunch ,L

47
 For the right arch half:
⎛ H − y0,R ⎞
β haunch ,R = arctan ⎜⎜ w,R ⎟⎟ [4-15]
⎝ LR − x0,R ⎠
xinner , haunch ,R = x0, R + RR cos β haunch ,R [4-16]
yinner , haunch ,R = y0,R + RR sin β haunch ,R [4-17]
LR − x0,R
d haunch ,R = − RR [4-18]
cos β haunch ,R

dhaunch,L

(xinner,haunch,L,yinner,haunch,L )

βhaunch,L

(0,0)
Figure 4.3. Haunch location.

4.2. Locate the upper and lower tangent points for each arch half and determine the
depths at those locations (Figure 4.4). For stability considerations, the tangent point
depths should not exceed six times the width if the arch is laterally braced by decking
fastened directly to the arch or by girts or purlins placed at frequent intervals. When
such lateral bracing is lacking, the depths at the tangent points should not exceed five
times the width. If the depths at the tangent points are excessive, the geometry should
be revised so that the depths are within the prescribed limits. Adjustments may need
to be made to the arch width, angles of taper and/or the radius.

48
For the left arch half:

xLT ,L = x0,L − RL cos α leg ,L [4-19]

y LT ,L = y0,L + RL sin α leg ,L [4-20]

xUT ,L = x0,L − RL sin φb,L [4-21]

yUT ,L = y0,L + RL cos φb,L [4-22]

The angles between a horizontal line and the radial lines to the tangent points are:

β LT ,L = α leg ,L [4-23]
βUT ,L = 90D − φb,L [4-24]

The tangent point depths are determined from the following equations:

LL + xLT , L LL + xLT , L
d LT , L = = [4-25]
cos β LT , L cos α leg , L
⎡( LL + xUT ,L ) tan φt ,L + H w,L − yUT ,L ⎤ cos φt ,L
dUT ,L = ⎣ ⎦ [4-26]
cos ( 90 − βUT ,L − φt ,L )
D

(xUT,L,yUT,L)
dUT,L

d LT,L

βUT,L
(x LT,L,y LT,L)
βLT,L

(x0,L,y 0,L)

Figure 4.4. Tangent point locations (only left arch half shown for clarity).

49
 For the right arch half:

xLT ,R = x0,R + RR cos α leg ,R [4-27]

y LT ,R = y0,R + RR sin α leg ,R [4-28]
xUT ,R = x0,R + RR sin φb,R [4-29]
yUT ,R = y0,R + RR cos φb,R [4-30]

The angles between a horizontal line and the radial lines to the tangent points are:

β LT ,R = α leg ,R [4-31]
βUT ,R = 90 − φb,R
D
[4-32]

The tangent point depths are determined from the following equations:

LR − xLT ,R LR − xLT ,R
d LT ,R = = [4-33]
cos β LT ,R cos α leg ,R
⎡( LR − xUT ,R ) tan φt ,R + H w,R − yUT ,R ⎤ cos φt ,R
dUT ,R = ⎣ ⎦ [4-34]
cos ( 90 − βUT ,R − φt ,R )
D

4.3. If arches will be manufactured with detached haunches, choose the maximum depths,
dmax,L and dmax,R, through the curved segments and locate the points of discontinuity
on the back of each arch half. The depth through the curved segment must be chosen
such that it is equal to or larger than both the upper and lower tangent point depths.

For the left arch half:

The angle between a horizontal line and a radial line to the lower point of
discontinuity is:

⎛ LL + x0, L ⎞
β LD , L = cos −1 ⎜⎜ ⎟⎟ [4-35]
⎝ RL + d max, L ⎠

50
 The x-coordinate of the upper point of discontinuity is:

− ⎡⎣ tan φt , L ( H c − y0, L ) − x0, L ⎤⎦

xUD , L =
⎛ 1 ⎞
⎜⎜ ⎟⎟
⎝ cos φt , L
2
⎠
⎡ 1 ⎤⎡ 2 [4-36]
⎡ tan φt , L ( H c − y0, L ) − x0, L ⎤ − ⎢ 2 ( ) ( ) ⎤
2 2 2
⎣ ⎦ ⎥ ⎢ x + H − y − R + d
⎣ cos φt , L ⎦ ⎣ 0, L c 0, L L max, L
⎦⎥
−
⎛ 1 ⎞
⎜⎜ ⎟⎟
⎝ cos φt , L ⎠
2

The angle between a horizontal line and a radial line to the upper point of
discontinuity is:

⎛ x0, L − xUD , L ⎞
βUD , L = cos −1 ⎜⎜ ⎟⎟ [4-37]
⎝ RL + d max, L ⎠

For the right arch half:

The angle between a horizontal line and a radial line to the lower point of
discontinuity is:

⎛ LR − x0, L ⎞
β LD , R = cos −1 ⎜⎜ ⎟⎟ [4-38]
⎝ RR + d max, R ⎠

The x-coordinate of the upper point of discontinuity is:

⎡ tan φt , R ( H c − y0, R ) + x0, R ⎤

xUD , R = ⎣ ⎦
⎛ 1 ⎞
⎜⎜ ⎟⎟
⎝ cos φt , R ⎠
2

⎡ 1 ⎤⎡ 2 [4-39]
⎡ tan φt , R ( H c − y0, R ) − x0, R ⎤ − ⎢ 2 ( ) ( ) ⎤
2 2 2
⎣ ⎦ ⎥ ⎢ x + H − y − R + d
⎣ cos φt , R ⎦ ⎣ 0, R c 0, R R max, R
⎦⎥
+
⎛ 1 ⎞
⎜⎜ ⎟⎟
⎝ cos φt , R ⎠
2

The angle between a horizontal line and a radial line to the upper point of
discontinuity is:

51
 ⎛ xUD , R − x0, R ⎞
βUD , R = cos −1 ⎜⎜ ⎟⎟ [4-40]
⎝ RR + d max, R ⎠

4.4. Locate points along the inside face of the arch corresponding to sections of interest
using the following equations:

Left arch half, below lower tangent:

(x + L − d b′,L )
yinner ,L =
L

tan α leg ,L
for: (−L L + d b′,L ) ≤ x ≤ xLT ,L and α leg ,L > 0 [4-41]

Left arch half, in curved segment:

xinner , L = x0,L − RL cos β L for: βUT ,L ≤ β L ≤ α leg ,L [4-42]

yinner ,L = y0,L + RL sin β L for: βUT ,L ≤ β L ≤ α leg ,L [4-43]

Left arch half, above upper tangent:

y inner , L = x tan φb,L + H c − d c′ for: xUT ,L ≤ x ≤ 0 [4-44]

Right arch half, above upper tangent:

y inner , R = − x tan φb,R + H c − d c′ for: 0 ≤ x ≤ xUT ,R [4-45]

Right arch half, in curved segment:

xinner ,R = x0,R + RR cos β R for: α leg ,R ≤ β R ≤ βUT ,R [4-46]

yinner ,R = y0,R + RR sin β R for: α leg ,R ≤ β R ≤ βUT ,R [4-47]

Right arch half, below lower tangent:

yinner ,R =
(L R − d b′,R − x )
for: xLT ,R ≤ x ≤ ( LR − d b′,R ) and α leg ,R > 0 [4-48]
tan α leg ,R

4.5. With sections of interest in the arch identified and located, determine the depth at
each section (taken perpendicular to the laminations) and determine the coordinates
of the outer face and centerline of each section. For arches with detached haunches,
maximum depth limits, dL,max and dR,max, chosen by the designer, establish the depth
through the curved segments.

52
The depth (perpendicular to the laminations) in the left arch half, at any section
below the haunch is given by:

LL + xinner , L
dL = ≤ d max, L [4-49]
cos β L

(for points below the lower tangent, β L = α leg ,L )

The depth (perpendicular to the laminations) in the left arch half, at any section
above the haunch is given by:

⎡( LL + xinner , L ) tan φt , L + H w, L − yinner , L ⎤ cos φt , L

dL = ⎣ ⎦ ≤ d max, L [4-50]
(
cos 90 − β L − φt , L
D
)
(for points above upper tangent, β L = 90D − φb,L )

The coordinates of the outer face of any section in the left arch half are given by:

xouter ,L = xinner ,L − d L cos β L [4-51]

youter ,L = yinner ,L + d L sin β L [4-52]

The coordinates of the centerline of any section in the left arch half are given by:

d L cos β L
xcenterline,L = xinner ,L − [4-53]
2
d sin β L
ycenterline,L = yinner ,L + L [4-54]
2

The depth (perpendicular to the laminations) in the right arch half, at any section
above the haunch is given by:

⎡( LR − xinner , R ) tan φt , R + H w, R − yinner , R ⎤ cos φt , R

dR = ⎣ ⎦ ≤ d max, R [4-55]
(
cos 90 − β R − φt , R
D
)
(for points above upper tangent, β R = 90D − φb,R )

53
 The depth (perpendicular to the laminations) in the right arch half, at any section
below the haunch is given by:

LR − xinner , R
dR = ≤ d max, R [4-56]
cos β R

(for points below the lower tangent, β R = α leg ,R )

The coordinates of the outer face of any section in the right arch half are given by:

xouter ,R = xinner ,R + d R cos β R [4-57]

youter ,R = yinner ,R + d R sin β R [4-58]

The coordinates of the centerline of any section in the right arch half are given by:

d R cos β R
xcenterline ,R = xinner ,R + [4-59]
2
d sin β R
ycenterline ,R = yinner ,R + R [4-60]
2

4.6. Calculate the moment of inertia and section modulus for each section.

5. Calculate the base reactions and forces at the peak connection. Distributed loads are
assumed to act on the outer faces of the arch, and base reactions and peak connection
forces act at the arch centerline. Figures 4.5 and 4.6 show the free-body diagrams used
for determination of the base reactions and peak connection forces. The resulting
equations follow.
LL LR

ωry,L
ωry,R

Pcx

ωrx,R
ωrx,L Pcy

Hc

ωwx,R H w,R
H w,L ωwx,L ωwy,L ωwy,R

R x,L R x,R

R y,L R y,R

L L - db,L/2 L R - db,R/2

Figure 4.5. Free body diagram of arch.

54
 LL LR

ωry,L
ωry,R

Cy
Cx Cx Pcx

Cy ωrx,R
ωrx,L Pcy

Hc H c - dc/2

ωwx,R H w,R
H w,L ωwx,L ωwy,L ωwy,R

R x,L R x,R

R y,L R y,R

L L - db,L/2 L R - db,R/2

Figure 4.6. Free body diagram of arch separated at crown connection.

⎡ ωwx ,R H w2 ,R ωrx ,R ( H c2 − H w2 ,R ) ωry ,R LR ( LR − d b′,R ) ⎤

⎢ + + ⎥
⎢ 2 2 2 ⎥
⎢ ω L ( L + 2L − d ′ ) ω ( H 2 − H 2 ) ω H 2 ⎥
⎢ + ry ,L L L R b,R
−
rx , L c w,L
− wx , L w,L
⎥
1 ⎢ 2 2 2 ⎥
Ry ,L =
⎛ d b′,L d b′,R ⎞ ⎢ ⎛ d b′,R ⎞ ⎛ d c′ ⎞ ⎛ d b′,R ⎞⎥
L
⎜ L − + L − ⎢ + − − − + ω − − ⎥
2 ⎟⎠ ⎢ cy ⎜⎝ R 2 ⎟⎠
cx ⎜ ⎟ w,L ⎜ R wall , L ⎟
R P L P H wy , L H L x
⎝ 2 ⎝
c
2⎠ ⎝ 2 ⎠⎥
⎢ ′ ⎥
⎢ −ωwy ,R H w,R ⎛ xwall ,R − LR + d b,R ⎞ ⎥
⎢⎣ ⎜ 2 ⎟⎠ ⎥⎦
⎝
[4-61]

⎡ ωwx ,L H w2 ,L ωrx ,L ( H c2 − H w2 ,L ) ωry ,L LL ( LL − d b′,L ) ⎤

⎢ + + ⎥
⎢ 2 2 2 ⎥
⎢ ω L ( L + 2L − d ′ ) ω ( H 2 − H 2 ) ω H 2 ⎥
⎢ + ry ,R R R L b,L
−
rx , R c w,R
− wx ,R w,R ⎥
1 ⎢ 2 2 2 ⎥
Ry ,R =
⎛ d b′,L d b′,R ⎞ ⎢ ⎛ d ′
b,L ⎞ ⎛ d′ ⎞ ⎛ d ′ ⎞⎥
⎜ LL − 2 + LR − 2 ⎟ ⎢ + Pcy ⎜ LL − ⎟ + Pcx ⎜ H c − c ⎟ + ωwy ,R H w, R ⎜ LL − b,L + xwall ,R ⎟ ⎥
⎝ ⎠⎢ ⎝ 2 ⎠ ⎝ 2⎠ ⎝ 2 ⎠⎥
⎢ ′ ⎥
⎢ −ωwy ,L H w,L ⎛ − xwall ,L − LL + d b,L ⎞ ⎥
⎢⎣ ⎜ 2 ⎟⎠ ⎥⎦
⎝
[4-62]

55
 ⎡ ⎛ d b′,L ⎞ ⎛ H w,L d c′ H w2 ,L ⎞ ⎤
⎢ R y ,L ⎜ LL − − ωwy ,L H w,L ( − xwall ,L ) − ωwx ,L ⎜ H c H w,L − − ⎟⎥
1 ⎢ ⎝ 2 ⎟⎠ ⎝ 2 2 ⎠⎥
Rx , L =
d c′ ⎞ ⎢ ⎥
⎜ H c − ⎟ ⎢ − ωrx ,L ( H c − H w,L )( H c − H w,L − d c ) − ωry ,L LL
⎛ ′ 2
⎥
⎝ 2 ⎠⎢
⎣ 2 2 ⎦⎥
[4-63]

⎡ ⎛ d b′,R ⎞ ⎛ H w,R d c′ H w2 ,R ⎞ ⎤
⎢ R y ,R ⎜ LR − − ωwy ,R H w,R ( xwall ,R ) − ωwx , R ⎜ H c H w,R − − ⎟⎥
1 ⎢ ⎝ 2 ⎟⎠ ⎝ 2 2 ⎠⎥
Rx , R =
d c′ ⎞ ⎢ ⎥
⎜ H c − ⎟ ⎢ − ωrx ,R ( H c − H w,R )( H c − H w,R − d c ) − ωry ,R LR
⎛ ′ 2
⎥
⎝ 2 ⎠⎢
⎣ 2 2 ⎦⎥
[4-64]

C y = ωwy ,L H w,L + ωry ,L LL − R y ,L [4-65]

C x = −ωwx ,L H w,L − ωrx ,L ( H c − H w,L ) − Rx ,L [4-66]

6. Determine forces and moments and corresponding stresses on each section. Arches act
as combined axial and flexural members. General free-body diagrams for both arch
halves at sections both below and above the haunches are shown in Figures 4.7 through
4.10. The diagrams and corresponding equations are general for all cases of uniform
loading with point loads applied at the crown. For other load cases, free-body diagrams
and corresponding equations specific to the load case should be developed for the
analysis.
M P
V
βL

ωwx,L ωwy,L
y outer,L y centerline,L

R x,L

R y,L
Figure 4.7. Free body diagram of left arch half below the haunch.

56
For the left arch half at any section below the haunch, the following equations were
derived from the free-body diagram shown in Figure 4.7. For sections in the leg below
the lower tangent point, the same equations apply with β L = α leg ,L .

⎛ d′ ⎞
M = R y ,L ⎜ xcenterline ,L + LL − b,L ⎟ − Rx ,L ycenterline, L
⎝ 2 ⎠
[4-67]
⎛ ⎞
− ωwx ,L youter ,L ⎜ ycenterline ,L − outer ,L ⎟ − ωwy ,L youter ,L ( xcenterline, L − xwall ,L )
y
⎝ 2 ⎠

P = sin β L ( Rx ,L + ωwx ,L youter ,L ) + cos β L ( R y ,L − ωwy ,L youter ,L ) [4-68]

P sin β L − Rx ,L − ωwx ,L youter ,L

V= [4-69]
cos β L

-xouter,L

ωry,L

Cx

Cy
Hc - youter,L ωrx,L

M
P V
βL

Figure 4.8. Free body diagram of left arch half above the haunch.

For the left arch half at any section above the haunch, the following equations were
derived from the free-body diagram shown in Figure 4.8. For sections in the arm above
the upper tangent point, the same equations apply with β L = 90D − φb,L .

⎛x ⎞ ⎛ d′ ⎞
M = ωry ,L ( xouter ,L ) ⎜ outer ,L − xcenterline ,L ⎟ − C y ( xcenterline,L ) − C x ⎜ H c − c − ycenterline,L ⎟
⎝ 2 ⎠ ⎜ 2 ⎟
⎝ ⎠ [4-70]
⎛ H + youter ,L ⎞
− ωrx ,L ( H c − youter ,L ) ⎜ c − ycenterline,L ⎟
⎝ 2 ⎠

57
P = cos β ⎡⎣ Py ,L − C y − ωry ,L xouter ,L ⎤⎦ − sin β ⎡⎣ Px ,L + C x + ωrx ,L ( H c − youter ,L )⎤⎦ [4-71]

− P cos β L − ωry ,L xouter ,L − C y

V= [4-72]
sin β L

P M
V

βR

ωwy,R ωwx,R y outer,R

y centerline,R

R x,R

R y,R

Figure 4.9. Free body diagram of right arch half, below the haunch.

For the right arch half at any section below the haunch, the following equations were
derived from the free-body diagram shown in Figure 4.9. For sections in the leg below
the lower tangent point, the same equations apply with β R = α leg ,R .

⎛ d′ ⎞
M = R y ,R ⎜ LR − b,R − xcenterline ,R ⎟ − Rx ,R ycenterline, R
⎝ 2 ⎠
[4-73]
⎛ ⎞
− ωwx ,R youter ,R ⎜ ycenterline ,R − outer ,R ⎟ − ωwy ,R youter ,R ( xwall ,R − xcenterline ,R )
y
⎝ 2 ⎠

P = sin β R ( Rx ,R + ωwx ,R youter ,R ) + cos β R ( R y ,R − ωwy ,R youter ,R ) [4-74]

P sin β R − Rx ,R − ωwx ,R youter ,R

V= [4-75]
cos β R

58
 xouter,R

ωry,R

Cy
Cx Pcx

Pcy ωrx,R H c - y outer,R

M
V P
βR

Figure 4.10. Free body diagram for right arch half, above the haunch.

For the right arch half at any section above the haunch, the following equations were
derived from the free-body diagram shown in Figure 4.10. For sections in the arm above
the upper tangent point, the same equations apply with β R = 90D − φb,R .

⎛ d′ ⎞
M = Pcx ⎜ H c − c − ycenterline ,R ⎟ − Pcy ( xcenterline,R ) − C y ( xcenterline, R )
⎝ 2 ⎠
⎛ d′ ⎞ ⎛ ⎞
− C x ⎜ H c − c − ycenterline ,R ⎟ − ωry ,R ( xouter ,R ) ⎜ xcenterline ,R − outer ,R ⎟
x
[4-76]
⎜ 2 ⎟ ⎝ 2 ⎠
⎝ ⎠
⎛ H + youter ,R ⎞
− ωrx ,R ( H c − youter ,R ) ⎜ c − ycenterline ,R ⎟
⎝ 2 ⎠

P = cos β R ⎡⎣ Pcy + C y + ωry ,R xouter ,R ⎤⎦ − sin β R ⎡⎣ Pcx − C x − ωrx ,R ( H c − youter , R )⎤⎦ [4-77]

Pcy − P cos β R + ωry ,R xouter ,R + C y

V= [4-78]
sin β R

7. Determine the adjusted design values (allowable stresses) for each section. Allowable
stresses in compression, bending, and shear, must be determined for each of the sections

59
 chosen for evaluation. Tabulated stresses should be adjusted by all appropriate
adjustment factors. For tapered sections outside of the tangent points, where high-grade
material on the face is removed, exposing lower grades toward the core, a reduction in
the reference flexural stress is appropriate (Appendix A).

Fbx′ = FbxCDCM Ct ( CL or CV ) CI Cc [4-79]

Fvx′ = FvxCDCM Ct [4-80]
Fc′ = FcCDCM Ct CP [4-81]
Ft′ = Ft CDCM Ct [4-82]
Frt′ = Frt CDCM Ct [4-83]
Frc′ = FrcCDCM Ct [4-84]
E x′ = E xCM Ct [4-85]

Out-of-plane buckling of arches should be considered between points of lateral support

through application of the column stability factor, CP. In-plane buckling of the straight
segments of the arch should also be considered through application of the column
stability factor, CP. These segments should be analyzed as tapered columns modeling the
large end as fixed and the small end as pinned for arch legs and modeling both ends as
pinned for arch arms.

It is not necessary to evaluate in-plane buckling in the curved segment of Tudor arches,
because secondary moments due to the eccentricity caused by buckling will generally be
small compared to the bending moment due to the curved shape. Therefore, for in-plane
buckling in the curved segment, the column stability factor is equal to unity (CP = 1.0).

8. Calculate combined stresses on each section. Arches act as combined axial and flexural
members, so stress interactions must be considered. The combined effect of axial
compressive stresses and flexural stresses at each section must satisfy the inequality:

2
⎡ f ⎤ ⎡ f bx ⎤
⎢ c ⎥ +⎢ ⎥ ≤ 1.0 [4-86]
⎢⎣ Fc′ ⎥⎦ ⎢⎣ Fbx′ ⎡⎣1 − ( f c FcE )⎤⎦ ⎥⎦

The following inequalities must be satisfied for sections stressed in combined bending
and tension:

ft f
+ bx* ≤ 1.0 [4-87]
Ft ′ Fbx

f bx − f t
≤ 1.0 [4-88]
Fbx**

60
 The abrupt change in section at the haunch causes stress concentrations that can be
accounted for by multiplying the bending stress calculated using the flexure formula by
an empirical bending stress shape factor, Kθ. For the haunched section, this factor can be
calculated using the following equation:

⎛ 180D − θ ⎞ ⎛ 90D − φt ⎞
Kθ ,haunch = 1 + 2.7 tan ⎜ ⎟ = 1 + 2.7 tan ⎜ ⎟ [4-89]
⎝ 2 ⎠ ⎝ 2 ⎠

For sections between the haunch and the tangent points, it is permissible to estimate
bending stress shape factors, Kθ,leg and Kθ,arm, by interpolating between the value of
Kθ,haunch at the haunch and a value of 1/CI at the tangent point, using the angle, β, as the
index value for interpolation:

⎛ β − β LT ⎞ ⎛ 1 ⎞ 1
Kθ , leg = ⎜ ⎟ ⎜ Kθ ,haunch − ⎟ + for β LT ≤ β ≤ β haunch [4-90]
⎝ β haunch − β LT ⎠ ⎝ CI ⎠ CI
⎛ βUT − β ⎞ ⎛ 1 ⎞ 1
Kθ , arm = ⎜ ⎟ ⎜ Kθ ,haunch − ⎟ + for β haunch ≤ β ≤ βUT [4-91]
⎝ βUT − β haunch ⎠ ⎝ CI ⎠ CI

If a detached haunch is used, two points of discontinuity will be created on the back face
of the arch. At the lower point of discontinuity, the bending stress shape factor, Kθ, can
be calculated using Equation [4-92]. At the upper point of discontinuity Kθ can be
calculated using Equation [4-93].

⎛β ⎞
Kθ , LD = 1 + 2.7 tan ⎜ LD ⎟ [4-92]
⎝ 2 ⎠
⎛ 90D − φt − βUD ⎞
Kθ ,UD = 1 + 2.7 tan ⎜ ⎟ [4-93]
⎝ 2 ⎠

For sections between the points of discontinuity and the tangent points, Kθ can be
determined by interpolation using Equations [4-90] and [4-91], except with the
appropriate angle through the point of discontinuity used in place of βhaunch. Interpolation
between the two points of discontinuity can be used to calculate Kθ for intermediate
sections or the larger value can be used.

9. Evaluate radial stresses in the curved segment. Loads that result in an increase in the
radius of curvature cause radial tension stresses in the curved segment(s). Loads causing
a decreased radius of curvature result in radial compression stresses. These stresses must
satisfy the following inequality:

6M
Fr′ ≥ f r = K r [4-94]
bd 2

61
 The radial stress shape factor at the haunch can be calculated as:
2
⎛d ⎞ ⎛ 180D − θ ⎞
K r ,haunch = 0.29 ⎜ c ⎟ + 0.32 tan1.2 ⎜ ⎟ [4-95]
⎝ Rm ⎠ ⎝ 2 ⎠

Linear interpolation between a value of K r ,haunch at the haunch and a value of 0 at the
tangent points can be used to estimate factors for the sections between the haunch and the
tangent points.

For arches with detached haunches, Kr can be calculated for the section through the lower
point of discontinuity using Equation [4-96] and can be calculated for the section through
the upper point of discontinuity using Equation [4-97]. Interpolation can be used to
estimate factors for the other sections between the tangent points.
2
⎛ d ⎞ 1.2 ⎛ β ⎞
K r ,haunch = 0.29 ⎜ LD ⎟⎟ + 0.32 tan ⎜ LD ⎟ [4-96]
⎜R ⎝ 2 ⎠
⎝ m , LD ⎠
2
⎛ d ⎞ 1.2 ⎛ 90 − φt − βUD ⎞
D
K r ,UD = 0.29 ⎜ UD ⎟⎟ + 0.32 tan ⎜ ⎟ [4-97]
⎜R 2
⎝ m ,UD ⎠ ⎝ ⎠

10. Use the principal of virtual work to calculate arch deflections. The deflections should
be compared to any limits required for the design. In the absence of specified limits, the
limits in Table 1.1 are recommended for arch design.

10.1. Calculate the reactions for a virtual load, P1, applied downward at the peak.

1 ⎡ ⎛ d ′ ⎞⎤
R y ,L = ⎢ P1 ⎜ LR − b,R ⎟ ⎥ [4-98]
⎛ d b′,L d b′,R ⎞ ⎣ ⎝ 2 ⎠⎦
⎜ LL − 2 + LR − 2 ⎟
⎝ ⎠
1 ⎡ ⎛ d ′ ⎞⎤
Rx , L = ⎢ R y ,L ⎜ LL − b,L ⎟ ⎥ [4-99]
⎛ d c′ ⎞ ⎣ ⎝ 2 ⎠⎦
⎜ Hc − ⎟
⎝ 2 ⎠
1 ⎡ ⎛ d ′ ⎞⎤
R y ,R = ⎢ P1 ⎜ LL − b,L ⎟ ⎥ [4-100]
⎛ d b′,L d b′,R ⎞ ⎣ ⎝ 2 ⎠⎦
⎜ LL − 2 + LR − 2 ⎟
⎝ ⎠
1 ⎡ ⎛ d ′ ⎞⎤
Rx , R = ⎢ R y ,R ⎜ LR − b,R ⎟ ⎥ [4-101]
⎛ d c′ ⎞ ⎣ ⎝ 2 ⎠⎦
⎜ Hc − ⎟
⎝ 2⎠
C y = − Ry ,L [4-102]
C x = − Rx , L [4-103]

62
10.2. Calculate the moment on selected sections caused by the virtual load, P1.

For the left arch half at any section below the haunch:

⎛ d′ ⎞
m1 = R y ,L ⎜ xcenterline,L + LL − b,L ⎟ − Rx ,L ycenterline, L [4-104]
⎝ 2 ⎠

For the left arch half at any section above the haunch:

⎛ d′ ⎞
m1 = −C y xcenterline ,L − C x ⎜ H c − c − ycenterline, L ⎟ [4-105]
⎝ 2 ⎠

For the right arch half at any section below the haunch:

⎛ d′ ⎞
m1 = R y ,R ⎜ LR − b,R − xcenterline ,L ⎟ − Rx ,R ycenterline,R [4-106]
⎝ 2 ⎠

For the right arch half at any section above the haunch:

⎛ d′ ⎞
m1 = −C y xcenterline ,R − C x ⎜ H c − c − ycenterline,R ⎟ − P1 ( xcenterline, R ) [4-107]
⎝ 2 ⎠

10.3. Calculate the vertical displacement at the peak for each load combination.

1 Mm1s
Δ1 =
P1
∑ E x′ I
[4-108]

10.4. Calculate the reactions for a virtual load, P2, applied rightward at the peak.

1 ⎡ ⎛ d ′ ⎞⎤
R y ,L = ⎢ − P2 ⎜ H c − c ⎟ ⎥ [4-109]
⎛ d b′,L d b′,R ⎞ ⎣ ⎝ 2 ⎠⎦
⎜ LL − 2 + LR − 2 ⎟
⎝ ⎠
1 ⎡ ⎛ d b′,L ⎞ ⎤
Rx , L = − [4-110]
d c′ ⎞ ⎢⎣ y ,L ⎜⎝ L 2 ⎟⎠ ⎥⎦
R L
⎛
⎜ Hc − ⎟
⎝ 2⎠
1 ⎡ ⎛ d c′ ⎞ ⎤
R y ,R = P ⎜ H − ⎟ [4-111]
⎛ d b′,L d b′,R ⎞ ⎢⎣ ⎝ 2 c
2 ⎠ ⎥⎦
⎜ LL − 2 + LR − 2 ⎟
⎝ ⎠

63
 1 ⎡ ⎛ d ′ ⎞⎤
Rx , R = ⎢ R y ,R ⎜ LR − b,R ⎟ ⎥ [4-112]
⎛ d c′ ⎞ ⎣ ⎝ 2 ⎠⎦
⎜ Hc − ⎟
⎝ 2⎠
C y = − Ry ,L [4-113]
C x = − Rx , L [4-114]

10.5. Calculate the moment on selected sections caused by the virtual load, P2.

For the left arch half at any section below the haunch:

⎛ d′ ⎞
m2 = R y ,L ⎜ xcenterline ,L + LL − b,L ⎟ − Rx ,L ycenterline,L [4-115]
⎝ 2 ⎠

For the left arch half at any section above the haunch:

⎛ d′ ⎞
m2 = −C y xcenterline, L − C x ⎜ H c − c − ycenterline, L ⎟ [4-116]
⎝ 2 ⎠

For the right arch half at any section below the haunch:

⎛ d′ ⎞
m2 = R y ,R ⎜ LR − b,R − xcenterline ,R ⎟ − Rx ,R ycenterline,R [4-117]
⎝ 2 ⎠

For the right arch half at any section above the haunch:

⎛ d′ ⎞ ⎛ d′ ⎞
m2 = P2 ⎜ H c − c − ycenterline ,R ⎟ − C y xcenterline ,R − C x ⎜ H c − c − ycenterline, R ⎟ [4-118]
⎝ 2 ⎠ ⎝ 2 ⎠

10.6. Calculate the horizontal displacement at the peak for each load combination.

1 Mm2 s
Δ2 =
P2
∑ E x′ I
[4-119]

10.7. Calculate the reactions for a virtual load, P3, applied rightward at the haunch on
the left arch half.

− P3 H w,L
R y ,L = [4-120]
⎛ d b′,L d b′,R ⎞
⎜ LL − 2 + LR − 2 ⎟
⎝ ⎠

64
 1 ⎡ ⎛ d′ ⎞ ⎛ d′ ⎞⎤
Rx , L = ⎢ R y ,L ⎜ LL − b,L ⎟ − P3 ⎜ H c − c − H w,L ⎟ ⎥ [4-121]
⎛ d c′ ⎞ ⎣ ⎝ 2 ⎠ ⎝ 2 ⎠⎦
⎜ Hc − ⎟
⎝ 2 ⎠
P3 H w,L
R y ,R = [4-122]
⎛ d b′,L d b′,R ⎞
⎜ L − + L −
2 ⎟⎠
L R
⎝ 2
⎛ d′ ⎞
R y ,R ⎜ LR − b,R ⎟
⎝ 2 ⎠
Rx , R = [4-123]
⎛ d c′ ⎞
⎜ Hc − ⎟
⎝ 2⎠
C y = R y ,R [4-124]
C x = − Rx , R [4-125]

10.8. Calculate the moment on selected sections caused by the virtual load, P3.

For the left arch half at any section below the haunch:

⎛ d′ ⎞
m3 = R y ,L ⎜ xcenterline,L + LL − b,L ⎟ − Rx ,L ycenterline ,L [4-126]
⎝ 2 ⎠

For the left arch half at any section above the haunch:

⎛ d′ ⎞
m3 = C y ( − xcenterline ,L ) − C x ⎜ H c − c − ycenterline ,L ⎟ [4-127]
⎝ 2 ⎠

For the right arch half at any section below the haunch:

⎛ d′ ⎞
m3 = R y ,R ⎜ LR − b,R − xcenterline ,R ⎟ − Rx , R ycenterline ,R [4-128]
⎝ 2 ⎠

For the right arch half at any section above the haunch:

⎛ d′ ⎞
m3 = −C y xcenterline ,R − C x ⎜ H c − c − ycenterline ,R ⎟ [4-129]
⎝ 2 ⎠

10.9. Calculate the horizontal displacement at the left haunch for each load combination.

1 Mm3s
Δ3 =
P3
∑ E x′ I
[4-130]

65
10.10. Calculate the reactions for a virtual load, P4, applied leftward at the haunch on
the right arch half.

P4 H w,R
R y ,L = [4-131]
⎛ d b′,L d b′,R ⎞
⎜ LL − 2 + LR − 2 ⎟
⎝ ⎠
⎛ d′ ⎞
R y ,L ⎜ LL − b,L ⎟
⎝ 2 ⎠
Rx , L = [4-132]
⎛ d c′ ⎞
⎜ Hc − ⎟
⎝ 2⎠
− P4 H w,R
R y ,R = [4-133]
⎛ d b′,L d b′,R ⎞
⎜ LL − 2 + LR − 2 ⎟
⎝ ⎠
1 ⎡ ⎛ d′ ⎞ ⎛ d′ ⎞⎤
Rx , R = ⎢ R y ,R ⎜ LR − b,R ⎟ − P4 ⎜ H c − c − H w,R ⎟ ⎥ [4-134]
⎛ d c′ ⎞ ⎣ ⎝ 2 ⎠ ⎝ 2 ⎠⎦
⎜ Hc − ⎟
⎝ 2⎠
C y = − Ry ,L [4-135]
C x = − Rx , L [4-136]

10.11. Calculate the moment on selected sections caused by the virtual load, P4.

For the left arch half at any section below the haunch:

⎛ d′ ⎞
m4 = R y ,L ⎜ LL − b,L + xcenterline ,L ⎟ − Rx ,L ycenterline,L [4-137]
⎝ 2 ⎠
For the left arch half at any section above the haunch:

⎛ d′ ⎞
m4 = C y ( − xcenterline,L ) − C x ⎜ H c − c − ycenterline ,L ⎟ [4-138]
⎝ 2 ⎠

For the right arch half at any section below the haunch:

⎛ d′ ⎞
m4 = R y ,R ⎜ LR − b,R − xcenterline ,R ⎟ − Rx ,R ycenterline,R [4-139]
⎝ 2 ⎠

For the right arch half at any section above the haunch:

⎛ d′ ⎞
m4 = −C y ( xcenterline, R ) − C x ⎜ H c − c − ycenterline, R ⎟ [4-140]
⎝ 2 ⎠

66
10.12. Calculate the horizontal displacement at the right haunch for each load
combination.

1 Mm4 s
Δ4 =
P4
∑ E x′ I
[4-141]

10.13. Calculate the reactions for a virtual load, P5, applied perpendicular to the arch
arm at a point located between 50-60% of the distance from the upper tangent to the
peak on the left arch half. (This is an estimation of the location of maximum
deflection.)

⎛ d′ ⎞
(
P5 cos φb , L ⎜ LR − b , R − xP5 ,centerline ⎟ − P5 sin φb , L yP5 ,centerline
2
)
Ry , L = ⎝ ⎠ [4-142]
db′, L d′
LL − + LR − b , R
2 2

⎛ d′ ⎞
(
P5 cos φb , L ⎜ LL − b , L + xP5 ,centerline ⎟ + P5 sin φb , L yP5 ,centerline
2
)
Ry , R = ⎝ ⎠ [4-143]
db′, L d′
LL − + LR − b , R
2 2
C y = Ry , R [4-144]

⎛ d′ ⎞
− Ry , R ⎜ LR − b , R ⎟
⎝ 2 ⎠
Cx = [4-145]
d ′
Hc − c
2
Rx , L = − P5 sin φb , L − Cx [4-146]

Rx , R = −Cx [4-147]

10.14. Calculate the moment on selected sections in the left arch half, between the upper
tangent point and the peak, caused by the virtual load, P5.

For sections below the load point:

⎛ d′ ⎞
m5 = Ry , L ⎜ LL − b , L + xcenterline, L ⎟ − Rx , L ( ycenterline, L ) [4-148]
⎝ 2 ⎠

For sections above the load point:

⎛ d′ ⎞
m5 = −Cx ⎜ H c − c − ycenterline, L ⎟ − C y xcenterline, L [4-149]
⎝ 2 ⎠

67
 10.15. Calculate the deflection at the selected point (relative to a line between the upper
tangent and peak) for each load combination. Include only segments between upper
tangent and peak.

1 Mm5 s
Δ5 =
P5
∑ Ex′ I
[4-150]

10.16. Calculate the reactions for a virtual load, P6, applied perpendicular to the arch
arm at a point located between 50-60% of the distance from the upper tangent to the
peak on the right arch half. (This is an estimation of the location of maximum
deflection.)
⎛ d′ ⎞
(
P6 cos φb , R ⎜ LR − b , R − xP6 ,centerline ⎟ + P6 sin φb , R yP6 ,centerline
2
)
Ry , L = ⎝ ⎠ [4-151]
db′, L db′, R
LL − + LR −
2 2

⎛ d′ ⎞
(
P6 cos φb , R ⎜ LL − b , L + xP6 ,centerline ⎟ − P6 sin φb , R yP6 ,centerline
2
)
Ry , R = ⎝ ⎠ [4-152]
db′, L d′
LL − + LR − b , R
2 2

C y = − Ry , L = Ry , R − P6 cos φb, R [4-153]

⎛ d′ ⎞ ⎛ d b′, L ⎞
( ) ( )
P6 cos φb , R xP6 ,centerline − P6 sin φb , R yP6 ,centerline − Ry , R ⎜ LR − b , R
2 ⎟ − R y,L ⎜ L
L −
2 ⎟⎠
Cx = ⎝ ⎠= ⎝
d′ d′
Hc − c Hc − c
2 2
[4-154]
Rx , L = −Cx [4-155]

Rx , R = − P6 sin φb , R − Cx [4-156]

10.17. Calculate the moment on sections in the right arch half, between the upper tangent
point and the peak, caused by the virtual load, P6.

For sections below the load point:

⎛ d′ ⎞
m6 = Ry , R ⎜ LR − b , R − xcenterline , R ⎟ − Rx , R ( ycenterline, R ) [4-157]
⎝ 2 ⎠
For sections above the load point:
⎛ d′ ⎞
m6 = −Cx ⎜ H c − c − ycenterline, R ⎟ − C y ( xcenterline, R ) [4-158]
⎝ 2 ⎠
68
 10.18. Calculate the deflection at the selected point (relative to a line between the upper
tangent and peak) for each load combination. Include only segments between upper
tangent and peak.

1 Mm6 s
Δ6 =
P6
∑ Ex′ I
[4-159]

11. Estimate deflections due to seasoning in service. The thickness of the laminations will
change due to moisture content fluctuations in service. Typically, arches dry in service,
resulting in shrinkage of the laminations. Shrinkage of the laminations results in the
change in shape and displacement of the crown and haunches illustrated in Figures 4.11
and 4.12. As the lamination thickness becomes less, the angle η increases. The
percentage change in angle η is approximately equal in magnitude to the percentage
change in thickness of the laminations.

η+Δη

Figure 4.11. Change in shape of arch half due to seasoning.

11.1. Estimate the expected change of angle, Δη , based on the expected equilibrium
moisture content of the arches in service. A 5% decrease in moisture content
corresponds to 1% decrease in thickness and a 1% increase in angle η . Unless
specified otherwise, the average moisture content at the time of manufacturing
can be estimated as 12%.

1
Δη = − ( MCin−service − 0.12 )η [4-160]
5

69
11.2. Calculate the length, aL or aR, of a line from the base to the crown for each arch
half. The length is calculated for the arch after the moisture content change using
the law of cosines (Figure 4.13).

2
⎛ ⎞ ⎛ ⎞
⎟ cos ( 90 + φt , L − Δη L )
LL LL
aL = H 2
+⎜ ⎟ − 2 H w, L ⎜ [4-161]
w, L
⎜ sin ( 90 − φt , L ) ⎟ ⎜ sin ( 90 − φt , L ) ⎟
⎝ ⎠ ⎝ ⎠
2
⎛ ⎞ ⎛ ⎞
⎟ cos ( 90 + φt , R − Δη R )
LR LR
aR = H 2
+⎜ ⎟ − 2 H w, R ⎜ [4-162]
w, R
⎜ sin ( 90 − φt , R ) ⎟ ⎜ sin ( 90 − φt , R ) ⎟
⎝ ⎠ ⎝ ⎠

90-φ t,R

90-φ t,L

90+φ t,R
90+φ t,L

Angles before dimensional change

90-φ t,R +Δη R

90-φ t,L+Δη L

90+φ t,R -Δη R

90+φ t,L -Δη L

Angles after dimensional change

Figure 4.12. Change in outside arch geometry due to moisture content change.

70
 L R/sin(90-φ t,R)
L L/sin(90-φ t,L)

90-φ t,R +Δη R

90-φ t,L+Δη L

aR 90+φ t,R -Δη R

aL
90+φ t,L -Δη L
H w,R
w,L

γL γR

L L +L R

Figure 4.13. Dimensions of seasoned arch for use in displacement calculations.

11.3. Calculate the angle, γ , between a horizontal line and the line from base to crown
for each arch half (Figure 4.13).

⎡ a L2 − aR2 + ( LL + LR )2 ⎤
γ L = cos ⎢−1
⎥ [4-163]
⎢⎣ 2a L ( LL + LR ) ⎥⎦
⎡ a R2 − a L2 + ( LL + LR )2 ⎤
γ R = cos ⎢
−1
⎥ [4-164]
⎢⎣ 2a R ( LL + LR ) ⎥⎦
11.4. Locate the coordinates of the crown,( xc,yc) after moisture content change.

xc = LR − a R cos γ R = a L cos γ L − LL [4-165]

yc = a R sin γ R = aL sin γ L [4-166]

11.5. Determine the displacement of the crown in each direction due to the change in
moisture content.

Δxc = xc [4-167]
Δyc = yc − H c [4-168]

71
 11.6. Determine the angle of rotation,ψ , of each arch leg due to the change in moisture
content (Figure 4.14).

⎡ LL sin ( 90 + φt , L − Δη L ) ⎤
ξ L = sin −1 ⎢ ⎥ [4-169]
⎣⎢ aL sin ( 90 − φt , L ) ⎦⎥
⎡ LR sin ( 90 + φt , R − Δη R ) ⎤
ξ R = sin −1 ⎢ ⎥ [4-170]
⎢⎣ aR sin ( 90 − φt , R ) ⎥⎦

ψ L = 90D − ξ L − γ L [4-171]
ψ R = 90D − ξ R − γ R [4-172]

ξL ξR
ψL ψR

γL γR

Figure 4.14. Rotation of arch legs due to seasoning.

11.7. Determine the horizontal displacement of the haunches. A negative value indicates
outward displacement.

Δxhaunch ,L = H w,L sinψ L [4-173]

Δxhaunch ,R = H w,R sinψ R [4-174]

11. Design and detail connections. Shear plates are typically used to transfer the vertical
reaction forces at the peak connection. Bolts with steel side plates are used to transfer tension
forces between members at the peak. The base connection typically consists of a bearing seat to
resist outward thrust, with a bolt to resist inward thrust and uplift.

72
Example 4-1: Design of Tudor Arch

Given: A building with outside dimension of 50 ft by 60 ft is to be constructed using Southern

Pine Tudor arches spaced at 15 ft.

Preliminary design of the arches was performed in chapter 2. The trial geometry is
shown in Figure E4.1.

11.0 in

24 ft

12 ft 10.5 ft 12 ft

14.5 in
25 ft 25 ft

Figure E4.1. Trial arch geometry with αarm = 2.5° and αleg = 4°.

Parameters chosen for the trial arch geometry are as follows:

b = 6.75 in
H w, L = H w, R = 144 in = 12 ft
H c = 288 in = 24 ft
db′, L = db′, R = 14.5 in
d c′ = 11.0 in
RL = RR = 126 in = 10.5 ft
α leg ,L = α leg ,R = 4D
α arm ,L = α arm ,R = 2.5D

The arch has continuous lateral bracing along the wall and roof.

The building has no shear walls at the ends, so the arches will be assumed to support all
vertical and lateral loads acting parallel to the planes of the arches.

Loads: the loads have been determined as shown in Figures E4.2 through E4.5. The
wall dead loads are assumed to act at the outer face of the arch.

73
 225 plf

150 plf 150 plf

Figure E4.2. Dead loads.

375 plf

Figure E4.3. Balanced snow load.

585 plf
113 plf

Figure E4.4. Unbalanced snow load.

74
 -123 plf
-84 plf

-84 plf -123 plf

77 plf -114 plf

Figure E4.5. Wind Load.

Material Properties: Fbx+(at tangents) = 2000 psi Fbx-(d>85% tangent depth) = Fbx+
Fbx+(at ends) = 1700 psi Fbx-(d< 85% tangent depth) = 1000 psi
Fc = 1400 psi Ex min = 0.84(106) psi
Ex = 1.6(106) psi Ey min = 0.84(106) psi
Fvx = 215 psi Frt = 70 psi
Fc⊥x = 650 psi Frc = 650 psi
Ft = 1000 psi

Wanted: Evaluate the trial geometry and re-design as appropriate. Determine the deflections of
the arch at the haunch (horizontal), at the peak (vertical) and at a point slightly above the
midway point between the upper tangent and the peak. Consider the following load
combinations:

(1) D + S (balanced snow)

(2) D + S (unbalanced snow)
(3) D + 0.75W + 0.75S (balanced snow)
(4) 0.6D + W

In practice, additional load combinations may need to be considered.

Solution: Using a spreadsheet to perform the calculations, analysis of the arch designed in
chapter 2 shows that the arch is adequate to support all of the considered load
combinations. The reactions calculated based on the trial geometry are shown in Table
E4.1. The combined bending and compression stress ratios for each of the load
combinations are shown in Table E4.2. Shear stress ratios were less than 40%
throughout the arch for all load combinations. Radial compression stresses ratios were
less than 15% for all load combinations. Radial tension was developed in the windward
arch half for the load combination 0.6D + W with a stress ratio of 3%.

75
Table E4.1. Reaction forces (lb) for selected load combinations.
Load R y ,L Rx , L Cy Cx Ry ,R Rx , R
D + S (balanced) 16800 7530 0 -7530 16800 7530
D + S (unbalanced) 13130 7210 -2880 -7210 19180 7210
D + 0.75S +0.75W 12360 4680 520 -4610 12670 6750
0.6D + W 1660 -550 690 640 2070 2210

2
⎛ f ⎞ f bx
Table E4.2. Combined bending and compression stress ratios, ⎜ c ⎟ + .
⎝ Fc′ ⎠ Fbx′ (1 − f c FcEx )
Left Arch Half Right Arch Half
Lower Upper Upper Lower
Leg Arm Arm Leg
Haunch Haunch Haunch Haunch
D + S (balanced) 0.652 0.871 0.697 0.438 0.438 0.697 0.871 0.652
D + S (unbalanced) 0.631 0.849 0.805 0.670 0.810 0.555 0.817 0.617
D + 0.75S +0.75W 0.298 0.403 0.323 0.232 0.321 0.461 0.543 0.409
0.6D + W 0.022 0.025 0.023 0.041 0.279 0.175 0.162 0.126

Some observations from the spreadsheet output summarized in Table E4.2 include:

1. The maximum combined stress ratio due to the load combination D + S

(balanced snow) occurs at a section between the lower tangent and the
haunch. The maximum combined flexural and compressive stresses were
approximately 87% of the allowable combined stresses, indicating that the
lower tangent point depth could be possibly be less.
2. The maximum combined stress ratio in any section above the haunch due to
any load combination was less than 82%, indicating that the depth of the arm
(away from the peak) could be reduced by decreasing the taper angle.
Reducing the depth of the upper tangent point might also allow a reduction of
depth at the lower tangent point. (The lower tangent point depth was increased
to be within 10% of the upper tangent point depth).
3. The D + S (unbalanced snow on right side of roof) load combination produced
the highest vertical reaction force at the crown connection (2870 lb). The
0.6D + W load combination induced 640 lb of tension at the crown
connection and a horizontal reaction of -550 lb at the left support. These
values will be needed for connection design. No uplift occurred at the base.

Based on observations 1 and 2, the taper in the arm of the arch was reduced to 1.9
degrees, and the taper in the leg was reduced to 3.8 degrees, and the arch was reanalyzed.
The redesigned arch would save an average of approximately 2 inches of depth along the
length of arm and a small amount in the leg, resulting in a total savings of 4% of the
volume of the original arch. The redesigned arch is illustrated in Figure E4.6.

76
 11 in

24 ft

12 ft 10.5 ft 12 ft

14.5 in
25 ft 25 ft

Figure E4.6. Redesigned arch with αarm = 1.9° and αleg = 3.8°.

Analysis of this arch using a spreadsheet to perform the calculations shows that the arch
is adequate to support all of the considered load combinations. The reactions calculated
based on the revised geometry are shown in Table E4.3. The combined stress ratios for
each of the load combinations are shown in Table E4.4. As for the first trial geometry,
the shear stresses and the radial stresses were small.

Table E4.3. Reaction forces (lb) for selected load combinations.

Load R y ,L Rx , L Cy Cx Ry ,R Rx , R
D + S (balanced) 16800 7530 0 -7530 16800 7530
D + S (unbalanced) 13130 7210 -2880 -7210 19170 7210
D + 0.75S +0.75W 12360 4680 520 -4610 12670 6750
0.6D + W 1660 -550 690 640 2070 2210

2
⎛ f ⎞ f bx
Table E4.4. Combined bending and compression stress ratios, ⎜ c ⎟ + .
⎝ c⎠
F ′ F ′
bx (1 − f c FcEx )
Left Arch Half Right Arch Half
Lower Upper Upper Lower
Leg Arm Arm Leg
Haunch Haunch Haunch Haunch
D + S (balanced) 0.673 0.911 0.847 0.554 0.554 0.847 0.911 0.673
D + S (unbalanced) 0.651 0.888 1.000 0.766 0.881 0.658 0.856 0.637
D + 0.75S +0.75W 0.308 0.423 0.392 0.247 0.405 0.571 0.567 0.422
0.6D + W 0.022 0.025 0.029 0.047 0.326 0.221 0.168 0.129

77
 Some observations from the spreadsheet output include:

1. The load combination D + S (unbalanced snow on right side of roof) resulted

in a combined stress ratio of 100% in the upper haunch, indicating that the
taper is optimal. The maximum combined stress ratio in the lower haunch
was 91%, however, further experimentation with reduced taper or reduced
base depth resulted in subtle changes in geometry which caused sections
above the haunch to be overstressed.
2. The load combination D + S (unbalanced snow on right side of roof) produced
the highest vertical reaction force at the crown connection (2880 lb). The
0.6D + W load combination induced 640 lb of tension at the crown
connection and a horizontal reaction of -540 lb at the left support. These
values will be needed for connection design. No uplift occurred at the base.

Base Connection Design

The outward horizontal thrust at the bases will be resisted by a bearing seat connection.
The required height of the back plate can be determined using the maximum horizontal
force at the base:

Rx 7530 lb
lb = = = 1.73 in
Fc′⊥b ( 650 psi )( 6.75 in )

A minimum height of 5 inches is typically specified.

The inward horizontal thrust at the base will be resisted by a bolt in double shear. The
value will be taken from the NDS for the condition of a double shear connection with ¼
inch steel side plates. A 3/4 in. diameter bolt loaded perpendicular to the grain of the
member has a reference design value of 2000 psi (NDS 2005). It can be shown by
calculation that Mode III yielding governs the capacity of the bolt. Adjusting for
duration of load gives a value of 3200 psi. This is more than adequate to resist the
required 550 lb outward reaction caused by the 0.6D + W load combination. The
fastener is actually loaded at an angle of ( 90° − αleg ) = 86.2° to the grain. The capacity
will be slightly higher than that under perpendicular to grain loading, but the difference is
small and not worth calculating. A smaller bolt could have been used, but it is common
practice to use ¾ in. or 1 in. bolts for arch connections.

Even though no uplift was calculated at the base of the arch, it is considered good
practice to place the bolt a minimum distance of 7D from the end of the member. This
will locate the bolt 5.25 inches above the base of the arch.

The height of the back plate will be selected as 7 in. Assuming 36 ksi steel (Fb = 24 ksi)
and a ¼ in. positive tolerance on the seat width, the required thickness of the back plate
can be calculated as follows (bearing plate is modeled as a beam with pinned ends and
uniform load):

78
 3Rb′ 3 ( 7580 lb )( 6.75 in + 0.25 in )
t= = = 0.49 in
4hFb 4 ( 7 in )( 24000 psi )

A 1/2 in. thick back plate will be required.

The member will have to be checked for the notch effect when the horizontal component
of the base reaction is outward ( Rx < 1 ). The shear force to be resisted at the base,
perpendicular to the laminations, is calculated as:

V = − Rx cos α leg If no uplift is present (Ry ≥ 1)

V = − Rx cos α leg − Ry sin α leg If uplift is present (Ry < 1)

Therefore:

V = − Rx cos α leg = − ( −550 lb ) cos ( 3.8° ) = 548 lb

Assuming the bolt is placed 6 in. from the outside face of the arch to minimize potential
problems with shrinkage between bolt and bearing seat (Figure E4.7), the effective depth,
de, and the total depth, d, of the section through the bolt hole can be calculated as:

xh 6 in
de = = = 6.01 in
cos α leg cos ( 3.8° )

d = d e + ( db′ − xh ) cos α leg + yh sin α leg

d = 6.01 in + (14.5 in − 6.0 in ) cos ( 3.8° ) + ( 5.25 in ) sin ( 3.8° ) = 14.8 in

d
de

y h = 5.25 in

α leg = 3.8o

xh = 6.0 in
db' =14.5 in
Figure E4.7. Bolt placement and dimensions for notched shear calculation.

79
 The shear resistance of the member at the connection can be calculated as:

2 2
⎡2 ⎤ ⎡d ⎤ ⎡2 ⎤ ⎡ 6.01 in ⎤
Vr′ = ⎢ Fvx′ bd e ⎥ ⎢ e ⎥ = ⎢ ( 215 psi )(1.6 )( 6.75 in )( 6.01 in ) ⎥ ⎢ = 1534 lb
⎣3 ⎦ ⎣ d ⎦ ⎣3 ⎦ ⎣ 14.8 in ⎥⎦

The capacity of the member at the connection is adequate with the bolt placed 6 in. from
the outside face (1534 lb > 548 lb).

One 3/4 in. bolt with ¼ in. steel side plates will be chosen for the connection. The
connection will be detailed with the bolt positioned 5.25 inches from the base of the arch
and 6 inches from the outside face.

Peak Connection Design

The shear forces at the peak will be transferred by two 4 in. shear plates as determined
previously. Because the arm taper angle was changed from 2.5° to 1.9°, the angle of load
to the grain will be increased, and the capacity of the connectors will be reduced and
should be re-evaluated. The required spacing and edge distances will also be reduced, so
the previously chosen spacing, edge distance, and geometry factor can be used
conservatively.
⎛ 12 ft ⎞
φt = arctan ⎜ ⎟ = 25.6°
⎝ 25 ft ⎠
α arm = 1.9°
90° − φt − α arm = 90° − 25.6° − 1.9° = 62.5°

Hankinson’s formula can be used to determine the reference capacity for a single shear
plate loaded at an angle to the grain of 90° − φt − α arm .

PQ90
N=
P sin ( 90° − φt − α arm ) + Q90 cos 2 ( 90° − φt − α arm )
2

N=
( 4360 lb )(1824 lb )
( 4360 lb ) sin ( 62.5° ) + (1824 lb ) cos 2 ( 62.5° )
2

N = 2082 lb

The adjusted capacity (dry-use, normal temperatures, snow load) of the connection will
be:

nN ′ = 2 ( NCD CΔCg ) = 2 ( 2082 lb )(1.15 )( 0.68 )(1.0 ) = 3260 lb ≥ 2950 lb ∴ OK

The shear capacity of the members at the crown must also be evaluated for the notch
effect. The effective depth, perpendicular to the laminations, must be determined and the

80
shear capacity of the member determined. This shear capacity must be compared to the
resultant shear force from the reaction forces at the crown. The depth of the notched
member is calculated as:

⎛ inside groove diameter ⎞

d c′ e = d c′ − E + ⎜ ⎟
⎝ 2 ⎠
3.62 in
d c′ e = 11.0 in − 2.75 in + = 10.1 in
2

The effective depth measured perpendicular to the laminations is:

dc e = ce =
(
d ′ cos φt (10.1 in ) cos 25.6
D
) = 9.1 in
cos (α arm ) cos 1.9D( )
The depth of the full section measured perpendicular to the laminations is:

d c′ cos φt (11.0 in ) cos ( 25.6D )

dc = = = 9.9 in
cos (α arm ) cos (1.9D )

The member shear capacity at the connection is:

2 2
⎡2 ⎤ ⎡d ⎤ ⎡2 ⎤ ⎡ 9.1 in ⎤
Vr′ = ⎢ Fvx′ bd c e ⎥ ⎢ c e ⎥ = ⎢ ( 215 psi )(1.15 )( 6.75 in )( 9.1 in ) ⎥ ⎢ = 8560 lb
⎣3 ⎦ ⎣ dc ⎦ ⎣ 3 ⎦ ⎣ 9.9 in ⎥⎦

The shear force (perpendicular to the member) to be resisted by the member at the peak is
calculated as:

( ) ( )
V = C y cos φb + C x sin φb = ( 2880 lb ) cos 25.6D + 1.9D + ( 7210 lb ) sin 25.6D + 1.9D = 5880 lb

Because Vr′ = 8560 lb ≥ V = 5880 lb , the notched shear capacity of the members at the
trial crown connection is adequate.

The tension force of 640 lb will be transferred using bolts and ¼ in. steel side plates. A
single 3/4 in. bolt has a reference capacity of 2000 lbs when loaded perpendicular to grain
and 3480 lb when loaded parallel to grain. The capacity can be further increased by the
load duration factor of 1.6. It is clear that a single 3/4 in. bolt on each arch half will be
adequate to transfer the required tension force. A smaller bolt could have been used, but
it is common practice to use ¾ in. or 1 in. bolts for arch connections.

The arch is adequate to resist all of the loads considered.

81
Deflections Due to Loads

The deflections should be checked against applicable limits. Applying the limits of Table
1.1, the following deflection limits can be determined (Table E4.5):

Table E4.5. Deflection limits for arch example.

Haunch Arms Crown
With Brittle H 144 in l arm 245 in LL + LR 600 in
w
= = 1.2 in = = 0.7 in = = 1.7 in
Finishes 120 120 360 360 360 360
With Flexible H w 144 in l arm 245 in LL + LR 600 in
= = 2.4 in = = 1.4 in = = 3.3 in
Finishes 60 60 180 180 180 180

Deflections due to the considered load combinations were estimated using the principle
of virtual work and are shown in Table E4.6. (Calculated deflections are shown to
nearest 0.01 in. to facilitate comparison with a spreadsheet developed by the reader,
however, this level of precision is not justifiable for design. It is generally sufficient to
estimate displacements to the nearest 0.1 in.)

Table E4.6. Deflections (in.) for selected load combinations.

Load Δ cy Δ cx Δ arm , L Δ arm , R Δ x ,L Δ x ,R
D + S (balanced) 0.72 0 0.31 0.31 -0.37 -0.37
D + S (unbalanced) 0.69 -0.85 -0.46 1.04 -1.26 0.54
D + 0.75S +0.75W 0.70 0.31 0.29 -0.09 -0.03 -0.69
0.6D + W 0.28 0.42 0.11 -0.40 0.29 -0.59

If the structure will be finished with flexible materials, the arch deflections will be within
the acceptable limits. However, if the structure is finished with brittle materials, the
deflection of the right arch arm under the unbalanced snow load exceeds the limit.
Deflections at the crown and haunches can be decreased by increasing the size of the arch
or through the use of a diaphragm and shear wall system as described in the next chapter.
However, the use of a diaphragm and shear wall system will also cause a redistribution of
moments in the arch. The change in moment distribution may increase or decrease the
deflections in the arms.

Deflections Due to Seasoning

Additionally, deflections will be affected by seasoning of the arch in service. Assuming

the arch will equilibrate to a moisture content of approximately 7%, the deflections due to
seasoning can be estimated following the procedures of Section 11.

1 1
Δη L = Δη R = − ( MCin− service − 0.12 )η L = − ( 0.07 − 0.12 )( 58.7° ) = 0.587D
5 5

82
 2
⎛ ⎞ ⎛ ⎞
⎟ cos ( 90° + φt , L − Δη L )
LL LL
aR = aL = H w2 , L + ⎜ ⎟ − 2 H w, L ⎜
⎜ sin ( 90° − φt , L ) ⎟ ⎜ sin ( 90° − φt , L ) ⎟
⎝ ⎠ ⎝ ⎠
2
⎛ 300 in ⎞ ⎛ 300 in ⎞
(144 in ) − 2 (144 in ) ⎜ cos ( 90° + 25.64° − 0.587° )
2
aR = aL = +⎜ ⎟
⎜ sin ( 90° − 25.64° ) ⎟ ⎜ sin ( 90° − 25.64° ) ⎟⎟
⎝ ⎠ ⎝ ⎠
aR = aL = 414.8 in

⎡ aL2 − aR2 + ( LL + LR )2 ⎤ −1 ( 414.8 in ) − ( 414.8 in ) + ( 600 in )

⎡ 2 2 2
⎤
γ R = γ L = cos ⎢ −1
⎥ = cos ⎢ ⎥
⎢⎣ 2aL ( LL + LR ) ⎥⎦ ⎢⎣ 2 ( 414.8 in )( 600 in ) ⎥⎦
γ R = γ L = 43.68D

xc = aL cos γ L − LL
xc = ( 414.8 in ) cos ( 43.68° ) − 300 in = 0
Δxc = xc = 0

yc = aL sin γ L = ( 414.8 in ) sin 43.68D = 286.5 in

Δyc = 286.5 in − 288 in = −1.5 in (downward)

⎡ LL sin ( 90 + φt , L − Δη L ) ⎤
ξ R = ξ L = sin −1 ⎢ ⎥
⎢⎣ aL sin ( 90 − φt , L ) ⎥⎦
⎡ ( 300 in ) sin ( 90° + 25.64° − 0.587° ) ⎤
ξ R = ξ L = sin −1 ⎢ ⎥ = 46.62
D

⎣⎢ ( 414.8 in ) (
sin 90 ° − 25.64° ) ⎦⎥

ψ R = ψ L = 90° − ξ L − γ L = 90° − 46.62° − 43.68° = −0.30°

Δxhaunch , R = Δxhaunch , L = H w, L sinψ L = (144 in ) sin( −0.30D ) = −0.75 in (outward)

The designer will have to determine if the deflections due to seasoning are acceptable. If
they are determined to be excessive, increasing the angles of taper in the arm and leg will
reduce the angle η , leading to slight decreases in displacements. Placing additional
restrictions on the moisture content of laminations used in the manufacture of the arch is
an effective method that may be considered. However, the specification of lower
lamination moisture contents may require additional time and expense for manufacturing.

Consideration should be given to specifying that the end walls be built with the peak at a
lower elevation to accommodate the anticipated downward deflection of the arch crown
due to seasoning, so the ridge of the structure does not appear to sag after the arches
season.

83
Conclusion

This chapter presented equations for the design of Tudor arches. Arch geometry was established
and free body diagrams were used to derive formulas for moment, shear, and axial force at any
section in the arch under common load assumptions. Requirements and recommendations for
analysis of stresses were given and a method of estimating displacements was presented.
Efficient use of this information requires the development of computerized software or
spreadsheets. Once created, however, the software or spreadsheets can be used to efficiently
design Tudor arches.

84
 Chapter 5
Arch-Diaphragm Interaction

Glulam arches are often used in conjunction with a diaphragm and shear walls to support the
design loads on a structure. Diaphragm and shear wall systems are typically stiff relative to
structural glued laminated timber arches, so the diaphragm acts as a partial support to the
arch. The use of a diaphragm and shear walls can significantly reduce the overall deflection
of an arch, resulting in structurally efficient design. The following additional steps can be
followed to evaluate the effects of a diaphragm supporting the arch.

Nomenclature

Ac cross sectional area of each diaphragm chord (in2)

Cx horizontal component of pin force at arch crown
Cy vertical component of pin force at arch crown
D dead load
E modulus of elasticity of segment (psi)
Ec modulus of elasticity of diaphragm chords (psi)
Ga apparent diaphragm shear stiffness from nail slip and panel shear
deformation (kip/in)
I moment of inertia of segment (psi)
K x ,a arch stiffness in the horizontal direction for load applied at peak (lb/in)
K x ,d diaphragm stiffness in the horizontal direction (lb/in)
K y ,a arch stiffness in the vertical direction for load applied at peak (lb/in)
K y ,d diaphragm stiffness in the vertical direction (lb/in)
l length of diaphragm perpendicular to plane of arch (ft)
m1 moment in segment caused by P1 (in-lb)
m2 moment in segment caused by P2 (in-lb)
Px horizontal support load (positive rightward) (lb)
Py vertical support load (positive downward) (lb)
P1 vertical unit load applied at the peak for virtual work calculation (lb)
P2 horizontal unit load applied at the peak for virtual work calculation (lb)
Rx ,L horizontal reaction force at left base (lb)
Rx , R horizontal reaction force at right base (lb)
Ry ,L vertical reaction force at left base (lb)
Ry ,R vertical reaction force at right base (lb)
S snow load
s segment length for virtual work calculation (in.)
sa tributary width for loads on arch (arch spacing) (ft)
u distance from chord splice to nearest support (ft)
85
 v allowable unit shear in the diaphragm (lb/ft)
W wind load
Wx projected width of diaphragm parallel to direction of load (ft)
Wy projected width of diaphragm in vertical direction (ft)
δ x ,d deflection of diaphragm in horizontal direction at the design shear (in.)
δ x ,shearwall horizontal displacement from deformation in the shear wall
δ y ,d deflection of diaphragm in vertical direction at induced unit shear (in.)
Δ1x deflection of the arch in the x-direction due to applied loads, assuming no
diaphragm support
Δ1 y deflection of the arch in the y-direction due to applied loads, assuming no
diaphragm support
Δ arm, L displacement of left arch arm
Δ arm , R displacement of right arch arm
Δc diaphragm chord splice slip at the induced unit shear, v, in diaphragm.
This value can be estimated as 0.03 in. for tension chord splices and 0.005
in. for compression chord splices (APA Engineered Wood Handbook,
2002).
Δ cy vertical displacement at peak
Δ cx horizontal displacement at peak
Δx horizontal displacement of peak (in.)
Δ x ,L horizontal displacement at left haunch
Δ x ,R horizontal displacement at right haunch
Δy vertical displacement of peak (in.)
ωrx ,L horizontal distributed load on roof arm of left arch half (lb/in)
ωrx ,R horizontal distributed load on roof arm of right arch half (lb/in)
ωry ,L vertical distributed load on roof arm of left arch half (lb/in)
ωry ,R vertical distributed load on roof arm of right arch half (lb/in)
ωwx ,L horizontal distributed load on wall leg of left arch half (lb/in)
ωwx ,R horizontal distributed load on wall leg of right arch half (lb/in)
ωwy ,L vertical distributed load on wall leg of left arch half (lb/in)
ωwy ,R vertical distributed load on wall leg of right arch half (lb/in)

Procedure

The supporting effect of a diaphragm on a three-hinged arch can be modeled by linear spring
elements attached to the peak of the arch (Figure 5.1). The diaphragm does not add any
flexural stiffness to the individual arch members.

86
 ωry,L
ωry,R

Px = Kx,dΔ

ωrx,R
ωrx,L

Py = Ky,dΔ
ωwx,R
ωwx,L ωwy,L ωwy,R

Rx,L Rx,R

Ry,L Ry,R

Figure 5.1. Spring model representing diaphragm support on arch.

The system is statically indeterminate, so the diaphragm must be designed, and a trial
geometry for the arch must be chosen prior to determining the spring forces. Once the spring
loads are determined for a given loading, the arch can be analyzed using the equations
developed in the previous chapter.

The analysis is performed for an arch at the mid-span of the diaphragm, which will receive
the least amount of support from the diaphragm and deflect the most. The spring forces on
this arch can be determined using the following procedure:

1. Design the roof diaphragm and shear walls to support all of the lateral roof loads and
an estimated 5/8 of the lateral wall loads. Although the arches will reduce the load
on the diaphragm, the added design iterations and complexity necessary to determine
the magnitude of the load reductions is generally not justified.

2. Determine the stiffness of the diaphragm in each direction, Kx,d and Ky,d at the point of
maximum deflection. For structural panel diaphragms the stiffness can be estimated
as follows.

δ x ,d =
5vl 3
+
0.25vl
+
∑ ( uΔ c ) + δ [5-1]
x , shearwall
8Ec AcWx 1000Ga 2Wx
2vWx sa
K x ,d = [5-2]
l δ x ,d

δ y ,d =
5vl 3
+
0.25vl
+
∑ ( uΔ c ) [5-3]
8 Ec AcWy 1000Ga 2Wy

87
 2vWy sa
K y ,d = [5-4]
l δ y ,d

3. Calculate the stiffness of the arch, Ka, for point loads applied at the crown in each of
the horizontal and vertical directions using virtual work.

Apply a virtual unit load, P1, at the crown in the vertical direction. Calculate the
vertical deflection at the peak and the corresponding stiffness of the arch. (For this
analysis, the virtual load and the “real” load are equal).

m12 s
P1Δ y = ∑ [5-5]
EI
P1
K y ,a = [5-6]
Δy

Apply a virtual unit load, P2, at the crown in the horizontal direction. Calculate the
horizontal deflection at the peak and the corresponding stiffness of the arch. (For this
analysis, the virtual load and the “real” load are equal).

m22 s
P2 Δ x = ∑ [5-7]
EI
P2
K x ,a = [5-8]
Δx

4. Calculate the support forces from the diaphragm on the arch crown for each load
combination.

−Δ1 x K x ,a K x ,d
Px = [5-9]
K x ,a + K x ,d
−Δ1 y K y ,a K y ,d
Py = [5-10]
K y ,a + K y ,d

5. Analyze the arch using the procedures of the previous chapter with the support forces
determined in step 4 of this chapter applied at the crown of the arch in addition to all
of the other loads. Support forces from the diaphragm will be different for each load
combination.

88
Example 5-1: Arch-Diaphragm Interaction

Given: The building from Example 4-1 will be modified to have shear walls at the ends and a
structural panel diaphragm. The structural diaphragm (designed elsewhere) will consist of
7/16 OSB panels with 8d nails spaced at 6 inch intervals along all panel edges and 12
inches in the field of the panel. One 2 x 6 N2 SP chord will be used along each edge of
the roof, and two 2 x 6 chords (or equivalent glulam) will be used at the ridgeline.

Wanted: Design the diaphragm and calculate its stiffness.

Determine the effect of the shearwall and diaphragm system on the arches from Example
3-1 using the same load combinations. Neglect the deflection of the shear walls.

Solution: From the Special Design Provisions for Wind and Seismic (AF&PA 2005), the
following values can be obtained for the diaphragm:

lb
v = 255
ft
kip
Ga = 14.0
ft

The stiffness properties of each 2 x 6 chord are:

A2 x 6 = (1.5 in )( 5.5 in ) = 8.25 in 2

Ec = 1.6 (106 ) psi

Assuming that chord splices will occur at every arch:

∑ ( uΔ ) = (15 ft + 30 ft + 15 ft )( 0.03 in ) + ( 0.005 in )(15 ft + 30 ft + 15 ft )( 0.005 in )

c

∑ ( uΔ ) = 2.1 ft-in.
c

Determine diaphragm stiffness in the x-direction.

+∑
5vl 3 0.25vl ( uΔ c )
δ x ,d = + + δ x ,shearwall
8Ec AcWx 1000Ga 2Wx
5 ( 255)( 60 ) 0.25 ( 255)( 60 )
3
2.1
δ x ,d = + + +0
(
8 1.6 (10 6
) ) (8.25) 50 1000 (14.0 ) 2 ( 50 )

δ x ,d = ( 0.0522 + 0.273 + 0.021) in

δ x ,d = 0.346 in

2vWx sa 2 ( 255 lb/ft )( 50 ft )(15 ft ) lb

K x ,d = = = 18.42 (103 )
l δ x ,d ( 60 ft )( 0.346 in ) in

89
Determine diaphragm stiffness in the y-direction.

δ y ,d =
5vl 3
+
0.25vl
+
∑ ( uΔ c )
8 Ec AcWy 1000Ga 2Wy
5 ( 255)( 60 ) 0.25 ( 255)( 60 )
3
2.1
δ y ,d = + + +0
(
8 1.6 (106 ) ) (( 2 ) 8.25)12 1000 (14.0 ) 2 (12 )

δ y ,d = ( 0.1087 + 0.273 + 0.0875) in

δ y ,d = 0.469 in

2vWy sa 2 ( 255 lb/ft )(12 ft )(15 ft ) lb

K y ,d = = = 3.26 (103 )
l δ y ,d ( 60 ft )( 0.469 in ) in

Determine stiffness of the arch in x and y directions for point loading at the crown

Using a spreadsheet to apply the principle of virtual work, the stiffness of the arch in the
x- and y-directions is determined to be:

lb
K y ,a = 4970
in
lb
K x ,a = 5340
in

Calculate the support forces on the arch for each load combination

−Δ1 y K y ,a K y ,d −Δ1 y ( 4970 lb/in )( 3260 lb/in )

Py = = = (1969 lb/in ) ( −Δ1 y )
K y ,a + K y ,d 4970 lb/in + 3260 lb/in
−Δ1x K x ,a K x ,d −Δ1x ( 5340 lb/in )(18420 lb/in )
Px = = = ( 4140 lb/in )( −Δ1x )
K x ,a + K x ,d 5340 lb/in + 18420 lb/in

Table E5.1. Calculation of support forces.

Load Δ1y (in) Δ1x (in) Py (lb) Px (lb)
D + S (balanced) 0.72 0 -1418 0
D + S (unbalanced) 0.69 -0.85 -1360 3520
D + 0.75S +0.75W 0.70 0.31 -1370 -1290
0.6D + W 0.28 0.42 -560 -1720

90
With the support forces determined, the arch is reanalyzed according to the procedures of
Chapter 4

Table E5.2. Reaction forces (lb) for arch with diaphragm support.
Load R y ,L Rx , L Cy Cx Ry ,R Rx , R
D + S (balanced) 16090 6800 710 -6800 16090 6800
D + S (unbalanced) 10750 4740 -500 -4740 20190 8260
D + 0.75S +0.75W 12300 4610 580 -4550 11360 5390
0.6D + W 2210 20.0 140 64.0 960 1060

Table E5.3. Deflections (in.) for selected load combinations.

Load Δ cy Δ cx Δ arm, L Δ arm, R Δ x ,L Δ x ,R
D + S (balanced) 0.44 0 0.46 0.46 -0.22 -0.22
D + S (unbalanced) 0.42 -0.19 0.06 0.82 -0.42 -0.01
D + 0.75S +0.75W 0.42 0.07 0.30 0.19 -0.14 -0.29
0.6D + W 0.17 0.09 -0.01 -0.16 0.01 -0.19
2
⎛ f ⎞ f bx
Table E5.4. Combined bending and compression stress ratios, ⎜ c ⎟ + .
⎝ Fc′ ⎠ Fbx′ (1 − f c FcEx )
Left Arch Half Right Arch Half
Lower Upper Upper Lower
Leg Arm Arm Leg
Haunch Haunch Haunch Haunch
D + S (balanced) 0.606 0.819 0.726 0.513 0.513 0.726 0.819 0.606
D + S (unbalanced) 0.425 0.578 0.570 0.386 0.789 0.828 0.988 0.734
D + 0.75S +0.75W 0.304 0.417 0.385 0.251 0.257 0.401 0.444 0.333
0.6D + W 0.015 0.029 0.046 0.033 0.126 0.074 0.065 0.054

Some comparisons can be made with the design from the previous chapter where
identical arches were used, but shear walls and a diaphragm were not used:

1. The combined flexural and compressive stresses due to the load combination
D + S (balanced snow) decreased in sections below the haunch, but increased
in sections above the haunch.
2. For the load combination D + S (unbalanced snow on right side of roof), the
diaphragm support caused reductions in stress ratios on the left arch half, but
the maximum stress ratios through the haunch of the right arch half increased
as a result of the diaphragm support. The combined stress ratio increased
from 86% to 99% in sections between the lower tangent and haunch of the
right arch half.
3. The tension at the crown connection due to the 0.6D + W load combination
was reduced to 64 lb (as compared to 640 lb without the diaphragm support).
The horizontal base reaction, Rx,L, is no longer negative, so no forces will be
transferred through the bolt at the base.

91
Conclusion

This chapter has presented a model for consideration of system effects between a Tudor arch and
a roof diaphragm. In the example provided, the arch deflected much less than the same arch
without diaphragm support. In general, the deflections and stresses were reduced by the use of
the diaphragm and shear walls in conjunction with the arches. In some sections, however, the
stresses increased when the diaphragm support was present. The deformed shape of the arch
changed due to the support forces at the peak. Therefore, it may not be conservative in all cases
to size the arch as a single member, then use it in conjunction with a diaphragm and shear wall
system. The complete system should be modeled as accurately as possible for the analysis of the
arch members.

92
 Chapter 6
Seismic Design

The design of structures to resist seismic loads is somewhat different than design for wind and
gravity loads. Ductility of the structure becomes an important consideration. Yielding
mechanisms in the vertical components (walls and frames) of the structure reduce the loads
transmitted to the rest of the structure. Simplified methods have been developed to account for
the non-linear inelastic yielding of the structure while using elastic analysis. These methods
require the use of seismic coefficients specific to the structural system. In seismic-critical zones,
special detailing may be required to ensure ductile yielding can occur.

Glulam arches can be used to resist seismic forces alone or in conjunction with a
shearwall/diaphragm system. This chapter will present information for designing the arch as the
primary lateral force resisting system. When used in conjunction with shearwalls and a
diaphragm, the lower seismic design coefficients for the arch or shearwalls should be used. The
model presented in Chapter 4 can be used to analyze the interaction of the two systems.

Nomenclature

b width
C/D ratio of fastener capacity to demand
Cd deflection amplification factor
Cd penetration factor
Ceg end grain factor
CM wet service factor
Cst metal side plate factor
Ct temperature factor
Cx horizontal component of pin force at arch crown
Cy vertical component of pin force at arch crown
D dead load
d depth
de effective depth
dt b ratio of tangent point depth to width
fc compression stress from applied loads
Fc* partially adjusted compression stress
Fvx′ allowable shear stress
hw wall height
I importance factor
KF format conversion factor
n number of fasteners
Pr parallel to grain reference rivet capacity

93
 QE lateral seismic load effect
Qr Qr perpendicular to grain reference rivet capacity
R response modification factor
Rx horizontal reaction force at base
Rx , L horizontal reaction force at left base
Rx , R horizontal reaction force at right base
Ry vertical reaction force base
Ry ,L vertical reaction force at left base
Ry ,R vertical reaction force at right base
S snow load
SDS design spectral response acceleration parameter
v distributed seismic shear
V seismic base shear
Vr′ shear resistance of member at connection
W effective seismic weight of structure
xh horizontal distance from back of arch to bolt hole at base connection
yh vertical distance from floor to bolt hole at base connection
Z reference lateral design value for a single fastener
α leg leg taper angle
Δ allowable horizontal displacement (drift)
Δ x ,L horizontal displacement of left arch half at haunch
Δ x ,R horizontal displacement of right arch half at haunch
λ time effect factor
φz LRFD resistance factor for fasteners
ρ reliability factor for redundancy
ω distributed seismic weight of structure
Ω overstrength factor

Seismic Coefficients

Two sets of seismic coefficients have been established for glulam arch systems. One set of
values is limited to zones of low seismic hazard (seismic categories A, B, and C) and requires no
special detailing. The other set of coefficients is required for zones of high seismic hazard
(seismic categories D, E, and F) and requires special detailing to increase the likelihood of
ductile yield mechanisms. The seismic design coefficients for each system are shown in Table
5.1.

94
Table 6.1. Seismic Design Coefficients for Glulam Arches.
Seismic Force Resisting System R Ω Cd
1. Special glulam arch 2.5 2.5 2.5
2. Glulam arch not specifically detailed for
seismic resistance - limited to seismic design
categories A, B and C 2.0 2.5 2.0

Detailing Requirements

For timber structures, ductility is obtained through fastener yielding and localized wood crushing
mechanisms at connections. Failure of the timber members themselves generally produces
undesirable brittle failure mechanisms. For the special glulam arch system, the detailing
requirements discussed below provide additional assurance that the timber components of the
arch system will not fail prematurely before yielding can occur at the base of the arch. This is
achieved by sizing the member to resist the over-strength load combinations of ASCE 7 or
ensuring that the members have more capacity than required to develop the nominal strength of
the base connection. To promote yielding, the fastener(s) at the base are designed to resist the
seismic loads without considering the over-strength load combinations.

The following provisions are taken from a paper titled “Special Requirements for Seismic
Design of Structural Glued Laminated Timber (Glulam) Arch Members and Their Connections
in Three-hinge Arch Systems” in Part 3 of the 2009 NEHRP National Earthquake Hazard
Reduction Program Recommended Seismic Provisions for New Buildings and Other Structures
(FEMA P750).

Glulam arch systems shall comply with recommended detailing in AITC 104-2003
Typical Construction Details, requirements of the 2005 National Design
Specification® for Wood Construction (NDS®) including Appendix E, ASCE 7-
05 Minimum Design Loads for Buildings and Other Structures, and the applicable
building code.

Arch members and arch member connections shall be in accordance with the
requirements of the 2005 NDS including Appendix E Local Stresses in Fastener
Groups.

In addition, special glulam arch systems shall be in accordance with the Sections
6.1 through 6.7.

6.1 Connection requirements

Connections that are part of the special glulam arch seismic force resisting system
shall be in accordance with requirements of NDS Chapter 10 for Mechanical
Connections and additional requirements of this Section.

6.1.1 Arch Base: Arch base connections shall utilize a steel shoe assembly in
accordance with AITC 104 Typical Construction Details. Timber rivets or dowel-
type fasteners such as thru-bolts or lag screws shall attach the arch to the shoe.

95
Dowel-type fasteners shall be chosen such that the expected yield mode is Mode
III or Mode IV as defined in NDS. Timber rivet connections shall be designed to
ensure that the expected strength limit state is characterized by rivet capacity.

6.1.2 Arch Peak: Connection of the arch at the peak shall utilize shear plates,
bolts, steel dowels, or metal side plates or combination thereof in accordance with
AITC 104 Typical Construction Details.

6.2 Nominal connection capacity

The nominal capacity of a connection shall be determined in accordance with the
following:

a. For dowel type fasteners: n x Z(KF)(λ)(CM)(Ct)(Ceg) where n is the

number of fasteners, Z is the reference lateral design value for a single
fastener, and KF, l, CM, Ct, and Ceg are adjustment factors specified in
NDS for format conversion, time effect, wet service, temperature and
end grain, respectively.

b. For timber rivets: (Pr or Qr) x (KF)(λ)(CM)(Ct)(Cst) where Pr is parallel

to grain reference rivet capacity, Qr is perpendicular to grain reference
rivet capacity, and KF, λ, CM, Ct, and Cst are adjustment factors
specified in NDS for format conversion, time effect, wet service,
temperature and metal side plate, respectively.

c. For split ring and shear plate connectors: n x P x

(KF)(λ)(CM)(Ct)(Cd)(Cst) or n x Q x (KF)(l)(CM)(Ct)(Cd) where n is the
number of fasteners, P is the reference design value parallel to grain
for a single split ring connector unit or shear plate unit, Q is the
reference design value perpendicular to grain for a single split ring
connector unit or shear plate unit and KF, λ, CM, Ct, Cd and Cst are
adjustment factors specified in NDS for format conversion, time effect,
wet service, temperature, penetration and metal side plate,
respectively.

6.3 Member requirements

Arch members that are part of the special glulam arch seismic force resisting
system shall meet requirements of NDS and requirements of this Section.

6.3.1 Slenderness: The ratio of tangent point depth to breadth ( dt b ) shall not
exceed 6, based on actual dimensions, when one edge of the arch is braced by
decking fastened directly to the arch, or braced at frequent intervals as by girts or
roof purlins. When such lateral bracing is not present, dt b shall not exceed 5.

6.3.2 End grain bearing: At the arch base, end grain bearing shall be on a metal
plate with sufficient strength and stiffness to distribute the applied load. At

96
moment splices, end grain bearing shall be on a metal plate when fc > (0.75)( Fc* )
as required in accordance with NDS 3.10.1.3.

6.3.3 Compression perpendicular to grain: Compression perpendicular to grain,

induced at the arch base, shall be by a metal plate with sufficient strength and
stiffness to distribute the applied load.

6.4 Member resistance

6.4.1 Moment, Tension, Compression, and Shear: The arch member shall be
designed to resist moment, tension, compression, shear, and applicable
combinations of these induced by seismic forces determined in accordance with
load combinations of Section 12.4.3.2 of ASCE 7 (load combinations with over-
strength) but need not exceed forces resulting from strength at connections
determined in accordance with Section 6.4.2 (a).

6.4.2 Member Resistance at Connections: The member shall be designed for limit
states of net section tension rupture, row tear-out, group tear-out as defined in
NDS Appendix E, and shear in accordance with NDS 3.4.3.3 due to the seismic
forces as determined by the lesser of:

a. The nominal connection capacity determined in accordance with

Section 6.2 for LRFD, or the nominal connection capacity determined
in accordance with Section 6.2 divided by 1.35 for ASD.
b. The required capacity resulting from load combinations of Section
12.4.3.2 of ASCE 7 (load combinations with over-strength).

6.5 Transfer of forces to the arch members

The diaphragm, members and connections shall be sized to transfer of out-of-
plane wall and roof forces into the arch.

6.6 End fixity

In accordance with assumed pinned behavior of a 3-hing arch, determination of
reaction and arch member forces is based on assumed idealized pin behavior at
the arch peak and base. Actual detailing may introduce partial moment fixity at
reactions, and consideration shall be given to the effect of such fixity on member
and connection response.

6.7 Arch Moment Splice

Arch moment splices shall utilize a metal bearing plate (when required), metal
side plates, shear plates, bolts, steel dowels, timber rivets or combination thereof
in accordance with AITC 104 Typical Construction Details. Design forces for
determining the size and number of fasteners shall be based on load combinations
of Section 12.4.3.2 of ASCE 7 (load combinations with over-strength) but need
not exceed the member design force based on forces resulting from strength at
connections (see Section 6.4.1 and 6.4.2 (a)).

97
Drift Limits

In addition to the capacity checks and detailing as required above, arches must be stiff enough to
avoid excessive lateral displacement (drift) under seismic loads. The applicable drift limits are
adapted from ASCE 7-05 and depend on the Seismic Use Group as shown in table 6.2.

Table 6.2. Allowable drift at haunch of Tudor arch.

Occupancy Category I or II III IV
Allowable Drift 0.020 hw 0.015 hw 0.010 hw

These drift limits are calibrated based on LRFD loads and must therefore be calculated using the
LRFD seismic load combinations. However, the stresses in the arch do not need to be checked
using the LRFD load combinations when performing an ASD analysis.

Seismic Load Distribution

A Tudor arch resists lateral and vertical loads (in the plane of the arch) primarily through
bending. As such, the distribution of the loads throughout the arch has a significant effect on the
deflected shape and on the stress distribution in the arch. It is, therefore, not appropriate to
concentrate the seismic mass of the structure at a single location for the analysis; the load (mass)
should be distributed as it occurs along the length of the arch.

Procedure

Generally speaking, the seismic analysis is done after the arch has been designed and
proportioned to meet other design criteria. The arch is then analyzed to determine if it meets the
seismic design criteria. The following steps are applied to analyze the arch.

1. Design the arch and connections to resist all loads and deformations other than
seismic.
2. Apply the seismic load combinations without over-strength factors and verify that the
fasteners at the base connection are adequate to resist the applied loads and are
designed to yield in a ductile failure mode (Mode III or Mode IV for bolts, rivet
failure for timber rivets).
3. Determine the ratio of capacity/demand of the connection (C/D).
4. Satisfy one of the following options. (If the demand on the fastener is zero under the
seismic loads, option b. must be followed.):
a. Amplify all forces on the arch by multiplying by (C/D)*(1/φz) and verify that
the combined member stress ratios do not exceed 1.0 and that member failure
modes at connections (row and group tear-out, notched shear, etc.) do not
occur. (This provision ensures that the member capacity is higher than the
nominal connection capacity).
b. Apply the seismic load combinations with over-strength factors (ASCE 7-05,
Section 12.4.3.2) and verify that the combined member stresses do not exceed

98
 1.0 and that member failure modes at connections (row and group tear-out,
notched shear, etc.) do not occur.
5. Apply the LRFD seismic load combinations (without the redundancy factor, ρ) and
check drift against the applicable limits.

Example 6-1: Seismic Analysis of Tudor Arch

The Equivalent Lateral Force Procedure will be used to analyze the arch from Example 4-1 for
seismic loads.

Given: SDS=1.0. The building is classified in occupancy category III (I=1.25) and is located in
seismic design category D. Snow load is not required to be considered as part of the
mass of the structure for determining lateral seismic loads. The system has sufficient
redundancy for the reliability factor to be taken as ρ = 1.0.

Wanted: Analyze the arch under seismic loading.

Solution: Using the Equivalent Lateral Force procedure, the seismic base shear is given by:

IS DSW
V=
R

The total shear forces on the structure must result in this base shear, which is caused by
the mass of the structure subject to seismic acceleration. For analysis of the tudor arch
system, the weight of the structures is distributed as it occurs in the structure and the
lateral distributed loads on the arch are determined by:

IS DS ω
ν=
R

Because the building is classified as occupancy category III, the importance factor is
1.25. For seismic design category D, R = 2.5 and special detailing is required. The
design acceleration coefficient was given as SDS = 1.0 in the problem statement. Because
the snow load is not included in the seismic mass, ω is simply the distributed dead weight
of the structure.

1.25 (1.0 ) ωD
ν= = 0.5ωD
2.5

Therefore, the lateral seismic loads (QE) are calculated as 50% of the vertical dead load
and are distributed as shown in Figure E6.1.

99
112.5 plf -112.5 plf

75 plf -75 plf

Figure E6.1 Lateral Seismic Loads (load from left).

The bolted connection at the base must be adequate to support the following load
combinations from ASCE 7-05, Section 12.4.2.3 and be designed for Mode III or IV
yield mechanisms.

(1.0 + 0.14S DS ) D + 0.7ρQE

(1.0 + 0.105S DS ) D + 0.525ρQE + 0.75S
( 0.6 − 0.14S DS ) D + 0.7ρQE
Reaction forces for these load combinations are shown in Table E6.1.

Table E6.1. Reaction forces (lb) for arch subject to seismic load combinations.
Load Combination R y ,L Rx , L Cy Cx Ry ,R Rx , R
(1.0+0.14SDS)D+0.7ρQE 7610 1610 850 -3190 9320 4760
(1.0+0.105SDS)D+0.525ρQE+0.75S 14600 5460 640 -6640 15880 7820
(0.6-0.14SDS)D + 0.7ρQE 2560 -290 850 -1290 4270 2860

From Example 4-1, a ¾ in. bolt was used at the base. Its capacity was determined to be
3200 lb. The demand on this bolt under the seismic load combinations is 290 lb. The
ratio of capacity to demand is C/D = 3200/290 = 11.0.

The arch members and the peak connection must be adequate to resist the overstrength
load combinations from ASCE 7-05, Section 12.4.3.2 or be adequate to resist the seismic
loads amplified by a factor of (C/D)*(1/φz)=11.0/0.65 = 16.9. Because the seismic
loading places very low demand on the fastener relative to its capacity, it will be more
efficient to ensure that the arch is adequate to resist the overstrength load combinations
from ASCE 7-05, 12.4.3.2.

100
 The overstrength load combinations from ASCE 7-05, 12.4.3.2 are as follows:

(1.0 + 0.14S DS ) D + 0.7ΩQE

(1.0 + 0.105S DS ) D + 0.525ΩQE + 0.75S
( 0.6 − 0.14S DS ) D + 0.7ΩQE
The reactions calculated for the overstrength load combinations are shown in Table E6.2.

Table E6.2. Reaction forces (lb) for seismic load combinations with overstrength factor.
Load Combination Ry ,L Rx , L Cy Cx Ry ,R Rx , R
(1.0+0.14SDS)D+0.7ΩQE 6330 -750 2130 -3190 10590 7130
(1.0+0.105SDS)D+0.525ΩQE+0.75S 13640 3690 1600 -6640 16830 9600
(0.6-0.14SDS)D + 0.7ΩQE 1290 -2650 2130 -1290 5550 5220

Analysis of this arch using a spreadsheet to perform the calculations shows that the arch
is adequate to support all of the considered overstrength load combinations (Table E6.3).

Table E6.3. Combined stress ratios for seismic overstrength load combinations.
Left Arch Half Right Arch Half
Lower Upper Upper Lower
Leg Arm Arm Leg
Haunch Haunch Haunch Haunch
(1.0+0.14SDS)D+0.7ΩQE 0.033 0.037 0.209 0.424 0.641 0.694 0.597 0.444
(1.0+0.105SDS)D+0.525ΩQE+0.75S 0.247 0.340 0.250 0.493 0.644 0.861 0.815 0.605
(0.6-0.14SDS)D + 0.7ΩQE 0.154 0.203 0.342 0.362 0.662 0.542 0.429 0.321

The analysis indicates that no uplift occurs at the base of the arch. No tension forces are
transferred between arch halves at the peak of the arch. The shear plate connection at the
peak of the arch must be sufficient to transfer 2130 lb. The base of the arch must be
detailed to transfer 2650 lb from the bolt to the member considering the notch effect.

Base Connection

The shear force to be resisted at the base of the arch, perpendicular to the laminations, is
calculated as:

V = − Rx cos α leg If no uplift is present (Ry ≥ 1)

V = − Rx cos α leg − Ry sin α leg If uplift is present (Ry < 1)

Therefore:

V = − Rx cos α leg = − ( −2650 lb ) cos ( 3.8° ) = 2640 lb

101
In Example 4-1, the capacity of the member to resist the shear from the bolt was
determined to be 1534 lb with the bolt placed 6 inches from the bearing face. This is not
sufficient to resist the required seismic shear load of 2640 lb. Moving the bolt toward the
inside face of the arch (away from the bearing seat) will reduce the notch effect and
increase the member’s capacity to resist the shear force.

Assuming the bolt is placed 8 in. from the outside face of the arch (Figure E6.2) to
minimize potential problems with shrinkage between bolt and bearing seat, the effective
depth, de, and the total depth, d, of the section through the bolt hole can be calculated as:

xh 8 in
de = = = 8.02 in
cos α leg cos ( 3.8° )

d = d e + ( db′ − xh ) cos α leg + yh sin α leg

d = 8.02 in + (14.5 in − 8.0 in ) cos ( 3.8° ) + ( 5.25 in ) sin ( 3.8° ) = 14.9 in

d
de

y h = 5.25 in

αleg = 3.8o

xh = 8.0 in
db' =14.5 in
Figure E6.2. Bolt placement and dimensions for revised base connection.

The shear resistance of the member at the connection can be calculated as:

2 2
⎡2 ⎤ ⎡d ⎤ ⎡2 ⎤ ⎡ 8.02 in ⎤
Vr′ = ⎢ Fvx′ bd e ⎥ ⎢ e ⎥ = ⎢ ( 215 psi )(1.6 )( 6.75 in )( 8.02 in ) ⎥ ⎢ = 3600 lb
⎣3 ⎦ ⎣ d ⎦ ⎣3 ⎦ ⎣ 14.9 in ⎥⎦

The capacity of the member at the connection is adequate with the bolt placed 8 in. from
the bearing seat (3600 lb > 2640 lb). It is important to note, however, that placing the
bolt a long distance from the bearing seat is not recommended, because a split can form
as the member shrinks (see AITC 104). If possible, the bolt should be placed within 6
inches of the bearing seat, or further detailing must be considered. In this case, the hole
in the timber member will be oversized by 1/8 in. (instead of the standard 1/16 in.

102
 oversize) to accommodate additional shrinkage. The hole in the steel plate will be
oversized by a standard 1/16 in.

Peak Connection

In Example 4-1, the adjusted capacity of the two 4 in. shear plates was determined to be
3220 lb for snow load, and the notched member shear capacity of the member at the peak
was calculated as 8,000 lb for snow load. For seismic load duration, the resistances will
be 1.6/1.15 = 1.39 times higher resulting in capacities of 4480 lb for the connectors and
11,130 lb for the members. Therefore, the peak connection is adequate to transfer the
2130 lb of shear from the overstrength load combinations.

Displacement (Story Drift)

In addition to evaluation of the stresses on the arch, a calculation of story drift is typically
required. For the case of arches in a structure classified in Occupancy Category III, the
displacement at the top of the wall must be limited to no more than 1.5% of the wall
height (Table 5.2). The allowable drift is calculated as:

Δ ≤ 0.015hw = 0.015 (144 in ) = 2.16 in

The LRFD load combinations should be used when calculating displacements for
comparison with this drift limit. The results of the calculations are shown in Table E6.4.

Table E6.4. Displacements (in.) at haunches under LRFD seismic load combinations.
Load Combination Δ x ,L Δ x ,R
(1.2+0.2SDS)D+QE+0.2S 0.59 in. -1.09 in.
(0.9-0.2SDS)D+QE 0.74 in. -0.94 in.

The displacements of the haunches are well below the prescribed limit of 2.16 in.

The arch is adequate to resist the seismic loads.

Conclusion

This chapter presented requirements and procedures for the analysis of tudor arches as the
primary seismic resisting system. In general, seismic analysis is performed as a design check
after the arches are designed for all other requirements. In seismic zones D, E, and F, special
detailing is required to minimize the occurrence of brittle failures. In addition to stress
assessments, the arches must be sufficiently stiff to prevent excessive displacement under
seismic forces. These concepts were illustrated with an example problem.

103
104
 Chapter 7
Specifying Arch Geometry

Once the design has been completed, the requirements must be communicated from the designer
to the laminated timber manufacturer. The arch geometry, species and grade, and any special
lay-up requirements must be specified. This chapter illustrates the parameters that should be
specified to communicate the required arch geometry to the manufacturer in a clear, succinct
manner. The parameters specified in this chapter correspond to the lay-out procedure illustrated
in Appendix B.

The dimensions illustrated in Figure 7.1 should be communicated to the laminated timber
manufacturer. Dimensions d1 and d2 are measured perpendicular to the outside of the arch at the
tangent points.

dc"

A2

d2
Hc

A1
R

Hw
d1

y0

db'
x' 0
L/2
Figure 7.1. Parameters required for Tudor arch specification (only left half shown for clarity).

105
Dimensions d1 and d2 and their corresponding locations A1 and A2 can be calculated with the
following equations:

A1 = H w, L − yLT , L [7-1]
d1 = d LT , L cos α leg , L [7-2]

A2 =
(xUT , L − dUT , L cos α arm, L ) sin φt , L + LL
[7-3]
cos φt , L

d 2 = dUT , L cos α arm, L [7-4]

The center of curvature can be located with the following equations:

x0′ = LL − x0, L [7-5]

y0 = y0,L [7-6]

The depth at the crown, measured perpendicular to the roof is calculated as:

cos φb , L
d c′′ = d c [7-7]
cos φt , L

106
Appendix A
Arch Lay-up

A standard arch lay-up is divided into 7 zones as shown in Figure A1. The zones are numbered
starting from the inside face of the arch.

Figure A1. Requirements for a standard arch lay-up.

Grade Requirements by Zone

Zones 1, 5, and 6 require the highest grade of laminations in the lay-up corresponding to the
tabulated outer tension zone requirements in AITC 117.

Zones 2 and 4 use the 2nd highest grade of laminations in the lay-up, corresponding to the
tabulated inner tension zone requirements in AITC 117.

Zones 3 and 7 use the lowest grade of laminations permitted in the lay-up, corresponding to the
core zone grade tabulated in AITC 117.

107
Zone Thicknesses

The thicknesses of zones 1 and 2 are determined based on the lay-up requirements from AITC
117 and the total depth at the lower tangent point (dLT) or at the upper tangent point (dUT),
whichever is greater.

Zones 4 and 5 are required to be as thick or thicker than zones 2 and 1, respectively, at all
sections between the upper and lower tangent points.

Zone 6 is required to be sufficiently thick to ensure that depth d2 is a minimum of 1.2 times the
larger of dLT or dUT.

Zones 4, 5, and 6, are permitted to taper off along the outer face of the arch as shown.

Effect of Taper on Reference Design Values

For tapered arches, the taper sawn on the outside faces of the arch will result in the removal of
high grade material, necessitating consideration in design. The flexural design value and
modulus of elasticity are both affected by the removal of high grade material.

Tapered Face in Flexural Compression

Where the tapered face is stressed in compression and high grade material is removed, reduced
design values can be obtained from AITC 117-2004, Table A3. These reduced reference design
values are permitted to be applied to the end section of the arch, and linear interpolation
(between the reduced reference design value at the end section and the full reference value at the
nearest tangent point) is used to assign positive flexural design values for intermediate sections.

Tapered Face in Flexural Tension

Where the tapered face is stressed in tension and the core zone is exposed, it is recommended
that the flexural reference stress be reduced by 50%. For sections where part of the tension
zones are removed, but the core zone is not exposed, the reference design value for negative
flexure can be set equal to the reduced positive design value at the same section. The depth of
the tension zone can be conservatively estimated as 15% of the tangent point depth for most
combinations, so the 50% reduction should generally be applied to sections shallower than 85%
of the tangent point depth.

Modulus of Elasticity

Because a significant portion of the arch will generally have core-grade laminations on the outer
face, a reduced modulus of elasticity is recommended for the entire arch. The reduced moduli
from AITC 117-2004, Table A3 are sufficient for arch design.

108
Alternate Arch Lay-ups

The standard arch lay-up is generally the most economical method of manufacturing an arch. It
uses high-strength material sparingly, placing lower grade material where stresses are typically
lower. Reductions in reference design values in segments where high-grade material is removed
by tapering generally have minimal impact on the design of the arch. However, there may be
cases where the designer requires full design values at all sections of the arch. For these cases,
the designer should specify the appropriate uniform-grade lay-up from AITC 117, Table A2 or
require that the tension zone requirements are maintained along the full length of the arch.
Either of these options may increase the cost of the arch, depending on the grades of lumber
involved.

109
110
Appendix B
Arch Lay-out

One method of laying out a tudor arch using the parameters specified in Chapter 7 is presented
here. Four basic steps are used to establish the arch geometry. All depths are measured
perpendicular to the outside of the arch to facilitate lay-out using a tool such as a carpenter’s
square.

111
112
113
114
115
 American Institute of Timber Construction
7012 South Revere Parkway • Suite 140 • Centennial, CO 80112
Phone: 303/792-9559 Fax: 303/792-0669 http://www.aitc-glulam.org

QUALITY
INSPECTED

THE SYMBOL OF QUALITY

IN ENGINEERED TIMBER