High Mast Light Pole HD Bolt Fatigue4

4th Draft unchecked work in progress - redacted
Safety risks associated with the use of un-preloaded anchor bolts

to fix High Mast light Poles
By A Weighell BSc CEng MIMechE
Date 30 May 21
Revision 4th draft for discussion, Unchecked, Work-In-Progress.

Place names removed for confidentiality
Note!
This reports has been written for personal interest and to address the
author's concern for public safety.
This report has not been motivated, financed or otherwise encouraged by any other
party, either private or commercial.
All drawings and illustrations have been produced by the author using his own
resources unless attributed otherwise.
high mast light pole hd bolt fatigue4.docx Page 1 of 39 Date Printed: 31 May 2021 09:27
Contents
1 Summary ......................................................................................................................................................3
2 Objective ......................................................................................................................................................3
3 Terms ............................................................................................................................................................4
4 Introduction ..................................................................................................................................................4
5 Risk of death and injury ..............................................................................................................................5
6 Author's concerns about HMLP design ....................................................................................................5
7 Inspection .....................................................................................................................................................6
7.1 Inservice inspection of anchor bolts to confirm design intent ........................................................................6
7.2 Inservice inspection of anchor bolts to predict failure ...................................................................................6
7.3 Inservice inspection of welds .........................................................................................................................6
8 HMLP Design ...............................................................................................................................................6
9 Observation of publicly accessible HMLP installations by the author ..................................................8
9.1 Levelling nuts .................................................................................................................................................8
9.2 Anchor bolt corrosion.....................................................................................................................................8
10 Installation ..................................................................................................................................................11
10.1 Assumed installation procedure ..................................................................................................................11
11 Nut installation accuracy ..........................................................................................................................12
11.1 Anchor bolt load distribution ........................................................................................................................12
11.2 Relationship between anchor bolt load distribution and nut turn ................................................................12
12 Wind loads .................................................................................................................................................14
13 HMLP Fatigue calculations .......................................................................................................................16
13.1 Fatigue loading ............................................................................................................................................16
13.2 BS EN 1993-1-9 BS EN 1993-1-9 Eurocode 3: Design of steel structures - Part 1-9: Fatigue. [Ref 4] ......17
14 Finite element Analysis ............................................................................................................................17
14.1 Introduction ..................................................................................................................................................17
14.2 Anchor bolt stress below the levelling nut (Tensile load side of HMLP) .....................................................18
15 HMLP with fully grouted based plate and pre-loaded anchor bolts .....................................................20
15.1 FEA model ...................................................................................................................................................20
15.2 Discussion of results....................................................................................................................................21
16 Stresses in HMLP base plate and mast...................................................................................................25
16.1 FEA models .................................................................................................................................................25
16.2 Discussion of results....................................................................................................................................25
17 References .................................................................................................................................................27
Appendix A Anchor bolt axial stress range BS EN 1993 Part 1-9: Fatigue ..............................................28
Appendix B Wind turbine failure ...................................................................................................................30
B.1 Introduction ..................................................................................................................................................30
Appendix C Proposed solution .....................................................................................................................30
C.1 Fatigue design .............................................................................................................................................30
C.2 Proposed corrective design solution ...........................................................................................................30
C.3 Features of support ring (See C.1) ..............................................................................................................31
Appendix D An automotive equivalent of un-preloaded anchor bolts? ...................................................33
Appendix E Anchor bolt stress range with structural grout and anchor bolt preload ...........................34
1 Summary
This report discusses the author's concerns about the design of High Mast Light
Poles (HMLPs) using stand-off base plates with un-preloaded anchor bolts.
The same concerns apply equally to other applications using similar base plate
designs that are also subject to repeated loads which induce stresses in the anchor
bolts in excess of the anchor bolt endurance limit. e.g. Small wind turbines.
Fatigue failure of un-preloaded anchor bolts would cause the structure to collapse.
Collapse has occurred in several small wind turbines (< 50 kW).
The author believes that HMLPs are a particular concern because of their location.
e.g. Alongside motorways, at road junctions, airports, car parks etc. Collapse of
HMLPs with heights up to 35 m would create a very high risk of death and injury.
The author's concerns include:
a) The difficulty of accurately predicting the safe design life of stand-off base plates with
un-preloaded anchor bolts subject to repeated and dynamic loads.
b) The difficulty of accurately confirming that the basis of design for un-preloaded
anchor bolts has been achieved or is maintained throughout the life of the structure.
c) Higher fatigue stresses in the lower part of the HMLP caused by flexure of the base
plate when compared with that of fully grouted base plates restrained using
preloaded anchor bolts. (Fatigue cracking of welds in the lower part of HMLPs has
occurred at several installations [Ref 7]).
d) The absence of any requirement in the HMLP design standard ILE TR07 to consider
fatigue failure of the structure and anchor bolts. (ILE TR07 was replaced by PLG07
in 2013. To be confirmed - does PLG07 now include requirements to consider
fatigue?).
2 Objective
To explain why the author does not believe that it is possible to undertake realistic
fatigue design calculations for un-preloaded anchor bolts necessary to justify the safe
long term design of High Mast Light Poles (HMLPs) using stand-off base plates.
3 Terms
The terms and their definitions below apply to this report only.
Terms listed together are intended to be used interchangeably.
Term Abbreviation Description
Anchor bolt Partially or fully threaded rod cast into a concrete foundation for
Anchor rod purpose of locating and restraining HMLPs.
Hold down bolt
Bolt preload Imposing a tensile load in a bolt during installation for purpose of
Bolt post-tension ensuring joint tightness and resistance to fatigue failure when
subject to dynamic load. The term "preload" is often used in
mechanical engineering applications. The term "post-tension" is
used in civil engineering applications.
High Mast Light Pole HMLP Vertical cantilever structure used to support of CCTV or large light
High Mast Lighting Tower HMLT fittings for illumination motorways, carparks, airports where large
numbers of smaller lamp posts are not practicable. Supported
using a circular base plate flange with pre-cast anchor bolts.
Anchor bolts normally arranged as a "stand-off" base plate.
Lamp post A vertical cantilever structure used to support CCTV cameras or
single light fittings for illumination of streets or pedestrian areas.
Lamp posts are smaller than HMLPs. Most lamp posts are
supported by "planting" the lower part of its structure directly into
the ground or foundation. i.e. Supported in the same way as a
fence post.
Stand-off base plate (See Figure 3) A base plate supported above the top of concrete level.
All vertical downwards and upwards loads are carried entirely by
the anchor bolts using levelling and top nuts. The gap between
top of concrete and underside of base plate might or might not be
filled with "cosmetic" grout after anchor bolt tightening.
The "cosmetic" grout does not carry any significant structural
load. (See xx)
4 Introduction
High Mast Light Poles (HMLPs) are a type of "lamp post" used to support larger,
multiple light fittings at higher elevations than lamp posts typically used for street
lighting. (See Figure 1, Figure 2). HMLPs are typically between 20 m and 35 m high.
HMLPs are typically located in areas of high occupancy or high road traffic.
Typical locations include:
- Adjacent to some motorways.
- Motorway and road junctions.
- Motorway service areas.
- Air Ports, Car parks
Figure 1 Typical High Mast Light Pole Figure 2 High Mast Light Pole located in retail park
5 Risk of death and injury

The location of most HMLPs, combined with the high occupancy of adjacent areas
within a radial distance equal to their height, means that collapse would cause a very
high risk of death and injury.
6 Author's concerns about HMLP design
Note The author accepts that un-preloaded anchor bolts used for HMLPs do not fail very
often. The author was informed that anchor bolts have been reused when some
HMLPs have been replaced during upgrades. The author's concern is that absence
of frequent failure is not a reasonable design justification for future safe operation -
especially when not supported by corroborating calculations and logical reasoning.
The author's concerns about HMLP design that make risk of collapse impossible to
predict are as follows.
a) Unknown installation loads in individual anchor bolts. i.e. A need to make design
assumptions that cannot be measured, substantiated or monitored during the HMLP
operating life.
b) Unknown bond strength between concrete and individual anchor bolts.
i.e. More effective axial bonding between anchor bolt and concrete would create a
stiffer anchor bolt that would attract a higher load compared with a less effectively
bonded neighbour.
c) High stress range in each anchor bolts even if installation loads could be accurately
measured and monitored.
d) Cyclic bending loads applied to threaded anchor bolts for which design acceptance
data is not available. i.e. Fatigue design acceptance data is only available for cyclic
axial loading.
e) The HMLP anchor bolt design below the levelling nut is impossible to inspect or
prove by load testing or by checking of nut tightness.
f) Most HMLP installations are subject to high levels of corrosion. e.g. Moisture, road
salt, dirt or even completely buried in soil and vegetation.
g) Higher stresses imposed on base plate and in welds connecting base plate to the
mast caused by increased flexing when compared to a fully grouted and preloaded
anchor bolts. i.e. HMLPs using stand-off base plates are fully supported by anchor
bolts acting as point point restraints rather than by directly and uniformly supported
on grout.
h) Reduced stiffness of HMLP caused by use of stand-off base plate and un-preloaded
anchor bolts. i.e. The axial and lateral stiffness of the anchor bolts is integral to the
overall stiffness of the HMLP structure. A reduced overall stiffness is expected to
reduce natural frequency and increase the dynamic load factor. An increased
dynamic load factor will increase loads and hence stresses created during the
application of variable wind loads.
7 Inspection
7.1 Inservice inspection of anchor bolts to confirm design intent
The author does not believe that there are any effective methods to predict the actual
loads in individual anchor bolts below the levelling nuts. Such data is essential to
confirm that each HMLP installation is in accordance with all assumptions and
calculations used to justify its design.
7.2 Inservice inspection of anchor bolts to predict failure
The author does not believe that there are any simple methods of inspection or
testing that would give advance warning of imminent collapse due to fatigue failure of
anchor bolts below the levelling nut. For example, checking anchor bolt nut tightness
has no effect on bolt forces below the levelling nut. Anchor nut tightness could not
be used to confirm that the anchor bolt had not completely fractured below the
levelling nut.
7.3 Inservice inspection of welds
In contrast to anchor bolts, welds at the base of most but not all HMLPs are easily
accessible for normal visual and NDT / examination. However, the base of some
HMLPs is buried in grass, soil, compost, tarmac and other materials which would
prevent monitoring and inspection of welds and anchor bolts. The age of some
covering materials would indicate that many HMLP installations have not been
inspected for many years. (See Figure 8).
8 HMLP Design
HMLPs are normally much taller and larger than lamp posts.
The primary design difference between lamp posts and HMLPs is their method of
anchorage.
Lamp posts are normally "planted" or cast into the ground or foundation in the same
way as a fence post.
HMLPs are connected to anchor bolts cast into a concrete foundation. The HMLP is
fitted with a circular base plate flange. Matching anchor bolt holes are drilled in the
flange in a uniformly spaced circular pattern. (See Figure 3, Figure 5 below).
Anchor bolts are typically cast directly into concrete without any type of sleeving **.
i.e. The anchor bolts are in direct contact with the concrete.
** Based on the author's observations of HMLPs mostly situated in xxxx. HMLP

installations in other parts of xxxx and yyyy appear similar.
HMLP mast
Gussets
Top nut
Concrete Base plate Flange

Foundation
Levelling nut
Anchor bolt
Figure 3 Typical design of HMLP base plate showing anchor bolts
Figure 4 Typical HMLP base plate
Figure 6 Close-up view

(with approx. 25 mm clearance between concrete
and underside of nut).
Figure 5 Typical HMLP base showing anchor bolts
9 Observation of publicly accessible HMLP installations by the author

9.1 Levelling nuts
When inspection has been possible, all HMLPs have been supported using
"levelling nuts" ***. See Figure 3 above. Most levelling nuts appear to be typically
located at a distance above the concrete equal to the anchor bolt diameter.
Some installations have been observed with little or no gap between the top of
concrete and the underside of the levelling nuts.
*** Some installations are impossible to inspect because they are either completely
covered with grass, compost, soil, tarmac etc. (See Figure 8, Figure 11, Figure 12).
In other cases, the levelling nuts are too close to the top of concrete to allow
inspection of the threads below the levelling nut. It would also be impossible to
confirm that the anchor bolt has not snapped below the levelling nut. (See Figure 7).
9.2 Anchor bolt corrosion
Some anchor bolts inspected by the author were corroded such that all evidence of
thread form was lost. Corroded anchor bolt diameters in some cases as measured
by the author were less than that corresponding to the effective stress area.
(Measurement using vernier callipers). (See Figure 12, Figure 13, Figure 14).
Note The author was able to observe five HMLPs from the same retail park installation
before and after removal following prolonged opposed action by the author ****.
Two of the five HMLPs are shown in Figure 11, Figure 12, Figure 13. The anchor
bolts of one of the five were buried under tarmac. Three of the five were buried and
fully covered in at least 150 mm soil and compost with standing water during wet
weather. One of the five HMLPs was installed above ground. (None of the buried
installations showed evidence of corrosion following cursory visual inspection.
Burial appeared to have been very effective in preventing corrosion.
However, the author would not recommend burial as a method of preventing
corrosion. i.e. Soil conditions at other sites could be very different and very much
more corrosive).
**** Removal of the HMLPs was because one HMLP had a defect that had existed for the
full life of the installation. i.e. Several top nuts were engaged by less than half of
their depth. Some nuts were engaged by less than two thread pitches!
Understanding or acknowledgement of the author's concern about anchor bolt fatigue
was not acknowledged or commented on by the XXX, the local authority or the owner
(xxxxx). Further details of this installation are shown in CROSS Safety Report 610
(Ref 9).
NOTE! CROSS is a confidential safety reporting system operated by the Institution of
Structural Engineers. It was NOT the author's intent to keep the report confidential.
The author had previously limited success after contacting several universities,
organisations including the xxx, and other experts, with the objective of obtaining
"moral" and technical support. CROSS and Highways England took a positive and
constructive interest.
Figure 7 xxxx services xxx westbound located Figure 8 xxx Service xxx eastbound
close to the main service area buildings.
Figure 9 xxxx Figure 10 xxx

(Installation not unusual) (Installation not unusual)
Figure 11 Car park, xxxx Figure 12 Car park, xxxx

(Installation not unusual) (Now removed following prolonged action by author -
installation not unusual)
Figure 13 Car park, xxxx Figure 14 Car park, xxxx

(HMLP now removed). The installation is not unusual). (HMLP now removed) The installation is not unusual.
(Same HMLP as Figure 13)
Estimated 20 m radius collapse

zone assuming no material
ejected to a distance greater
than the HMLP height.
Figure 15 Aerial view of HMLP Figure 13, Figure 14 after removal (Courtesy Google Maps)
The HMLP was located in centre of the hatched area to the left side of the blue temporary lighting generator shown above.
10 Installation
10.1 Assumed installation procedure
The author is not familiar with the precise methods used for installation of HMLPs.
However, some assumptions can be made about installation sequence following
completion of the foundations. For example.
Step 1 All levelling nuts are placed on the anchor bolts.
Step 2 Alignment of the levelling nuts before installation of the HMLP.
Option 1 (Fully determinate). This could involve 3 levelling nuts located at 120 deg
apart with all remaining nuts at lower levels thus not taking part in the HMLP levelling
process.
Option 2 (Indeterminate). This could involve levelling all nuts together.
Step 3 Placement of the HMLP
Step 4 Fitting of the top nuts.
Step 5 Adjusting the levelling nuts to plumb the HMLP.
Option 1 Adjusting one of the 3 active levelling nuts, then adjusting the second active
levelling nut.
Option 2 Adjusting more than one levelling nut at each adjustment step.
Step 6 Bringing all remaining levelling nuts into contact with the HMLP base plate and
progressively tightening nuts not used in the levelling process.
Option 1 would appear to be the simplest and most likely method of installation and
plumbing of the HMLP.
11 Nut installation accuracy
11.1 Anchor bolt load distribution
Is the design intent and calculation based on an assumption that in the absence of
wind, each anchor bolt carries an equal axial load below the levelling nut?
Is so, to achieve this objective, both levelling and anchor nuts for each anchor bolt
need to be adjusted relative to every other anchor bolt.
How accurately do the nuts need to be aligned relative to each other to ensure that
each anchor bolt carries the same axial load?
There are no simple methods that can be used to ensure or confirm that each anchor
bolt carries an equal compressive load when installed.
Load distribution in each anchor bolt is therefore entirely reliant on the judgement of
the installer.
HSE Report RR1081 (Ref 8) includes the following under the heading
"Key Messages" "….. Slight variations in the use of levelling nuts, underbase
grouting, and sequence and level of torqueing can intentionally but radically
alter the load transfer mechanisms and, in particular, fatigue resistance. …"
Note HSE Report RR1081 (Ref 8) also compares the method of support using stand-off
base plates used for small wind turbines with "Other structures - Lighting / Gantry
bases…"
11.2 Relationship between anchor bolt load distribution and nut turn
The following is a grossly simplistic calculation is for purpose of comparison and
illustration. (See Table 1 below). This calculation assumes that each anchor bolt is
fitted with a frictionless sleeve over its full embedded length. This assumption is for
purpose of establishing the most optimistic limiting condition. (The author is unaware
that anchor bolt sleeves have been used on any HMLP installation).
The purpose of the calculation below is to estimate the relationship between nut-turn
and force generated in the anchor bolt below the levelling nut.
For purpose of discussion, let it be assumed that the installer is able to install nuts to
an accuracy of 0.25 turns when attempting to create equal force in each anchor bolt
below the levelling nut.
Table 1 Change of bolt force caused by nut turn

Description Data Units Comments
Anchor bolt size M27
Thread pitch 3 mm
Elastic modulus 205 GPa
Effective anchor bolt length 680 mm Maximum possible length if frictionless in concrete
Anchor bolt tensile diameter 24.17 mm ISO metric thread form
Anchor bolt tensile cross section 459 mm2 Under estimate of cross section for axial stiffness calculation
Anchor bolt unthreaded cross 573 mm2 Over estimate of cross section for axial stiffness calculation
section
Anchor bolt axial stiffness - 138 kN/mm Using anchor bolt tensile area
Minimum
Anchor bolt axial stiffness - 173 kN/mm Using unthreaded area
Maximum
Nut turn 0.25 Assumed accuracy achieved by judgement and "feel".
Min anchor bolt force for nut turn 104 kN Underestimate because axial stiffness used is too low.
Max anchor bolt force for nut turn 129 kN Overestimate because axial stiffness used is too high.
Mean anchor bolt force for nut turn 117 kN Based on anchor bolt cross section equal to mid way between tensile
and unthreaded cross section.
Discussion of calculation
a) The effective anchor bolt length used to calculate the anchor bolt axial stiffness is
likely to be too long. i.e. A fully cast-in threaded anchor bolt without sleeving will not
be effectively frictionless unless the bond between concrete is fully broken or non-
existent. Reducing the effective anchor bolt length will increase its calculated axial
stiffness and thus the calculated anchor bolt force for the nut turn assumption.
The author has not seen any evidence in at least twenty HMLP installations to justify
that any anchor bolts are axially loose in their foundations **. All anchor bolts
observed have appeared to be fully bonded to the concrete foundation at their top of
concrete level. It is difficult to believe that the concrete-to-anchor bolt bond has not
at least partially sheared. i.e. FEA calculations with fully bonded anchor bolts would
indicate that the concrete bond at the top of the anchor bolt will shear.
However, such dis-bonding has not been observed by visual inspections - at least
visually obvious cracking around the anchor bolts has not been observed.
(** When visual inspection has been possible).
Conclusion - The variations of axial load in adjacent anchor bolts created by
judgement based installation is likely to be very much higher than implied by the
above calculation.
b) The accuracy of nut-turn used in the calculation is no more than a crude estimate
based on practical experience. The actual values might be much higher or much
lower? In the author's experience, a judgment based accuracy of less than 0.125 nut
turns does not seem credible. At the other extreme, an accuracy of 0.5 nut turns
would seem easily achievable given the relatively high thread pitch.
Conclusion - The calculated anchor bolt force based on nut turn could be very
inaccurate.
12 Wind loads
Wind is the most significant load imposed on HMLPs. The HMLP foundations must
be designed to resist maximum wind loads as pseudo-static loads with the
appropriate load factor.
Typical unfactored design wind loads are shown in xxxx vendor data [See Ref xx].
Typical unfactored wind loads for HMLPs of similar dimensions used in Figure 3 -
Figure 6 above are as follows.
Table 2 Typical vendor wind data
Description Value Units Comments
HMLP model number xxxx 35 m high
Head area 3.2 m2
Wind speed 45 m/2
Horizontal shear 11 kN
Over turning moment 259 kNm
The following is an approximate simplistic "ball-park" calculation to compare with

Table 2 above.
Table 3 Approximate calculation to confirm range vendor wind data
Head area 3.2 m2 [Ref xx]
cd Head 0.5 Author's estimate
Wind velocity 45 m/s [Ref xx]. Assumed constant over height for simplicity.
Air density 1.24 kg/m3 Typical value for air
Wind pressure 1256 Pa Calculated using data above
Column Height 35 m [Ref xx]
Wind force on head 2.0 kN Calculated using data above
Over turning moment - Head 70 kNm Calculated using data above
Mean diameter of column 400 mm Author's estimate
Cd column 0.6 Author's estimate
Column area 14 m2 Calculated using data above
Wind force on column 11 kN Calculated using data above
Over turning moment - column 185 kNm Calculated using data above
Total over turning on base 255 kNm Calculated using data above
Total shear force on base 13 kNm Calculated using data above
Table 4 Approximate calculation of anchor bolt axial load and axial stress
Overturning moment at HMLP base due to wind 259 kNm Typical value. [See Table 2] (Note 1).
Number of anchor bolts 10 For example HMLP data. i.e. xxxx [Ref xx]
Pitch Circle Diameter (PCD) 838 mm For example HMLP data. i.e. xxxx [Ref xx]
Anchor bolt Size M27 For example HMLP data. i.e. xxxx [Ref xx]
Anchor bolt stress diameter 24.17 mm ISO thread form
Anchor bolts stress area 459 mm2 ISO thread form
Maximum anchor bolt force 123628 N Linear anchor bolt load distribution. (Note 4).
Maximum anchor bolt axial stress 269 MPa Corresponding to maximum anchor bolt load.
(Note 2, 3).
Notes
1. Excludes dynamic load factor (DLF). A DLF will increase anchor bolt forces and stresses because of the rapid
application and changes of magnitude and direction of wind forces. e.g. Gusting.
2. Excludes stress concentration factor (SCF) due to threads. An SCF is likely to be at least 2.8. i.e. The maximum axial
stress could be 3 times higher than calculated without consideration of an SCF
3. Excludes anchor bolt bending stresses induced by a) Lateral wind loads and b) Deflection of the anchor bolts due to
flexing of the HMLP base plate.
4. Calculated assuming a perfectly rigid based plate with a linear anchor bolt load increase from a central neutral axis.
Table 5 Approximate calculation of anchor bolt bending stress

Lateral force due to wind 11 kN Typical value. [See Table 2] (Note 1).
Number of anchor bolts 10 For example HMLP data. i.e. xxxx [Ref xx]]
Pitch Circle Diameter (PCD) 838 mm For example HMLP data. i.e. xxxx [Ref xx]
Anchor bolt Size M27 For example HMLP data. i.e. xxxx [Ref xx]
Anchor bolt stress diameter 24.17 mm ISO thread form
Anchor bolts section modulus 1386 mm3 ISO thread form
Bending length 30 mm
Maximum shear load per anchor bolt 1100 N Assume uniform load distribution.
Bending moment (assuming encastre restraint) 16.5 Nm
Bending stress 12 (Note 3)
Notes
1. Excludes dynamic load factor (DLF). A DLF will increase anchor bolt forces and stresses because of the rapid
application and changes of magnitude and direction of wind forces. e.g. Gusting.
2. Excludes stress concentration factor (SCF) due to threads. An SCF is likely to be at least 2.8. i.e. The maximum axial
stress could be 3 times higher than calculated without consideration of an SCF
3. Excludes anchor bolt bending stresses induced by deflection of the anchor bolts due to flexing of the HMLP base plate.
13 HMLP Fatigue calculations

13.1 Fatigue loading
Wind loads are variable and cyclic. Wind loads are applied at different times and in
different directions. Fatigue calculations must be undertaken to ensure that no part
of the HMLP is at risk of failure by repeated applications of loads.
The only exception would be if the stress range at every location was always below
the endurance limit of the material for every load application.
The following data is required to complete a fatigue calculation.
a) Input data
Description Comments
Stress range Calculated or measured based on imposed load. Note 1
Number of cycles at each stress range Calculated or measured based on imposed load. Note 1
Minimum stress Unknown for un-preloaded parts of the anchor bolts.
The minimum stress depends on the combination of imposed load with installed
stresses. As discussed above, installed stresses in the anchor bolts are unknown
and cannot be measured.
Maximum stress Unknown for un-preloaded parts of the anchor bolts.
The maximum stress depends on combination of imposed load with installed
stresses. As discussed above, installed stresses in the anchor bolts are unknown
and cannot be measured.
Mean stress The minimum and maximum stresses are unknown. Hence, the mean stress is
unknown.
Thread fatigue stress history - axial Calculated or measured. Normal design practice is to pre-load anchor bolt
threads to reduce anchor bolt stress range to an absolute minimum.
Thread fatigue stress history - bending Calculated or measured.
Anchor bolt bending is caused by
a) Lateral loads applied to the exposed section of anchor bolt below the levelling
nut.
b) Flexure of the HMLP base plate.
Such a calculation would be very complex and would need confirmation by
testing i.e. The load distribution is difficult if not impossible to predict.
Normal design practice is to ensure that bolt threads are not subject to dynamic
bending loads and thus avoid the need to do such a calculation. The author is
unaware of any other object or system that relies on the need to know the design
life of bolt threads in bending.
b) Allowable design data for design confirmation

Thread fatigue life - axial Obtained from design codes. For example Ref 4 BS EN 1993-1-9 Eurocode 3:
Design of steel structures - Part 1-9: Fatigue.
Thread fatigue life - bending Not available. i.e. Ref 4 Table 8.1 states "Anchor bolts and rods with rolled or cut
threads in tension". The only available design option using Ref 4 would be to
consider all stresses as tensile for purpose of fatigue life calculation.
13.2 BS EN 1993-1-9 BS EN 1993-1-9 Eurocode 3: Design of steel structures - Part 1-

9: Fatigue. [Ref 4]
a) Extract from BS EN 1993-1-9 Table 8.1
Detail Construction Detail Description Requirements
Category
50 size effect 14) Anchor bolts and rods with 14)  to be calculated using the tensile stress
for rolled or cut threads in tension. area of the anchor bolt. Bending and tension
t > 30 mm: For large diameters (anchor resulting from prying effects and bending
ks=(30/t)0.25 bolts) the size effect has to be stresses from other sources must be taken into
taken into account with ks account.
For preloaded anchor bolts, the reduction of
stress range many be taken into account.
b) Extract from BS EN 1993-1-9 Annex A

Annex A [normative] - Determination of fatigue load parameters and verification
formats
A.1 Determination of loading events
(1) Typical loading sequences that represent a credible estimated upper bound of all service
load events expected during the fatigue design life should be determined using prior
knowledge from similar structures, see Figure A.1 a).
A.2 Stress history at detail
(1) A stress history should be determined from the loading events at the structural detail
under consideration taking account of the type and shape of the relevant influence lines
to be considered and the effects of dynamic magnification of the structural response, see
Figure A.1 b).
(2) Stress histories may also be determined from measurements on similar structures or
from dynamic calculations of the structural response.
14 Finite element Analysis
14.1 Introduction
A simple linear static finite element analysis (FEA) has been completed using the
unfactored wind load used in section 12 above.
The wind load has been applied as though it was slowly applied from zero to
maximum and then slowly removed. i.e. The effect of a DLF and oscillations
(swing back) have not been included. The calculations will therefore under estimate
the stresses induced in the HMLP structure and anchor bolts.
Note! The following calculation is based on a design maximum wind speed of 45 m/s.
Results for other wind speeds could be estimated in proportion to the wind velocity
squared.
Wind velocity Factor
45 1
40 0.79
35 0.6
30 0.44
25 0.31
20 0.2
15 0.11
10 0.05
14.2 Anchor bolt stress below the levelling nut (Tensile load side of HMLP)
Calculated stresses are shown in Figure 18 below.
Description Value Comments
Max tensile stress ~300 MPa FEA (approximate nodal selection)
Min tensile stress ~ 180 MPa FEA (approximate nodal selection)
Mean tensile stress ~ (300 - 180)/2 = 240 MPa Based on 25 mm effective diameter
Bending stress ~ (300 - 180)/2 = 60 MPa Based on 25 mm effective diameter
Corrected mean tensile stress ~ 260 MPa Manual calculation 269 MPa Table 4 (Note 1)
Corrected mean bending stress ~ 67 MPa Manual calculation 12 MPa Table 5 (Notes 1, 2)
Notes
1. Calculated stresses above are based on an anchor bolt diameter of 25 mm. A larger diameter than that of the
equivalent stress area was used to better represent the stiffness of an M27 anchor bolt. The effective stress diameter of
an M27 thread is 24.17 mm. i.e. The calculated axial stress above will be 7% too low. The calculated bending stress
above will be 11% too low.
2. The manual calculation is expected to be very inaccurate because of the following:
a) The manual bending stress calculation is based on a beam that has a length nearly equal to its depth.
Classical bending theory is a poor representation of this geometry. Accurate results are not expected.
b) The manual calculation is based on an idealised ~ 30 mm long encastre beam to represent the exposed thread length
below the levelling nuts. This idealisation is not expected to accurately represent the interaction between the concrete
and the anchor bolts. i.e. The anchor bolts are being held in place with concrete that has an elastic modulus equal to
11% of the steel anchor bolts.
c) The manual calculation does not consider the effects of flexing of the base plate. The anchor bolts are connected to the
base plate and thus will be deflected by the same amount.
Figure 16 Half symmetry FEA model of example 35 m high HMLP

Lateral force and bending moment at base created by imposing the equivalent force FY = 5.5 kN at a elevation of 23.515 m
above the underside of the HMLP base plate. (Equivalent total load for full model of 11 kN, 259 kNm)
Figure 17 Axial stress (Z) results for load applied right to left
(Visibility of concrete foundation turned off for clarity)
Underside of levelling nut
Locations where calculated stress

exceeds 300 MPa.
See Note 1 below
This horizontal line is 3 mm

above top of concrete
Anchor bolt modelled 25 mm

diameter
Figure 18 Axial stress at 3 mm above top of concrete (tension side anchor bolt)
The concrete foundation is present in calculation. Visibility of the concrete has been turned off in this view for clarity.
Note 1 Areas of the anchor bolt shown in grey have calculated axial stresses greater than 300 MPa. Values greater than
300 MPa are not expected to be realistic because they are close to a perfectly sharp internal corner thus creating a
singularity.
15 HMLP with fully grouted based plate and pre-loaded anchor bolts
15.1 FEA model
The same FEA model as used for un-preloaded anchor bolts, with same imposed
loads was used to calculate comparable stresses and deflections.
The only difference is that the anchor bolts are modelled as preloaded beam
elements. Solid elements were used for the un-preloaded FEA model because
contact between the anchor bolts and concrete foundation are required to maintain
lateral position of the model. In contrast, lateral position of the preloaded model is
maintained by friction between the HMLP base plate and the foundation.
In reality, the anchor bolts would maintain lateral position if the preload friction was
lost. This model uses a preload of 125 kN per anchor bolt. i.e. 1250 kN of preload
for 10 anchor bolts. A nominal coefficient of friction / stiction of, say 0.3 would
provide a resistance of 500 kN. The maximum lateral load due to wind for this model
is 11 kN. i.e. A safety factor against lateral slippage of about 40.
15.2 Discussion of results

Results with preloaded anchor bolts indicate that this HMLP design is probably not at
risk from fatigue failure of the anchor bolts. i.e. In the absence of dynamic load
factors, the maximum anchor bolt stress range is less than the endurance limit even
at the design wind load of 45 m/s. Dynamic load factors should be confirmed before
the above statement can be treated as fact.
Stress levels in the HMLP structure are slightly lower than for the un-preloaded FEA
model. Lower stresses in the HMLP structure might be expected because the HMLP
base plate is prevented from flexing by full contact with the concrete foundation.
Figure 19 FEA model with grout

Anchor bolts are modelled as preloaded rods of 24.2 mm diameter. i.e. Diameter corresponding to the anchor bolt thread
axial stress area. Preload = 125 kN per anchor bolt. Axial preload stress = 272 MPa.
Foundation mesh excluded for clarity. Separation contact with friction modelled between HMLP and foundation.
Half symmetry model with load corresponding to a wind speed of 45 m/s. i.e. Loads equivalent to a total force of 11 kN at
an elevation of 23515 mm above bottom of HMLP base plate. (5.5 kN for half symmetry model).
Figure 20 Von Mises stress with preloaded anchor bolts and full wind load
Figure 21 Anchor bolt axial stress results for Subcase 1 - Application of anchor bolt preload and
gravity.
i.e. As installed. No wind.
Figure 22 Anchor bolt axial stress results for Subcase 2 - Application wind load.
Figure 23 Vertical (Z) displacement caused by application of anchor bolt preload
Figure 24 Vertical (Z) displacement with pre-load and wind wind load
Underside of base plate does not move upwards thus indicating that the HMLP remained in contact with the foundation.
16 Stresses in HMLP base plate and mast

16.1 FEA models
The purpose of this section is to compare two FEA calculations rather than to obtain
accurate stress values. i.e. The mesh used for this model is too large for accurate
stress calculations. Particularly, the mesh size needed to accurately calculate peak
stresses which are required for fatigue analysis.
The mesh size and loading in each model is identical. The only difference is that one
model uses a stand-off base plate with un-preloaded anchor bolts. (See Figure 25).
The second model uses a fully grouted base plate with preloaded anchor bolts.
(See Figure 26).
Note A mast wall thickness of 6 mm has been used in the absence of vendor data. It is
possible that the mast wall thickness might be less than 6 mm. e.g. 4 mm or 5 mm.
16.2 Discussion of results
When compared with the fully grouted base plate with preloaded anchor bolts, the
stand-off base plate model shows:
a) More areas of high stress than the fully grouted model.
b) Higher stresses in the base plate.
The differences in calculated stress are assumed to be caused by the difference in

method of support. i.e. The stand-off base plate is effectively point supported.
The only external resistance to base plate deflection is the bending strength of the
anchor bolts.
In contrast, a fully grouted base plate with preloaded anchor bolts is continuously
supported on the grout. The grout will also resist torsional deflection of the base
plate due to eccentric load transfer from the mast wall.
Figure 25 Von Mises stress with stand-off base plate (See Figure 16)
Figure 26 Von Mises stress for grouted base plate and preloaded anchor bolts (See Figure 19)
17 References
Title Location / File Comments
1. High Mast Light Pole Foundation https://civilweb-spreadsheets.com/other-design-
Design spreadsheets/high-mast-light-pole-design-
spreadsheet/high-mast-light-pole-foundation-
design/
2. xxxx
3. xxxx xxxx Contains dimensional and load
data.
4. BS EN 1993-1-9 Eurocode 3: Current UK structural steelwork
Design of steel structures - design standard for fatigue
Part 1-9: Fatigue assessment.
5. xxx xxxx
6. xxx xxxx
7. NCHRP Report 718 Fatigue nchrp_rpt_718.pdf Transportation Research Board
loading and design methodology Washington DC 2012
for high mast lighting towers. RJ (Applicable to US installations
Connor et al only?)
8. HSE report RR1081 Review of https://www.hse.gov.uk/research/rrpdf/rr1081.p
small wind turbine construction df
instructions and specifically for
structural supports and
foundations
9. CROSS Safety Report 610 https://www.cross-safety.org/uk/safety- Original report written by
High mast light poles removed information/cross-safety-report/high-mast-light- A Weighell (Personal cost and
from UK site poles-removed-uk-site-610 public interest motivation).
Appendix A Anchor bolt axial stress range BS EN 1993 Part 1-9: Fatigue
References
1. Eurocode 3: Design of steel structures - Part 1-9: Fatigue
Reference value of the fatigue strength
for anchor bolts in tension at cycles.
[1, Table 8-1]
Reference number of cycles for
Reference number of cycles for
Reference number of cycles at cut-off

limit
Allowable stress range ( < ) with

[1, 7, (3)]
Fatigue limit for constant amplitude stress

ranges at the number of cycles
Allowable stress range ( < )

[1, 7, (3)]
Direct stress range for nominal stress

spectra with stress ranges above and
below the constant amplitude fatigue
limit
Values of Direct Stress Range at various

cycle counts
Appendix B Wind turbine failure

B.1 Introduction
Several small wind turbines (~50 kW) supported using the same design as that used
for HMLPs (un-preloaded anchor bolts with levelling nuts) have collapsed.
e.g. Bradworthy (Ref xx). (Reported by the BBC on 29 Jan 2013).
Appendix C Proposed solution
C.1 Fatigue design
It is impossible to design threaded anchor bolts to resist fatigue failure caused by
dynamic and repetitive loads unless either
a) The threaded section is preloaded or
b) The stress range is below the anchor bolt material endurance limit. i.e. ~20 MPa.
(See Appendix A above).
Option b) is impractical because the permitted allowable stress range would be
negligible thus requiring very large anchor bolts. (The use of large anchor bolts
(> 50 mm) and thick base plates (50 mm - 75 mm) to minimise stress range appears
to have been a design approach used in the USA for some installation).
A practical and commercially viable solution is required to create a low and
predictable risk to public safety.
One solution would be to convert existing stand-off base plates to structurally grouted
base plates with preloaded anchor bolts.
The existing anchor bolts and HMLP base plates in most if not all cases should be
large suitable for such conversion without modification. i.e. Modification or
replacement would not be required to HMLP, the anchor bolts or the foundation.
C.2 Proposed corrective design solution
1 Fit segmented a support ring to the circumference of HMLP base plate. (See Figure
27 below). Ensure the underside of the support ring is coated with release agent to
prevent bonding when the grout is cast.
2 Use the jacking anchor bolts in the support ring to carry the weight of the HMLP while
maintaining alignment and plumbness of the HMLP. Sacrificial load spreader feet will
be fitted to each jacking anchor bolt. i.e. The jacking anchor bolt feet will be left in
place when the jacking anchor bolts are withdrawn following grouting).
3 Lower the levelling nuts to create a gap between the top of the levelling nut and the
underside of the HMLP base plate.
4 Insert a sponge "washer" above the levelling nut. The sponge would create a
compressible gap in the grout to prevent the levelling nut resisting anchor bolt
tension during later preloading. Coat the levelling nuts with a release agent or plastic
tape to prevent bonding with the grout.
5 Fit plastic drain tubes to allow drainage of any water that might collects inside the
HMLP.
6 Fill the gap between foundation and underside of HMLP base plate with load bearing
non-shrink grout of suitable compressive strength. Monitor grout from the inside of
the HMLP to ensure that no air bubbles are left under the HMLP base plate.
7 Leave the temporary support collar in place until the grout has set and achieved its
design compressive strength.
8 Remove the support ring.

9 Add "Bellville" spring washers to each anchor bolt in turn. Clean bolt threads and
add anti-corrosive lubrication.
Note The existing anchor bolts are not sleeved. It is likely that anchor bolts might still be
bonded to the concrete. The concrete bond might break in the future and thus
release the preload. Bellville washers will reduce the loss of preload if the
concrete-to-anchor bolt bond breaks during operation. Bellville washer might also
provide some visual indication that preload has been lost.
10 Pre-load the anchor bolts by tightening the nuts to the specified torque.
C.3 Features of support ring (See C.1)
1 High torsional rigidity. The support ring will be formed as a hollow section to ensure
maximum torsional rigidity to resist eccentric loading.
2 Rotation resistant "Spigot and Sockets" on each end of the support ring segments to
maintain alignment and torsional continuity.
3 2, 3 or 4 support ring segments to allow fitting and removal. (3 segments will
probably the the cheapest and most practicable option).
4 Narrow lip at bottom of ring to engage with HMLP base plate. The lip should be as
narrow as practicable to maximise the volume of grout. (See C.2).
5 Support ring design strength and jacking anchor bolt spacing to allow operation with
a single jacking anchor bolt removed to gain access to loosen tight levelling nuts.
Anchor bolt with fully

compressed Bellville
spring washer
Jacking bolts with

2, 3 orr 4 segment support ring detachable support feet
Figure 27 HMLP base plate supported on a torsionally stiff multi-segment support ring
(Support ring design copyright A Weighell)
Support ring HMLP base plate supported

on support ring lip
Figure 28 Close up view of HMLP flange supported on lip projecting from bottom of support ring
(Design copyright A Weighell)
Appendix D An automotive equivalent of un-preloaded anchor bolts?

During the late 1960s and early 1970s, it was common practice among some
teenagers in the UK to "widen" the wheels of their "Mini" cars. i.e. Increasing the
track width in the hope of cornering at higher speeds. (This practice became less
common by the mid-1970s because insurers enforced rules on what was classed as
an unauthorised non-factory modification).
Which method would require the same fatigue calculation methods and unjustifiable
design assumptions as HMLPs with un-preloaded anchor bolts?
Which method would YOU trust with you life?
(The author is unaware of wheel widening that did not use spacers and wheel hub
stud bolts that were not preloaded over their full length).
Figure 29 Widened car wheel using a spacer
Figure 30 Widened car wheel using nuts rather than spacers?
Appendix E Anchor bolt stress range with structural grout and anchor bolt preload

High Mast Light Pole HD Bolt Fatigue4

Uploaded by

Copyright:

Available Formats

High Mast Light Pole HD Bolt Fatigue4

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

High Mast Light Pole HD Bolt Fatigue4

Uploaded by

Copyright:

Available Formats

4th Draft unchecked work in progress - redacted

Safety risks associated with the use of un-preloaded anchor bolts

By A Weighell BSc CEng MIMechE

Revision 4th draft for discussion, Unchecked, Work-In-Progress.

5 Risk of death and injury

** Based on the author's observations of HMLPs mostly situated in xxxx. HMLP

Concrete Base plate Flange

Figure 3 Typical design of HMLP base plate showing anchor bolts

Figure 4 Typical HMLP base plate

Figure 6 Close-up view

9 Observation of publicly accessible HMLP installations by the author

Figure 9 xxxx Figure 10 xxx

Figure 11 Car park, xxxx Figure 12 Car park, xxxx

Figure 13 Car park, xxxx Figure 14 Car park, xxxx

Estimated 20 m radius collapse

Table 1 Change of bolt force caused by nut turn

The following is an approximate simplistic "ball-park" calculation to compare with

Table 5 Approximate calculation of anchor bolt bending stress

13 HMLP Fatigue calculations

b) Allowable design data for design confirmation

13.2 BS EN 1993-1-9 BS EN 1993-1-9 Eurocode 3: Design of steel structures - Part 1-

b) Extract from BS EN 1993-1-9 Annex A

Figure 16 Half symmetry FEA model of example 35 m high HMLP

Underside of levelling nut

Locations where calculated stress

This horizontal line is 3 mm

Anchor bolt modelled 25 mm

15.2 Discussion of results

Figure 19 FEA model with grout

Figure 23 Vertical (Z) displacement caused by application of anchor bolt preload

16 Stresses in HMLP base plate and mast

The differences in calculated stress are assumed to be caused by the difference in

Reference number of cycles for

Reference number of cycles for

Reference number of cycles at cut-off

Allowable stress range ( < ) with

Fatigue limit for constant amplitude stress

Allowable stress range ( < )

Direct stress range for nominal stress

Values of Direct Stress Range at various

Appendix B Wind turbine failure

8 Remove the support ring.

Anchor bolt with fully

Jacking bolts with

Support ring HMLP base plate supported

Appendix D An automotive equivalent of un-preloaded anchor bolts?

Figure 29 Widened car wheel using a spacer

Figure 30 Widened car wheel using nuts rather than spacers?

You might also like