Introduction

This manual describes the Dynatest Elmod 6 program package. ELMOD is an acronym for Evaluation of
Layer Moduli and Overlay Design.
The program is especially useful for engineers responsible for the maintenance and rehabilitation of a
road or airfield network. Using FWD or HWD measured deflection basins; the program can automatically
carry out a complete structural analysis and overlay design, based on user-defined design parameters.
Elmod 6 allows the user to choose a set of parameters (meteorological, material characteristics, loads
etc.) to be used in the calculations, which customizes the analyses according to relevant local conditions.
Elmod 6 performs 3 major tasks.

• First, the program calculates the modulus of each layer in two, three, four or five layer pavement
systems using either the "Radius of Curvature" - Odemark-Boussinesq transformed section
approach, the "Deflection Basin Fit" method normally used with numerical integration techniques
or using the "FEM/LET/MET" option which allows the user to select either the Finite Element
Method, Linear Elastic Theory or the Method of Equivalent Thicknesses. The backcalculation
provides the apparent moduli for the as-measured deflections at each FWD or HWD test point,
and taking the non-linearity of the subgrade (or all layers with FEM) into consideration. Elmod 6
can also provide a theoretical estimate of the equivalent depth to a rigid layer from the measured
deflections.

• Next, the as-measured moduli are adjusted to reflect conditions representative of each season
specified for the design period, for up to 12 seasons. For each season, asphalt moduli are
calculated as a function of temperature, while moduli of unbound materials (including the
subgrade) are a function of the time of year in relation to spring thaw or wet period.
• Finally, Miner's law is used to sum the structural and/or functional pavement damage caused
during each season, by each load (up to 24 design vehicles can be defined with a total of up to
100 wheel configurations), based on user-defined damage (fatigue, roughness or rutting)
relationships. The program then calculates the expected remaining life of the pavement, and the
overlay thickness required, using a specified overlay material for a given design period. A Life
Cycle Cost Analysis (LCCA) option is also available. This makes use of an incremental-recursive
prediction of future pavement condition and ranks different maintenance strategies according to
costs (including user and capital costs if desired).

The program is ideal for routine structural evaluation of pavement systems using Dynatest FWD or HWD
deflection data. It also forms an integral part of both the extremely advanced Dynatest Pavement
Management System (DMS), the Dynatest AIRPORTS Pavement Management System, which can
include pavement structural evaluation at the network level and the Dynatest Performance and Economic
Rating System (PERS), which is compatible with the LCCA option. This option can significantly reduce the
subjectivity involved in the prediction of pavement performance based mainly on visual surface condition.

Limitations and Considerations

Elmod 6 essentially performs three tasks in evaluating FWD or HWD measured deflection basins. First, it
backcalculates the as-measured E-moduli of up to five layers, the apparent moduli based on the
measured deflection basins. Second, it adjusts the as-measured moduli to seasonal moduli which reflect
local conditions, based on user-defined seasonal data. Third, the residual life and overlay requirements
are calculated, based on stresses and strains generated in the seasonal pavement structures by user-
defined design loads using user-defined damage relationships.
To obtain the highest precision with Elmod 6, the structure to be evaluated should preferably meet the
following conditions:
The structure should contain only ONE stiff layer (E1/Esubg>5). If the structure contains more than one stiff
layer they should be combined into one layer for the purpose of structural evaluation, or the modulus of
one of the layers should be fixed.

Moduli should be decreasing with depth (Ei/Ei+1>2).

The thickness of the upper (stiff) layer (H1) should be larger than half the radius of the loading plate.

When testing near a joint or a large crack, or on gravel surfaces, the structure should be treated as a two-
layer system.

If the structure does not comply with these limitations the Elmod 6 program may still be used, but the
precision will not be as good. Typically, Elmod 6 provides excellent results unless extremely unusual
conditions are encountered. For difficult structures where the above conditions do not apply, the
FEM/LET/MET routine may be used. The deflection basin matching routine provides an indication of how
well the back-calculated moduli simulate the measured pavement response.
The moduli of the pavement layers are calculated from the deflection basin measured during Falling
Weight Deflectometer (FWD) tests.

When the "Deflection Basin Fit" option is selected, Odemark's layer transformation approach is used with
Boussinesq's equations to calculate deflections, and an iterative procedure is used to determine those
moduli that result in the same deflections as measured. This approach has been found to be a good
approximation to the generalized Burmister equations under the following two conditions (Ullidtz & Peattie:
"Pavements analysis by programmable calculators", Transportation Engineering Journal of ASCE, vol.
106, No. TE5, Sept. 1980):
1. The layer thickness should be more than half the radius of the loading plate.
2. The modular ratio of two adjacent layers (Ei/Ei+1) should not be less than 2.
The stresses, strains and deflections calculated through the Odemark-Boussinesq method have been
found to agree as well as, or better, with measured stresses and strains in real pavement structures than
calculations performed with the generalized Burmister equations (e.g. CHEVRON or similar programs),
even when the two conditions above are NOT fulfilled.
In all methods, Poisson's ratio is assumed to be 0.35 (except in the FEM/LET/MET option). In most cases,
however, Poisson's ratio has little influence on the calculated deflections and stresses. Strains are
influenced by Poisson's ratio, and if they are used for design purposes it is very important to use the same
Poisson's ratio that was used for deriving the critical values. If the Shell criteria is used this is not a
problem, since a Poisson's ratio of 0.35 was assumed in deriving this criteria.
All materials are assumed to be homogeneous, isotropic and linear elastic, except the subgrade which is
assumed to exhibit non-linear response defined by:

This non-linearity is typical for cohesive (fine-grained) soils.

In many cases real subgrades will be layered or will have moduli that vary with depth due to overburden
pressure, changes in moisture content etc. The impact load from the FWD will also create an effect that is
similar to a non-linearity. All backcalculation options in Elmod 6 treat these variations as non-linearity.
This is not quite correct, but it is considerably better than neglecting the variation of subgrade modulus
with depth and distance from the load. If these variations are not considered, very large errors may result.
This non-linear feature in Elmod 6 typically results in very good agreement between measured and
calculated deflection basins. It also generally removes the so-called "compensating layer effect" which can
introduce large errors in the moduli back-calculated assuming only linear elastic materials.
Elmod 6 does not evaluate the non-linear parameters of granular base or subbase materials. If this is
important, the FEM/LET/MET moduli can be used instead. The moduli of the granular materials calculated
with Elmod 6 correspond to the centerline of the loading plate and to the stresses used in the FWD test.
Since the non-linearity of granular materials is not considered, gravel roads (without surfacing) should be
considered as two layer systems.
If the pavement system contains a stabilized material which is cracked in blocks that are
of the same size or larger than the loading plate the system should also be considered
to be a two layer system, because the deflections in the neighborhood of the loading
plate cannot be correctly evaluated with elastic theory when the layer is not continuous.
When two or more materials are combined to form one layer the moduli should preferably be of the same
order of magnitude. For pavements with thin asphalt surfacing, it may therefore be advantageous to
estimate the modulus of the asphalt, from temperature and amount of cracking. In this case the modulus
of an intact asphalt material must be known at a reference temperature. The asphalt modulus can be
estimated using various techniques. The program will then calculate the modulus at the temperature of
testing (using the relationship defined in the parameter file), and as a function of the amount of cracking of
the asphalt layer (as input by the FWD operator).
Different asphalt materials, such as wearing course, binder course and base course should always be
combined into one layer for the purpose of calculating the existing moduli.
It is important to recall that the generalize Burmister equations are based on continuum (or solid)
mechanics and are only valid for a linear elastic layered system under a static load. Pavement materials
are not solid, many materials are particulate, and they tend to be non-linear elastic with viscous, visco-
elastic and plastic deformations in addition to the elastic deformations. The Odemark-Boussinesq
approach used in Elmod can consider the very important non-linearity of the subgrade, and in studies
where measured stresses and strains have been compared to values calculated using different
procedures, the Odemark-Boussinesq approach has been found to give a considerably better agreement
with measured values than the Burmister equations.

Compatibility With Previous Versions

Elmod 6 can read and convert Elmod 5 databases. Since Elmod 6 has a changed database format, it is
necessary for Elmod 6 to create an Elmod 6 database from the Elmod 5 project.
If you from Project → Open select an Elmod 5 project database, you will get the following message:

Click Yes to proceed, and you will be asked for location and name for the Elmod 6 project database:
As a default Elmod suggests the same folder, and name of the project database with _E6 added to the
name.
When you click Save the new database will be created and opened ready to be used.

Databases and parameter files from Elmod versions 1 to 4, cannot be read automatically by Elmod 6.
Contact Dynatest in case you need assistance with converting old data.

Sample Data Files

Included with your installation are sample data files located in your Elmod 6 program folder. These files
can be used also when Elmod 6 is in Demo mode.

46SA1AL1.FWD is from a flexible pavement, and can be used with both SI units and US Customary units.

SI_PCC.MDB is from a PCC slab pavement to be used with SI units.

US_PCC.MDB is the same as the above, but with US Customary units.

46SA1AL1.FWD
This example file is a SHRP section, tested six times at different times of the year - July 1994 to
November 1994.

SHRP Section ID: 460804

Location: South Dakota

State Highway 1804 Eastbound 9 miles
North of Pollock

Pavement Structure:
178 mm (7") Asphalt
305 mm (12") Granular Base
Clayey Silt Subgrade
Test Location Information
Number of Test Points: 150
Section Length: 152.4 m (500 ft)
Test Interval: 7.62 m (25 ft)
4 drops each test point - average of 16, 4 different loads
approx. 378 kPa (6000 lb)
567 kPa (9000 lb)
755 kPa (12000 lb)
1000 kPa (16000 lb)
Plate Radius: 150mm (5.91 in)

Geophone spacing: 0,203,305,457,610,914,1524 mm (0,8,12,18,24,36,60 in)

o
Test date Points Temp Range C
940714 1-9 33-35 91-95
940715 10-49 25-52 77-125
940812 50-79 24-37 75-99
940927 80-111 14-27 57-81
941025 112-131 8-15 46-59
941122 132-149 -10 to -1 14-30
o
The previous day’s average temperatures in C, for use with BELLS’ equation were:

July 13 19.5
July 14 23
August 11 21
September 26 7
October 24 4
November 21 -1.5

SI_PCC.MDB, US_PCC.MDB
These data are from an airfield pavement and collected with an HWD.

The pavement consists of PCC slabs with thickness of 300 mm (12 in.). Slabs have been tested both at
nd rd
center and joints. At the joints the 2 and 3 geophone were positioned at each side of the joint, which
makes it possible to evaluate the Load Transfer Efficiency (LTE).

At each test point 3 drops were performed with 3 different load levels. This can be used to estimate
possible voids under the plates, and results can be viewed with the Plot option (see Raw Data for
details.)

Registration
Each installation on a new computer requires a request for a license file for the specific computer to make
Elmod6 run in full mode. When starting the Elmod program, a registration form will automatically pop up,
if no license is present for the computer. This registration form must be used when requesting the license
from Dynatest. When the window no longer pops up automatically, you can access it from: Help →
Registration
If you do not wish to register at this point, you can press the OK button, and the program will continue to
the main menu. The first 5 days after installation that you start the program without a license file, you can
use it with all facilities. Thereafter the program will convert to "Demo mode". In "Demo mode" Elmod will
only work with the DEMO files located in the Elmod6 program folder (see Sample Data Files.) All
functionalities in Elmod will be available also in Demo mode.

If you would like to request the license file, then press the button "Mail to Dynatest". You will then be
asked to enter the name of your company or institution, and to identify your home country from a list.

Next you will get a window, as shown below, where you can accept or reject the license conditions.
Please use the scroll button at the right to see all the text in the license agreement.
If you accept the agreement the program will start your default e-mail service, and prepare a new e-mail
with certain information placed in the header lines. All you have to do is send the e-mail. Dynatest will mail
back the license file, normally within 24 hours. If you choose not to accept the agreement, then the
program will convert into a DEMO version.

If one of these situations occurs:

- You don’t have e-mail service on the computer.

- The e-mail service does not start automatically.
- The wrong e-mail service is starting.
- The header line is empty.

Then please use the other button "I don’t have email facilities on this computer". This procedure will save
the necessary information to a text file "Mail_to_Dynatest.txt" in the C:\Dynatest\Elmod6 folder. Next you
can attach this file to an email to Dynatest using your normal e-mail service.

Main Menu
Navigating in the Menus

The screen below shows the startup menu for Elmod 6:

Some of the menu items may be disabled, if they have no relevance at the current stage, e.g. you can not
view a data file before you have opened one, and you can not do a design before you have done a
backcalculation.
Following is a short description of each menu item.

Project
Clicking Project brings up this submenu:

Select Database to manage project databases.

Open opens a browser to locate a previously created project database.
Create new opens a browser to allow selection of location and name for a new project database
Compress database is a facility to compact and repair existing databases. It is only enabled if no project
database has been opened.
For help on these topics, see Projects and Files

Select Datafile to manage files in the project database.

Open from database is used when an existing file in the project database should be the current file to
process.
Import FWD file to database opens a file browser to locate FWD files to import.
Import from several directories can be used to locate FWD files to import if they are located in different
folders.
Merge files is a facility to combine two imported files into one.
Rename file allows you to give the currently opened file a new name.
For help on these topics, see Projects and Files and Data Files

Select Batch to run backcalculation for a range of files in one operation. All selected files must be present
in the current database.
For help on this topic, see Batch Runs

Select Delete Options for access to various possibilities to delete analysis results and files.
For help on this topic, see Deleting Results and Files

Select Recent Files for fast access to previously opened files.

These can be accessed even if no project is open and even if they belong to another project than the
current one. The accessed file’s project database will automatically become the current one.

Review Data
Clicking Review Data gives access to view and edit information from an opened data file. Temperature
curves can be viewed and edited.
For help on this topic, see Viewing and Editing

Structure
Click Structure to enter pavement thicknesses and materials, and to define subsections. Further it gives
access to define GPR thickness data, and to input information for joint calculations.
For help on this topic, see Structural Data

Moduli
Click Moduli to get access to start a backcalculation of layer moduli. You have access to Radius of
curvature method, to Deflection Basin Fit method, and to the optional module FEM/LET/MET.
For help on this topic, see Estimating Moduli

Plot
Click Plot to access the facilities for presentation of deflections data and results of calculations. There are
facilities for graphics, printed reports and export to various file types.
For help on this topic, see Plot

Parameters
Click Parameters to view and edit a parameter setup for designs. If no project is open you have access to
the default parameter setups, otherwise you have access to the parameter setups stored in the current
project database, and the possibility of saving edited setups both as default and as project specific.
For help on this topic, see Parameters

Design
Click Design to perform either overlay design or new design.

Overlay design is used to calculate required overlay thickness on existing layers.

New design is used to calculate completely new structures, not using any backcalculation results.
To carry out a design you have to select a parameter setup and to enter information about layer material
types and traffic
For help on these topics, see Design Calculation and

LCCA
Click LCCA to get access to the optional module LCCA (Life Cycle Cost Analysis). This is only available if
you have license for this module.
For help on this topics, see Life Cycle Cost Analysis

PCN
Click PCN to calculate Pavement Classification Numbers for Airport facilities. This is only available if you
have license for this module.
For help on these topics, see PCN Calculations and PCN/ACN Method

Options
Clicking Options brings up this submenu:

Edit Aircraft Database is used to add, delete and modify information in this database delivered as a part
of the Elmod software package. The database contains information about aircraft details, and is used in
connection with setting up load configurations for aircrafts in Parameters, and as the source for ACN
values, when using the PCN facility.
For help on this topics, see Edit Aircraft Database

Click Edit Geophone Setup to get access to select/deselect geophones used for input to the
backcalculation.
For help on this topics, see Edit Geophone Setup

Click Edit Temperature Table to make changes to the defined asphalt modulus / temperature relations
defined as tables.
For help on these topics, see Edit Temperature Table and Temperature Sensitivity

Settings
Click Settings to access some overall settings for Elmod6, like language and unit system. Correct settings
must be entered before creating a new project database.
For help on this topics, see Settings

Help
Clicking Help brings up this submenu:

Help File brings up this Help system.

Registration gives access to apply for license file.

For help on this topics, see Registration

About gives information about program version and licenses.

Navigating in the Menus

With the Mouse

When moving the mouse cursor over a menu item (label or related icon), you will automatically see the
submenu if there is one, and you can continue to move the mouse to the submenu item you want to go to.
Submenu items can have a new submenu level. In this manner you can reach any menu item or submenu
item without clicking the mouse. First when you have highlighted the desired item, you need to click to
activate the item.
The advantages are that you can move fast around in the menus and save a lot of mouse clicks. The
drawback is that you need to be a bit more careful with the mouse movements, because a selected
submenu window will disappear if you move the cursor outside the submenu window. This drawback is
reduced a bit by making the submenu items larger, which also allows more space for the related icons.

From the keyboard.

The whole menu system can be controlled without using the mouse. Instead the key with the first letter of
the menu label can be pressed (not case sensitive).

To open a database you would then press: pdo and <Enter>

(p for Project, d for Database, o for Open and <Enter> to activate)

You will see that the mouse cursor is following your choices, so it’s easy at any time in the process to let
the mouse take over.

Settings
Select Settings from the main menu to access this facility.

This allows setting a number of general parameters in Elmod, and it must be checked that the settings are
correct before creating a new database.

General

Settings on this tab are defaults for any new database that is created Changes will be reflected for both
new and existing databases, except Printer Setup which needed to be checked for each run, as it may be
changed by the use of other applications.

Language contains a selection of languages which can be used in the user interface. Elmod is delivered
with English as standard, and prepared for translation into other languages if feasible. In the current list of
languages only few of the choices are active, and they may not be updated in the current release, as this
will only happen in cooperation with clients demanding translations.
Units. Elmod can work with either the SI Metric system or with the US Customary system. Switching
between the two systems of units can only be done when no project database is open. Once a project
database is created with the defined unit system, it cannot be changed within the project database.

Weather Report Site indicates a link to a website where temperature information is available. In the
Estimate Moduli window, there is an option to click the Link button, when using Bell’s equation for
correcting surface temperature to mid-depth asphalt temperature. Clicking this button will open the
website to allow searching for temperature information for the specific area and period.

Tooltips activated. Elmod6 comes with a comprehensive tooltips functionality, giving information boxes
when moving the mouse pointer over menus, buttons and text fields. This is useful for new users of
Elmod6, but after a while when getting experienced with the functionalities you may wish to switch it off,
as they can be annoying if you don’t need them.

Printer Setup. This allows changing the default printer As all reports now goes into Microsoft Word or
Microsoft Excel you can just as well set the default printer in these applications. The only exception is the
print graph facility in the PLOT program, where the graphics are sent directly to the printer.

Technical

If a data file is open the settings on this tab will always be saved to the current project. By checking Save
also as permanent the changes will also be save for use with new projects created. If no data file is
open changes will be saved as permanent changes.

Sort data by chainage. This is used when importing FWD files. Usually it is most convenient to have the
data sorted by chainage because pavement structure data are entered by chainage and some plotting
features like sectioning and XY graphs only works if stationing is in increasing order. One possible reason
to keep the original order of the test points is if the file contains measurements in both directions of a road,
and the two directions have to be analyzed separately. In this case the sorting can be unchecked and the
Review data facilities can be used to split the file into two more easily than if the data were sorted.
Automatic insertion of seed values. In the Structural data window where you define the pavement
structure, Elmod will use this setting to determine if seed moduli should be inserted automatically when
selecting materials. If it is switch on, Elmod will insert the materials reference moduli as defined in the
Parameter setup. The seed moduli are used as start values in the iteration procedures for calculating the
layer moduli.

Use moduli restriction. In the Parameter setup you can set a minimum and maximum allowable modulus
for each material. If you switch on the restriction Elmod6 will use these values in the backcalculation of
moduli, and not allow the modulus to exceed these values.

Distress reduction factors. These values relate to a project database, and can only be set when a
project database is open. They can be relevant to use if FWD tests have been performed on cracked
surfaces, and the FWD operator during the tests have indicated this in the datafile using the facility in the
Field program where he can indicate moderate cracking or severe cracking. When the surface is cracked
it may be impossible or irrelevant to try to backcalculate the modulus of the top layer. Instead the modulus
of the top layer can be fixed to an estimated value, and Elmod will then reduce this fixed value with the
factors entered for moderate and severe cracking.

Load transfer. When joints have been tested in relation to jointed concrete pavements, the load transfer
can be calculated. The load transfer efficiency can be calculated according to two different methods. Both
methods make use of the deflections from the two geophones positioned at each side of the joint, and a
bending factor from the mid-slab test using the deflections from the same geophones. Westergaard’s
equation provides the actual load transfer between the two slabs, whereas the FHWA equation is the
deflection ratio.

D refers to deflection, J and M to joint and mid-slab and 1 and 2 to first and second geophone chosen for
the joint calculation.

Em factor. This factor’s default value is 1. The Em factor is a design factor. The "measured Em"
corresponds to the dynamic FWD load. If the subgrade design criterion was based on a static plate
loading test, and you know that the dynamic modulus is 1.3 times the static modulus, for example, then
you would enter this factor, and the design part of the program will use the static Em. Similarly if you have
been using results of a triaxial test, and you know that they give 0.3 times the dynamic FWD value (as
claimed by AASHTO) then you would enter a factor of 1/0.3 = 3.33.

k factor.
A similar factor can be applied to the k – value, used with Westergaard’s equations, and also used for
reporting the subgrade category with PCN for rigid pavements. Notice the "dynamic" k-value is considered
to be less than the "static" derived k-value, in which case the entered factor must be less than 1.

Colors
Changes on this tab will be valid for both current project and for new created projects.

You can select colors for various objects in the windows. For window background color you can specify a
graduation by setting different top and bottom color. Level 1 refers to the Main menu window, and Level 2
refer to windows that open on top of the main window, likewise Level 3 refers to windows that opens on
top of Level 2 windows. Click in the boxes to change the colors. Click the Change to factory settings
button to reset colors to the default settings as they came with the installation.

Projects and Files

Data Files
Create, Import and Open
Viewing and Editing
Deleting Results and Files

All tools for creating and opening project files, and for importing and opening FWD files are found in the
Main menu item Projects

To work in Elmod6 with data collected from the FWD equipment, you have to create a project file with one
or more imported FWD files. The project file is a Microsoft Access database, and once FWD data have
been imported to the project file, all Elmod6 work is based on this database. This means that the FWD
data files will not be used anymore. It is recommended to keep the FWD files in a safe place for later
reimport in case the project file should get lost or corrupted.

The idea is to have measurements from one project collected in one project file. This may be a collection
of files from an airfield, from a single road, or from multiple roads belonging to the same contract. A
reasonable number of files in one project database would be from 1 to 50 or 100, but there is no upper
limit, except that the size of the database could grow to an unmanageable size or the performance would
slow down at some stage.
Data Files
The user can open and analyze all standard FWD files stored by a Dynatest FWD or HWD regardless of
field program version or file format. A standard dialog box enables choice of directory and subdirectory
location. By default, all files having extensions as FWD, F10, F20, F25 and MDB are shown. If another
extension has been selected by the user e.g. DAT, these can be displayed in the file dialog boxes by
selecting "All files" in the File Type Box.

Elmod 6 can analyze deflection basins comprising up to a maximum of 16 sensor readings.

Portland Cement Concrete pavements can be analyzed using Elmod 6, but require specific data
collection procedures if Elmod 6 has to calculate joint conditions. These procedures are outlined in the
Concrete Data Considerations section.

Create, Import and Open

To create a new project database select: Project → Database → Create New
This will open the File Dialog Box:

The standard extension for Elmod project databases is .mde. You must not try to change this. These
windows could look a bit different depending on the version of your operating system.

Browse to the appropriate location, enter a name for your project file and click Save. If the filename
already exists, you will be asked if you want to open the existing file. If you select No, you will be asked if
you want to overwrite the existing Project database.

The program will now create the new project database, which is empty except that the existing default
parameter setups and temperature files have been copied to the new database. The new project
database will automatically become the current opened database, so you don’t have to open it before
importing files. In the Windows Title Bar you will see the path to your database:
You are now ready to import FWD files.

To open a project database select: Project → Database → Open

This will open the file dialog box:

When you browse you will see a list of *.MDE files (the standard extension for Elmod project databases).
Select the appropriate project database by clicking the filename, and click Open.

Notice: If you select an Access database that is not created with Elmod version 5 or 6 the program
might crash.

The database is now the current project and the path to the database is displayed in the Windows Title
Bar. You are now ready to open a file, or to import new files.

If the project database only contains one file, this file will be opened automatically. If there are more files
you will get the list of files for selection.

To import FWD files select: Project → Datafile → Import FWD File to Database

Use this option to import a single file or to import multiple files from the same directory.

You will get the file open dialog box:

Files with standard extensions will be displayed from the selected directory. Notice that also the MDB
format from the Dynatest Field Program FWDwin can be imported.

You can now select a single file or a range of files to be imported. To select a range of files use the
standard Windows functionality <Shift> or <Control>.
If you only select a single file, you will have the opportunity to change the name, when you have clicked
the OK button:

Next, the import will be carried out, and the last imported file will be the current file as you will see in the
Windows Title Bar:

Also the file will be added to the list of recent files displayed in the Project submenu. Next time you open
Elmod 6 you just have to click the filename from this submenu, and the project and file will be opened
automatically.

If you select Project → Datafile → Import From Several directories you will have a more convenient
way of importing files one by one located in different directories, and you have the possibility for each file
to change the filename.

To make another file from the current database the current file select:
Project → Datafile → Open

You will then be presented for a list of all imported files in the current project database:

Click the desired file and click OK.

Viewing and Editing

File Splitting
Temperature Curves

Select Review Data to view and edit data imported from the FWD file.
Be aware that the FWD raw file, used for import, will not in any way be affected by changes. This also
means that a later import of the file to the same name, will overwrite any changes done to the data. You
will then see the following information displayed in the window:

At the top of the window you have general information from the file setup, and in the table below, the
actual measured data are listed. Only white cells in the table can be edited. If you need to correct plate
radius or a geophone distance, you have to make these corrections to the FWD raw file before import. A
simple text editor can be used to edit the raw file, but you have to be careful not to spoil the format of the
file (like line length and number of lines).
The geophone distances are calculated values based on the geophone positions as defined in the file. By
clicking the button View geophone positions you will see a graph of the geophone setup. Notice that all
data are sorted according to geophone distances when the file is imported. So the columns with
deflections may look different from what you see if you open the FWD raw file in a text editor.
The usual reasons for editing data are that you need to make corrections to station identifiers, to
temperature readings or because the data of a certain test point has been corrupted.
To correct a data field, you simply left click the field, and enter the new value. Press <Enter> or click
another field to accept. Changes will not be saved before you press the button Save New Data.
To remove all data for a test point press <Alt> and left click the station data field in the Chainage column.
If you want to keep the information, you can instead eliminate a test point from the calculations by
changing the station to its negative value. Then the point will not be within the section used for
calculations.
View file info button gives access to view general information from the file, like operator id, section id,
and possible comments made during the testing.
In the grid there are to columns named JointID and PosID. These are only relevant, if the special setup
for testing on PCC slab joints has been used during testing. See Concrete Data Considerations for
further details on the numbers.
Insert asphalt temperatures is a small facility for automatically entering asphalt temperatures into the
file. This can be useful in case the existing values in the Asphalt temperature column are not correct, or if
they were not entered during the field testing. You will get these input fields:

Enter Point No. for first and last station where temperature should be inserted, and enter the desired
temperature. Click Insert to transfer the temperature to the asphalt column, and you can continue with a
new range and a new temperature. Changes are not saved before you press the button Save New Data.
It is not a requirement that you have asphalt temperatures. Alternatively Bell’s equation can be used to
estimate the asphalt temperature from the measured surface temperature and user input for last day’s
average air temperature. Input regarding this is entered in the Moduli window, see Estimating Moduli
Split file is an option for dividing the current file into two new files. The button will be enabled when you
have marked points to be moved to a second file. See File Splitting for help on this.
Temperature graphs opens a window with graphical presentation of the recorded temperatures, and
various options. See Temperature Curves for help on this topic.

Temperature Curves
Viewing and Editing

This option is available from the menu Review Data by clicking the Temperature Graphs button.

In the following example you see temperature curves for air and pavement surface.
The asphalt temperature was not recorded.

You now have the option of seeing the effect of getting the asphalt temperature by using Bell’s equation.
This requires information about average air temperature the day before FWD testing and thickness of the
asphalt layer. When you have entered this inforrmation click Bell, and the results will be included as a
new curve in the graphics.
You can refresh the asphalt temperature graph for other input values to see the effect.

Interpolate is a facility which can be used with the manually recorded asphalt temperaures. In the field
during testing the usual procedure is to measure the temperature in the middle of the asphalt layer,
whenever it is expected that the temperature has changed significantly. For test points measured between
the temperature measurements, the asphalt temperature is in the file stored as the last recorded asphalt
temperature. When clicking Interpolate the program will do a linear interpolation between test points
where asphalt temperature has changed in the file, and calculated the asphalt temperature for all points
between these. The interpolation can be set to either based on distance bewteen points or time spacing
between points. The results of the calculation will be added to the graph, nad it is following possible to
transfer the values to the Revew data window, by clicking the Transfer button. From the Review Data
window you must use Save to permanently save the new temperature data.

Notice that the points are sorted according to time of measurements.

Options
Selecting Options from Main Menu gives access to various settings and functionalities. You will have the
following submenu items:

Edit Aircraft Database

Edit Geophone Setup
Edit Temperature Table

Edit Aircraft Database

The purpose of the Aircraft database is to have all aircraft data organized for easy access in Elmod
Design and PCN calculations. If the data for an aircraft are complete, you only have to load the name of
the aircraft into the parameter file, and all relevant data will be imported. It is following possible to change
any of the data if necessary.

Select Options → Edit Aircraft Database to access this facility.

A window will open with a list of aircrafts stored in the database. Click an aircraft to view or edit the
information. Click Add New button to add a new aircraft type, where all data fields are blank. You can also
create a new aircraft type based on information of an existing aircraft, by clicking Copy data to new
aircraft button on the tab General.

If you select an existing aircraft from the list, you will get the present information in the database for the
specified aircraft.:

If you edit or inserts new aircraft types you may not want to fill in all the information. For that reason the
data fields are marked with small colored boxes to indicate which information that is important to which
part of the program.

If you enter new information, you can select which unit system you want to use in the option box at the
upper, right corner of the window. This works independently of the unit system you are using in general for
your project databases.

General
The name or type you enter in the Aircraft field must be unique. The button Copy data to new aircraft
type is used to create a copy of all the information for the loaded aircraft to be saved with a new aircraft
type. This is useful when creating a new entry with similar information as one of the existing.
If the name of manufacture is not in the standard list supplied, you can enter a new name.

In Diagram you can connect to a bitmap showing the gear configuration of the aircraft. Clicking View
diagram opens a little window with the selected type:

The letters W, A and T refers to wheel distance, axle distance and tread.

Wheels
On the tab Wheels you will find a table with wheel configuration input for each gear.
This information is used in the Elmod design procedures to calculate the load magnitude and position for
each tire.

The Wheel name must be unique, so possible conflicts with other aircraft’s wheel types can be avoided.

If Center wheel is set to Yes it indicates the configuration does not have a duplicate, symmetrical to the
aircrafts center line, which is the case for nose wheels and body wheels. If it is set to No as for wing
gears, Elmod will distribute the load indicated in the Load % column on two symmetrically positioned legs.
Tread indicates the distance from the centerline of the aircraft to the center line of the wheel configuration.
In the Elmod design a testline offset can be entered, and the difference between the tread and the testline
offset is, for each aircraft defined in the fleet mix, used to give the position of the load compared to the
position of the line for which critical strains and stresses are calculated.

To edit any of the data you have to click the appropriate name in the Wheel name column, and then click
Edit wheel:

If data that are critical for the design are not entered, Elmod will not include the wheel in the design
calculation, however it may still be accessible for the PCN/ACN calculations.

ACN Details
ACN values for Weight, min can be left blank if they are not relevant. If they are present then Elmod is
able to interpolate ACN values to any Weight entered in the facility in Plots.

Edit Geophone Setup

Select Options → Edit Geophone Setup to reduce the number of active geophones used in the analysis.
There can be various reasons to do this:

• Several geophone distances are the same, which is not allowed in the analysis.
• A geophone has failed during testing.
• Experimental or research related reasons.

In this example 15 geophones have been used during testing:

To deactivate a geophones channel, click on the green area representing the channel, and it will turn red.
Clicking on a red deactivated channel will make it green and active again.

Moving the mouse over the geophones will give a text box with channel information, as seen in the figure.

If several geophones have the same distance to the center, like in the example, you will have to
deactivate one, or more, geophones. In this case at least 6 geophones:
Click save to accept the changes. After calculations it is always possible to go back and change to
another setup of geophones, but earlier analysis results will be deleted. If you want to keep results with
different setups, you can import the file to the database multiple times, and give each file a unique name
in the database.

Edit Temperature Table

Select Options → Edit Temperature Table to view and/or edit an existing temperature file defining the
asphalt modulus/asphalt temperature relationship. These are the data used if Factors is selected in the
Parameter Setup for a material’s Temperature Sensitivity .

First you will have a list of existing files (tables) in the database. If you have a project database open, the
list reflects the files in the project database, otherwise you will have the list from the default database.

Click the file and the OK button

You will now get the content of the table in a window:

From here you can edit each cell for temperature and factor. The factor refers to 1 which should be set for
the reference temperature.

Click Delete if you want to remove the table from the database.

Click Save to save the table. You will be asked to give a name for the temperature table:

Use Paste to copy data from the clipboard into the table. You can use the Copy function from a Microsoft
Excel spreadsheet to get the data into the clipboard. The structure of the data must be as shown below.
Structures
GPR Data
Jointed Pavements

Backcalculation and design programs need to know the composition of the pavement structure before any
calculations can be done. This involves specifying the number of layers in the pavement structure, and
their respective thickness.

Select Structure from the Main menu to get the input screen:
In this example the minimum requirements of input data have been entered in terms of layer thicknesses.
In this case a 3-layer structure has been defined, where the subgrade is the third layer with infinite
thickness. The first time you open the window the From and To fields are automatically filled in with the
lowest and the highest chainage from the file.

Sections
If the deflection data was obtained on a road with varying layer thickness or number of layers, the road
can be divided into subsections corresponding to each unique structural pavement section. This is done
by first changing the end station in the To field to the end of the existing section, and then click Add
section after No. 1. If 99 was entered as end of section 1 in the example, this will result in following
window:
and you can then enter input information for section 2.

At the top you can switch between the sections, and you can continue defining the number of sections
using the buttons Add section… and Delete section…

The maximum number of sections you can define is 20. If you have a need for defining more than 20
sections, you will have to first split the file into two files, see Viewing and Editing

Thicknesses
Pavement layer thicknesses are typed in the Thickness fields. Elmod 6 can handle up to 5 layers,
including the subgrade. Thicknesses are entered from the top down. Layer 1 is assumed to be the surface
course etc. Since the subgrade is infinitely thick, no thickness information is entered for this layer. For
example, if the pavement structure is composed of a surface course, base, and subgrade, thickness
values would be entered for Layers 1 and 2 only. All other layer thickness would be left blank. [In this
case, Layer 3 would be assumed by Elmod 6 to correspond to the subgrade.]

Seed Moduli
These input fields can have two purposes. Either to indicate a modulus only to be used as a seed
modulus for the Deflection Basin Fit method used in the backcalculation process or to fix the modulus of a
given layer. How it is used is determined from the input settings to the Deflection Basin Fit method, see
Estimating Moduli

The backcalculation process is sensitive to the seed moduli, and the more realistic the seed moduli are,
usually the better results are produced by the iterations process used in the Deflection Basin Fit
procedure. If no seed moduli are entered, then the Deflection Basin Fit method will calculate seed values
by using the Radius of Curvature method.

Sometimes it can be necessary to use the seed modulus as a fixed modulus This is useful when a layer is
too thin to backcalculate (typically less than 75-100 mm or 3-4 inches), or if the program is backcalculating
unrealistic modulus values for a given layer. If it is used with asphalt materials, the modulus can be
defined at a given temperature, which can be entered, otherwise it is assumed to be the modulus at the
temperature recorded during the FWD testing

The "E2/E3" and similar fields are used to assign a modulus ratio of two (or more) adjacent pavement
layers. These are typically needed when the Radius of Curvature (based on the Odemark/Boussinesq
transformed section approach or theory) backcalculation method is used, as this method of
backcalculation can only handle up to three independent layers. If the user specifies more than three
layers, Elmod 6 will prompt the user to enter information for these fields, even though this information will
be disregarded by the program when the "basin fitting" method of backcalculation is used (it will only be
used for estimating seed moduli). The user should use engineering judgment when entering modular
ratios. A "default" number appears in each field when the appropriate button is clicked. The default value
is calculated using the relationship derived by Dormon and Metcalf ("Design curves for flexible pavements
based on layered system theory", Highway Research Record No. 71, 1965), as follows:

where Eg is the modulus of the granular layer of thickness hg (mm) above a subgrade with modulus Es.

For the purpose of backcalculation a "layer" may consist of several materials. In most cases, asphalt
materials should be combined into one layer. It may be advantageous in certain cases to combine an
asphalt layer with a gravel layer or with a stabilized layer, depending on the stiffness of this second layer,
to better simulate the theoretical conditions on which the approach is based.

Max depth to rigid layer

At the bottom of the dialog box, the Max depth to rigid layer field appears. This is used to activate the rigid
layer calculations in Elmod 6. Leaving the field blank excludes this option and Elmod 6 considers an
infinitely thick non-linear subgrade.
The presence of rock, or a stiff layer, at shallow depth will greatly influence the deflection basin. The
program considers this influence by first calculating the equivalent depth to rock using the measured
deflections, and then carrying out the analyses for the layers above the rigid interface.
If this option is chosen, the output will contain a table showing the calculated equivalent depth to the stiff
layer. The maximum depth that can be input is 9999 in. (or mm), but a more reasonable depth to use is
200 to 300 in. (5000 to 7500 mm). A fairly common default value used by other back-calculation programs
is 240 in. (about 6000 mm).

Note that Elmod 6 does NOT use the input depth to fix the location of the stiff layer. Basically, the input
depth is used as the limit below which the stiff layer is considered (by the user) to not influence the
measured deflections. Elmod 6 calculates the equivalent depth of an apparently stiff layer, compares this
calculated depth with the input depth and if it is less than the input depth uses the calculated depth to
perform the analysis. If the calculated depth is greater than the input depth, the analysis reverts to
consideration of a semi-infinite, non-linear subgrade.

It is possible, however, to let Elmod 6 use the input as a fixed value. To do this you must check the Fix
depth to rock option on the backcalculation window, see Estimating Moduli

The equivalent depth to bedrock can be calculated as:

where,
f1 is a correction factor equal to 0.8 except for the first layer,
where it is 0.9 for a two layer structure or 1.0 for
a multi-layer structure

n is the number of layers (excl. bedrock).

Hi is the thickness of layer i

Ei is the modulus of layer i

Es and Hs are modulus and thickness of the subgrade material above bedrock.
PCC is layer No.
This is used to indicate layer number for a PCC layer, counted as 1 from the top. This is relevant if PCN is
to be used or in case of jointed pavements with joint tests.

Use PCC joint ID numbers

Check this if testing includes jointed pavements, see Jointed Pavements

Use GPR data

Check this if you want to operate with point related thickness data instead of section related thickness
data, see GPR Data

Use Parameter Setup.

It is possible to define the material type for each layer together with thickness information. To do this you
must first specify which parameter setup you want to use for the project. This must be the same as you
intend to use for the design calculations. If you enter materials types here, you don’t have to enter them
again as input in the design calculation window.

One advantage of entering the information before backcalculation is that reference moduli can be
calculated as a part of the backcalculation procedure, and these can be reported from the PLOT facility as
the E-moduli graphs and tables (Selectable from Graph Options). The reference moduli for asphalt
materials are the moduli at the reference temperature, and for unbound materials the modulus at a
modulus factor = 1.

When the Parameter Setup is chosen from the combo box, you can select the materials types from the list
popping up when clicking in the material fields:

Depending on your settings the seed moduli fields are automatically filled in with the reference modulus of
each material. This is an option that can be turned on/off in Settings

Clicking on the computer icon next to the Material caption will copy the materials and seed moduli to the
other defined sections.

View Structure
Clicking this button opens a window with graphics of the defined sections and their layer definitions:
This can be used to check whether the complete file is covered by sections as intended. Further by
moving the mouse cursor over the layers, you can see the layer information input. The hatch for each
layer is chosen in the parameter setup, and can only be shown on the graphics if the material types are
entered.

GPR Data
Structures

As an alternative to entering section related thickness data, it is possible to enter thickness data related to
points, as obtained from GPR measurements or from core samples. The GPR option is working on file
level, meaning you cannot both have section thickness and GPR thicknesses connected to the same file.
From the Structures window you will check the Use GPR Data checkbox. This will enable the
Import/Edit button, and the Thickness fields will be disabled:

Click the Import/Edit button to get an input window for the thickness data:
You can have as many entries as you like. The chainage column should reflect the stationing of the
thickness data. During the backcalculation Elmod will, for each FWD test point, search for the nearest
GPR data point, and use the thicknesses entered for this. Thicknesses must be entered consecutively
from top to bottom in the same way as for section related thickness data. You can enter points with
various number of layers and do backcalculation for those. But for design you must have the same
number of layers for each section. You can handle this by defining the sections in the Structural Data
window, according to the changes in the number of layers. This can also be handled in a more automatic
way, using the sectioning features under Plot

Example of grid filled with point related thicknesses:

You can enter or edit the contents in the grid by clicking the appropriate cell.
Please notice that using points with different layer thickness within a section may cause problems if you
later want to determine representative structural values for the section. This means that a section with
varying layer thickness can not be exported to the LCCA module.

Import new data from text file

Use this facility to get thickness data into the grid, if you have already organized the input in a text file.
Clicking the button will open a file browser and you can select the appropriate file. There is no limitations
to the name of the text file, but the format of the text file must be like in this example from Notepad:

where you have the first column with stationing, and following 4 columns with thickness data for layer 1 to
4. In case of no layer a 0 must be entered. The values must be separated with spaces or <Tab>.

Paste from Excel

You can also copy data from a Microsoft Excel sheet into the table. Enter the data in Excel as seen in the
example below and copy the fields inside the frame to the clipboard. Then click the button Paste from
Excel to get them into the table. Notice that you don’t enter zero’s in the fields not used, unlike when
importing from a text file.

When you click Save all information in the grid is stored to the Elmod database and connected to the
proper file, so they are available next time you open this file.

If the information about layers in the GPR table does not match the existing section definition, a blinking
icon with a cut symbol will appear next to the To field:
If you click the icon Elmod will automatically create the necessary sections, each having uniform number
of layers. In the example 3 sections will be created, and can be viewed clicking the View Structure
button:

Jointed Pavements
Structures

To be able to calculate for joint conditions, you have to use a field setup and procedure as described in
Concrete Data Considerations

Then to activate the analysis in Elmod you have to check the Use PCC joint ID numbers checkbox in the
Structures window.

This will enable some options:

Verify Plates
Click this button to get an overview of which joint measurements that will be connected to which center
slab measurements in the analysis. For calculating k-values at joints the program will need the E-value as
calculated at the center of the slab. The program will search for the closest slab center measurement
using the stationing of the test points. This means the distance from the joint to the corresponding center
must be less than the distance from the joint to any other slab center. This has to be considered during
testing.
The overview gives a line for each joint measurement, with an indication of the corresponding center
location. If it can be seen that the program will connect a joint to the wrong slab, this has to be corrected
by editing the station for either the joint test or the slab center test to reduce the distance. This can be
done in Viewing and Editing. The overview gives also information about the location of the joint test.

Change channels for joint calculations

In the data fields below the button, you will see which two geophones Elmod will use for calculating load
transfer between the slabs. Elmod will usually be able to detect the correct ones automatically, looking for
the geophones in distance 200 mm and 300 mm from the center. But if they are indicated as 0, or the
geophone setup has been altered, because a "star bar" with multiple geophones at the same distances
has been used, then it might be necessary to change the values. You can do this by clicking the button
Change channels for joint calculations and you will get the Geophone setup window:
You can now change the geophones used for joint calculation, by editing the two data fields. The blue
rectangle will move according to your changes. When clicking Save or Cancel you will return to the
Structural data window.

Estimating Moduli
When pavement structures have been entered it is possible to perform the backcalculation of layer moduli
selecting Moduli from the Main menu.
Select Drops
This allows the user to select which drop to analyze. More than one drop may be selected, in which case
each single drop will be analyzed separately

Select Sections
The Select Section panel allows the user to select which section to analyze. See Structures and
Sectioning for details about creating sections

Temperature Status
This gives an overview of which temperatures that are available in the data file, and the minimum and
maximum recorded temperatures. From the Viewing and Editing feature you have access to see
graphics of the recorded temperatures.

Asphalt Temperature
Here you decide which should be used as the asphalt temperature in the calculations, making it possible
to correct the asphalt modulus for temperature to be used with design. Please see Temperature
Sensitivity for details about using asphalt temperatures. If manually recorded mid-depth asphalt
temperatures are present in the file, select the Use as recorded option. In the Temperature Status
frame you can see if asphalt temperatures are present. If only surface temperatures are recorded you can
select either to use them directly as the asphalt temperature (not recommended), or you can use them as
input to Bell’s equation for calculating the asphalt temperature as a function of surface temperature and
previous day average air temperature. In the latter case you must enter this average temperature in the
temperature field. Clicking the Link button opens a website with worldwide temperature measurements.
Which website to open is defined in Settings. You can select the Not required option if there is no
temperature sensitive materials in the pavement structure, or if the backcalculation results are not going to
be used for design.

Fix depth to Rock

Checking this option will regard the input from the Structures window as the exact depth to bedrock,
consequently Elmod will not calculate a depth within the maximum range given.

Deflection Basin Fit Options

This frame contains settings used when running the Deflection Basin Fit method. At the top of the frame
there are options to fix the seed moduli entered in the Structures window. This means that the layer
modulus will not be calculated, and the iteration procedure will fit the other moduli based on this. C and n
refers to the subgrade modulus and the non-linearity coefficient. If you want to fix n, a value for n must be
entered in the field to the right.
You can check the minimize % difference option if the iteration should use this for matching the
deflections rather than minimizing the absolute difference which is used as default.
Old moduli as seeds will use previous results as seed values for the new calculation.
Offset and Steps defines the steps in the iteration procedure. Offset defines how wide a range of moduli
should be searched in next iteration, and steps defined how may "points" within that range which should
be picked for further iterations. This goes for all defined layers. The higher the values the more detailed
the iteration. If steps = 1 then only actual value +/- offset will be picked for next iterations, if steps = 2 then
also actual value +/- offset/2 will be picked etc.

Backcalculation Methods
You can select either Radius of Curvature or Deflection Basin Fit or FEM/LET/MET if included. The
Radius of Curvature method is very similar to (but not exactly the same as) the method used in earlier
DOS-based versions of ELMOD, all of which were based on the Odemark-Boussinesq method of
equivalent thicknesses (MET). The Deflection Basin Fit option also utilizes Odemark-Boussinesq
methods, but the convergence criteria is based on degree of fit between the measured and calculated
deflection basins including input from all geophones used. Clicking one the buttons Radius of Curvature
or Deflection Basin Fit will start the calculations, while the FEM/LET/MET button will open a window with
input settings for further processing.

For each new calculation results from earlier calculations for the same section and drop will be
overwritten. To preserve previous calculations you can create several sections for the same length of
road.

Reference Moduli.
As default these will always be calculated when running the deflection basin fit option, assuming that
material types are entered in Structures. If FEM/LET/MET has been used for backcalculation they are not
calculated, and you can click this button to have them calculated, and these can be reported from the
PLOT facility as the E-moduli graphs and tables (Selectable from Graph Options). The reference moduli
is for asphalt materials the modulus at the reference temperature, and for unbound materials the modulus
at modulus factor = 1.

Radius of Curvature
Related topic: Estimating Moduli

The "radius of curvature" or "Elmod 3" option uses the radius of curvature along with the actual or
apparent non-linear subgrade properties to determine moduli within the pavement system. Initially, the
subgrade material properties, stiffness and non-linearity, are calculated using the deflections from the
outer geophones. The "radius of curvature" from the central geophones can be used to assess the
stiffness of the upper pavement layer. The stiffness of remaining layers is then calculated based on the
overall pavement response to the applied load. This ensures that the proposed pavement structure results
in the correct central deflection under the measured load.

Deflection Basin Fit

Estimating Moduli

The basin fit option methodology starts with a set of estimated moduli for the pavement structure. The
theoretical deflection bowl for this pavement structure is calculated. The error between the measured
deflections and calculated deflections is then assessed. The moduli in the structure are then
increased/decreased by a small amount (typical 10%), and if the error in either of these deflection bowls is
less than the original deflection bowl this is taken to be a better solution. This process is iterated until a
minimum in error between the calculated and measure deflection bowls are found.

There are some limitations to backcalculation of layer moduli. Several pavement structures (different
combinations of layer stiffness) may result in the same deflection bowl and the process may output any
one of these solutions. The user can assist the process by specifying suitable limits for layer stiffness and
seed values for the iterative process.

For 2-layer systems (a layer on a subgrade) this problem does not arise. For 3-layer systems it may occur,
particularly if the moduli of the two pavement layers are similar or if the modulus of the intermediate layer
is close to the modulus of the subgrade. It may also occur with thin layers. For 4 and 5-layer systems the
same deflection bowl can often be obtained with different combinations of layer moduli. For these cases
the deflection information is not sufficient to determine the layer moduli, and additional information is
needed such as limitations to the individual layer moduli or a fixed modulus for one or more layers. The
alternative is to reduce the number of layers by combining similar layers.

Batch Runs

A batch-processing feature has been included in Elmod 6 to simplify and automate the backcalculation
process. [Please note that overlay design can not be run in batch mode.] This feature allows the
computer to process many deflection data files with minimal interaction by the user. Batch processing is
useful when large numbers of deflection basins are to be analyzed to determine moduli alone. It allows
the user to run Elmod 6 overnight or over weekends while the computer is normally not in use. The batch
processing feature also provides the ability for several computers to access data on a network server and
perform backcalculation simultaneously, thus speeding up the backcalculation process. This requires that
you have a multi-user license, and that you split your files into several projects. The following steps outline
the process of running the Elmod 6 program in batch mode:

Projects → Batch

First the program checks if structural data have been entered for the files present in the project database.
If some files have no input for thickness data, the following message box will appear:

If you select NO because you want the possibility to enter thickness data for these files, you will get the
following window:
Click the arrows to select/deselect single files, or click the buttons at the bottom to select/deselect all files.

When you click OK to continue, you will now be asked to enter structural data for each file:
See Structures for details.

When all structural data have been entered, you can select files for the batch run:
Use the select/deselect options as described above, and click OK to enter final input:
You will have options to run Radius of Curvature or Deflection Basin Fit

You can now start the calculations by clicking OK.

FEM/LET/MET
FEM Plot
Edit Mesh

This option is only available if a license has been issued by Dynatest.

This option cannot be used with jointed pavements.

With this option backcalculation may be carried out either with the Finite Element Method (FEM), Linear
Elastic Theory (LET) or the Method of Equivalent Thickness (MET). FEM makes use of a modified
version of an axial symmetric finite element program, originally developed by Wilson at University of
California (Duncan et al., 1968). LET makes use of the Waterways Experiment Station’s program
(WESLEA) and MET is similar to the method used in the Elmod option "Basin fit", but with a simpler use of
adjustment factors. Where Elmod applies adjustment factors directly on the calculated responses, MET
makes use of the traditional Odemark adjustment factors directly on the layer thicknesses. The moduli
backcalculated by the two methods may, therefore, be different.

It is important to notice that this option is for backcalculation only, not for calculation of remaining life or
overlay design. The moduli calculated with this option are for the test conditions, without any adjustment
for temperature or seasonal effects. Therefore temperatures are not checked. The program may,
however, be used to calculate the stresses and strains under a single load. If all the points of a section are
evaluated using the Fit all points option, the moduli will be stored in the database and can be used for
calculation of remaining life and needed overlay design in Elmod.

When the option FEM/LET/MET is selected on the Estimating Moduli screen, the main input and output
screen is shown, for example as in Figure 1. The default data shown on this screen, refers to the first drop
and the first section selected on the Moduli screen in Elmod. The chainage will be from the lowest to the
highest value for the first section, and the two numbers in the From drop and the to input boxes at the top
of the screen will both be equal to the first drop number selected on the Moduli screen. Both chainage
and drop numbers may be changed by the user. If, for example, the From drop value is changed to 1,
then the FWD test data used will be the average values for all of the drops 1 to 4.

When opening this option, the test data corresponding to the first chainage and the first drop number will
automatically be imported. The contact stress is shown in a text box as is the radius of the loaded area
(Plate radius). For backcalculation none of these values should be changed. For forward calculation of
pavement response, different values of contact stress, radius of loaded area and an offset distance input,
which can be set in Options, may be entered.

The measured deflections are shown in a column next to the column giving the distances to the
geophones. In Figure 1 a deflection of 640.0 is measured at the center geophone (distance 0). In SI units
deflections are in µm, distances in mm and moduli are in MPa. The contact stress is in MPa, but output
stresses (response) are in kPa and output strains are in µstrain. The corresponding US units are:
deflections in mil, distances in inch and moduli in ksi. Contact stress is ksi, but output stress is psi.

The layer thicknesses are those corresponding to the section and the moduli are the default values
entered in the Structures form (in Elmod), next to the thicknesses. If no default moduli have been entered
in Elmod, the program will suggest some values. Poisson’s ratio is 0.35 for all layers, as a default, but
may be changed. Default values are also used for the minimum (Emin) and maximum modulus (Emax),
of each layer.
Figure 1. Main input and output screen for FEM/LET/MET.

The column with the heading n indicates the non-linearity of the materials. With FEM all layers may be
non-linear, with LET none of the layers and with MET only the subgrade may be non-linear. With a
negative non-linearity (n is negative, cohesive materials), n will be the exponent in the relationship:

where C is a constant equal to the value in the E column,

σ1 is the major principal stress, and
p is atmospheric pressure.
With FEM the vertical stress, σz, may be used instead of σ1, by checking the Use zigz with FEM box on
the Options screen.

With a positive non-linearity (granular materials), n is the exponent of:

Where θ is the hydrostatic stress (mean normal stress or first stress invariant, 1/3 times the bulk stress).

The deflections may be calculated by pressing the command buttons: FEM, LET or MET. Before FEM can
be used, however, the finite element mesh has to be set up. This may be done by pressing the command
button Generate FEM mesh with, where the number of columns and rows is given in the text box. A
maximum of 100×100 columns and rows may be used. The program suggests a distance to the outermost
node. This may be selected or overwritten. If no mesh is generated, the program will create a mesh based
on default values.

Figure 2 Input box for maximum distance in finite element mesh.

The mesh is saved in the active Microsoft Access database in the tables "Column" and "Row". The default
mesh may be edited in the database.
Backcalculation
The layer moduli, the non-linearity and the depth to bedrock may be varied in the backcalculation. The
checkbox next to the parameter determines whether a parameter is varied or kept constant. In Figure 1,
for example, the moduli of the three layers are all varied, whereas the non-linearities are kept constant
(equal to 0 for the first two layers and -0.01 for the third). In this example no thickness has been entered
for the subgrade which is, therefore, treated as semi-infinite. With LET all layers are treated as linear
elastic and with MET only the subgrade may be non-linear, as mentioned above. With FEM the depth to
bedrock will not be varied.

The option buttons below FEM, LET or MET are used to select the backcalculation method. If
backcalculation is to be done for a single point only, the button Fit point is pressed.

In the example in Figure 3, Linear Elastic Theory was chosen. The resulting calculated deflections are
shown in the column under LET. The deflections are also shown in an output grid which may be activated
by selecting Show → Details, see figure 4. Here the deflections are shown at the top of each layer in the
structure. The two output lines at the bottom of the grid show the difference between the measured and
calculated deflections, first the percentage difference ({calculated-measured}/measured * 100%) and then
the absolute difference (calculated-measured), in µm for SI units and mil with US units. As a default, the
program will attempt to minimize the Root Mean Square (RMS) of the absolute difference. The check box
minimize % in Figure 3 may be used to minimize with respect to relative difference (percentage).
Figure 3 Example of backcalculation using LET.

Figure 4 Details of deflections

The two text boxes at the top of Figure 3, next to RMS show of the Root Mean Square (RMS) difference
as absolute difference (abs) and percentage difference (%). In the example the absolute RMS was 4.04
µm and the percentage was 1.80. It may be noticed in Figure 3 that the modulus of layer two is lower than
the modulus of the subgrade, a frequent problem when the subgrade is assumed to be linear elastic.

Plotting deflections and surface moduli

Select Show → Graph to plot the deflections and the surface moduli, for measured as well as for
calculated values.

In Figure 5 moduli have also been calculated with MET, assuming a non-linear elastic subgrade. The
results from MET are compared to those from LET below:

Layer Modulus LET Modulus MET

1 2147 1858
2 119 175
3 127 112
With MET the subgrade was non-linear n = -0.02. The modulus at the top of the subgrade, corresponding
to the FWD load, is not shown during back-calculation but can be obtained by selecting Show →
Responses

Figure 5. Plot of measured, LET and MET values.

The response is given for each interface and each layer. The top line is for interface number 1 and layer
number 1 (1/1), the next line for interface number 1 and layer number 2 (1/2), etc. The tangential stress
(zigt), the radial stress (zigr) and the vertical stress (zigz) are given followed by the corresponding strains.
The last column is the modulus at the position where stresses and strains are calculated. For a non-linear
subgrade it may be recalled that the modulus to be used for calculation of deflections is approximately (1-
2n) times the modulus used for strain.

Figure 6. Response under FWD load.

Additional controls
A number of additional input boxes may be used to control the backcalculation. The input box Use first
may be used to limit the number of geophones used for the backcalculation. Sometimes, particularly on
weak pavements, the deflections recorded by the outer geophones are judged to be unreasonable, and it
may be preferable to leave them out. If for example only the first five geophones were to be used, the
number 5 would be entered in the Use first input box.

The menu Calibrate is used to calibrate FEM and MET against LET, which provides the mathematical
exact solution for a layered linear elastic system. First the response is calculated using LET (with an
appropriate subgrade modulus if the subgrade is non-linear). Then FEM is used on the linear elastic case,
and appropriate calibration factors are determined to get the same response as with LET. Then FEM is
used on the non-linear case and with the calibration factors to determine the response in the linear case,
and finally MET is used on the non-linear case and the calibration factors needed for MET to give the
same response as FEM are determined.

Pressing the button Fit point will result in a basin fit with the method selected (FEM, LET or MET), but the
results will not be stored in the database. To store results in the database the button Fit all points must
be used. This will process all of the data points within the given chainage range. Similarly for responses. It
is normally advisable to manually modify the moduli until a reasonably good agreement is obtained
between measured and calculated deflections, and then store these values as seed values using the
Save seed values button.

When FEM is used the results of the last iteration are stored in the tables "Displacements" and "Stress" in
the database.

When a new test point is selected with Select point the thicknesses and moduli are refreshed with the
seed values.

More options are available from the Options menu:

Figure 7. Options

If a value (integer) is entered in Alternate field optimization of the RMS differences will be attempted for
the number of times entered here. If 1 is entered, the program will first minimize with respect to the
absolute difference, it will then use the resulting parameters (moduli, non-linearity or depth to bedrock) as
seed values for a minimization with respect to relative difference (%), and finally redo the absolute
minimization starting from the results of the relative.

MET will use a standard set of adjustment factors, unless a value is given for the adjustment factor, in
which case this will be used for all layers.

If a value is given for the Ei/Ei+1 (the ratio of two consecutive moduli), the equivalent thickness will not be
calculated if the ratio is less than this value. Below this ratio the actual layer thickness is used directly.
This is used only with MET

The input for Factor on E1 for response is not used for backcalculation, but only for calculation of
response. In calculating the response, the modulus of the first layer will be multiplied by the value entered
here. This allows, for example, for a lower asphalt modulus under a slow moving rolling wheel load than
under the FWD impact load.

MET will as default use a point load for calculation of response off the centerline of the load, unless the
box for adjusting for point load is checked.

The number of iterations used by the Finite Element Method is set in an input box, and a checkbox use
zigz with FEM may be used to let the modulus be calculated based on the vertical stress rather than on
the major principal stress.

The value given in the Search width input box controls how much the parameters are varied. A large
value will search with larger steps.

Use tail for C and n seeds calculates seed values for these based on the deflections from the outer-most
geophones.

References

Duncan, J.M., Monismith, C.L. & Wilson, E.L. "Finite Element Analysis of Pavements", Highway Research
Record 228, 1968.

FEM Plot
FEM/LET/MET

The results of the last Finite Element calculation may be plotted with FEMplot.
The types of response that may be plotted are given in the list of options in the upper left hand corner.
Stress is in kPa or psi and strain is in µstrain.

Initially the central part of the mesh is plotted, but you may zoom in or out by using the "In" or "Out"
command buttons.

The loading plate is indicated as a gray rectangle and the layer interfaces by white lines.

The initial plot uses a scale from the "maximum" to the "minimum" value found in the calculation.
"Maximum" should be interpreted as worst, and will be the highest negative value for example for the
tangential strain. The scale may be changed by manually entering the maximum and minimum values.
When clicking a point within the mesh area the coordinates of the midpoint of the closest cell are shown,
as well as the response in that cell. The position of the cursor arrow within the mesh is given in textboxes
next to R and Z and a textbox next to the Z textbox (not shown in the figures) give the value of the
response at the present cursor position.

The Figure below has been zoomed in on the shear stress, and the bold line above the graph indicates
that the shear stress at a radial distance of 144.5 mm and a depth of 45.8 mm is –312.4 kPa.

Edit Mesh
FEM/LET/MET
When a mesh has been generated the positions of rows and columns may be changed, new rows and
columns may be inserted or rows and columns may be deleted. There must be an equal number of rows
and columns, so if for example a row is deleted, a column must also be deleted.

Columns that correspond to the plate radius or a geophone distance and rows that correspond to a layer
interface cannot be edited.

When entering "Edit mesh" about half of the mesh is shown. It is possible to zoom in or out on the mesh
by pressing the command buttons "In" and "Out" respectively.

To work with columns, the option "Column" at the top of the screen should be chosen. "Row" is to the left
of the screen.

Moving, inserting or deleting is selected from the options in the upper left corner.

If "Column" and "Move" are selected, for example, left clicking within the mesh area will select the closest
(editable) column and turn it green. Releasing the left button will move the column to the position of the
curser. If the curser is moved to another position within the mesh area, while the left button is pressed,
and released, the column will be moved to this position. The textboxes R and Z keeps track of the cursors
position within the mesh.

If you accidentally activate a column and do not want to move it, then move the cursor out of the mesh
area before releasing the left button.

It is also possible to move the column to a position indicated in the R textbox. To do this, activate the
column you want to move, by clicking the left button close to the column, then move the cursor outside of
the mesh area before releasing the left button, and input the radial distance to which you want to move
the column in the R text box. Then press "Move to".

If "Row" and "Insert" are selected, left clicking within the mesh area releasing the left button will insert a
row at the position where the left button was released. To avoid inserting a row, move the cursor outside
of the mesh area before releasing the left button.

After inserting a row, a column must be inserted as well.

The mesh cannot have more than 100×100 rows and columns.
When deleting a row or column, a message box is first shown, asking you to confirm that you want to
delete the row or column. After deleting a row, a column must be deleted as well.

Design Calculation
Overlay Design
New Design

Overlay design requires that you have performed backcalculation generating elastic moduli for the
pavement layers. You can design based on the whole structure or part of the structure by removing top
layers.

New design does not make use of FWD results or backcalculation. You can define a structure and
calculate the design thickness of the top layer.

Overlay Design

An overlay design can only be performed after moduli have been calculated.

Select Design → Overlay Design to access this facility. Following window will open:
Sections with backcalculation results are listed in the grid. In the column Chosen you can mark sections
to be included in the design by clicking in the field.

If parameter file and material types were already entered in the Structural input (see Structural Data),
then this information is shown in the grid, and the name of parameter file used. If the fields with the layer’s
material type are empty, you have first to select which parameter file to use, and then by clicking in the
fields identify the material type for each layer from the popup list.
If you have multiple sections, you can enter the material type for the first section, and following click the
button Copy first structure to copy this to the other sections.

Base design on
Here you select if any existing layers should be removed before design of new overlay. When you select
an option, you will see in the grid which layers are removed indicated by a grayed font color.
If you select Existing pavement then you have an option to mill a part of the top layer, by entering a
value in the frame Milled from layer 1. The value entered must be smaller than the thickness of the top
layer.
If you select Existing unbound layers then you can in the same way have a part of the first granular
layer removed.
If you select Existing subgrade then you can specify an improvement to the subgrade moduli determined
from the backcalculation by entering a percentual increase. Entering a value of 25 means the existing
moduli will be multiplied with a factor of 1.25

Add new layers

Here you define which layer(s) to add on the top. In the usual case where you design for an overlay of the
existing pavement, you will enter one bound layer type. Click in the combo box to specify the design
layers material type. The reference modulus for this material will automatically be inserted in the Modulus
column, and you have following possibility to change this value.

In the case where you wish to do a reconstruction based on the existing subgrade you would normally
also add new granular layers on the top. Elmod will only calculate design values for the top layer, so you
will have to enter both thickness and modulus of the other new layers added, like shown in this example:

The new construction must not have more than 6 layers, with maximum 2 bound layers included.

Traffic
Traffic must be entered as the total passages of all vehicles per year. The number of passages for each
vehicle is then entered in the fields below as percentages of this value.

These are the values directly used in the deterioration functions defined in the parameter file. Issues
regarding multilane roads, equivalence factors, traffic growth etc. must be considered before entering the
value. Facilities are available for this through the buttons in the upper right corner of the Traffic frame. For
help on this see:
Forecasting of Traffic
AADT to ESA
WIM to ESA

If multiple vehicles have been defined in the parameter file, only the Forecasting facility is available.

For airfields you have usually defined a fleet mix of the most critical aircrafts using the facility, as
illustrated in this example:

If a normal distribution for the lateral wander of loads was selected in the parameter file, the offset from
the centerline to the line where the damage should be evaluated must be entered. Only one distance from
the centerline will be evaluated in each design. In this case the critical aircrafts have a tread around
10.000 mm, and the offset should then be around 5.000 mm. You can repeat the design with various
offset values to check if the design is sensitive to the change.
With the Import Traffic button you can select an earlier defined traffic distribution from the same
project.The processed files will be listed, and you can select from which one you will use the traffic
information. The aircraft mix defined must be the same. The button is only enabled when you have
multiple vehicles defined.

If the structure has been defined with jointed concrete as top layer, k-values are calculated for both center
and joints during the backcalculation. In the design window you will have an option for improving k-values
at joints before overlaying:

If you select Subsealing the design calculation will assume the k-values at joints are improved to be the
same as the values under slab center.

Click the OK button to commence the design calculation.

Earlier created designs will not be overwritten, meaning you can view design results in the Plot feature for
various traffic input or other changes.

The button View Plot is a shortcut to the plot of the design results.

New Design

With this facility you can design a new pavement structure, and get the required thickness of the
top layer calculated. No results from the FWD testing or backcalculations are used, so the
facility is available even you have no file opened. You still have to create a database as usual.

Select from menu Design → New Design and you will have following window:
Select parameter setup
Select the parameter file that is valid for your design calculation. You have to select your
parameter file, before you can enter material types in the pavement structure. For setting up
parameters, see Parameters

Add Layers
Here you enter your pavement structure from bottom to top. Only the thickness of the top layer
can be calculated. First layer to add is the subgrade material. The material’s reference modulus
is automatically inserted, and if desired you can change this value. Next you can add the
pavement layers with their respective thicknesses (remember to enter the thickness to the right
of the layer name), and at last the top layer for which the thickness is to be calculated.

Use Monte Carlo Simulation

Check this to base the design on a Monte Carlo Simulation. You will have the following input
window next to the Add Layers window:
This will simulate a variation in the layer moduli according to a logarithmic normal distribution.
Each simulation will generate a set of moduli randomly but with probabilities that follow the
distribution given by the standard deviation factors. The standard deviation factor is 10 raised to
the standard deviation of the logarithms, also called the geometric standard deviation. The
number of runs in the simulation, and the standard deviation factor for each layer must be
entered. The result for each simulation can be viewed from the Plot facility, and the average
and standard deviation will be reported in the Results frame.

Select Method
Responses under the design loads can be calculated either according to MET (Method of
Equivalent Thicknesses) or LET (Linear Elastic Theory).

Traffic
Please refer to description of input in topic Overlay Design .

Results
When you have clicked Start Design the calculation will be carried out, and the results can be
viewed in the Results frame:

If a Monte Carlo simulation has been used, The thickness is given as the average over the
number of simulations and with the corresponding standard deviation.

Forecasting of Traffic
Overlay Design
New Design

This facility is available from the design windows, and can be used to take traffic growth into consideration
by calculating the yearly amount of traffic to be entered for the design.

This is particularly useful if large variations in growth rate between different load types are expected. If for
example a certain aircraft is expected to be phased out whereas a new type will increase in number, the
percentage distribution over the design period can be determined using this calculator. For each type of
load (aircraft) the number of loads during the first year in the design period must be entered. If forecasts
are available these may be entered, or a general growth rate may be input.
In the example above the number of aircraft have been entered for year 1 as well as forecasts for year 10
and 20. After pressing Calculate the percentage distribution and the average number of loads per year is
calculated. The calculated data will be transferred to the design window by pressing Transfer.

A simpler example is shown below, where you enter traffic in ESALs for first year and then applies a
general growth rate during the design period:

AADT to ESA
Overlay Design
New Design

If the design standards require you use ESALs, you can use this tool. Otherwise you may consider
defining each load separately in the parameter setup to avoid load conversions.
This facility is available from the design windows, and can be used to convert traffic data given for
different traffic categories into yearly ESALs. The facility is only available if you have defined exactly 1
vehicle with 1 wheel in the parameter setup with the load entered as the equivalent standard load.

In the example you see traffic numbers entered for 4 different vehicle categories:

It is assumed that the standard load is defined as a 10 T axle, and the traffic is given for 10 T, 5 T, 3 T and
th
1 T axle loads. The conversion factors need to be calculated first. In the example it is assumed that the 4
power of damaging effect is valid, implying that 10.000 passages by a 1T axle have the same damaging
effect as 1 passage by a 10 T axle. The conversion factors can then be derived from:
4
Cj = (Lj/10 T)

Where
Cj = conversion factor for load j
Lj = axle load in tonnes for load j

Even though it is preferable to enter the individual loads rather than ESALs, these facilities may still be
useful for determining the effects of lateral wander, slope, number of lanes and the growth rate. The
"Growth of Commercial Vehicles" may be indicated by a factor or by a percentage growth. If a percentage
growth rate is chosen, the program will calculate the growth factor, which is the ratio between the mean
annual number of loads and the number of loads in the first year.

If you enter values in the Get from…. input fields, the correction factor will be automatically inserted.
These correction factors are reasonable default values, but may not be applicable for all situations. Any
change to the values in the window will be immediately reflected in the calculated Average number of
ESA per year. When finished click Transfer to move the value to the design window. Be aware that your
input to the calculation will be lost when the window closes.

WIM to ESA
Overlay Design
New Design

If the design standards require that you use ESALs, you can use this tool. Otherwise you may consider
defining each load separately in the parameter setup to avoid load conversions.

This facility is available from the design windows, and can be used to convert traffic data given for
different traffic load categories into yearly ESALs. The facility is only available if you have defined exactly
1 vehicle with 1 wheel in the parameter setup with the load entered as the equivalent standard load.

WIM (Weight In Motion) data comes from traffic counting stations also recording the weight of each axle
passage.

In the example you see traffic numbers entered for 4 different vehicle load types.:

It is assumed that the standard load is defined as a 10 T axle. The standard axle load must be entered in
the input file Standard Axle Load in kN (or lbf). The power of conversion must be entered. In the
th
example it is assumed that the 4 power of damaging effect is valid, implying that 10.000 passages by a
1T axle will have the same damaging effect as 1 passage by a 10 T axle.

Wim period refers to the length of the counting period in days.

Even though it is preferable to enter the individual loads rather than ESALs, these facilities may still be
useful for determining the effects of lateral wander, slope, number of lanes and the growth rate. The
"Growth of Commercial Vehicles" may be indicated by a factor or by a percentage growth. If a percentage
growth rate is chosen, the program will calculate the growth factor, which is the ratio between the mean
annual number of loads and the number of loads in the first year.

If you enter values in the Get from…. input fields, the correction factor will be automatically inserted.
These correction factors are reasonable default values, but may not be applicable for all situations. Any
change of values in the window will be immediately reflected in the calculated Average number of ESA
per year. When finished click Transfer to move the value to the design window. Be aware that your input
to the calculation will be lost when the window closes.

Life Cycle Cost Analysis

This option is only available if a license has been issued by Dynatest.

Select LCCA from the Elmod Main menu to access the LCCA module:

If a database has not been opened yet, or if no E-values have been calculated for the file, then only the
first option is available.

The first option is used only for opening a previously created database in LCCA. When you click this
option you will get a file browser to select which LCCA database you want to open, and LCCA will then
start.

The two other options are used when you want to transfer new Elmod data to an LCCA database. You
can then select whether you want to use a previously created LCCA database, or if you want to create a
new. When you click OK you will, if option 2 is chosen, be asked to browse for the LCCA database to use,
and if option 3 is chosen, you will have to use the browser to select a folder and enter the name of the
new LCCA database.

You will then get the following data transmission window.

If you have selected to create a new database the window will look like this:
In the grid you will see the actual sections you have defined for the file, and the number of layers is filled
in for each section. For each section you want to transfer, you must fill in the white empty "Layer" fields
with material types, if they are not already filled in with information entered in the Structures window. You
can do this by clicking inside the empty field, and a list with the materials will pop up, from which you
select which material type should be inserted.

The list of materials you can select comes from the Elmod parameter setup you are using. You can select
this setup from the combo list box to the left in the window. So remember to select the correct Elmod
parameter setup, before filling in materials. In the example above you see a grid with data filled in.

Remember to click in the column Chosen for the sections that should be included in the LCCA database.
When you fill in layer materials for several sections, you can often make the job a little easier by using the
button Copy first structure. This will copy your entrances for the first line to the other sections.

In the LCCA Parameter Setup combo list box, you have the list of setups currently available in the LCCA
database. When you are creating a new LCCA database the list will be from the template database. From
the LCCA module you have the possibility of creating new template parameter setups, which can then be
selected when creating new LCCA databases.

Below the combo list box in the Copy to new name field you can enter a name for a new parameter setup
to be created. If you leave it blank then Elmod will write the new information to the setup chosen in the
combo list box, overwriting any existing information, otherwise if you enter a name, then Elmod will create
a new setup, based on the selected LCCA setup, and the information transferred from Elmod.

In the field Total traffic per year you must enter the total number of axle passages for the loads defined
in the Elmod parameter setup. When transferred the number will be recalculated to passages per day for
each load type.

Finally in the Transfer frame, you can select which information to transfer. Section data contain the
section values including the representative elastic moduli, calculated from the current backcalculation
results. Checking Parameter setup will let Elmod transfer material properties and seasonal information to
the LCCA parameter setup. The load information from the Elmod parameter file will always be used for
calculation of traffic per load type. For this reason it is necessary always to select the correct Elmod
parameter setup, even if you have chosen not to transfer parameter data.
If you have selected to transfer data to an existing LCCA database, then the data transfer window will look
slightly different:

As default the transfer of Parameter setup has been unchecked, because usually you would already have
created a valid LCCA parameter setup you want to use. Below the Elmod sections you will now see a new
grid with the sections created earlier in the LCCA database.

You now have the possibility of either creating new sections or to transfer section data to existing
sections. If you check Write to existing sections you must following define the link between the Elmod
sections and the destination LCCA sections as listed in the grid at the bottom. You do this by filling in
SecID values in the column Link to. When you click a field in this column you can select the SecID
number from a list.

If you have checked Keep existing materials then Elmod will automatically fill in the layer materials from
the information in the LCCA database, otherwise you have to add them yourself.

You can also define whether you want to Keep existing traffic information. If you choose to transfer new
traffic information, then notice that the load type definition in the selected Elmod parameter setup, must
match the Traffic types defined in the LCCA database.

Below is an example of information filled in for writing to an existing LCCA database. In this case only
information for the first 3 sections is transferred, and only new elastic moduli will be transferred.
When linking to existing sections then the pavement information will be added to the existing information,
in case the year of FWD testing is different from the year of pavement structure data in the LCCA
database.

PCN Calculations
PCN/ACN Method

This option is only available if a license has been issued by Dynatest.

PCN calculations can be run, whenever backcalculations have been carried out for the data file. Results
from the PCN calculations can be viewed and compared to ACN values using the Plot features.

LCN (Load Classification numbers) will be calculated at the same time as the PCN values, and can be
selected for viewing in Plot under Graph Options when displaying PCN graphs. For more details about
LCN see LCN Method

Similar to the Elmod Design procedure, you have to select which Parameter Setup to use for the PCN
calculation. Two standard setups are supplied with the Elmod package:

SI_PCN-Flex for flexible pavements (US_PCN-Flex for US Customary units)

SI_PCN-Rigid for PCC pavements (US_PCN-Rigid for US Customary units)

(Notice: In case of PCC pavements, you must have indicated PCC layer number in the Structures
window ).

The two standard setups are constructed to meet the ICAO definitions and guidelines in the best possible
way. The criteria are also in line with the default setups in the previous WinPCN software. You are able to
run PCN calculations with any parameter setup, but be aware that this can lead to results that are far
from being in accordance with the ICAO guidelines, and may not be compared to standard ACN
values or other PCN reporting.

The major issues in the standard setup’s are:

- SI_PCN-Flex (US_PCN-Flex) is setup to use only subgrade stress as criteria.
- SI_PCN-Rigid (US_PCN-Rigid) is setup to use only stress at bottom of PCC as criteria.
- Load configurations are not used with the PCN calculations
- Seasonal setup and temperature sensitivity can be modified according to the specific conditions and
materials.

For viewing and modifying Parameter Setup’s see Parameters

Select the appropriate Parameter Setup from the dropdown list box.

Below you have a grid with the sections for which backcalculation have been done:

Right click in the "Layer" fields to select materials for the pavement structure and click in the "Chosen"
field to include sections for PCN calculations.

Enter values for Unrestricted number of loads, which refers to the total number of passages over the
design period. Notice this is different from the design features, where you enter traffic per year.

ICAO operates with the term "Unrestricted usage" which relates to the actual loading on the airfield. This
should not be confused with the 10000 operations used for determining the ACN of an aircraft. Pavements
deteriorate gradually under the effects of loading and climate. Both the size of the loads and the number
of load repetitions are important for the rate of deterioration. The PCN of a given pavement structure will,
therefore, depend not only on the pavement structure itself, but also on the expected number of load
repetitions. If "unrestricted operations" corresponds to a large number of load repetitions, the PCN will be
lower than if it corresponds to a more limited number of repetitions.

Subgrade categories gives access to modify the relation between CBR values and subgrade elastic
modulus. This is required to report the subgrade category according to ICAO’s definition.
This is a table which simply translates the CBR limit values into E-values. The default values are based on
the relationship:

(Powell et al., 1984)

Pass to Coverage ratio can be entered if the traffic relates to the total number of passes of a design
aircraft on the runway. Pass to coverage factors are given for most aircrafts listed in the aircraft database,
see Edit Aircraft Database

PCN/LCN upper limit is used to restrict the calculated PCN/LCN to a maximum value. Depending of the
design traffic entered and the strength of the pavement the calculated PCN values can get infinitely high,
and it has usually no relevance to report PCN values in ranges that are higher than the ACN range, which
has a maximum of about 140. If the field is left blank then no restriction is used. When plotting the values
with the Plot features, it is also possible from there to set a maximum reporting value from Graph
Options.

You can now press Ok to start the calculations. PCN calculations that were carried out previously will not
be overwritten. Use Plot features to view results, and to compare with ACN values for specific aircraft
types.

Plot
Graph Settings
Graph Options
Raw Data
Backcalculation
Design
PCN
Sectioning
Tools

Whenever a file is open, you can select Plot from the main menu, to get access to various graphs and
report options.

The Plot function brings up a screen with the Plot Main menu:

If no design has yet been performed for this file, the Design option will be disabled. Similar for the
Backcalculation and PCN options.

Graph Options and Sectioning will be enabled when a graph has been selected for viewing. The same
is the case for some of the submenu items in Tools.

Follow the links above for detailed information of the menu options.
General
When a graph type has been selected by choosing one of the submenu items to Raw Data,
Backcalculation, Design or PCN a graph will appear, based on the most recent calculation:

Graphs will usually only display data for 1 section at a time. To access data for other sections click Graph
Options in the Plot Main menu and select the appropriate section from the section grid. Alternatively you
can use Next and Prev. buttons to scroll through multiple sections or multiple designs or multiple PCN
calculations.

For most of the graph types you have statistics available in the lower right corner of the window. N is the
number of points included. The two next columns are the mean value and the standard deviation (factor)
either based on a normal distribution or a lognormal distribution. The latter is used for elastic moduli and
residual life.
When the graph opens the complete section is included in the statistics, but you can select to see the
statistics for any part of the section by "clicking and dragging" from a start point to and end point on the
graph. If you only want information for one point you can just click the point.

Click Save to save the statistics to the database. Click Report to create a printable page with the results.
You will be asked if you want to print all saved statistics or only the present one.

You can copy and paste both the table data and the statistics to other Windows applications like Excel or
Word. To do this click the upper left corner in the table to select the whole table, and then press <Ctrl> +
C to copy to the Clipboard. In the example above you will click in the field saying "Station" for the table
data, or the field "Layer" for the statistics. You can also limit the range to individual columns by clicking the
column header or to individual cells by marking the range of cells. To mark multiple cells or columns, hold
down the left mouse button and drag the mouse to define the area. In any case stationing and column
headers will automatically be included when copying.

Graph Settings
Plot

At the top of the Plot window you have a series of icons:

Clicking one of these icons opens a window to modify the layout, axis settings and texts on the graphics:

For further help on each tab please use the built-in help by clicking the Help button.
Notice that the changes you do are only temporary. Next time you open a graph the defaults will be used.

Graph Options
Plot

This refers to the menu item Graph options in the Plot Main menu, and the item is enabled whenever a
graph has been selected for viewing.

A window will open, and the options available in the window will depend on which graph type is currently
being displayed. The options dialog makes it possibel to change various settings for the graph, and also to
select among multiple sections or drops if present. Following are the Option dialog shown for 4 different
graph types:
Graph Type
You can select X-Y or Line. This refers to how points are plotted along the X-axis. In X-Y graph the points
are plotted according to the X-value (Station), while in Line graph the points are evenly distributed over
the axis. If points are not sorted according to stationing, the X-Y graph option is disabled. When you
display graphs based on measurements where the Station identification does not correspond to distances
(like Plate ID numbers) you should select the Line option. If you wish to use the sectioning facility Graph
Type must be set to X-Y.

Data sorted by chainage

If the test points during the measurements have not been measured with increasing stationing order from
beginning to the end of the file, you have the option of displaying the data either sorted or not sorted. This
could be the case if a section has been measured in both directions, one at a time, and stored in the same
file. If not checked, you can only have Graph Type = Line.

Accumulated difference
Please see Accumulated Difference

Show all sections

This means that for a file subdivided into sections, you can check this and see the graphics for the
complete file. Otherwise you will only have one section at a time.

Include only
This option is available for some graphs showing results for PCC including joint measurements. You can
select to show all points, or only points at centers or at joints.
Sections, Designs
If multiple sections are defined, you can select from the grid which section you want to have displayed.
You have the same possibility for a range of graphs types, when different drop No. has been used in the
analysis. For Design graphs you can select among all designs saved to the database. To select a section
or Design No., you have to click in the "Check" column on the appropriate line.

Deflections
This option is available in the graphs "Deflections" and "Surface Moduli". If you don’t want to have all
geophones presented on the graph, you can uncheck the appropriate checkboxes.

Drop No.
If multiple drops have been performed during testing, you can select which one to use for the graph.

Normalize to
Checking this option will normalize the deflections to the specific load level you enter. This is based on a
linear extrapolation.

Layers
For graphs presenting results for multiple layers (like the "Elastic Moduli" graph"), you can select which
layers to include in the graphics.

View Moduli as
As default the graph shows the results of the backcalculation based on the measured deflection bowls. If
you want to have the moduli reported at reference conditions you can select this option. It is only available
if pavement materials have been entered before the backcalculation, see Structures

For the "Pavement Response" graph you have additional options:

Selection
Here you can select whether the graph should show the results for multiple seasons based on 1 design
load, or multiple loads based on 1 season. Depending on your setting, you can then define which design
load or which season to use in the rightmost range of option buttons.

Interface
This refers to the interface number in the pavement structure for which you would like to have responses
reported. 1/2 indicates bottom of layer 1 and top of layer 2 and so on. In this case the reported values will
be bottom of layer one for transverse and longitudinal responses and top of layer two for vertical
responses.

Response Type
Check which type of response you want to have displayed in the graphics. If you create a report based on
the graph, you will have all responses included automatically.

ACN Options
Select the aircraft type for which you want ACN values from the dropdown list box. The plot data field will
automatically show max weight value. Change this if you wish to have ACN’s for another weight. You can
only enter weights between min and max weight.

For Allowable tire pressure you can select one of the categories. It is usually an input from the airport
authority, and is used only for reporting with the results. Any choice will be saved as a new default value,
for new calculations.

Usually ACN is reported according to one of the 4 subgrade categories, but you can have ACN calculated
as a function of the subgrade modulus, by linear interpolation between the subgrade category limits.
Normally PCN is reported with a maximum of 120, but you can enter your own upper limit by changing the
value in the PCN/LCN upper limit data field. The value will be saved as a new default value.

If you want a graph of LCN values instead of PCN/ACN then check the Plot LCN values field.

Accumulated Difference
Plot
Graph Options

Accumulated difference is a graph option for most of the possible graph types you have in Plot .

When you check the checkbox in Graph Options, the graph will convert to show the accumulated
difference for each curve in the graph.

For each point the accumulated difference will be calculated as the sum of the previous point’s value
minus the mean value for the section. It can be expressed as shown below, where the accumulated
difference for point No. N is:

The boundaries of the uniform sections occur at the locations where there is a significant change in the
slope of the accumulated difference. An example is shown below using sample data file for flexible
pavements:

The method is based on the AASHTO cumulative difference approach (See Appendix J of the AASHTO
Design Guide).
Raw Data
Plot
Graph Options

This refers to the Plot Main menu item Raw Data

Selecting this menu item gives access to a collection of graphs, which are available before any
calculations have been done with Elmod.

Deflections
This graph will display a curve for each active geophone with the deflection values. Additionally the
applied load level can be seen, for which the values refers to the right-hand Y-axis. The load is expressed
as the stress level under the loading plate when using SI units. For US units the load is given in lbf.
If the file is divided into subsections, you will as default only have values for one section. Select Graph
Options from the menu to change to another section or to view values for the complete file. You also
have the possibility of selecting another drop No.

Surface Moduli
This graph will display a curve for each active geophone with the calculated surface moduli. You will have
the same options as for the Deflection graph.

Notice that statistics are based on a lognormal distribution.

A more convenient way to see the surface moduli is to view the graphs point by point, where surface
moduli is plotted vs. geophone distance. These graphs you can get by clicking in the graph area or in the
table. This will show the point closest to your indication:
From here you can scroll through all points using the navigation buttons below the graph.

Step one point back or forward and stay there.

Scroll automatically backwards or forwards through all points.

Scroll fast automatically backwards or forwards through all points.

Go to first point or last point.

Fixed x–scale is used to indicate that the maximum value on the x-axis must remain the same for all
points. This makes it easier to compare from point to point. When the x-axis is not fixed, the maximum x-
axis value will be scaled for each point.

Include only abnormal basins when scrolling through the points, will have the result that only graphs for
points which have unexpected shapes will be included. This can typically be used to detect difficulties for
the backcalculation, and can have various reasons (deep stiff layers, non-elastic characteristics of upper
materials, present of cracks, poor friction between bound layers, defect geophones or geophones out of
calibration). It may happen that basins that look normal still turns up because the test for abnormality is
sensitive to small changes in the slopes between geophones.

Clicks points to remove is a facility to disable specific geophones to be used in the analysis. It will only
apply to the single point. The geophones closest to the center cannot be disabled. The changes will not
be saved unless you click the Save button. Before saving you can Undo changes which refreshes the
graph to the state when you entered. Cancel will disable the possibility to save any changes. If you have
saved changes, you will see disabled geophones marked in the table with pink background color. If you
later want to enable the geophones again, you can select the Eo plot again and press the Reset button. If
you want to disable the same geophone for all points in the file you should use the Edit Geophone Setup
option
For details about how surface moduli are calculated see Surface Moduli
Voids
This option is mainly useful when analyzing results from measurements on concrete slabs. If the test
points have been measured with several drops using various drop heights, it is possible to detect possible
voids under the plates. The graph displays the part of the deflection which is due to voids. Since the
values also are influenced by non-linear properties of the materials, this has to be considered when
evaluating the results. A general rule says that if the values are higher than about 50 microns, it is likely
due to influence of voids.
The values are calculated as the Y-intercept for a linear regression line in a deflection vs. load plot.

In this example you see clear indications for voids for the measurements at the joints, for the last part of
the section.
Statistics are not available for this graph type.

Thicknesses
The graph showing the input thicknesses is most useful when operating with GPR Data or other point
related thickness data. The graph will for each FWD test point indicate the thicknesses that will be used in
the backcalculation. From the graph it is easily seen if there are various number of layers over the section.
It should then be considered to divide into subsections with uniform number of layers. This is a
requirement if the backcalculated E-values are going to be used for design. The following example shows
a section where point related data has been entered:
You can get this layout by selecting an "Area" graph.
The thicknesses are displayed from the top of the pavement, assuming the surface has constant altitude.
This might of course not be the truth.
If only one out of multiple subsections is displayed, you can change this in Graph Options.

Backcalculation
Plot
Graph Options

This topic refers to the Plot Main menu item Backcalculation

Selecting this menu item gives access to a collection of graphs, which are available when
backcalculations have been done with Elmod. Some of the graphs refer only to jointed pavements.

E-Moduli
This graph displays the calculated moduli as a curve for each layer, including the subgrade. You can only
view results from one section and one drop at a time. To view other sections select Graph Options from
the Menu or use the Prev. and Next buttons below the graph. When opened for the first time the graph
display results for the most recently calculated section.
Notice that statistics are based on a lognormal distribution.

Non-linearity
This graph displays the nonlinear constants C and N for the subgrade. The curve for the exponent N (the
"degree" of non-linearity) refers to the secondary Y-axis. Please see Non-linearity for more information.
Notice that statistics for C-values are based on a lognormal distribution.

Depth to bedrock
This option is only enabled if the data field Max depth to stiff layer in the Structures window has been
filled with a value before the backcalculation. For each test point where a depth to bedrock has been
found within the maximum depth, the calculated equivalent depth will be marked. For test points where no
influence have been found, the value will be indicated as "0". See Equivalent Thickness for more
information
Statistics are not available for this graph type.

Deflection comparisons
This graph is useful for getting an overview of the "Goodness of fit" for the backcalculation. For each test
point the difference between each geophone reading and the recalculated deflection based on the
backcalculated E-values is displayed with a line. In Graph Options from the menu you can select
between absolute difference and percentage difference. If the "Deflection Basin Fit" method has been
used for the evaluation, you will also see a line for the RMS (Root Mean Square) value based on the
deflection differences. The RMS values refer to the secondary Y-axis.
Following is an example from the sample data file:

Joint Conditions → Load Transfer

This option is available if backcalculation has been carried out on jointed pavements as outlined in
Concrete Data Considerations
Two curves are displayed, one for Differential Deflection and one for Load Transfer. Differential Deflection
refers to the primary Y-axis, and Load Transfer to the secondary Y-axis.
Both values reflect the difference in deflection between the two geophones positioned at each side of the
joint. Differential deflection is calculated as the difference normalized to an American Standard Axle (18
kips), while the Load Transfer is calculated as the deflection for the outer geophone, divided by the mean
value of the two deflections and expressed in percentage. This means that if the load transfer is 100 %,
then the 2 plates will each carry 50 % of the load.

Joint Conditions → k - values

This option is available if backcalculation has been carried out on jointed pavements as outlined in
Concrete Data Considerations

As default 3 curves are displayed. One for the k – value at the joint, one for the k – value at the center of
the plate, and one for the ratio between the values which refers to the secondary Y-axis.

In Graph Options it is possible to deselect curves for better viewing. The ratio curve is useful for
comparing joint conditions with slab center conditions, and can often be used as an indicator of possible
slab curling.

The k – value (the modulus of subgrade reaction) is calculated according to Westergaard’s equations for
PCC slabs on a Winkler foundations.
Notice that statistics are based on a lognormal distribution.

Joint Conditions → Moduli

This option is available if backcalculation has been carried out on jointed pavements as outlined in
Concrete Data Considerations

This Graph will display the moduli at the joints for the slab and the subgrade. The modulus of the slab is
assumed to be the same as for the slab center. Since the calculations are based on Westergaard, no
layers between the slab and the subgrade can be considered.
Notice that statistics are based on a lognormal distribution.

Design
Plot
Graph Options

This topic refers to the Plot Main menu item Design

Selecting this menu item gives access to a collection of graphs, which are available when designs have
been done with Elmod.

Contrary to what is the case for backcalculations, all designs will be saved, and results can be viewed with
this option.

Seasonal Moduli
First phase in the design procedure is the calculation of the design moduli or seasonal moduli according
to the seasonal parameters set up in the parameter file. These values can be viewed in this option.
The Graph will display a curve for each season for one selected layer. The layer number to be viewed can
be chosen in Graph Options. as well as results from earlier designs.
The statistics show the results at the reference temperature for bound layers, and at the reference
condition for unbound layers

Pavement Response
Second phase in the design procedure is the calculation of stresses and strains under the design loads for
each season. These values can be viewed with this option.
In Graph Options from the menu you can select response type and interface. See Graph Options for
more details. You have the possibility of seeing either the curves for multiple loads or for multiple
seasons.

Life and Overlay

Third phase in the design procedure is the calculation of the residual life and the overlay requirements
according to the design life setup in the parameter file.
The residual life is presented for the most critical layer and damage type. To view the remaining life for all
layers see the graph for Critical Layer or select Print Report from the Tools menu.
Notice that you can skip the calculation of residual life for a layer, if you in the parameter setup for
materials deselect both fatigue cracking and permanent deformation as distress mode.
This graph will display one curve for residual life referring to the secondary Y-axis, and one curve for
required overlay referring to the primary Y-axis. If the residual life is more than 40 years the value will be
set to 40 years. Theoretically you can have infinite residual life times calculated according to the criterias
and calculated responses and the traffic, but it has no practical meaning to report this. If you for some
reason would like to see the values, you can select Print Report from the Tools menu.

Notice that statistics for Life is based on a lognormal distribution.

Critical Layer
This graph presents a curve of residual lifetimes for each layer. This gives the possibility to compare the
layers, and see which layer is the most critical.
Statistics is not available for this graph type.

PCN
Plot
Graph Options
PCN/ACN Method
LCN Method

This refers to the Plot Main menu item PCN

First time you open the PCN graph after a PCN calculation you will only see the PCN graph itself. To add
an ACN graph you have to click Graph Options and select the appropriate aircraft type. For details about
how to select ACN values and other settings, see Graph Options
Notice that for measurements on slabs that include joint testing, the PCN is not calculated for the joints
because ACN/PCN is only defined at the center of PCC slabs. Use the backcalculation results to evaluate
the joint conditions.

In the result grid you will see the reporting code on the form R/B/W/T.
st
1 letter refers to pavement type:
F for flexible
R for Rigid (using Winkler foundation and k-values for subgrade category)
nd
2 letter refers to subgrade category:
A: High
B: Medium
C: Low
D: Very low
 rd
3 letter refers to tire pressure category (as entered in Graph Options)
W: High
X: Medium
Y: Low
Z: Very low
th
4 letter refers to type of evaluation:
T: Technical (default for evaluation with FWD and Elmod)
U: By experience of Aircrafts using the facility.

Sectioning
Plot

This topic refers to the Plot Main menu item Sectioning.

This sectioning feature can be used in two ways. Either to view and save subsection statistics or to
permanently divide the section into uniform subsections with respect to pavement structure or/and layer
properties.
The Sectioning feature is available when a graph is selected, and data are sorted by chainage.
Click Sectioning and Start and you will have a moveable subdivider line to the left of the primary Y-axis:

The yellow box at the top of the line indicates the station of the first point on the right hand side of the line.
You can drag the line to the desired position by placing the mouse curser inside the yellow box, press the
left mouse button and move the mouse. While you are dragging, the box indicates which point that will be
the start point of the subsection created to the right of the line.
You can continue this procedure if you want to create more subsections. Each time a new subsection is
created the window with statistics will reflect the changes, and have a new TAB with information about the
new subsection. Following is an example with 5 subsections created, and statistics for each subsection:
The statistics can now be save or printed with or without earlier saved statistics for other graphs using the
Save and Report buttons.
If you do not want to save the subdivisions permanently you can just click Sectioning and Stop and the
structure of the section with all the results will stay unchanged.
If you want to save the new subsections as individual sections, similar to the procedure used in
Structures you can select Sectioning and Save new sections. This will have a major impact on the
database, since all results will be reorganized to reflect the new sectioning. The PLOT window will close
automatically.

Tools
Plot

This topic refers to the Plot Main menu item Tools

Tools provides a range of options for exporting and printing results. Some options may be disabled for
some graph types.
Most of the options in the menu refer to the currently displayed graph.

Export to Excel
This opens a window with settings to the export:

Check the data types you want to have exported and click Export.
All selected data for the current section will now be transferred to Excel, which will automatically open and
organize the data into the sheets Raw data, Backcalculation and Life and Overlay.
Notice if you only need data relating to one graph, you can copy the data from the table below the graph,
see Plot for details. It may be necessary to first open the application to which you want to paste the data
from the Clipboard

Create Report
This will create a report in Microsoft Word based on the data for the current graph. The report may in
some case include more data than listed with the graph. The report will automatically open Word, from
where it can be printed.
Following is an example of a report of calculated E-moduli:
To print the report click the printer symbol. To export data click the envelope symbol, which will bring up a
window with choices for format of the file.

Print Graph
This will immediately send the currently displayed graph to the default printer. The graph will be fitted to
the printed page.

Graph to Clipboard
This will copy the current Graph to the Clipboard, from where it can be pasted into other applications like
Microsoft Word or Excel. It may be necessary to first open the application to which you want to paste the
data from the Clipboard

Report parameters
This gives the possibility to create a report based on the parameter files used for the current project. A
Word document will open with values used for Seasons, Materials and Loads.

Seasons
Parameters

For design purposes, the year may be divided into seasons, with a maximum of 12 seasons. The program
calculates moduli corresponding to each season, with respect to seasonal temperature for the asphalt
layer and the time elapsed since spring thaw or wet period for the unbound materials (including the
subgrade).
The remaining life and associated overlay design is then based on the cumulative amount of damage
occurring during each season for all design loads using Miner's law.
The user subdivides the year into seasons. For each season the length in weeks is entered, as well as the
week to be considered characteristic of that season. The week number is counted from the beginning of
the year. The characteristic week is the week in which all the loads of the season would result in the same
damage as when the loads are spread out over the season. By subdividing a season Elmod 6 may be
used, through a trial and error process, to accurately determine which week is characteristic. Total
number of weeks must add up to 52.
st
The first season will start 1 of January, which means that you would need to divide a season into two
seasons if it overlaps this date.

If loads are evenly distributed over the year you can click the Even distribution button and Elmod will
calculate the accurate percentage values to put into the load per season fields.

When you modify seasonal temperatures or factors on the Material Types tab, you have graphics
available to see the effect of the settings for each season. From this graphics you also have the possibility
of adjusting the characteristic week inside each season.

Material Types
Parameters
Temperature Sensitivity
Design Criterias

When selecting the Material Types tab you get the possibility to see and edit all information in the setup
that relates to each of the defined materials.
In the list box to the left you have the current list of materials, and the information on the tab relates to the
highlighted item in the list. You can click in the list to view the other material’s properties.

Create new material button adds a new material to the list. You will be asked to assign a name for the
material (must be unique for the setup), and the new material will be added with properties which are a
copy of the currently selected material. You can proceed by changing any values for the new material.
Nothing will be saved before you click one of the save buttons on the right hand side, but all changes will
be stored in memory even if you select other material types to view and edit.

Delete current material removes the material from the list..

Material Name replicates the highlighted name from the list box. It is not possible to change the name
due to reference conflicts. Alternatively you can create a new material with the desired name, and delete
the existing material.

Material Type can be selected from the dropdown box. The information is used in the design procedures
to determine the number of bound layers.

Ref. Modulus. The reference modulus of the material is used as the design value in case the material is
chosen as the design material. The value is also used to optionally be inserted as seed value in
Structures window.

Temperature dependant. This defines whether the material modulus is sensitive to temperatures. If it is
checked the tab Temperature dependency will be enabled for definition of the modulus/ temperature
relationship. The title of the last tab will be Seasonal temperatures for defining temperature variation
over the year. If "Temperature dependant" is not checked the last tab will show Seasonal factors for
defining modulus variation over the seasons due to other reasons, like frost/thaw periods or rainy seasons
affecting the condition of the unbound materials.

Fatigue Cracking and Permanent Deformation

This defines which criteria should be used for determining the residual life of the material. The details for
selecting the criteria are on the related tabs below. Usually you operate with fatigue cracking for bound
materials and permanent deformation for unbound materials. It is possible to have them both switched off,
in case it is not desirable to have these materials properties influencing the design calculations.

Hatch. Select a pattern to be associated with this material. Click on the pattern to open a window for the
possible selections.

Min. and Max Modulus

This input can be used to force the backcalculation to produce moduli results within the given range. Be
aware that this will influence the moduli calculated for the remaining layers as well.

For details on the settings on the tabs see:

Fatigue Cracking and Permanent Deformation

Temperature Dependency
Seasonal Temperatures
Seasonal Factors

Fatigue Cracking and Permanent Deformation

Material Types

These two tabs are enabled depending on the choice for using Fatigue Cracking relations or Permanent
Deformation relations in the checkboxes above the tabs.

In the relation N is the traffic and E the modulus of the material. The constants A, B, C and Eo can
following be modified in the fields below the equation

If the constants reflects a known criteria, then name of the criteria is shown below the constants. If desired
another criteria can be selected from this dropdown box, which is prefilled with various criterias developed
and used in various regions. The list of items in the dropdown box depends on the choice for using fatigue
relations or permanent deformations and the choice for using strain or stress dependency.

It is possible (but not necessary) to assign a name to a user specified criteria. Enter the desired constants
in the input fields, and the name will change to Unknown criteria. Write the name of the criteria in the
drop down field and press Add to list.

The list is global, and does not belong to the specific parameter setup. This means changes in the
standard list of criterias will be accessible from all projects and parameter setups. Changing the contents
of the lists will not affect any design results or parameter setups. Only the values in the white input fields
are saved with the setup.

On the Fatigue Cracking tab you can select if Longitudinal or Transversal responses should be used in
the criteria. The longitudinal responses are the most critical.
Define the materials Design life. The design life together with the yearly amount of traffic determines the
number of axle passages in the criteria functions. Notice that the required design life with respect to
different types of deterioration (as well as for different types of materials) may be different.

By clicking View the relation will be shown graphically:

For equations that are moduli dependant, the graph uses as default the Ref. Modulus as the value for E
in the relation. If you want to see the graph for other values, you can change E in the upper left corner and
press Graph to refresh.

If you uncheck the Clear check box, it is possible to compare the relation with other criterias, by closing
the graph window, select another criteria and open graph window again. In the example below is shown
the procedure used twice to compare 3 different criterias graphically.
If you don’t want logarithmic Y-axis, you can uncheck the Log y check box.

Temperature Dependency
Material Types
Temperature Sensitivity

This tab is enabled if the checkbox Temperature dependent is checked.

Temperature dependency is usually used with bitumen bound materials, and allows the asphalt modulus
to be corrected to a reference temperature. This is important to do before any design because the
temperature usually vary a lot over the day and over the year, and because the stiffness of the bitumen is
very sensitive to the temperature.

The design procedure first converts the asphalt modulus from asphalt temperature recorded during testing
to the modulus at the reference temperature as entered in the Ref. Temperature input field. There next
this modulus at reference temperature is used to calculate the seasonal moduli, according to the
temperature given for each season, see Seasonal Temperatures .

On this tab you can choose between 3 methods for temperature correction:

Factors
This will refer to a table created, which assigns a factor to each temperature in the temperature range. In
the dropdown box you can select which table should apply to the current material. To create and edit
these tables see Edit Temperature Table .

Semi–logarithmic
If the dependency can be fitted to a logarithmic relation, you have possibility to define the equation:

Notice that the range for which the equations is valid must be positive. When temperatures exceed the
range Elmod will use the from or to value to calculate the correction factor.

Exponential
If the dependency can be fitted to an exponential relation, you have possibility to define the equation:

When temperatures exceed the range Elmod will use the from or to value to calculate the correction
factor.

View
Clicking the button brings up a window with a graphic presentation of the temperature-modulus correction.
In the example the graph for all 3 options are shown. This is possible by unchecking Clear, close graph
rd
window, select another option and click View again. This is then repeated to also show 3 option.

Seasonal Temperatures
Material Types

This tab is available if the material has been defined to be temperature dependant in the checkbox above
the tabs.

The purpose is to define each season’s design temperature. The seasonal design temperature should be
fitted to a value that will result in a damage representing the average over the season.

The seasonal temperatures can either be set manually or calculated by a sinusoidal

relationship between time over year and asphalt temperature.

Manual
Selecting this option will give an input field for each defined season. In this case 3 seasons have been
defined in Seasons
If you click View you will have a graphic presentation of the seasons and chosen temperatures:

Sinusoidal
You have to enter three values to define the relationship in terms of the yearly minimum and maximum
temperature, and in which week No. the maximum temperature occurs.

The temperatures will then be calculated according to

where

T is the temperature in week No. W

T1 is the maximum temperature during the year
T2 is the minimum temperature during the year
st
W is the week No. (counted from January 1 ), and
W 0 is the week No. corresponding to maximum temperature.

Click View to see graphics for the calculated temperature variation over the year:
The characteristic week for each season is indicated at the top, and the resulting seasonal temperatures
are shown the vertical lines.

You have possibility to change the charactestic week by dragging the indicator at the top towards left or
right. The effect to the calculated seasonal temperature is displayed immediately by the vertical line
moving up or down. The changes will also be reflected on the tab Seasons

Seasonal Factors
Material Types

This tab is enabled if the checkbox Temperature dependent is unchecked.

For materials that are not temperature sensitive the Elmod 6 user has the opportunity to consider the
(probable) springtime or wet season weakening of any, or all, of the unbound layers in the pavement
section.

The purpose of the facility is then to determine the factor used for each season. A factor of 1 will be the
reference condition., and the seasonal modulus of each season is then calculated as the seasonal factor
multiplied with the modulus at the reference condition. The modulus at reference condition is calculated
first as the backcalculated modulus divided by the factor belonging to the week where the FWD
measurement took place. Interpolation between the factors entered for the seasons and the characteristic
weeks will determine the factor belonging to the week of FWD testing.
You can choose between 4 methods for entering the factors.

None
This will indicate the modulus is constant over the year, and the backcalculated moduli are used as design
value for all seasons.

Factors

You will have an input field for each season to manually enter the factor.

Sinusoidal

You have to enter three values to define the relationship in terms of the yearly minimum and maximum
factor, and in which week No. the maximum factor occurs.

The factors will be calculated according to:

where
Exponential

You have to enter three values to define the relationship in terms of the yearly minimum factor, in which
week No. the minimum factor occurs, and the exponent (must be negative).

If the pavement is subject to frost exposure, the moduli of unbound materials will decrease sharply during
the spring thaw, depending of the frost susceptibility of the materials. The moduli will then slowly
regenerate. This modulus variation may be approximated by an exponential relationship:

E = [ 1 - (1- R) x exp(A x W)] x Emax

where
E is the modulus W weeks after the minimum modulus
R is the ratio between minimum and maximum modulus
A is a constant
Emax is maximum modulus at complete recovery.

The ratio, the exponential constant, and the week of minimum modulus can be entered in the data fields.

View
Click the button to graphically view the effect of your settings. This is an example of the effect of
calculating factors according to a sinusoidal relationship:
The indicators on the top show the characteristic week.

You have possibility to change the charactestic week by dragging the indicator at the top towards left or
right. The effect to the calculated seasonal temperature is displayed immediately by the vertical line
moving up or down.

Be aware of that if you also use a sinusoidal relationship to determine asphalt temperatures for the
seasons, the same characteristic week No. will be used. In this case it might be an alternative to set the
asphalt temperatures manually.

Vehicles and Wheels

Parameters

The design traffic and load characteristics is defined on the tabs Vehicles and Wheels.

In the design procedures Elmod will include all defined vehicles, and for each vehicle Elmod will include
the wheels specified on the Vehicle tab. It is possible to define a traffic mix of up to 24 vehicle types
including a total of up 100 wheel types.

Often ESALs is used to characterize the design traffic. In this case only 1 vehicle including 1 standard
wheel is going to be used. However, it is preferable to use the best approximation to the real traffic mix of
loads, as the derivation of the equivalent standard loads depends on the design criteria (transfer
functions), and may be different for different layers.
In the list to the left, you will see which vehicles that are included. You can add a new vehicle by clicking
Create new vehicle. If the option Copy current is chosen you will be asked for a name of the new
vehicle, and this will be created with a copy of the data for the current one. Creating a new vehicle this
way will not influence the choices for selecting wheels. The list of available wheels will be the same, and if
you need to attach a wheel which is not in the list of wheels on the Wheel tab, you will have to shift to this
tab and add a new wheel.

If you define a new aircraft, you have the possibility to retrieve all necessary information from the aircraft
database. Select the option Aircraft list and click Create new vehicle. This opens a window for selecting
the aircraft. Select one and click OK. Elmod will now load the information required in both the Vehicle tab
and the Wheel tab.

Click Delete current vehicle to remove this from the list of vehicles. No changes will happen to the list of
wheels on the Wheel tab.

The example above shows input value for operating with an ESAL as design load.

Pass to coverage or normal distribution defines which method should be used to reduce the traffic
numbers due to the lateral distribution of loads. For roads the reduction is usually not considered,
meaning all wheels loads are in the same line. In this case select Pass to coverage and on the Wheels
tab, set the factors to 1.

For evaluation of Airport pavements the lateral distribution of the aircraft loads will be important. This may
be considered by entering Pass to coverage ratios for rigid and flexible pavements, or by using a normal
distribution of the aircraft loads. In this case the imported wheel base is used, whereas the user must
provide a value for the standard deviation of the lateral position of the particular aircraft at the facility to
be evaluated. A default value is provided by the program, but the actual lateral wander will vary
considerably, for example between a runway and a taxiway. When a normal distribution is used for the
overlay design, the stresses and strains will be calculated at one line, given by the offset from the
centerline (input in the Overlay Design screen) with the wheel at 10 points of the normal distribution,
each with a probability of occurrence relating to normal distribution of the traffic. The damage at the offset
line is summed using Miner’s law.

Vehicle name. This replicates the highlighted name from the list box. It is not possible to change the
name due to reference conflicts. Alternatively you can create a new vehicle with the desired name, and
delete the existing vehicle.

Total weight is the complete mass to be distributed over the axles and wheels. In case of ESAL’s it
should be the weight corresponding to the definition of the ESAL.

Dynamic Ratio defines a factor to be multiplied on the load. A factor of 1.2 means the load will be
increased by 20 % for design purposes.

Standard deviation relates to the lateral normal distribution of traffic as described above.

Wheel name. Here you enter each axle or wheel belonging to the vehicle, and for which responses
should be calculated in the design procedure. Clicking in the input fields brings up the selection of defined
wheels from the Wheels tab. This means the selection of wheels must be defined on the Wheels tab
before you can connect them to a vehicle.

% of load indicates how big part of the total weight applies to the wheel (or set of wheels). Usually you
have a set of wheels on the axles symmetrical positioned to the center line of the vehicle. The percentage
of load should then apply to the total load of both sets of wheels. On the Wheels tab you can see exactly
how many wheels are included in the definition. Load on each tire if further calculated to the right of the
input field, to avoid any misunderstandings.

In the list to the left you can see the defined wheels. Only wheels that are referred to from the Vehicle tab
are included in the design.
Select the wheel in the list for which you want to see or to edit properties. You can create a new wheel as
a copy of the current by clicking Create new wheel. You will be asked for a name of the wheel, which
must be unique for the setup. Click Delete current wheel to remove a wheel from list. You cannot delete
a wheel if it is referred to from the Vehicle tab.

Wheel name. This replicates the highlighted name from the list box. It is not possible to change the name
due to reference conflicts. Alternatively you can create a new wheel with the desired name, and delete the
existing wheel.

Pass to coverage. Defines the factors to be used for respectively flexible and rigid pavements. For
aircraft setups where data from the aircraft database are included, it is recommended to use the normal
distribution option (see above). If you use the Pass to coverage values in this case be aware of that
these only apply to the critical gear, for the other gears which may be defined the fields should be left
blank.

Centerline wheel. Check the field if there is no symmetrical positioned wheel (like the nose wheel for an
aircraft). You will see on the graphics exactly how it is defined.

Wheel Configuration. You can select between 6 configurations of tires making up the complete wheel or
gear. The graphics will show the outline of each.

Tire pressure. Enter the tire pressure for the wheel.

Wheel distance. Enter the distance between the centerlines of dual wheel configurations.

Axle distance. Enter the distance between the axles for tandem and tridem wheel configurations.

Tread. This defines the distance between the wheel sets positioned symmetrically to the center line. This
is used together with Offset to calculate the distance between the each line of traffic from the normal
distribution and to the line for which responses are calculated. Se more about using Normal distribution
above and the offset value entered in the Overlay Design window.

Below is shown input example for a setup with a fleet mix of four aircrafts:
Concrete Data Considerations
While Elmod 6 can analyze PCC data files there are special considerations that must be used during data
collection.

For jointed rigid pavements:

• Each slab to be evaluated must be tested at the center. In addition it may be tested at the joints
and corners for the purposes of load transfer evaluation or void detection.

• The slab must be identified by an integer number (e.g. 257). If the station is used for identification,
it must be input as an integer (i.e. Station 25.4 must be input as 254). The two digits following the
decimal point are used to identify the location of the test on the slab in accordance with the figure
below. Generally, at location XX, the slab center is identified by XX.1, while joints are identified by
XX.2Y and corners by XX.3Y, where Y can be used to identify which joint or corner is being
referred to as shown below. Note that Y must be even for transverse joints and odd for
longitudinal joints. If FWDWin is used with mdb data formats, it is no longer necessary to
identify the location by the station decimal. Instead the graphical facility in the FWDWin software
can be used to identify the slab location, and the information will be stored separately in the
database. Elmod 6 will be able to read this. You must however be aware of the fact, that Elmod
relates the information from slab center to slab joints. This means that when Elmod calculates for
joints it will search for the closest center test, and assume this to be at the same slab. The reason
is that Elmod uses the E-value calculated from the slab center test in the evaluation of the k-value
at the joints. You have the possibility to check whether Elmod relates slab centers and joints
properly, see Jointed Pavements

• When testing at a joint (or corner) the geophones at distances 8 in. (200 mm) and 12 in.(300 mm)
from the load center must be placed on either side of the joint, as shown below:
 • Climatic conditions will influence the load response of a concrete slab. Both temperature and
temperature (or moisture) gradients are important. For the purpose of evaluating PCC modulus
and foundation support (using the modulus of subgrade reaction) at the center and at joints, the
slab should, preferably, not be warped when FWD testing is carried out.
• When testing during periods with significant daily temperature variations, warping may be difficult to
avoid. In this case modulus of the PCC, as well as foundation support, should be determined from
night or early morning measurements at the slab center, while good contact between the slab
center and foundation may be expected. During night and early morning, joint condition will be at
its poorest. A more realistic measure of the foundation support at the joints can, however, be
obtained later during the day. Comparison of foundation support for tests conducted at both times
can be used for void detection applications.

Theory of Elasticity
Since the early sixties the so-called Analytical-Empirical, or Mechanistic-Empirical method has
been gaining popularity amongst pavement engineers. This method uses fundamental physical
properties and a theoretical model to predict the stresses, strains and deflections, i.e. the
pavement response, caused by a load on the pavement, see Figure 1. If the basic assumptions
with respect to materials and boundary conditions are correct, this method is valid anywhere
and may be used to correctly predict the response for any combination of loads, climatic effects
and materials.

Figure 1 Pavement response (stress, strain and deflection).

 Figure 2 Pavement performance (cracking, rutting and roughness).

In the second part of the method, use is still made of empirical relations, but these relations are
based on the pavement response. Cracking of asphalt is predicted from the maximum tensile
strain in the asphalt layer and permanent deformation is determined from the maximum
compressive strain in the unbound materials, for example. This is illustrated in Figure 2.
Presently much effort is put into reducing the empirical part of the method. Permanent
deformation, for example, may be calculated directly if the stress-permanent strain relationships
for the materials are known. In some methods, like the Distinct Element Method (DEM), elastic
and plastic displacements are calculated in the same process.

This section will briefly describe some of the essential elements of the theory of elasticity, which
is the most widely used model for pavement design. In most cases it is assumed that the
loading is static, that deformations are continuous (compatibility between normal and shear
strains) and that all materials are homogeneous, isotropic and linear elastic.

Elastic parameters

Subjecting a cube to a uniform stress in the direction z (vertical), σz, as shown in Figure 3, will
result in a change in length of the sides. When the deformations are very small compared to the
dimensions of the cube the strains will be equal to the relative change of length.

Figure 3 Elastic parameters for uniaxial stress.

If the material is linear elastic, then the ratio of vertical stress, σz, to vertical strain, εz, will be a constant,
the coefficient of elasticity or Young’s modulus (E). This is known as Hooke’s law. The ratio between the
horizontal strain (εx or εy) and the vertical strain is also a constant. The ratio will be negative, and the
positive value is known as Poisson’s ratio (ν). A stress in one direction is seen to produce a strain
proportional to -ν/E in a perpendicular direction.

A shear stress, τxz = τzx, will produce a shear strain, γxz = γzx. If the material is isotropic, the constant shear
modulus, G, will be:

The shear strain does not affect the normal strains.

For three-dimensional loading of an isotropic material Hooke’s law may be written by the above
equation for G and by:
with rotation of the subscripts. For the general anisotropic case E, ν, and G should take the same
subscript as the corresponding stress, and strain for ν.

Because few pavement materials are ideally elastic it is often useful to divide the stress into a hydrostatic
and a deviatoric component.

I1(σ) is the first stress invariant (independent of the orientation of the co-ordinate system), p is the mean
normal stress or the hydrostatic pressure, and {s} is the deviator stress tensor. The mean deviator stress
is derived from the second deviator stress invariant:

where τoct is the octahedral shear stress.

The strains used with p and q are the volumetric and deviator strain:

where γoct is the octahedral shear strain.

The ratio of hydrostatic stress to volumetric strain is called the bulk modulus, K = p/εv. The relationships
between K, G, E and ν are:
Stresses and Strains in a continuum

The theory behind the calculation of stresses and strains in a continuum will be illustrated using the plane
stress case (two dimensions). Much of the following is based on van Cauwelaert (1989).

For isotropic, linear elastic materials, the theory is based on three assumptions:

1. Equilibrium
2. Compatibility, and
3. Hooke’s law

The assumption of equilibrium leads to the following three equations (for plane stress):

Written in this format, it is also assumed that the load is static or that the materials are
weightless. For dynamic loads the inertial forces should be included.

The assumptions of continuity (or compatibility between normal and shear strains) leads to:

where ϕ is the shear (angular) strain.

Finally Hooke’s law for plane stress may be written as:

[For plane strain Hooke’s law is:

with z perpendicular to the xy-plane

The 3D Poisson’s ratio and modulus may be found from

To solve equations (2.1) to (2.7) a stress function φ is introduced. Stresses expressed as the
following differentials will satisfy the conditions of equilibrium, equation (2.1) to (2.3):

By replacing the stresses in equations (2.5) to (2.7) by their differentials and then substituting
the strains in equation (2.4) by the expressions (2.5) to (2.7), the following equation is obtained:

The solution to this equation is determined by the boundary conditions.

For the special case of a point load P and length a = 0 the integrals have a closed form solution:
For a semi-infinite half space with axial symmetric loading the solution becomes more complex.
For most boundary conditions the solution of the differential equation requires numerical
integration. The differential equation is solved for stresses, Hooke’s law is then used to
calculate the strains, and displacements are determined from an integration of strains.

For the special case of a point load P and for the centre line of a circular load that is
either uniformly distributed or transmitted through a perfectly rigid plate, closed form solutions
exist. These are given in the next section.

Boussinesq’s equations

Equations for calculating stresses, strains and displacements in a homogeneous, isotropic, linear elastic
semi-infinite space, with modulus E and Poisson’s ratio ν, loaded by a point load, P, perpendicular to the
surface, were given by Boussinesq in 1885. Below, some of Boussinesq’s equations for a point load are
given in polar co-ordinates as shown in Figure 4.

Figure 4 Notation in polar co-ordinates used with Boussinesq’s equations

It is interesting to notice that the vertical stress and the major principal stress are independent of
the elastic parameters. The modulus does not influence any of the stresses.

At the centerline of the load, the equations for vertical stress, strain and displacement reduce to:
These equations reveal an important difference between the variation with depth of stress and
strain, on one hand, and displacement on the other. Where the displacement is inversely
proportional to the depth, stress and strain are inversely proportional to the square of the depth.

One of the consequences of this difference is that surface deflection on a layered system is
poorly correlated to the stresses and strains in the individual layers. Surface deflection is,
therefore, a poor substitute for stresses and strains. As pavement deterioration is related to the
stresses or strains in the individual layers, it is unfortunate that surface deflection is still widely
used for evaluation of bearing capacity.

The deflection at the surface of the half-space is:

or inversely proportional to the distance from the load.

Figure 5 Comparison of uniformly distributed load to point load

Boussinesq’s equations for a point load are quite useful. The distance from the load does not
need to be very large before a point load produces almost the same response as a load
distributed over an area. In Figure 5 the surface deflection on a semi-infinite half space
calculated using ELSYM5 is compared to the deflection from equation (3.3). Modulus and
Poisson’s ratio were 100 MPa and 0.35, respectively. The load was 10 kN uniformly distributed
over a circular area with radius 150 mm. Already at a distance of just one diameter, the
deflection produced by a point load is quite close to the correct deflection
.
Closed form solutions exist for the centerline of a circular load that is either uniformly distributed
(3.4) or transmitted through a perfectly rigid plate (3.7). For the uniformly distributed load some
equations are given below:
The horizontal strain may also be calculated using Hooke’s law:

If the load is transferred via a completely rigid circular plate, so that the surface deflection, at all
points of the plate, is the same, then the theoretical stress distribution on an elastic medium will
be:
where σo is the mean value of the stress, a is the plate radius and r is the distance from the
center to the point where the surface stress is determined. For this loading condition a closed
form solution is available for the centerline:

The horizontal stresses and strains may be calculated using Hooke’s law.

Surface Moduli
On the subgrade the FWD may be used as a dynamic plate loading test and the modulus of the subgrade
may be calculated directly from Boussinesq’s equations, taking the stress distribution under the loading
plate into consideration and estimating Poisson’s ratio.

The theoretical stress distribution under a completely rigid plate on an elastic medium has infinitely large
values at the perimeter. This, of course, cannot occur in real pavement materials. On cohesive materials,
such as clay, the actual stress distribution will resemble the theoretical distribution, but for granular
materials the stress at the perimeter will be small, because the bearing capacity at the surface of an
unconfined granular material is small.

Figure 1 Typical stress distributions on granular and cohesive soils during plate loading tests.
The modulus of a semi-infinite half space may be calculated from:

where do is the deflection at the center of the circular load and f is a factor that depends on the
stress distribution:

Stress distribution F
Uniform 2
Rigid plate π/2
Parabolic, granular 8/3
Parabolic, cohesive 4/3

2
If neither stress distribution nor Poisson’s ratio is known, then the factor f(1-ν ) may range from 1 to 2.67.
This very large uncertainty can be reduced by measuring the deflections at different distances from the
load. As shown in Theory of Elasticity the deflections caused by a point load are very close to the
deflections under a circular load, for distances of more than two radii from the center of the load. The
modulus may, therefore, be calculated from:

where do(r) is the deflection at distance r from the load center.

By measuring the deflections at different distances from the load center it is also possible to check the
assumption that the measurements are done on a semi-infinite, linear elastic half space. If that is the
case, then the moduli calculated at all distances should be the same. If it is not the case, then the
deflections may be used in an inverse analysis (back calculation) to determine moduli or layer
thicknesses. Some of the variation in surface moduli may, however, be due to Non-linearity, gradually
changing moduli with depth or a rigid layer (bed rock).

The moduli calculated from the equations above are called "Surface moduli" because they are calculated
from the surface deflections.

If a rigid layer is present, then the approximate depth to the layer may be found by plotting the
deflection against the inverse of the distance from the center of the load.

The surface deflection from a point load on a semi-infinite half-space varies inversely
proportional to the distance from the load:

This is the same equation as equation (2) above. The deflection at a depth equal to the distance
from the load is almost equal to the deflection at the surface. Within 45° there may be a slight
compression or extension, depending on Poisson’s ratio, but it will be small compared to the
compression of the material below 45° (0 with Poisson’s ratio equal to 0.396447, compression
with Poisson’s ratio less than this value, otherwise extension).
 Figure 2 Deflection on a semi-infinite half-space

The deflection at the surface is, therefore, representative of the material below the 45° line. If
this material is rigid, the deflection will approach 0 (it will not be precisely 0 because introducing
a rigid layer changes the boundary conditions).

Figure 3. Deflection versus inverse of distance from load center.

Figure 3 shows an example for a three-layer structure with a rigid layer at a depth of 2450 mm
and the following characteristics:

Layer Thickness Modulus

1 150 mm 3000 MPa
2 300 mm 300 MPa
3 2000 mm 50 MPa
Poisson’s ratio was 0.35 for all layers. The system was loaded with a 50 kN load uniformly
distributed over a circular area with radius 150 mm, and the deflections were calculated using
ELSYM5, at distances 300 mm, 600 mm, 900 mm, 1200 mm, 1500 mm, and 1800 mm.
-1
The tangent to the plotted values intersects the abscissa at 1/r = 0.000384 mm , or r = 2603 mm, which is
6% larger than the true depth and 10% less than the Equivalent Thickness .

If there is a rigid layer at some depth, this should be considered in the backcalculation, but if the
variation of deflection with distance is due to other reasons, i.e. if there is no rigid layer, then
treating the variation as a result of non-linearity is preferable to introducing an artificial rigid
layer.

Equivalent Thickness
In 1943 Burmister presented a solution to the differential equations of the Theory of elasticity, for the
boundary conditions of a two-layer system. Since then a large number of computer programs have been
developed for calculating stresses, strains and displacements in layered elastic systems. Some of the
more well known are the programs developed by Shell (Bistro and Bisar), by Chevron (Elsym5),
Waterways Experiment Station (WES5), the Laboratoire Central des Ponts et Chaussées (Alize III) and
the Commonwealth Scientific and Industrial Research Organisation (Circly). The latter will even accept
cross-anisotropic materials.

All of these programs are based on linear elastic materials. Most pavement materials, however, are non-
linear elastic. To accommodate the non-linear characteristics, a number of finite element programs were
developed in the sixties. Modern finite element programs will accept a number of material models,
incorporating elastic, plastic and viscous strains. Some programs can handle discontinuities and dynamic
loads. The drawback is that many parameters may be needed to describe the complex stress-strain
relationships of the different materials, parameters that may be very difficult to determine in practice. A
considerable computer capacity is also required.

The finite element method still assumes the strains to be compatible, except for special elements that may
allow for discontinuities. It is doubtful whether this can be used to correctly describe particulate materials,
where the stresses are substituted by contact forces between the particles and where most of the
deformation derives from sliding and rolling of particles. To describe particulate materials Cundall (1978)
developed the Distinct (or Discrete) element method. This method puts an even harder strain on
computers, but allows for the description of elastic and plastic deformations in the same process.

Even when a model, eventually, has been developed that can correctly predict the stresses, elastic as
well as plastic strains, and failure in pavement materials, there will still be a need for a simplified model for
Pavement Management Systems (PMS). In PMS the deterioration of a large number of road sections,
over 20 years or more, with a large number of different maintenance strategies, needs to be calculated.
With the analytical-empirical approach, stresses and strains must be calculated millions of times, and this
is only possible with a very fast algorithm. This is what Odemark’s method (1949) provides.

Odemark’s method is based on the assumption that the stresses and strains below a layer depend on the
stiffness of that layer only. If the thickness, modulus and Poisson’s ratio of a layer are changed, but the
stiffness remains unchanged, the stresses and strains below the layer should also remain (relatively)
unchanged. The stiffness of a layer is proportional to:

where h is the thickness of the layer.

The transformation shown in Figure 1 should, therefore, not influence the stresses or strains in layer 2
provided that:
where he is known as the "equivalent" thickness (the method is also called the "method of
equivalent thicknesses" or MET).

Figure 1 Odemark’s transformation of a layered system

The transformed system in Figure 1 is a semi-infinite half space on which Boussinesq’s

equations may be used, but only for stresses, strains and displacements below the interface.

Odemark’s assumption is not mathematically correct. Whether the response of real pavements is closer to
that predicted using the theory of elasticity or using Odemark’s method remains to be seen. If it is
desirable to obtain results close to the theory of elasticity, when using Odemark’s method, a correction
factor, f, may be introduced. If, in addition, Poisson’s ratio is assumed to be the same for all layers (and in
practice Poisson’s ratio is seldom known with any degree of accuracy and may be larger than 0.5 for
some layers), the transformation may be written as:

Reasonably good agreement with the theory of elasticity is obtained with a correction factor of 0.8, except
for the first interface where a factor of 0.9 is used for a two-layer system and 1.0 for a multi-layer system.
0.3
If the thickness of layer one, h1, is less than the radius of the loaded area, a, then a factor of 1.1(a/h1)
will bring the horizontal tensile strain at the bottom of layer one, closer to that obtained with the theory of
elasticity.

If measured values of stresses and strains, in real pavements, are available, these should be used to
"calibrate" Odemark’s method, rather than values from the theory of elasticity.

For a multi-layer system the equivalent thickness of the upper n -1 layers with respect to the modulus of
layer n, may be calculated from:
where the latter version is a recursive equation well suited for programming purposes. Layers below layer
n are assumed to have the modulus En in the transformed system.

With no friction at the interfaces the equivalent thickness may be found from:

Deflections are calculated as the sum of the compression of the layers plus the deflection of the
subgrade. The compression of an individual layer is found as the difference between the
deflection at the top and the bottom of the layer in the transformed system. For the top layer the
transformed system is a half space with modulus E1.

With the correction factors given above Odemark’s method will give answers reasonably close to the
theory of elasticity provided that

• moduli are decreasing with depth (Ei /Ei +1 > 2), and
• the equivalent thickness of each layer is larger than the radius of the loaded
area.

If this is not the case (or if a better agreement is desired) the correction factors should be
changed and possibly expressed as functions of thicknesses and moduli.

Example of the Use of Odemark’s Method

A three layer system (subgrade is counted as a "layer") with the thicknesses and moduli shown
in Figure 2 is loaded by a uniformly distributed load with radius 150 mm and contact stress 0.7
MPa. Poisson’s ratio is assumed to be 0.35 for all materials.

Figure 2 Three-layer system used to illustrate Odemark’s method.

Numbers given in parenthesis were calculated using ELSYM5

1) Horizontal strain at the bottom of the asphalt:

The values determined using Odemark’s method range from 89% to 92% of the values from the
theory of elasticity (ELSYM5). Moduli can hardly be determined with this accuracy.

2) Vertical stress on subgrade

3) Surface deflection
Cundall, P.A. "BALL - A Program to Model Granular Media using the Distinct Element Method", Dames &
Moore, Advanced Technology Group, Technical Note No. TN-LN-13, 1978.

Odemark, N. "Undersökning av elasticitetegenskaperna hos olika jordarter samt teori för beräkning av
belägningar eligt elasticitetsteorin", Statens Väginstitut, meddelande 77, 1949.

Non-linearity
In North American literature two types of non-linearity are normally considered. For granular materials the
modulus is assumed to be a function of the bulk stress:

where MR is the "resilient modulus", i.e. the deviator stress (or dynamic stress) divided by the reversible
strain. θ includes static and dynamic stresses, and k2 is a positive value typically in the range 0.4 to 0.6.

For cohesive materials the non-linear relation is often written as:

where k2 is negative, i.e. the modulus is decreasing with increasing deviator stress.

The two effects may be combined to the following equation:

where pa is atmospheric pressure, and

τoct is the octahedral shear stress (see Theory of Elasticity)
It would be a reasonable assumption that the "k3-part" of this equation should be a function of the ratio (or
the difference) between actual sheer stress and permissible shear stress. As the shear stress approaches
the permissible shear stress, the modulus would be expected to decrease. If this is the case, the modulus
will depend not only on the stress level but also on the position within the structure.

Dynamic deflection testing, e.g. with a Falling Weight Deflectometer (FWD), often shows a variation of
deflections with distance that could be due to a non-linearity similar to that given by equation (2), or the k3
part of equation (3). .

Finite element calculations (Ullidtz, 1987) with a non-linearity of the form:

have shown that the non-linearity has very little influence on the stress distribution in a semi-infinite half
space. In equation (4) σ1 is the major principle stress from the external loading, i.e. excluding any static
stresses due to the weight of the material, and pa is a reference stress, often taken equal to atmospheric
pressure (0.1 MPa). The purpose of the reference stress is to avoid having units raised to a power
different from 1. C and n are constants (corresponding to k1 and k2). In the finite element calculations, n
was varied from 0 (linear elastic material) to -0.5.

That a gradual variation of the modulus has little influence on the stress distribution may not be that
surprising, considering that the modulus has no influence on the stress distribution in a linear elastic half
space.

It was also found that the strains and displacements in the non-linear elastic half-space could be
calculated using Boussinesq’s equations, with the modulus substituted by a non-linear function of the
major principal stress. For the center line, under a point load P, this gives:

The strain is calculated from Boussinesq’s equation by substituting E with Ez, i.e. the modulus at depth z
corresponding to σ1, and the deflection by substituting E by (1-2n)Ez. The latter is a kind of "surface
modulus", i.e. the linear elastic modulus that all the material below depth z should have to give the same
deflection at depth z. In other words, a plate loading test on the surface of a material with the modulus
described by equation (4) would give the surface modulus:

where σo is the uniformly distributed stress under the plate.

With n equal to -0.5 the surface modulus would be twice the modulus of the material immediately under
the plate.
To use Odemark’s transformation with a non-linear subgrade, the Equivalent Thickness of the pavement
layers, he,m, is calculated from:

where
f is the adjustment factor, often equal to 1 with a non-linear subgrade,
Em-1 is the modulus of the layer above the subgrade,
hm-1 is the thickness of this layer, and
he,m-1 is the equivalent thickness of the layers above this layer, without any
adjustment factors (f).

A large number of backcalculation programs have been developed, most of which are based on linear
elastic theory for the forward calculation. If the subgrade is non-linear elastic (or if the apparent modulus
varies with the distance from the load for other reasons) then using a linear elastic approach may result in
incorrect layer moduli. A typical outcome is that the moduli of the top layer and of the subgrade are
overestimated and that the modulus of an intermediate layer is underestimated. This is illustrated in the
following example.

The following pavement structure was assumed:

Layer No. Thickness Modulus

1 150 mm 4000 MPa
2 300 mm 300 MPa

The modulus of the subgrade was assumed to be non-linear, varying with the major principal stress, σ1:

All layers were assumed to have a Poisson’s ratio of 0.35.

First a Finite Element (FE) program (the FEM program in FEM/LET/MET) was used to calculate the
deflections at different distances for a load of 56.4 kN evenly distributed over a circular area with radius
150 mm. Then an inverse analysis was carried out using a program having the Waterways Experiment
Stations five layer linear elastic program (WES) as the forward calculation routine (the LET program in
FEM/LET/MET) ). One inverse analysis attempted to minimize the root mean square (RMS) of the
percentage difference in deflections (WES%) and another minimized the RMS of the absolute differences
(WESmy).

The backcalculated layer moduli are shown in the table below, given in MPa. The modulus given for the
FE calculation is at the top of the subgrade, at the center of the load. The "surface modulus" at that
position would be twice the value shown:

Layer FE WES% WESmy

1 4000 7348 6034
2 300 63 85
3 80 451 373

Minimizing the absolute deflection differences results in slightly better values than minimizing the
percentage difference, but the agreement with the moduli used for generating the deflections is poor. The
results shown are quite typical for what happens when the subgrade is non-linear, but is treated as if it
was linear. The modulus of the subgrade is overestimated and that of the intermediate granular material is
underestimated.

The deflection basins are compared in Figure 1 below, which shows the original deflection basin as
calculated with the FE program, and the two deflection basins calculated with the moduli derived from the
inverse analysis (backcalculation):

Figure 1 Comparison of deflection basins.

The "fit" between "measured" and calculated deflection is not very good. For the percentage minimization
the RMS is 8.0 %, and for the absolute differences 4.6 µ (micron). The differences are mostly due to the
outer deflections, as may be seen in the plot of the surface moduli, calculated using Boussinesq’s
equations for surface deflection on a semi-infinite half space.

Figure 2 Comparison of surface moduli.

If the modulus of the subgrade is a function of the major principal stress, e.g. like equation (4), or of the
deviator stress then the power n may be found from FWD tests done at different load levels. If, however,
the modulus is a function of the ratio of dynamic deviator stress to dynamic bulk stress, then both stresses
will change with the load level, so that changing the load level will have little effect on the displacements.

If the modulus of the subgrade is of the format given in equation (4) then the power n may also be found
from:
where Eo(r ) is the surface modulus at a distance r larger than the Equivalent Thickness of the pavement
layers with respect to the subgrade modulus. When the non-linearity is determined from equation (8) a
variation with position within the structure, caused for example by the modulus being sensitive to the ratio
of shear stress to permissible shear stress, is also captured. Using tests at different drop heights would
not capture this effect.

Assuming that the modulus of the subgrade varies with the major principal stress and calculating the
power, n, from surface deflections at different distances, using equation (8), appears to work quite well,
even when the subgrade only has an apparent non-linearity. The approach has been used on three
instrumented test pavements in the southern part of Sweden (Wahlman & Stahle, 1995). Stresses and
strains calculated from FWD tests were found to agree reasonably well with measured values.

One reason for the apparent non-linearity is the dynamic effect of the FWD. The impulse load produced
by the FWD travels through the pavement layers at different speeds for compression waves and shear
waves. This means that at any point in time only a part of the material is influenced by the load. This is
illustrated in Figure 3 where the relative force at different depths has been calculated, using a
compression wave. An impulse load with a duration of 30 msec (sine shaped) was assumed. The pulse
was assumed to travel at a speed of 300 m/sec, and the relative force was calculated at a distance of
1000 mm from the center of the impact. The relative force was calculated for delays of 15, 30 and 45
msec.

Figure 3 Relative force at different depths for impulse load.

To look into the problem that only part of the material is influenced by the impulse load at any point in
time, a backanalysis program was developed. The method used is simplified and assumes that the effect
of the impulse travels at the speed of the compression wave (in the different materials), but does not
consider any inertial effect. For small increments of time and at distances corresponding to the geophone
distances, the vertical strains are calculated (using Boussinesq’s equations and Odemark’s
transformation, corresponding to the "delayed" force) and summed to give the surface deflection as a
function of time.
To evaluate the validity of this simplified dynamic analysis, the program was used on some examples
provided by Mallela & George (1994). For three different pavement sections (two rigid and one flexible)
dynamic deflections under an impact load had been calculated using the three-dimensional Finite Element
program ABAQUS on a Cray X-MP/16 computer.

Deflections on the same structure were calculated using the simple dynamic program. For most distances
these deflections were slightly larger than the deflections obtained with ABAQUS, however static
deflections calculated with WES5 (and mathematically exact) were also slightly larger than static
deflections obtained with ABAQUS. This indicates that the simple dynamic analysis gives reasonable
results.

Some examples of dynamic backcalculation are shown in the following figures. The FWD measurements
were from three instrumented test sections in southern Sweden. The geophones were spaced at 300 mm,
from 0 mm to 1800 mm.

Figure 4 Measured and calculated deflections (test section I)

Figure 5 Measured and calculated deflections (test section II)

 Figure 6 Measured and calculated deflections (test section III)

The difference between moduli obtained from static and dynamic analysis was not systematic. When
measured stresses and strains were compared to theoretical values it was found that a static analysis,
assuming a non-linear subgrade, gave the best agreement.

An estimate of the static deflections may be found from:

For a linear elastic subgrade this would create an apparent non-linearity with a power of n ≅ -0.2, if
determined from the surface deflections.

For a gradual variation of modulus with depth on the form:

where zo is a constant, the deflection at depth z, dz, may be found from:

A more fundamental description of the non-linearity is often given by expressing the volume strain and
deviator strain (rather than modulus which is itself a function of stress and strain) as functions of the
hydrostatic and the deviator stress, or of the stress invariants:
where the first stress invariant, I1, and the second deviator stress invariant, J2, are as given in
Theory of Elasticity . A notation similar to equation (12) is often used to describe the plastic
strain-stress relationships.

Alternatively the bulk modulus, K, and the shear modulus, G, are given as functions of p and q. An
example of this is the generalized Boyce model (Boyce, 1976), revised by Balay et al. (1997):

where Ka , Ga and n are constants, pa is atmospheric pressure (100 kPa or 0.1 MPa), po is a
static hydrostatic stress, and pr and qr are the dynamic values (corresponding to the resilient
strains). The Boýce model is also a function of positions as well as on stress level.

For materials showing viscous or visco-elastic characteristics, it is necessary to also include the time (or
frequency) in the relationship.

Ullidtz, P. "Pavement Analysis", Developments in Civil Engineering, 19, Elsevier 1987.

Wahlman, T. & Stahle, F. "Öppna Överbyggnader och Dränerande Bärlager", Provväg Borrebackevägen
Malmö, 1991 - 1994, Rapport SBUF projekt 0115, BFR projekt 901035-0, Gatukontoret Malmö, 1995 (in
Swedish).

Mallela, J. & George, K.P. "Three-Dimensional Dynamic Response Model for Rigid Pavements",
Transportation Research Record No. 1448, Transportation Research Board, Washington, D.C. 1994.

Boyce, J.R. "Behaviour of granular materials under repeated loading", PhD thesis, University of
Nottingham, 1976.

Balay, J., Gomes Correia, A., Jouve, P., Hornych, P. & Paute, J-L. "Mechanical Behaviour of Soils and
Unbound Granular Materials, Modelling of Flexible Pavements – Recent Advances", Proceedings of the
th
8 International Conference on Asphalt Pavements, Vol. I, University of Washington, Seattle, Washington,
1997.
Temperature Sensitivity
As is well known in the pavement evaluation field, asphalt-bound materials are "temperature sensitive" in
the sense that the modulus of elasticity, or E-value, of the asphalt-bound layer is dependent on its
temperature.
In Elmod 6, there are two relationships, which matter in this respect:
a. For backcalculation purposes, what was the "weighted average" or representative temperature of
the asphalt layer at the time the FWD test was conducted?
b. For design purposes, how does the temperature of the asphalt change throughout the yearly
design seasons?
Both of the above considerations relate to the determination of the relevant modulus of
the asphalt layer, both at the time the FWD test was conducted and as used in the
design process subsequently.
If the temperature of the asphalt at the time of the FWD test is not known and the intention is to
backcalculate the various moduli in the pavement system, the derived modulus of the asphalt layer is not
particularly useful by itself, since the asphalt modulus is a function of the temperature within the layer at
the time of test. Therefore, to get started with the Elmod 6 backcalculation and overlay design processes,
it is vital that the temperature of the asphalt layer at the time of test is known.
Generally, one of two methods of determining the in situ asphalt temperatures is employed:
1. A hole is drilled, approximately to the middle of the asphalt-bound layer(s), or to one third of the
thickness, a small amount of fluid is poured into the bottom of the bore hole, and the temperature
is measured with a probe after the temperature of the liquid has stabilized (this usually takes 5
minutes or longer). Since this is a time-consuming process, in practice it is generally carried out
every hour or two, with the operator manually entering the same temperature (or his best estimate
of the present temperature) for each succeeding test point, until the next bore hole is drilled and
the asphalt temperature is measured again. There are several weaknesses associated with this
method, including but not limited to: a) the heat generated by the drill bit affects the measured
temperature, or the operator does not wait long enough for the temperature to truly stabilize; b)
too much or too little fluid is poured into the bore hole by the operator, which influences the
temperature of the fluid as subsequently measured; c) the distance to the middle of the asphalt
layer is not known at the time of test, thus the bore hole is either too deep or not deep enough; d)
the presence of shade or sunlight oftentimes shifts from station to station [due to clouds or trees,
etc.], thus the temperature variability from point to point is not revealed; e) the time lapse between
pavement temperature measurements is too great, or the weather changes too rapidly, so the
temperature of the asphalt layer changes appreciably before a new, manual temperature
measurement is made; and f) the accuracy of the temperature probe itself is inadequate. A good
FWD operator, however, may take account of most of these effects and may use the present
surface temperature and the previous mean daily temperature to give an estimate of the asphalt
temperature that is more accurate than any other method.
2. Based on the infrared (IR) temperature which can be measured automatically, in real time, while
the FWD tests are conducted, the average asphalt temperature is determined by the pavement
engineer as he/she runs the Elmod 6 program. Additional input data needed to carry out this
calculation are: 1) the time of day [present in the FWD data file]; 2) the thickness of the asphalt-
bound layer(s) [needed for backcalculation in any event]; and 3) the "average" local air
temperature over the past 24 hour period, i.e. the sum of previous day’s high and night’s low air
temperatures divided by 2. This method is called "BELLS" (from a method developed by Baltzer,
Ertman-Larsen, Lukanen and Stubstad), and it is an algorithm derived from the SHRP database,
from most of the continental United States and Canada, over a three year period of time.
Weaknesses associated with this method include: a) the SHRP database consists entirely of data
obtained between approx. 06:00 hrs and 17:00 hrs, thus the use of BELLS outside of this time
frame may result in some error; b) data from all worldwide latitudes and longitudes is not included
in the SHRP database, thus FWD data obtained elsewhere may not be strictly valid with BELLS;
c) the various asphalt layer thicknesses in the SHRP database is broad but finite, thus data from
very thin [less than approx. 45 mm or 1¾ in.] or very thick [greater than approx. 305 mm or 12 in.]
may not be valid, especially below the bottom end of this range; d) the average pavement
O O
temperature must be greater than 0 C or 32 F when FWD testing is carried out (no frozen
pavements); e) a film of moisture or frost on the pavement directly under the IR temperature
sensor can adversely affect the IR reading; and f) the accuracy of the infrared temperature unit
itself is inadequate or it has been miscalibrated.
Assuming the IR temperature was measured, enter the "previous 24 hour average air temp" on the Elmod
6 Estimating Moduli "Select Sections and Program Control Options" where prompted, if BELLS equation
is to be used. Otherwise, measure the approximate middepth asphalt layer temperature, in the field, and
have the FWD operator enter this temperature manually while testing. In this case, do not enter any
previous 24 hour average air temperature on the Program Control Options.
If an overlay design and/or remaining life calculation is/are to be carried out for an asphalt-surfaced
pavement, the temperature of the asphalt as a function of season also needs to be known. In Elmod 6,
this is accomplished by entering either a modulus ratio vs. temperature "look-up" table or an appropriate
equation, which calculates the asphalt modulus as a function of temperature. Several look-up tables may
be defined and edited through the Edit Temperature Table option.
In the prompts for the Parameters of Elmod 6, please choose one (or more if desired) of the asphalt-
bound materials and then select the type of temperature variation you wish to use. If you choose
"Factors", you will be prompted to select the appropriate look-up table:

If you choose either "Semi-log" or "Exponential", you will be prompted to enter a regression constant (or
slope), plus the maximum and minimum weighted average asphalt layer temperature encountered during
the year under "Local seasonal variations". By "weighted average" is meant the average asphalt
temperature normally encountered during the maximum and minimum temperature weeks. Extreme highs
or lows should not be used, as this will skew the data to the conservative or liberal side, depending on
how extreme the temperatures entered are.
Please enter, as well, the temperature range over which the selected equation is valid. This range of
temperatures should generally be greater than the local climatic range entered under "Local seasonal
variations", because FWD testing may, in fact, have been conducted on an even more extreme day, or
hour, than the normal weekly variations indicate.
When a material is temperature sensitive the yearly variation of temperatures must also be indicated
under Parameters The temperature may vary sinusoidal over the year or a temperature may be indicated
for the characteristic week of each season, in which case the program will interpolate for dates between
characteristic weeks.

Temperature data is available for anywhere in the world from HTTP://Hurricane.ncdc.noaa.gov

Design Criterias
Related topic: Design Criteria Parameters

Elmod 6 is delivered with some default equations for cracking and rutting. The user must change these
functions for remaining life and overlay calculations, to correspond to the materials, environment,
construction practices etc. for which the program is to be used. This section explains how to reformat
transfer functions for use in Elmod 6.

Transfer functions relate traffic loading to pavement deterioration. Wheel loads transfer stresses into the
pavement structure. Materials strains develop which are a function of the stresses applied by the loading
wheel(s) and the modulus of the various layers in the pavement structure.

Two types of pavement deterioration are modeled in Elmod 6- cracking of bound layers and permanent
deformation for unbound layers. Bound layers consist of materials that are composed of aggregates held
together by a binder such as asphalt or Portland cement. Bound materials can develop tensile stresses.
Unbound materials may be granular, composed of aggregates or particles that are compacted and held
together through (primarily) inter-particle friction, or cohesive (clay). Granular materials typically cannot
develop tensile stresses. An example of an unbound material is a granular base or fine-grained subgrade.

The primary distress mode for bound layers is usually fatigue cracking, which is related to the magnitude
of the repeated stresses or strains at the bottom of the pavement layer. Cracking in asphalt pavements
has been found to correlate well with repeated tensile strains. Cracking in Portland cement stabilized
layers has been found to correlate well with repeated tensile stresses.

The primary distress mode for unbound layers is usually through permanent deformation which results in
either rutting or poor ride quality (roughness), or both. Permanent deformation has been found to correlate
well with repeated vertical compressive strains at the top of the unbound layer.

Elmod 6 can be configured to mimic most mechanistic design procedures by incorporating the
appropriate transfer functions into the program. The transfer functions in Elmod 6 take the following form:

where

A, B, and C are constants which depend on the design procedure;

MN = millions of wheel load applications (Mload in the parameter setup);
E = modulus of the material;
Eref = a reference modulus which makes the transfer function unitless.

Note: For strain-based transfer functions the permissible strain is in microstrain, as indicated in the
parameter setup screen.

Most published transfer function can be reconfigured to Elmod 6 format. In this example, the Asphalt
Institute transfer functions for fatigue cracking and permanent deformation will be reformatted for use in
Elmod 6.

Fatigue Cracking (U.S. Customary Units):

The transfer function used in the Asphalt Institute’s MS-1 Pavement Design Manual is as follows:

where:
N = number of wheel load applications to failure due to cracking for asphaltic materials;
C = function of asphalt content, air voids;
= tensile strain at the bottom of the asphalt layer; and
E = modulus of the asphalt (pounds per square inch).

where Vb is the volume percent of bitumen, and

Vv is the volume of voids.

The first step in converting this formula to Elmod format is to simplify it and to include units, as follows:
This relationship tells us that if we have a material with a modulus of 1 psi and the strain was 1, then the
permissible number of loads would be 0.07958*C. A typical asphalt material will not have a modulus of 1
psi, but rather a value at the order of 500 ksi, and it is more handy to express strain as microstrain and the
number of load repetitions in millions (not to have very small or very large numbers). If MN is the number
of loads in millions and µεt is the strain in microstrain, we may rewrite the equation as:

This relationship tells you that a material with a modulus of 500 ksi and a C value of 1 will have a
permissible strain of 231 microstrain at 1 million load repetitions.

If you need the relationship in SI units, 500 ksi ≈ 3450 MPa. If you want to use 3000 MPa as a reference
-.26
modulus, the constant will be 231× (3000/3450) = 240.

This equation is ready to be entered into Elmod 6.

Permanent Deformation (U.S. Customary Units)

The Asphalt Institute equation for permanent deformation is as follows:

where N = Number of wheel load applications to failure in permanent deformation; and

= vertical compressive strain at the top of the unbound layer.
If we again prefer to use strain in microstrain and number of load applications in millions, this may be
rewritten as:

Now the equation must be rearranged to place the strain parameter on the left side:

This equation tells us that the permissible strain is 478 microstrain for 1 million load application,
independent of the modulus of the material.

The same equation can be used with SI units.

Note: For this type of transfer function, where the modulus of the material is not used, the value for C in
the Elmod 6 transfer function should be set to zero. This causes the (E/Eref) term to always equal 1.

Design Criteria Parameters

Design Criterias
Material Types

Below are given examples of criteria parameters developed for various regions, material types and failure
characteristics. Notice that the equations are based on loads in millions, so the constants can be entered
directly as they are into Elmod in parameter setup for materials.
Westergaard's Equations
Jointed rigid pavements offer a special problem because elastic layer theory cannot be used at the joints
or corners, because one of the boundary conditions is that layers should be of infinite horizontal extent. To
solve this problem Westergaard (1925) assumed that a rigid pavement could be considered as a slab on a
Winkler foundation. For a slab on a Winkler foundation the reaction (or vertical stress) is a constant, k,
times the deflection. The constant is called the "modulus of subgrade reaction".

The modulus of subgrade reaction is determined through plate loading tests at a standard plate radius of
15 inches (381 mm), and often to a stress of 10 psi (70 kPa) or to a deflection of 1 mm. The theoretical
relationship between the modulus of subgrade reaction, k, and Young’s modulus, E, is:

where a is the radius of the plate (15 inch) and f depends on the stress distribution under the plate, as
discussed in Surface Moduli . Please notice that the k-value is inversely proportional to the radius of the
loading plate. If measured with a rigid plate with a radius of 15 inch, the modulus of subgrade reaction in
kPa/mm is roughly 1.9 times the modulus in MPa. With a plate radius of 150 mm it would be about 4.8
time the modulus.

If k is derived from deflections measured on the top of the concrete slab, through backcalculation, the load
radius is no longer 15 inch. If the subgrade is an elastic half-space, an approximate value of k may be
found from:

where Em is Young’s modulus of the subgrade and he is the equivalent thickness of the concrete slab with
respect to the subgrade, calculated using Odemark’s transformation.

Elmod calculates the k-value from the measured deflection. This value may be quite different from the k-
value measured with a static plate loading test on the subgrade, as shown by equations 1 and 2 above. If
you want to use the k-value, corresponding to a plate loading test, for design or PCN calculation, you can
enter the ratio between the k-value from equation (1) and from equation (2) under Settings. This ratio
must correspond to the average equivalent thickness of the concrete slab with respect to the subgrade,
he. It is calculated from:

Westergaard’s original equations (first published in Denmark in 1923) have been modified many times by
different authors, partly to bring them into better agreement with elastic layer theory and partly to get a
closer fit to experimental data. The equations presented below were derived by Ioannides, Thompson and
Barenberg (1985) using the ILLI-SLAB finite element program.
Westergaard distinguish between "interior loading", I, defined as the case where the load is at a
"considerable distance from the edges", edge loading, E, and corner loading, C. The maximum tensile
stress at the bottom of a concrete slab with thickness h and modulus E, for interior loading, may be found
from:

The equations for interior loading are reasonably correct when the least slab dimension is more than three
to four times the radius of relative stiffness, l. Edge loading is defined as the case where "the wheel is at
the edge, but at a considerable distance from any corner". For this case the maximum bending stress may
be determined from:

For corner loading ILLI-SLAB was used to derive stress and deflection for a square loaded area with a
side length of c. The bending stress is:

0.32 0.59
occurring 2.41*c *l mm from the corner.
The deflection at the corner is:

The equations for edge loading are for the case of a free edge, i.e. no load transfer to any adjacent slab. If
there is load transfer between slabs, Westergaard (1947) suggests a "load transfer efficiency factor", j,
defined by the equation:

in which zj and z’j are the deflections of the adjacent slabs at any point of the joint, while ze and z’e are the
corresponding deflections that would occur at the same place if the joint had no capacity for load transfer.
Often the amount of load transfer measured, for example, during FWD testing is given as the ratio of the
deflection at the unloaded slab to the deflection at the loaded slab. This measure of load transfer, j’, is
related to Westergaard’s load transfer efficiency factor through:

At a full load transfer, where 50% of the load is transferred to the adjacent slab, j = j’ = 1.
If the load transfer is 50%, ¾ of the load will be on the plate loaded by the FWD and ¼ will be transferred
to the next plate. With Westergaard’s equation the load transfer efficiency factor is ½ or 50%, whereas the
deflection ratio would be 1/3 or 33%, not 50%.

The tensile stresses σj and σ’j at the bottom of the two slabs along the joint may be computed from:

where σe and σ’e are the corresponding stresses that would occur if the joint had no capacity for load
transfer.

Stresses caused by environmental effects are often determined explicitly for concrete pavements. They
may be caused by contraction due to curing or low temperature, by curling (or warping) due to
temperature or moisture gradients or by differential settlements. Slabs may also have built-in curling. If the
environmental and material data are available these semi-static stresses may be calculated, but it is not
certain how they should be combined with the dynamic traffic stresses, in order to predict deterioration of
the rigid pavement. A Goodman diagram (MacLeod, 1979) or Aas-Jakobsen’s equation (Herholdt et al.
1979) may be used.

With the modified Goodman diagram the permissible number of loads, N, may be found from:

where fc is the flexural strength

σd is the dynamic stress, and
σs is the (semi) static stress

Aas-Jakobsen’s equation is:

A decrease in mean temperature (or in moisture content) will cause tensile stresses in the slab,
due to friction between the plate and the foundation support. The mean tensile stress may be
found from:

where f is the coefficient of friction between the plate and the support, γc is the density of the concrete, g is
the acceleration of gravity and L is the length of the plate. The coefficient of friction is often larger than 1.
Frequently f = 1.5 is used, but it may be as high as 2.5.

At the bottom of the slab the tensile stress may be from 1.5 to 2.0 times the mean value.

A temperature gradient will cause tensile stress at the cool side of the slab, if the slab is restrained from
warping. The maximum tensile stress far from any edge will be:

where λ is the coefficient of thermal expansion and dt is the temperature difference between the top and
the bottom of the slab.

If the warping is not restrained it may be found from:

Westergaard, H.M.S, "Stresses in Concrete Pavements Computed by Theoretical Analysis", Proceedings,

Highway Research Board, Part 1, 1925

Ioannides, A.M., Thompson, M.R. & Barenberg, E.J. "Westergaard Solutions Reconsidered",
Transportation Research Record 1043, 1985

Westergaard, H.M.S. "New Formulas for Stresses in Concrete Pavements of Airfields", Proceedings,
American Society of Civil Engineers, Vol. 113, 1947

Macleod, D.R. "Considerations for Maintenance Strategies for Portland Cement Concrete Pavements",
Ph.D. Dissertation, University of California at Berkeley, 1979.

Herholdt, A.D., Justesen, C.F.P., Nepper-Christensen, P., & Nielsen, A. "Beton Bogen",
Cementfabrikkernes tekniske Oplysningskontor, Danmark 1979 (in Danish).

PCN/ACN Method
PCN Calculations
LCN Method
Aircraft Classification Number (ACN)

The Aircraft Classification Number (ACN) is defined by ICAO, using a "mathematically

derived single wheel load to define the landing gear/pavement interaction. This is done by
equating the thickness given by the mathematical model for an aircraft gear to the thickness for
a single wheel at a standard tire pressure of 1.25 Mpa". Boussinesq’s equations are used for
flexible pavements and Westergaard’s solution for a plate on a Winkler foundation for rigid
pavements.

For flexible pavements the thickness is determined from the CBR value, using the equation:

where
t is the thickness in cm,
DSWL is the single wheel load in kg, and
ps is the tire pressure (1.25 MPa).

The ACN is two times the derived single wheel load in 1,000 kg. The ACN is calculated by the aircraft
manufacturer for 4 subgrade categories (A: CBR > 13, B: 8 < CBR < 13, C: 4 < CBR < 8 and D: CBR < 4).
The ACN is specific to a particular aircraft and does not depend on the number of operations (the equation
above is for 10,000 coverages) or on the pavement structure (apart from the subgrade category).

For rigid pavements, a standard stress for reporting purposes is stipulated (σ = 2.75 MPa) only as a
means of ensuring uniform reporting of ACN. The working stress to be used for the design or evaluation of
pavements has no relationship to the standard stress for reporting. For rigid pavements the subgrade
categories depend on the modulus of subgrade reaction (the k value determined using a 750 mm
diameter plate). The four categories are: A: k > 120 kPa/mm, B: 60 kPa/mm < k < 120 kPa/mm, C: 25
kPa/mm < k < 60 kPa/mm and D: k < 25 kPa/mm.

In Elmod an aircraft database is included with ACN values reported for several hundreds aircraft types. In
the PLOT features, any of these aircrafts can easily be selected for graphical comparisons between ACN
values and calculated PCN values.

Pavement Classification Number (PCN)

"Aerodrome Design Manual – Part 3 – Pavements", Second Edition 1983, by the International Civil
Aviation Organization (ICAO) defines the Pavement Classification Number (PCN) as "A number
expressing the bearing strength of a pavement for unrestricted operations".

Pavements deteriorate gradually under the effects of loading and climate. Both the size of the loads and
the number of load repetitions are important for the rate of deterioration. The PCN of a given pavement
structure will, therefore, depend not only on the pavement structure itself, but also on the expected
number of load repetitions. If "unrestricted operations" corresponds to a large number of load repetitions,
the PCN will be lower than if it corresponds to a more limited number of repetitions.

Calculation of the PCN at an FWD test point has three steps. In step one the layer
moduli are derived from the deflection basin, at the conditions of the test. In step two the design
moduli are determined for each season considered in the design, and in step three the single
wheel load that will correspond to the damage criterion for the subgrade or for the PCC, at the
specified number of load repetitions, is derived and converted to PCN in the same way as for
ACN.

For backcalculation procedures and theories please refer to the chapters:

Estimating Moduli
Theory of Elasticity
Equivalent Thickness
Non-linearity

Design Moduli
Second step is to derive design moduli. The E-moduli calculated are based on the pavement condition
during the FWD testing. Both asphalt modulus and moduli of unbound layers will change over the year,
depending on the climatic variations.

The climatic variation in terms of asphalt temperatures and change in condition of unbound layers can be
setup in the parameter file, see Material Types
Elmod will calculate design moduli for asphalt layers and unbound layers for each defined season in the
year, in accordance with the setup chosen.

Damage Criteria
The damage is accumulated for all seasonal conditions using Miner’s law, and the PCN
is determined as two times that single wheel load, in thousand kilograms, that will correspond to
the criterion expressed as:

where E is the modulus.

N is the number of load repetitions, and
A, B, C and Ereference are constants.

For flexible pavements the criterion refers to the stress at the top of the subgrade, while for rigid
pavements the criterion used, is for the tensile stress at the bottom of the concrete slab. For rigid
pavements only tests at slab centers are considered.

The constants used in the criteria have to be entered in the parameter setup, see Material Types

LCN Method
PCN Calculations

The LCN is calculated using the procedure described in "Design and Evaluation of Aircraft Pavements",
PSA 1971. The LCN for a range of loads is given by:

Load in pounds Tire LCN

pressure
in psi
100000 120 100
90000 115 90
80000 110 80
70000 105 70
60000 100 60
50000 95 50
40000 90 40
 30000 85 30
20000 80 20
10000 75 10

According to the guide, the stress caused by a load is proportional to:

0.27
Load * area

where the loaded area is in square inch.

If a parameter x is defined as:

LCN will be approximately related to x through:

2
LCN = 18.518 x +87.011 x – 5.2032

The relationship is shown in the figure below: