Greenhouse Gas Emissions Model (GEM) v4.0 User Guide Vehicle Simulation Tool for Compliance with the Greenhouse Gas Emissions Standards and Fuel Efficiency Standards for Medium and Heavy-Duty Engines and Vehicles Phase 2 Technical Amendments


Greenhouse Gas Emissions Model (GEM)
v4-0 User Guide

Vehicle Simulation Tool for Compliance with
the Greenhouse Gas Emissions Standards
and Fuel Efficiency Standards for Medium
and Heavy-Duty Engines and Vehicles -
Phase 2 Technical Amendments



£%	United States

Environmental Protect
Agency

-------
Greenhouse Gas Emissions Model (GEM)

v4-0 User Guide

Vehicle Simulation Tool for Compliance with
the Greenhouse Gas Emissions Standards
and Fuel Efficiency Standards for Medium
and Heavy-Duty Engines and Vehicles -
Phase 2 Technical Amendments

Assessment and Standards Division
Office of Transportation and Air Quality
U.S. Environmental Protection Agency

United States
Environmental Protection
^1	Agency

EPA-420-B-22-024
July 2022

-------
Table of Contents

I.	Introduction	2

II.	Installation	2

II.A. Installation Instructions	2

II.B.	Contents of Installation Package	4

II.B.l. Sample Input Files	4

II.B.2.	GEM Executable	5

III.	Model Description	5

III.A.	GEM Architecture	5

III.A.1.	Drive Cycles and Cycle Average Engine Fuel Map	7

III.B.	Vehicle Parameters for Each Regulatory Subcategory	7

III.B.l. Tractor Vehicle Parameters	7

III.B.2.	Vocational Vehicle Parameters	10

IILB.3. Trailer Vehicle Parameters	13

IV.	GEM Input Files	15

IV.A.	Tractor Input Files	15

IV.B. Vocational Input Files	21

IV.C. Trailer Input Files	27

IV.D.	Supplemental Input Files	28

IV.D.1.	Engine Input File for Tractor and Vocational Vehicles	28

IV.D.2. Transmission Input File for Tractor and Vocational Vehicles	31

IV.D.3. Optional Powertrain Input File for Tractor and Vocational Vehicles	33

IV.D.4. Optional Axle Input File for Tractor and Vocational Vehicles	34

V.	GEM Output File Structure	35

V.A.	Standard GEM Outputs for Compliance	35

VI.	Running GEM	36

VI.A.	Preparing for GEM Simulation	36

VI.B. Running GEM from the Start Menu Icons	36

VI.C. Testing Input Files for Errors	37

VI.D. Cycle Average Engine Fuel Map for Tractor and Vocational Vehicles	39

VI.E. Running GEM from the Command Prompt and Advanced Options	40

VII.	Final Notes	42

Appendix A. GEM HIL Model	43

Appendix B. GEM Simulink Model	48

Appendix C. Changelog	51

1

-------
I. Introduction

The Greenhouse Gas Emissions Model (GEM) was first created by EPA as part of the "Heavy-Duty
Greenhouse Gas Emissions Standards and Fuel Efficiency Standards for Medium- and Heavy-Duty
Engines and Vehicles: Phase 1" rulemaking finalized in 2011. The model was developed to serve as
a means for determining compliance with EPA's GHG emissions and NHTSA's fuel consumption
Phase 1 vehicle standards for Class 7 and 8 combination tractors and Class 2b-8 vocational vehicles.

For the Phase 2 rulemaking, significant enhancements were made to the model. In addition to the
model released with the Notice of Proposed Rulemaking (NPRM), additional refinements were made
to the model based on public comments received from our NPRM and subsequent Notice of Data
Availability (NODA) releases.

This User Guide describes the Phase 2 GEM release, GEM P2v3.7, released after the Phase 2 technical
amendment (TA) NPRM. The following sections include installation instructions, a general model
description, and instructions for running the model, including a description of the necessary input and
resulting output files. A more detailed description of the model architecture and updates, including
changes made in each release are available in the docket associated with each release, particularly in
Chapter 4 of the Phase 2 RIA.

II. Installation

EPA developed Phase 2 GEM to be a forward-looking Matlab/Simulink-based model for heavy-duty
(Class 2b-8) vehicle compliance for the Phase 2 rulemaking. The model is a free, desktop computer
application provided as an executable to be operated on a single computer. Since it is provided as an
executable, the user is not required to have access to the Matlab/Simulink software packages.1 The
following minimum computer specifications are required for the model to run:

- Operating System: 64-bit Windows 7 or newer

- CPU: 2 GHz processor

- Memory: 4 GB of RAM

II.A. Installation Instructions

The downloadable installation file is available on EPA's website (see Figure 1) at:

https://www.epa.gov/regulations-emissions-vehicles-and-engines/greenhouse-gas-emissions-model-
gem-medium-and-heavy-duty.

A link to the most recent GEM installer is located on this page. The installer includes the model
executable, supporting software (MATLAB Runtime) and several sample input file templates. A copy
of the GEM User Guide is also available separately on the website for convenient reference. The
GEM executable, sample files and documentation require about 10 MB of free space. Users that do
not have the Matlab Compiler Runtime application installed will need about 700 MB of free space.

Currently, GEM is only available to computers using 64-bit Windows operating systems (Windows
7 and newer). To request a CD of this software instead of downloading, or to request assistance if
having trouble with accessibility of this software, please send an email request to OTAQ@epa.gov.

1 The Matlab and Simulink models that make up the GEM source code are available in the docket to the Phase 2
rulemaking. Please see Docket: EPA-HQ-OAR-2014-0827 available at: www.regulations.gov.

-------
f%, United States

^ 1wMM \ ErwironrnenBl Protecton
** Agency

Environmental Topics Laws & Regulations About EPA

Search EPA.gov

Regulations for Emissions from Vehicles and
Engines

Regulations for Emissions
from Vehicles and Engines
Home

On road

Nonroad

Greenhouse Gas

Greenhouse Gas Emissions Model
(GEM) for Medium- and Heavy-
Duty Vehicle Compliance

On this page:

• Overview of GEM

• Phase 2 GEM simulation model

• Phase 1 GEM simulation model V2.Q.1

Overview of GEM

GEM is a free, desktop computer application that estimates the
greenhouse gas (GHG) emissions and fuel efficiency performance of
specific aspects of heavy-duty (HD) vehicles. GEM is designed to
operate on a single computer.

Phase 2 GEM Simulation
Model

(Supporting the Final HD Vehicle GHG Emissions and Fuel Efficiency
Rules)

Related Topic

Regulations for
Greenhouse Gas
Emissions from
Commercial Trucks &
Buses

Advanced Lrght-Outy
Powertraln and Hybrid
Analysis (ALPHA) Tool

You will need Free
Viewers and Adobe
Reader to view some
of the files on this
page. See EPA's
About PDF page to
learn more.

The model documentation provides details on how to install and
use the model. The PDF file also contains the input files that were

used to determine the stringency of the final GHG Emissions Standards and Fuel Efficiency Standards
for Medium-duty and HD Vehicles. The downloadable installation file below contains the application
executable files for the active version of Phase 2 GEM for simulating vehicle compliance.

• GEM User Guide for Phase 2; Vehicle Simulation Tool for Compliance with GHG Emissions
Standards and Fuel Efficiency Standards for MP and HD Engines and Vehicles fPDF)

(52 pp, 1.0MB. EPA-420-B-16-0G7, July 2016)

• Download the executable version of GEM P2V3.0 setup x64 f EXE'i (i ps, 604 mb? July 2016)

+ Top of Page

Figure 1: EPA Website to Obtain GEM Installation Package

To install GEM, run the "GEM_P2v4.0_Setup_x64" executable to start the setup wizard that will
walk through the installation process. Users have the option of choosing a separate location for their
Phase 2 GEM installation but using the default folder as seen in Figure 2 is recommended. The
instructions throughout this User Guide assume the user has installed Phase 2 GEM in the default
locations with the default folder names.

Users have the option of choosing a separate location for their Phase 2 GEM installation but using
the default folder as seen in Figure 2 is recommended. The instructions throughout this User Guide
assume the user has installed Phase 2 GEM in the default locations with the default folder names.

-------
0

1 «=> I ®

Select Destination Location

Where should Phase 2 GEM be installed?

Setup will install Phase 2 GEM into the following folder.

To continue, dick Next. If you would like to select a different folder, dick Browse.

At least 6.4 MB of free disk space is required.

[ j | Cancel ]

Select Start Menu Folder

Where should Setup place the program's shortcuts?

Setup will create the program's shortcuts in the following Start Menu folder.
To continue, dick Next. If you would like to select a different folder, dick Browse.

aafcMRWtil

Browse...

Figure 2: Destination and Start Menu Folder for Phase 2 GEM Download

Phase 2 GEM requires the use of Matlab Compiler Runtime r2020a and the Microsoft Visual C++
2005 or 2008 Redistributable (x64). The installer will warn users if their computers do not have the
necessary software installed. The setup wizard will install Matlab Runtime Compiler R2014a
(version 8.3) if the box shown on the left side of Figure 3 is checked. A pop-up window will initiate
installation of this software and will remain displayed until installation is complete. For computers
that already have the Matlab Runtime installed, users can uncheck this box to skip this step. The
final screen message as shown below on the right side of Figure 3 appears when Phase 2 GEM
installation has completed.

jgi Setup - Phase 2 GEM

' I I-S3-!

Select Additional Tasks

Which additional tasks should be performed?

Select the additional tasks you would like Setup to perform while installing Phase 2
GEM, then dick Next.

Additional icons:

1 Cr eate a desktop i con

Dependencies

W\ Install Matlab Compiler Runtime

[ < Back 11 Next > ) | Cancel |

j|§J Setup - Phase 2 GEM

Completing the Phase 2 GEM
Setup Wizard

Setup has finished installing Phase 2 GEM on your computer.
The application may be launched by selecting the installed
icons.

Click Finish to exit Setup.

Figure 3: Installation Windows; Matlab Runtime Compiler R2014a is Required
II.B. Contents of Installation Package

Once installed, files are stored in the installation location selected (i.e., C-' Program Files\US
EPA\Phase 2 GEM4.0\ by default) and are also available from the Start Menu, under the folder
named "EPA Phase 2 GEM". The start menu folder contains shortcuts to the different operating
modes of GEM as well as documentation, the sample input files and an uninstaller.

1 I.B.I. Sample Input Files

Sample input files are stored with the Phase 2 GEM executable. The "Sample Input Files -
RELOCATE BEFORE USE" folder includes: sample vehicle input files for each of the three vehicle
regulatory subcategories; four folders with example engine, transmission, axle and powertrain input

-------
files; and files to generate cycle averaged fuel maps. A description of each of these input files and
instructions for running the model are provided in later sections of this guide.

Due to file permissions on most installations, GEM is not permitted to use the sample input files in
the location they are installed by default. The recommendation is for users to copy the contents of
the sample input files folder to a location where the user has read and write permissions to minimize
these issues. Users can then easily run GEM and the output results will be saved in the same folder
as the input data.

II.B.2. GEM Executable

EPA and NHTSA require tractor and vocational vehicle manufacturers to use the Phase 2 GEM
executable for demonstrating compliance with the CO2 and fuel consumption standards.2 However,
Phase 2 GEM does not offer a graphical user interface (GUI) for users to provide their vehicle
parameters. Instead, inputs are provided in a comma delimited (.csv) file. Results are available in a
generated report that can be viewed using either a text editor or spreadsheet. The following sections
will describe the model, the input files, and the output files in more detail.

III. Model Description

Phase 1 GEM was updated in order to meet Phase 2 rulemaking requirements. Phase 2 GEM
improves the fidelity of the Phase 1 model to better match the function of the simulated vehicles and
accurately reflect changes in technology for compliance purposes. Many of the modifications were
the result of numerous constructive comments from both the public and GEM peer reviews3. The
following sections describe the model with an emphasis on the additional vehicle parameters
available in the Phase 2 upgrade of the model. Users are directed to Chapter 4 of the RIA for more
detailed information regarding the model architecture and validation.

III.A. GEM Architecture

The GEM architecture is comprised of four systems: Ambient, Driver, Powertrain, and Vehicle as
seen in Figure 4. The Powertrain and Vehicle systems consist of one or more subcomponent models
and a description of the subcomponent models is available in Chapter 4 of the Phase 2 GHG RIA.

2 Trailer manufacturers will use a GEM-based equation but are not required to use GEM. For convenience, this User
Guide provides instructions for using GEM for trailers, but manufacturers cannot use GEM for demonstrating
compliance.

3"Peer Review of the Greenhouse gas Emissions Model (GEM) and EPA's Response to Comments," Docket # EPA-

420-R-15-009, June 2015.

-------
GEM Vehicle Mode /

ambient

Figure 4: Simulink Structure of GEM Vehicle Model

EPA and NHTSA are adopting additional regulatory subcategories to better represent the heavy-
duty vehicles. These subcategories are reflected in GEM with additional vehicle models. Phase 2
GEM also incorporates improvements to the duty cycles, including the addition of idle cycles for
vocational vehicles, and modified cruise cycles that account for changes in road grade. Specifically,
the agencies implemented the following key technical features into Phase 2 GEM along with many
others:

Upgraded engine controller which includes engine fuel cut-off during braking and
deceleration and a cycle average method to supplement the steady state fuel map for
transient simulation which can optionally be applied to the cruise cycles.

Upgraded transmission model which includes automatic and automated manual
transmissions. Optional transmission power loss input and torque converter properties are
also available.

Upgraded driver model with a distance-compensated driver that will drive the certification
drive trace over a prescribed distance regardless of increased drive time due to vehicle
under-performance.

Increased options for number of driven axles including the ability to include axle power
losses measurement data.

With these upgrades, the model can recognize most technologies that could be evaluated in both
engine and chassis dynamometers and is better able to reflect changes in technologies for
compliance purposes. See Chapter 4 of the RIA for more information about these upgrades.

-------
III.A.l. Drive Cycles and Cycle Average Engine Fuel Map

Phase 2 GEM utilizes three drive cycles including a transient cycle and two cruise speed cycles. The
transient mode is defined by California Air Resources Board (CARB) in their Highway Heavy-Duty
Diesel Transient (HHDDT) cycle. The cruise speed cycles are represented by two nominally
constant speed 65 mph and 55 mph cycles, each with varying road grade. For vocational vehicles
two additional idle measurements are utilized, one simulating parked idling operation and the other
idling in traffic. Each regulatory subcategory is assigned a specific set of drive cycle weightings.

The agencies recognize the limitation of the steady state engine fuel map for transient simulation
and are requiring that a cycle average fuel map be generated to supplement the steady state fuel map
for the transient cycle. Additionally, users have the option to apply the cycle average method to the
55 and 65 mph cruise cycles as well. A summary of the procedure to generate the cycle average
map(s) is provided in section VI.D of this Guide. A detailed description and justification for the
cycle average method can be found in Chapter 4 of the Phase 2 GHG RIA.

III.B. Vehicle Parameters for Each Regulatory Subcategory

GEM can simulate a wide variety of Class 2b to Class 8 vehicles. The key to this flexibility is the
component description files that can be modified or adjusted to accommodate a variety of vehicle-
specific information. Phase 2 GEM includes several variations to match potential powertrain
options and the regulatory subcategories in the Phase 2 rulemaking. Each regulatory subcategory is
associated with specific vehicle parameters and technology options.

Many key parameters in GEM are predefined, since those parameters are either hard to quantify due
to lack of certified testing procedures or difficult to obtain due to proprietary barriers. Examples of
these parameters include transmission shifting strategies and engine inertia. The values selected for
these parameters are a result of substantial testing by EPA, as well as confidential discussions with
engine, chassis and component manufacturers. Some default parameters have optional overrides,
which require additional testing.

Each vehicle category has a set of user-defined parameters. These parameters include vehicle
technologies or component attributes that impact CO2 emissions and fuel consumption but have the
potential to vary across manufacturers. Depending on the regulatory category, parameters such as
aerodynamic performance, tire rolling resistance, vehicle weight, engine fuel map, transmission
gear ratios, tire radius, or axle ratio can be changed as inputs by the user.

The sections to follow outline the regulatory vehicle categories that manufacturers may select in
GEM for compliance and summarize the predefined and user-defined and parameters applicable to
each subcategory as well as regulatory citations for associated testing procedures. Details on the file
formats for entering the user defined parameters are discussed in GEM Input Files.

III.B.l. Tractor Vehicle Parameters

GEM recognizes sixteen variations (regulatory subcategories) of combination tractors. Class 8
tractors can have day or sleeper cab configurations with low, mid, or high roof heights. Class 7
tractors are only available in a day cab configuration but also have low, mid, or high roof options.
GEM also recognizes Class 8 heavy-haul tractors with a single vehicle for all cabs and roof heights,
and six optional heavy Class 8 tractor subcategories to represent tractors with higher gross
combined weight rating (GCWR) that are designed for heavy-haul operation in Canada.

Within GEM, high roof tractors are simulated as pulling a standard box van. Mid roof tractors and
low roof tractors are simulated as pulling tank and flatbed trailers, respectively. The standard box

-------
van for high roof tractor simulations also includes a skirt, which impacts the modeled aerodynamic
drag area input, CdA. Note that in MY 2027 and later, GEM will subtract 0.3 m2 from the user's
CdA input value for high roof tractors to account for improved box trailer aerodynamics.

Table 1 through Table 3 summarize some of the predefined modeling parameters for default tractor
subcategories. The standard Class 8 payload is 19 tons. All Class 8 heavy-haul tractors, including
the optional heavy Class 8, have a payload of 43 tons. The Class 7 payload is 12.5 tons. Sleeper cab
tractors are assigned drive cycle weightings that are more representative of long-haul driving. Drive
cycle weightings for day cab tractors are more representative of short-haul driving with more
transient cycle operation.

Table 1: Class 8 Combination Tractor Predefined Modeling Parameters

Sleeper Cab

Day Cab

Heavy Haul

Roof Height

High

Mid

Low

High

Mid

Low

All

Total Weight (kg)

31978

30277

30390

31297

29529

29710

53750

Number of Axles

Payload (tons)

Drive Cycle
Weighting

CARB HHDDT

0.05

0.19

GEM 55 mph

0.09

0.17

GEM 65 mph

0.86

0.64

Table 2: Class 7 Combination Tractor Predefined Modeling Parameters

Day Cab

Roof Height

High

Mid

Low

Total Weight (kg)

22679

20910

21091

Number of Axles

Payload (tons)

12.5

Drive Cycle
Weighting

CARB HHDDT

0.19

GEM 55 mph

0.17

GEM 65 mph

0.64

Table 3: Heavy Class 8 Combination Tractor Predefined Modeling Parameters

Sleeper Cab

Day Cab

Roof Height

High

Mid

Low

High

Mid

Low

Total Weight (kg)

53750

52049

52162

53069

51301

51482

Number of Axles

Payload (tons)

Drive Cycle
Weighting

CARB HHDDT

0.05

0.19

GEM 55 mph

0.09

0.17

GEM 65 mph

0.86

0.64

Table 4 shows the predefined modeling parameters that are consistent across all tractor types. These
common parameters include ambient temperature, efficiencies and accessory powers. The
calculations for overall rolling resistance and the distribution of weight savings are also consistent
for all simulated tractors.

-------
Table 4: Common Predefined Modeling Parameters for All Simulated Combination Tractors

Electrical Accessory Power (W)

1200

Mechanical Accessory Power (W)

2300

Environmental air temperature (°C)

Weight Reduction (lbs)

Add l/3*weight reduction to pay load mass,

Subtract 2/3*weight reduction from the vehicle simulated mass

Trailer Tire Crr (N/kN)

6.0

Overall Tire Crr Calculation (N/kN)

= 0.425*Trailer Crr + 0.425*Drive Crr + 0.15*Steer Crr

GEM allows the user to modify or adjust performance information for certain components in order
to quantify the reduction in tractor fuel consumption and CO2 emissions. More information on the
specific inputs and the procedures to determine them for combination tractors are listed in section
IV.A.

In Phase 1, a default engine and transmission were applied to all GEM-simulated combination
tractors. Phase 2 GEM requires tractor manufacturers to supply an engine fuel map and specific
transmission information as separate input files to the model. Manufacturers also have the option to
use engine and transmission performance data obtained from a powertrain test in their GEM runs
and replace the engine and transmission files with a single powertrain file. Additionally,
manufacturers can optionally include data for transmission and/or axle power loss the appropriate
supplemental input file. These input files have specific requirements, available in the Supplemental
Input Files section of this Guide.

Table 5: Default Transmission Losses for Simulated Combination Tractors

MT or
AMT
Gearbox

Mechanical
Efficiency

9 or more gears

96% for low gears, 98% high gears,
except 100% for 1:1 gear ratio

fewer than 9 gears

100% for 1:1 gear ratio, 98% for rest of gears

Spin Loss

12.3-30.1 Nm, varies with speed

AT
Gearbox

Mechanical Efficiency

99.5% for 1:1 gear ratio, 98% for rest of gears

Spin Loss

40.5 - 65.2 Nm, varies with speed

The default transmission losses in Table 5 are used if transmission power loss information is not
provided within the transmission input file. The default losses are different whether the transmission
is a manual (MT), automated manual (AMT) or a planetary type automatic transmission (AT). The
"low gears" mentioned in the MT or gearbox efficiency of Table 5 only applies when the total gear
number of a transmission is greater than 9. In this type of transmission, the low gear efficiency of
96% will be used for gear number less than or equal to the greater of total number of gears divided
by two or the total number of gears minus 6. Taking a transmission with 10 gears for example, the
greater of 10/2 or 10-6, would be 5. Thus, gears 1-5 would have 96% efficiency.

Phase 2 GEM also accounts for additional technology improvements that reduce CO2 and fuel
consumption but are not easily captured in the vehicle simulation. These reduction values vary for
each technology. The effect of vehicle speed limiter, weight reduction and neutral idle technology
improvement inputs will impact the vehicle simulation. The remaining technologies improvements
are applied as post-process percent reductions to the results from the vehicle simulation. More detail
on the available technology improvement options for tractors and how to determine the inputs is
provided in section IV.A.

-------
III.B.2. Vocational Vehicle Parameters

GEM recognizes nine variations of vocational vehicles based on both vehicle weight class and duty
cycle. Class 8 vocational vehicles are considered heavy heavy-duty (HHD). Classes 6 and 7 are
medium heavy-duty (MHD), and Classes 2b-5 are considered light heavy-duty (LHD). As
demonstrated in Table 6, Table 7 and Table 8 weight, number of axles, aerodynamic drag area and
payload are the same for all of the vehicles within a weight class. Vehicles within each weight class
are further categorized using three duty cycles (Regional, Multi-purpose and Urban) by varying the
drive cycle weightings associated with each composite duty cycle.

The regulations describe the drive cycle weighting factors for each subcategory in 40 CFR
1037.510. Manufacturers should consult the regulations at 40 CFR 1037.510(c) as well as 40 CFR
1037.140(g-h) and 40 CFR 1037.150(z) for instructions on how to select the appropriate
subcategory in which to certify a vocational vehicle configuration. The reasoning behind the
regulations can be found in the preamble of the Phase 2 GHG Rule Section V.D. 1 .e.

Table 6: Vocational Heavy Heavy-Duty (Class 8) Vehicle Predefined Modeling Parameters

Regulatory Subcategory

HHD

Duty Cycle

Regional

Multi-Purpose

Urban

Total weight (kg)

19051

Aerodynamic Drag Area - CdA (m2)

6.86

Payload (tons)

7.50

Electrical Accessory Power (W)

1200

Mechanical Accessory Power (W)

2300

ARB Transient Drive Cycle Weighting

0.20

0.54

0.90

GEM 55 mph Drive Cycle Weighting

0.24

0.23

0.10

GEM 65 mph Drive Cycle Weighting

0.56

0.23

0.00

Parked Idle Cycle Weighting

0.25

0.25

0.25

Drive Idle Cycle Weighting

0.00

0.17

0.15

Non-Idle Cycle Weighting

0.75

0.58

0.60

Table 7: Vocational Medium Heavy-Duty (Class 6-7) Vehicle Predefined Modeling Parameters

Regulatory Subcategory

MHD

Duty Cycle

Regional

Multi-Purpose

Urban

Total weight (kg)

11408

Aerodynamic Drag Area - CdA (m2)

5.40

Payload (tons)

5.60

Electrical Accessory Power (W)

900

Mechanical Accessory Power (W)

1600

ARB Transient Drive Cycle Weighting

0.20

0.54

0.92

GEM 55 mph Drive Cycle Weighting

0.24

0.29

0.08

GEM 65 mph Drive Cycle Weighting

0.56

0.17

0.00

Parked Idle Cycle Weighting

0.25

0.25

0.25

Drive Idle Cycle Weighting

0.00

0.17

0.15

Non-Idle Cycle Weighting

0.75

0.58

0.60

10

-------
Table 8: Vocational Light Heavy-Duty (Class 2b-5) Vehicle Predefined Modeling Parameters

Regulatory Subcategory

LHD

Duty Cycle

Regional

Multi-Purpose

Urban

Total weight (kg)

7257

Aerodynamic Drag Area - CdA (m2)

3.40

Payload (tons)

2.85

Electrical Accessory Power (W)

500

Mechanical Accessory Power (W)

1000

ARB Transient Drive Cycle Weighting

0.20

0.54

0.92

GEM 55 mph Drive Cycle Weighting

0.24

0.29

0.08

GEM 65 mph Drive Cycle Weighting

0.56

0.17

0.00

Parked Idle Cycle Weighting

0.25

0.25

0.25

Drive Idle Cycle Weighting

0.00

0.17

0.15

Non-Idle Cycle Weighting

0.75

0.58

0.60

GEM also provides seven custom chassis subcategories for manufacturers that know the specific
end-use of their vehicles. These custom chassis subcategories are listed in Table 9, and are based on
the standard vocational subcategories. In contrast to the standard vocational subcategories, custom
chassis utilize default inputs for several of the parameters, particularly the engine and transmission.
Use of the custom chassis subcategories reduces the number of required inputs but also limits the
ability to apply some technology improvements. Further details on how to input data for these
subcategories is described in the Vocational Input Files section of this Guide.

Table 9: Vocational Custom Chassis Subcategories and Associated Vehicle in GEM

Custom Chassis Subcategory

GEM Simulated Vehicle

Emergency Vehicles

HHD Urban

Cement Mixers and Other Mixed-Use Applications

HHD Urban

Refuse Vehicles

HHD Urban

Coach Buses

HHD Regional

Transit Bus, Other Bus and Drayage Tractors

HHD Urban

Motor Homes

MHD Regional

School Bus

MHD Urban

Table 10 provides the predefined modeling parameters that are consistent across all vocational
vehicle types. These common parameters include ambient temperature, the calculations for overall
rolling resistance and how weight reduction is applied. Note that some factors which were common
across tractors vary with vocational weight class.

Table 10: Common Predefined Modeling Parameters for All Vocational Vehicles

Environmental Air Temperature (°C)

25

Weight Reduction (lbs)

Add l/2*weight reduction to payload mass,

Subtract l/2*weight reduction from the simulation vehicle mass

Overall Tire Crr (N/kN)

0.7*Drive Crr + 0.3*Steer Crr

GEM allows a user to enter performance information for certain components in order to model and
quantify improvements the manufacturer is making to its vehicles. Table 10 lists the user-defined
modeling parameters for vocational vehicles in addition to a reference to the applicable test methods

11

-------
in the regulation. Phase 2 GEM continues to allow vocational vehicle manufacturers to model their
vehicle's tire rolling resistances, but also requires axle configuration, axle ratio, and loaded size of
the tires for standard vocational chassis subcategories. The model also provides the option to
account for aerodynamic drag as a wind-averaged change in aerodynamic drag area (delta CdA).

For the non-custom chassis vocational categories Phase 2 GEM requires manufacturers to supply an
engine fuel map and specific transmission information for each distinct engine and transmission in
the vehicles being modeled. Manufacturers also have the option to use engine and transmission
performance data obtained from a powertrain test in their GEM runs and replace the engine and
transmission files with a single powertrain file. Additionally, manufacturers can optionally include
data for transmission and/or axle power loss the appropriate supplemental input file. These input
files have specific requirements, as will be discussed in the Supplemental Input Files section of this
Guide.

The default transmission losses in Table 11 are used if transmission power loss information is not
provided within the transmission input file. The default losses are different whether the transmission
is a manual (MT), automated manual (AMT) or a planetary type automatic transmission (AT). The
"low gears" mentioned in the MT or gearbox efficiency only applies when the total gear number of
a transmission is greater than 9. In this type of transmission, the low gear efficiency of 96% will be
used for gear number less than or equal to the greater of total number of gears divided by two or the
total number of gears minus 6. Taking a transmission with 10 gears for example, the greater of 10/2
or 10-6, would be 5. Thus, gears 1-5 would have 96% efficiency.

Table 11: Default Transmission Losses for Vocational Vehicles

MT or
AMT
Gearbox

Mechanical
Efficiency

9 or more gears

96% for low gears, 98% high gears,
except 100% for 1:1 gear ratio

fewer than 9 gears

100% for 1:1 gear ratio, 98% for rest of gears

Spin Loss

HHD

12.3 - 30.1 Nm, varies with speed

MHD

9.1 - 22.3 Nm, varies with speed

LHD

5.9 - 14.5 Nm, varies with speed

AT
Gearbox

Mechanical Efficiency

99.5% for 1:1 gear ratio, 98% for rest of gears

Spin Loss

HHD

40.5 - 65.2 Nm, varies with speed

MHD

26.2 - 42.1 Nm, varies with speed

LHD

23.5 - 37.9 Nm, varies with speed

Similar to the tractor model, Phase 2 GEM accounts for additional vocational vehicle technologies
that reduce CO2 and fuel consumption but are not easily captured in the vehicle simulation. These
reduction values vary for each technology. The vehicle speed limiter, weight reduction, neutral idle,
and start-stop options will impact the vehicle simulation. The remaining technologies improvements
are applied as post-process g/ton-mile or percent reductions to the results from the vehicle
simulation. Table 12 directs users to the corresponding regulation reference to determine
appropriate values to apply for each technology. These technology improvements are available to
all vocational subcategories.

-------
Table 12: Technology Improvement Options for Vocational Vehicle Manufacturers

Technology Improvement

Regulation Reference

Vehicle Speed Limiter (MPH or NA)

40 CFR 1037.520(d)

Delta Power Take Off Fuel (g/ton-mile)

40 CFR 1037.520(k) and 1037.540

Weight Reduction (lb)

40 CFR 1037.520(e)

Neutral Idle (Y/N)

40 CFR 1037.520(h)

Start-Stop (Y/N)

40 CFR 1037.520(h)

Automatic Engine Shutdown

40 CFR 1037.520(h)

Accessory Load (%)

40 CFR 1037.520©

Tire Pressure System (%)

40 CFR 1037.520©

Other (%)

40 CFR 1037.520©

III.B.3. Trailer Vehicle Parameters

The agencies are adopting an equation-based compliance approach for box trailer manufacturers
and they are not required to certify their trailers using GEM (see 40 CFR 1037.515). However, the
equations for each box trailer subcategory are based on the simulated trailers described in this
section. Note that non-box trailers do not use GEM or the GEM-based equation for compliance and
a discussion of non-box trailers is not included in this User Guide. The following description of the
trailer model as it applies to box trailers is included for informational purposes only.

The agencies are adopting a set of predefined modeling parameters to establish consistent tractor-
trailer models from which box van trailer manufacturers can compare their vehicle improvements.
GEM recognizes four variations of box van based on length. Long box vans (trailers that are longer
than 50-feet) are represented by either a 53-foot dry van or a 53-foot refrigerated van pulled by a
Class 8 high roof sleeper cab tractor in GEM and are given the same long-haul drive cycle
weightings as the Class 8 high roof sleeper cab tractors mentioned previously. GEM models all
short box vans (box trailers 50-feet in length and shorter) as a single-axle, solo 28-foot dry van or
refrigerated van pulled by a Class 7 high roof day cab tractor with a 4x2 drive axle configuration.

Table 13 summarizes the predefined modeling parameters for long and short box vans, respectively.
All long box vans are modeled with tandem axles and a payload of 19 tons and drive cycle
weightings that are more representative of long-haul driving (i.e., 86 percent at 65-MPH, 9 percent
at 55-MPH and 5 percent transient). All short box vans are modeled with a single axle, a payload of
10 tons, and drive cycle weightings more representative of short-haul driving with 64 percent at 65-
MPH, 17 percent at 55-MPH, 19 percent transient. The vehicle weight varies between dry vans and
refrigerated vans to account for the weight of the refrigeration unit, and weight also varies
proportional to the length of the trailer. The baseline CdA values were obtained from EPA's
aerodynamic testing.

-------
Table 13: Predefined Modeling Parameters for Box Trailers

Regulatory Subcategory

Long Box
Dry Van

Long Box
Refrigerated
Van

Short Box
Dry Van

Short Box
Refrigerated
Van

Tractor Type

C8 Sleeper Cab - High Roof

C7 Day Cab - High Roof

Total weight (kg)

31978

33778

18306

20106

Baseline CdA Values (m2)

6.0

6.0

5.6

5.6

Tractor Engine

15L 455 HP

11L 350 HP

Tractor Drive Axle Configuration

6x4

4x2

Number of Axles

5

3

Payload (tons)

19

10

CARB HHDDT Drive Cycle Weighting

0.05

0.19

GEM 55 mph Drive Cycle Weighting

0.09

0.17

GEM 65 mph Drive Cycle Weighting

0.86

0.64

Table 14 shows the predefined modeling parameters that are consistent across all trailer types.
Many of these common parameters are associated with the simulated tractor in the tractor-trailer
model. The calculations for overall rolling resistance and the distribution of weight savings are
consistent with the calculations for GEM-simulated tractors.

Table 14: Common Predefined Modeling Parameters for All Box Trailers

Gearbox Efficiency

100% for 1:1 gear ratio, 96% for lower gears or 98% for
rest of gears

Axle Drive Ratio

3.7

Electrical Accessory Power (W)

300

Mechanical Accessory Power (W)

1000

Loaded Tire Size (rev/mi)

512

Steer Tire Crr (N/kN)

6.54

Drive Tire Crr (N/kN)

6.92

Overall Tire Crr (N/kN)

= 0.425*Trailer Crr + 0.425*Drive Crr + 0.15*Steer Crr

Weight Reduction (lbs)

Add l/3*weight reduction to payload mass

GEM allows a user to modify or adjust performance information for certain components in order to
model and quantify improvements the manufacturer is making to its vehicles. The trailer program
has three user-defined parameters and one pre-defined technology improvement option that has a
specified reduction value associated with its use (see Table 15). Trailer manufacturers can change
their tire rolling resistance, aerodynamic drag and cumulative weight reduction. GEM applies an
additional percent improvement value to trailers that have tire pressure systems installed on their
simulated trailer. Separate percentage values apply for tire pressure monitoring systems (TPMS)
and automatic tire inflation systems (ATIS).

Table 15: User-Defined Modeling Parameters and Technology Improvement Options for Trailers

Modeling Parameter

Method of Determining Parameter

Trailer Tire Crr (N/kN)

40 CFR 1037.515(b)

Change in Aerodynamic Drag Area, ~ CdA (m2)

40 CFR 1037.515(c) and 40 CFR 1037.526

Weight Reduction (lb)

40 CFR 1037.515(d)

Tire Pressure System a

1.2% for ATIS, 1.0% for TPMS, 0 if no system

a Note 40 CFR 1037.515(a) specifies the tire pressure system values in decimal format, because trailers use a GEM-based equation
for compliance. The GEM input file described in this Guide uses percent-based input values for tire pressure systems.

14

-------
IV. GEM Input Files

As mentioned previously, Phase 2 GEM does not offer a graphical user interface (GUI) for users to
provide their vehicle parameters. Instead, inputs are exclusively provided to the model in a .csv file.
Phase 2 GEM executable can be initiated using the Start Menu, or a command prompt. The
following subsections describe the input file structures, which are consistent for each method of
running the model.

Some manufacturers may create a script to automatically generate their input files in CSV (comma
separated value) format. Others may wish to manually populate their files using a spreadsheet tool,
such as Microsoft Excel, to easily view the input fields in a column format. For illustration
purposes, the sample input data in this document are shown in tabular spreadsheet like format.

Users may use Microsoft Excel or any other text editor, to create or edit their input files. If using a
spreadsheet program be sure to save the files in CSV format.

The top of the input file for each of the regulatory categories has three lines that list the regulatory
category, manufacturer name and model year (see Sample 1). User entries in the second column
determine the processing that will be applied to the file. The first line contains the regulatory
category and must contain "Tractor", "Vocational" or "Trailer" as appropriate for the vehicles being
simulated. Manufacturer name can be in any format, but it should be consistent with other
regulatory documents from the manufacturer. Model year must be expressed as a four-digit number.

Sample 1: Input File Header Information

Regulatory Category

Tractor

Manufacturer Name

EPA

Model Year

2018

The subsequent lines of the input file list the model inputs with one simulation listed per line. The
first two columns contain a Run ID and the Regulatory Subcategory for each run. The Run ID is a
unique value that will be used to identify the run (e.g., vehicle VIN) and can be any combination of
letters, numbers and separators such as dash ("-"), periods ("."), or underscores ("_"). The
Regulatory Subcategory determines the vehicle and associated predefined parameters that are to be
simulated along with the standard that is applied for compliance. Additional detail is provided in the
following sections regarding the columns and requirements specific to each regulatory category.
Each is accompanied with snippets containing selected columns from the sample input files.

IV.A. Tractor Input Files

Tractor Sample 1 shows a sample of the first two columns for a tractor input file. Table 16 shows
corresponding subcategory identifiers for use in column two. The first 10 are the standard
regulatory subcategories. The final six (starting with HC8) are the optional heavy Class 8 tractor
subcategories that represent tractors designed for heavy-haul operation in Canada. Manufacturers
may optionally certify their tractors as heavy Class 8, using the subcategories shown starting in
MYs 2021. See Section III.B.4.a of the preamble to this rulemaking for a discussion of these
optional tractor subcategories. Note these tractors are separate from the C8 HH subcategory that is
used for U.S.-based heavy-haul tractors.

-------
Tractor Sample 1: Input File Run ID and Regulatory Subcategory Inputs

Run ID

Regulatory Subcategory

Unique Identifier

(e.g. C8_SC_HR)

Sample_l

C8 SC HR

Sample_2

C7 DC MR

Sample_3

C8 HH

Sample_4

HC8 SC LR

Sample_5

HC8 DC MR

Sample_6

C8 DC MR

Sample_6a

C8 DC MR

Sample_7

C8 SC LR

Sample_8

C8 SC HR

PT_Samplel

C8_SC_HR

Table 16: Tractor Input File Naming Convention for Tractor Regulatory Subcategories

GEM Input Name

Regulatory Subcategory Description

C8 SC HR

Class 8 Combination, Sleeper Cab - High Roof

C8 SC MR

Class 8 Combination, Sleeper Cab - Mid Roof

C8 SC LR

Class 8 Combination, Sleeper Cab - Low Roof

C8 DC HR

Class 8 Combination, Day Cab - High Roof

C8 DC MR

Class 8 Combination, Day Cab - Mid Roof

C8 DC LR

Class 8 Combination, Day Cab - Low Roof

C8 HH

Class 8 Combination, Sleeper Cab - Heavy Haul

C7 DC HR

Class 7 Combination, Day Cab - High Roof

C7 DC MR

Class 7 Combination, Day Cab - Mid Roof

C7 DC LR

Class 7 Combination, Day Cab - Low Roof

HC8 SC HR

Heavy Class 8 Combination, Sleeper Cab - High Roof

HC8 SC MR

Heavy Class 8 Combination, Sleeper Cab - Mid Roof

HC8 SC LR

Heavy Class 8 Combination, Sleeper Cab - Low Roof

HC8 DC HR

Heavy Class 8 Combination, Day Cab - High Roof

HC8 DC MR

Heavy Class 8 Combination, Day Cab - Mid Roof

HC8DCLR

Heavy Class 8 Combination, Day Cab - Low Roof

The next columns in the input file are for information regarding the tractor powertrain. The primary
items are the engine and transmission file names or the powertrain file name. A description of the
content of these supplemental input files is located at the end of this section. The text in the fields of
the input file must exactly match the file name, including the .csv extension for GEM to run
properly. In the Sample Input Files folder provided with GEM, the supplemental files are in
separate Axle, Engine, Powertrain and Transmission subfolders. GEM searches for the
supplemental files from the same location as the vehicle input file. As a result, the subfolder name
must be included in the File Name fields to direct the model to the files. If users have their engine,
transmission, powertrain and axle files in the same folder as the vehicle input file, only the filename
is needed.

16

-------
Tractor Sample 2: Input File Reference to Engine, Transmission, and Powertrain Input Files

Engine

Transmission

Data

Idle Speed
at CUT

Data

File Name

RPM

File Name

Engines\EPA_2018_D_GENERIC_455_trans_c...

600

Transmissions\EPA_AMT_10_C78_4490.csv

Engines\EPA_2018_D_GENERIC_350_trans_ c...

650

Transmissions\EPA_AMT_10_C78_4490.csv

Engines\EPA_2018_D_GENERIC_600_trans_ c...

600

Transmissions\EPA_AMT_10_C78_4490.csv

Engines\EPA_2018_D_GENERIC_600_trans_ c...

600

Transmissions\EPA_MT_13_C78_4543.csv

Engines\EPA_2018_D_GENERIC_455_trans_ c...

600

Transmissions\EPA_AMT_10_C78_4490.csv

Engines\EPA_2018_D_GENERIC_455_trans_ c...

600

Transmissions\EPA_AT_10_C78_8001.csv

Engines\EPA_2018_D_GENERIC_455_trans_ c...

650

Transmissions\EPA_AT_10_C78_8001.csv

Engines\EPA_2018_D_GENERIC_455_trans_ c...

600

Transmissions\EPA_AMT_10_C78_4490_power_l...

Engines\EPA_2018_D_GENERIC_455_all_cyc_...

600

Transmissions\EPA_AMT_10_C78_4490.csv

Powertrains\EPA_Sample_Powertrainl.csv

600

Powertrains\EPA_Sample_Powertrainl.csv

Manufacturers may choose to perform powertrain testing to obtain the engine and transmission
performance for their GEM simulations. In order to indicate to GEM that powertrain data is
provided, users would provide the same powertrain input file name in both the Engine and
Transmission file name fields of the tractor input file as seen in the last row of Tractor Sample 2.

Another powertrain parameter in this section is the engine idle speed. This allows calibrated idle
speeds to be set at the vehicle level without the need for multiple engine input files. The idle speed
in this column will be used to interpolate the idle fuel maps within the engine input file to adjust the
estimated fuel consumption and emissions. For more information on the appropriate value to enter
see 40 CFR 1036.510.

Tractor Sample 3: Input File Performance Parameters and User-Defined Vehicle Characteristics

Drive Axle

Drive
Axle

Drive Axle

Aerodynamic

Steer Axle
Tire

Drive Axle
1 Tire

Drive Axle
2 Tire

Drive
Axle Tire

Configuration

Ratio

Data

Aerodynamic
Drag Area
(CdA)

Rolling
Resistance
Level

Loaded
Tire Size

(e.g. 6x4)

File Name

mA2

N/kN

rev/mi

6x4

3.08

5.4

6.9

500

4x2

3.42

5.1

6.9

500

6x4

3.23

6.9

500

6x4

3.12

5.07

6.8

512

6x4

3.45

5.35

6.9

500

6x4

3.45

5.18

6.9

512

6x4

3.45

5.18

6.9

512

6x4 D

3.12

Axles\EPA_Axle.csv

4.88

6.9

500

6x4

3.08

5.4

6.9

500

6x4

3.08

5.4

6.9

500

Tractor Sample 3 shows the next columns containing the tractor performance parameters and
several vehicle characteristics that are user-defined in GEM. Procedures or guidance to determine

-------
the appropriate value for each of these parameters is available in Table 17, with further information
available in the preamble to the Phase 2 rulemaking.

Table 17: User-Defined Modeling Parameters for Class 7 and Class 8 Combination Tractors

Modeling Parameter

Method of Determining Parameter

Drive Axle Configuration

40 CFR 1037.520(g)(2)

Drive Axle Ratio

40 CFR 1037.520(g)(3)

Drive Axle data file (Optional)

40 CFR 1037.560

Aerodynamic Drag Area, CdA (m2)

See 40 CFR 1037.520(b)

Steer Axle Tire Rolling Resistance (N/kN)

40 CFR 1037.520(c)

Drive Axle 1 Tire Rolling Resistance (N/kN)

Drive Axle 2 Tire Rolling Resistance (N/kN)

Drive Axle Loaded Tire Size (rev/mi)

See 40 CFR 1037.520(c)

The Drive Axle Data File Name field points to the location and name of the optional axle file. Note,
if users do not use an optional axle file, then "NA" should be entered. The format of this field is
similar to the engine and transmission file fields. For tractors that have a single rear axle (i.e., 4x2
axle configuration), the users input an "NA" into the Drive Axle 2 Tire Rolling Resistance Level
field for those vehicles. Also, heavy-haul tractors (third row) have a default Aerodynamic Drag
Area (CdA) of 5.0 m2 within GEM, so users also input an "NA" into that field for heavy-haul
tractors.

There are limits associated with each user-defined input value. Drive Axle Configuration is a text
input and the allowable options are "6x4", "4x2", "6x4D", or "6x2". Vehicles with more than two
drive axles are instructed to use the "6x4" configuration in the model. All tractors with "6x2" axle
configurations are modeled with five axles with two steer tires, 4 non-drive tires and 4 drive tires.
All tractors with "6x4" axle configurations are modeled with five axles with two steer tires and
eight drive tires. The only difference in GEM between "6x2" and "6x4" axles is an additional 1
percent loss for "6x4" axles to account for the inter-axle losses. All tractors with "6x4D" axle
configurations are modeled as "6x2" axles on the cruise cycles and "6x4" axles on the transient
cycle. All tractors with a "4x2" axle are represented by a four-axle tractor with two steer tires and
four drive tires Table 18 shows the limits for the remaining six tractor inputs in the model. GEM
will produce an error if any of these values are out of the acceptable range and will round any
values beyond their specified decimal limits.

Table 18: Minimum and Maximum Limits for User-Defined Values in Tractor Input Files

User-Defined Parameter

Units

Number of
Decimals

Minimum
Value

Maximum
Value

Drive Axle Ratio

1.00

20.00

Aerodynamic Drag Area (CdA)

mA2

3.00

8.00

Steer Axle Tire, Rolling Resistance Level

N/kN

3.0

20.0

Drive Axle 1 Tire, Rolling Resistance Level

N/kN

3.0

20.0

Drive Axle 2 Tire, Rolling Resistance Level

N/kN

3.0

20.0 or NA

Drive Axle Loaded Tire Size

rev/mi

100

1000

The remaining columns in the tractor input file are for the optional technology improvements.
These technology improvement fields cannot be blank in the input file. The first three tractor

-------
technology improvements, shown in Tractor Sample 4, will directly impact the vehicle simulation.
Vehicle speed limiters reduce the maximum allowable speed of the vehicle during the simulation to
the user-specified value. Weight reduction reduces the overall vehicle weight (and increases
payload). If no weight reduction is used, then input "0". Neutral-idle reduces the fuel consumption
of the engine when a simulated automatic transmission vehicle is idling. If the vehicle includes
neutral-idle technology, then input "Y", otherwise enter "N".

Tractor Sample 4: Technology Improvements

Technology Improvement

Technology Improvement

Technology Improvement

Vehicle Speed Limiter

Weight Reduction

Neutral-Idle

MPH or NA

lbs

Y/N

NA

0

N

NA

100

N

NA

0

N

NA

0

N

NA

0

N

NA

0

Y

NA

0

Y

NA

0

N

NA

0

N

NA

0

N

The remaining technology improvements, shown in Tractor Sample 5, have specific percent
reductions that manufacturers will apply for the given technology fields. If the technology
improvement(s) is not applicable to the vehicle being simulated, then input "0". The Other field
may be used for several technologies, including results from any off-cycle testing that
manufacturers may perform. Reference 40 CFR 1037.520 for general information or regarding how
to determine percent values. All values of "Y", "N", or "NA" must be in UPPERCASE LETTERS.
Lowercase letters will produce an error.

19

-------
Tractor Sample 5: Technology Improvements with Pre-Defined Percent Improvements

Technology
Improvement

Technology
Improvement

Technology
Improvement

Technology
Improvement

Technology
Improvement

Intelligent
Controls

Accessory Load

Extended Idle
Reduction

Tire Pressure
System

Other

%

%

%

%

%

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Table 19: Technology Improvement Options for Tractor Manufacturers

Technology Improvement

Regulation Reference

Vehicle Speed Limiter (MPH or NA)

40 CFR 1037.520(d)

Weight Reduction (lb)

40 CFR 1037.520(e)

Neutral Idle, AT only (Y/N)

40 CFR 1037.660

Intelligent Controls (%)

40 CFR 1037.520(j)

Accessory Load (%)

40 CFR 1037.520(j)

Extended Idle Reduction, Sleeper Cabs Only (%)

40 CFR 1037.520(j)

Tire Pressure System (%)

40 CFR 1037.520(j)

Other (%)

40 CFR 1037.520(j)

Table 19 directs users to the corresponding regulation reference to determine appropriate values to
apply for each technology. Similar to the user-defined parameters, these technology improvements
also have limits. The format and limits for the technology improvements are shown in Table 20.
Input values with additional decimal places will be rounded to the appropriate precision. Input
values outside the minimum and maximum range specified will produce an error.

Table 20: Minimum and Maximum Limits for Technology Improvement Values in Tractor Input File

Modeling Parameter

Units

Number of
Decimals

Minimum

Maximum

Vehicle Speed Limiter

MPH or NA

1

54.0

65.0

Weight Reduction

lb

0

0

40,000

Neutral Idle, Automatic Transmissions Only

Y/N

-

-

-

Intelligent Controls

%

1

1.0

10.0

Accessory Load

%

1

1.0

10.0

Extended Idle Reduction, Sleeper Cabs Only

%

1

1.0

10.0

Tire Pressure System

%

1

1.0

10.0

Other

%

1

1.0

50.0

20

-------
IV.B. Vocational Input Files

Consistent with the other regulatory categories, the first two columns, as shown in the sample in
Vocational Sample 1, contain a Run ID and the Regulatory Subcategory for each run. For
vocational vehicles, there are nine standard and seven custom chassis regulatory subcategories in
GEM and Table 21 shows codes to use in the Regulatory Subcategory column. See Section V.B.2.b
of the preamble to this rulemaking for a discussion of these optional custom chassis subcategories.
For the standard regulatory subcategories, the R, M and U suffixes only alter the duty cycle
weighting. GEM will calculate results for each of the possible weighting to report in the output.

Vocational Sample 1: Input File Run ID and Regulatory Subcategory Inputs

Run ID

Regulatory Subcategory

Unique Identifier

(e.g. HHD_R)

Sample_l

HHD R

Sample_2

HHD M

Sample_3

LHD U

Sample_4

LHD M

Sample_5

LHD U

Sample_6

LHD U

Sample_7

MHD

Sample_8

LHD

CC_Sample_l

HHD CC RF

CC_Sample_2

HHD CC EM

CC_Sample_3

HHD CC CM

CC_Sample_4

HHD CC OB

CC_Sample_5

HHD CC CB

CC_Sample_6

MHD CC MH

CC_Sample_7

MHD CC SB

PT_Samplel

HHD M

PT_Samplel

HHD

21

-------
Table 21: Vocational Input File Naming Convention for Vocational Regulatory Subcategories

GEM Input Name

Regulatory Subcategory Description

HHDR

Heavy-Heavy Duty, Regional

HHDM

Heavy-Heavy Duty, Multipurpose

HHDU

Heavy-Heavy Duty, Urban

MHDR

Medium-Heavy Duty, Regional

MHDM

Medium-Heavy Duty, Multipurpose

MHDU

Medium-Heavy Duty, Urban

LHDR

Light-Heavy Duty, Regional

LHDM

Light-Heavy Duty, Multipurpose

LHDU

Light-Heavy Duty, Urban

HHDCCEM

Emergency Vehicles

HHDCCCM

Cement Mixers and Other Mixed Use Applications

HHDCCRF

Refuse Vehicles

HHDCCCB

Coach Buses

HHDCCOB

Transit Bus, Other Bus and Drayage Tractors

MHDCCMH

Motor Homes

MHDCCSB

School Bus

The next columns (3-5) in the input file are for the engine and transmission or the powertrain input
filenames as shown in Vocational Sample 2. A description of the content of these supplemental
input files is located in section IV.C. The text in the fields of the input file must exactly match the
file name, including the .csv extension for the code to run properly.

Vocational Sample 2: Input File Reference to Engine, Transmission, and Powertrain Input Files

Engine

Transmission

Data

Idle Speed at CUT

Data

File Name

RPM

File Name

Engines\EPA_2018_D_GENERIC_350_trans_c...

650

Transmissions\EPA_MT_10_HHD.csv

Engines\EPA_2018_D_GENERIC_350_trans_c...

650

Transmissions\EPA_AT_5_HHD_LU3.csv

Engines\EPA_2018_D_GENERIC_200_trans_ c...

750

Transmissions\EPA_AT_6_LHD_LU3.csv

Engines\EPA_2018_D_GENERIC_200_trans_ c...

750

Transmissions\EPA_AT_6_LHD_LU3.csv

Engines\EPA_2018_D_GENERIC_200_trans_ c...

750

Transmissions\EPA_AT_6_LHD_LU2.csv

Engines\EPA_2018_G_GENERIC_300hp_trans...

600

Transmissions\EPA_AT_6_LHD_LU2.csv

Engines\EPA_2018_D_GENERIC_270_trans_ c...

750

Transmissions\EPA_AT_6_MHD_LU3.csv

Engines\EPA_2018_D_GENERIC_200_trans_c...

750

Transmissions\EPA_AT_6_LHD_LU3.csv

Powertrains\EPA_Sample_Powertrainl.csv

600

Powertrains\EPA_Sample_Powertrainl.csv

Powertrains\EPA_Sample_Powertrain2.csv

600

Powertrains\EPA_Sample_Powertrain2.csv

In the Sample Input Files folder provided with GEM, the supplemental files are stored in separate
Axle, Engine, Powertrain and Transmission subfolders. GEM searches for the supplemental files

-------
from the same location as the vehicle input file. As a result, the subfolder name must be included in
the File Name fields, as shown in Vocational Sample 2, to direct the model to the files. If users have
their engine, transmission, powertrain and axle files in the same folder as the vehicle input file, only
the filename is needed. Manufacturers may choose to perform powertrain testing to obtain the
engine and transmission performance for their GEM simulations. In order to indicate to GEM that
powertrain data is provided, users would provide the same powertrain input file name in both the
Engine and Transmission File Name fields of the tractor input file, as seen in the last row of
Vocational Sample 2. Note that the custom chassis subcategories (rows 9-15 of the sample input
file) use default engines and transmissions within GEM and manufacturers input "NA" in those
fields.

The next columns of the sample input file are shown in Vocational Sample 3 and contain the
vocational performance parameters and vehicle characteristics that are user-defined in GEM. A
description of these parameters is given in Table 22 and additional information is available in
Section V.D of the preamble to the Phase 2 rulemaking. For vehicles that have a single rear axle
(i.e., 4x2 axle configuration), the users input an "NA" into the Drive Axle 2 Tire Rolling Resistance
Level field. As shown in Vocational Sample 3, custom chassis manufacturers only specify the drive
axle configuration and tire rolling resistance values; all other user-defined fields are marked "NA".

Vocational Sample 3: Input File Performance Parameters and User-Defined Vehicle Characteristics

Drive Axle

Drive
Axle

Drive Axle

Aerodynamic
Improvement
(Delta)

Steer Axle
Tire

Drive Axle 1
Tire

Drive Axle
2 Tire

Drive
Axle
Tire







Aerodynamic

Rolling

Rolling

Rolling

Loaded

Configuration

Ratio

Data

Drag Area

Resistance

Resistance

Resistance

Tire







(CdA)

Level

Level

Level

Size

(e.g. 6x4)

#

File Name

mA2

N/kN

N/kN

N/kN

rev/mi

6X4

3.73

NA

0

7.7

7.7

7.7

530

6X4 D

4.33

Axles\EPA_Axle.csv

0

7.7

7.7

7.7

530

4x2

4.09

NA

0

6.2

6.2

NA

500

4x2

4.09

NA

0

6.2

6.2

NA

500

4x2

4.09

NA

0

6.2

6.2

NA

500

4x2

4.09

NA

0

6.2

6.2

NA

500

4x2

3.8

NA

0.2

6.2

6.2

NA

500

4x2

4.09

NA

0

6.2

6.2

NA

500

6X4

NA

NA

NA

7.7

7.4

7.4

NA

6X4

NA

NA

NA

7.3

7.1

7.1

NA

6X4

NA

NA

NA

7.7

7.7

7.7

NA

4x2

NA

NA

NA

7.7

6.9

NA

NA

6x2

NA

NA

NA

7.7

6.8

6.8

NA

4x2

NA

NA

NA

7.6

7.5

NA

NA

4x2

NA

NA

NA

7.7

7.7

NA

NA

6x4

3.54

NA

0

6.9

6.9

6.9

512

6x4

3.54

NA

0

6.9

6.9

6.9

512

23

-------
Table 22: User-Defined Modeling Parameters for Vocational Vehicles (All Weight Classes)

Modeling Parameter

Method of Determining Parameter

Engine data file

40 CFR 1036.535 and 1036.540

Engine Idle Speed

40 CFR 1036.510

Transmission data file

40 CFR 1037.520(g)(1) and optionally 1037.565

Powertrain data file (Optional)

40 CFR 1037.550

Drive Axle Configuration

40 CFR 1037.520(g)(2)

Drive Axle Ratio

40 CFR 1037.520(g)(3)

Drive Axle data file (Optional)

40 CFR 1037.560

Aerodynamic Drag Area, Delta CdA (m2)

40 CFR 1037.520(m) and 40 CFR 1037.527

Steer Axle Tire Rolling Resistance (N/kN)

40 CFR 1037.520(c)

Drive Axle 1 Tire Rolling Resistance (N/kN)

Drive Axle 2 Tire Rolling Resistance (N/kN)

Drive Axle Loaded Tire Size (rev/mi)

See 40 CFR 1037.520(c)

There are limits associated with each user-defined input value. Drive Axle Configuration is a text
input and the allowable text is "6x4", "4x2", "6x4D", or "6x2". Vehicles with more than two drive
axles are instructed to use the "6x4" configuration in the model. All vehicles with "6x2" axle
configurations are modeled with five axles with two steer tires, 4 non-drive tires and 4 drive tires.
All vehicles with "6x4" axle configurations are modeled with five axles with two steer tires and
eight drive tires. The only difference in GEM between "6x2" and "6x4" axles is the additional 1
percent loss for "6x4" axles to account for the inter-axle losses. All vehicles with "6x4D" axle
configurations are modeled as "6x2" axles on the cruise cycles and "6x4" axles on the transient
cycle. All vehicles with a "4x2" axle are represented by four axles with two steer tires and four
drive tires.

The Drive Axle Data File Name field points to the location and name of the optional axle file. The
format of this field is similar to the engine and transmission file fields shown in Vocational Sample
2. Table 23 shows the limits for the next six vocational inputs in the model. GEM will produce an
error if any of these values are out of the acceptable range and will round any values beyond their
specified decimal limits. The aerodynamic improvement for vocational vehicles is measured as a
delta CdA and not the absolute CdA value used in the tractor program.

Table 23: Minimum and Maximum Limits for User-Defined Values in Vocational Input File

User-Defined Parameter

Units

Number of
Decimals

Minimum
Value

Maximum
Value

Drive Axle Ratio

1.00

20.00

Aerodynamic Drag Area (Delta CdA)

mA2

0.00

4.00

Steer Axle Tire, Rolling Resistance Level

N/kN

3.0

20.0

Drive Axle 1 Tire, Rolling Resistance Level

N/kN

3.0

20.0

Drive Axle 2 Tire, Rolling Resistance Level

N/kN

3.0

20.0 or NA

Drive Axle Loaded Tire Size

rev/mi

100

1000

The next columns in the vocational input file are for the optional technology improvements. These
technology improvement fields cannot be blank in the input file. Five of the next six vocational
technology improvements, shown in Vocational Sample 4, will directly impact the vehicle
simulation. Vehicle speed limiters reduce the maximum allowable speed of the vehicle during the
simulation to the user-specified value. Weight reduction reduces the overall vehicle weight (and
increases payload) as noted previously in Table 10. Neutral-idle reduces fueling when a simulated

-------
automatic transmission vehicle is idling, Start-Stop reduces fueling when the simulation comes to a
stop in the transient drive cycle and drive idle cycle, and Automatic Engine Shutdown reduces
fueling during the parked idle cycle.

Reference the preamble Section V.C.I.a.iv for a description of the three workday idle reduction
options. If the vocational vehicle will be built with a hybrid power take-off (PTO) and testing was
conducted according to 40 CFR 1037.540, the Delta PTO value obtained from that test procedure
may be entered. Please note these inputs are case-sensitive. All values of "Y", "N", or "NA" must
be in UPPERCASE LETTERS. Lowercase letters will produce an error.

Vocational Sample 4: Technology Improvements

Technology
Improvement

Technology
Improvement

Technology
Improvement

Technology
Improvement

Technology
Improvement

Technology
Improvement

Vehicle
Speed
Limiter

Delta PTO
Fuel

Weight
Reduction

Neutral-Idle

Start-Stop

Automatic

Engine
Shutdown

MPHorNA

g/ton-mile

lbs

Y/N

Y/N

Y/N

NA

0

0

N

N

Y

NA

0

0

N

N

Y

NA

0

0

N

N

N

NA

0

0

Y

N

N

NA

0

0

N

Y

Y

NA

0

0

Y

N

Y

NA

0

0

Y

N

N

NA

0

0

Y

N

N

NA

0

0

N

N

N

NA

0

0

N

N

N

NA

0

0

N

N

N

NA

0

0

N

N

N

NA

0

0

N

N

N

NA

0

0

N

N

N

NA

0

0

N

N

N

NA

0

0

N

N

Y

NA

0

0

N

N

Y

The remaining three technology improvements, shown in Vocational Sample 5, have specific
percent reductions that manufacturers will apply for the given technology fields. The "Other" field
may be used for several technologies, including results from any off-cycle testing that
manufacturers perform. See Table 12 and, generally, 40 CFR 1037.520 for the appropriate percent
values.

25

-------
Vocational Sample 5: Technology Improvements with Pre-Defined Percent Improvements

Technology
Improvement

Technology
Improvement

Technology
Improvement

Technology
Improvement

Intelligent Controls

Accessory Load

Tire Pressure System

Other

%

%

%

%

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Similar to the user-defined parameters, these technology improvements also have limits. The format
and limits for the technology improvements are shown in Table 24. Input values with additional
decimal places will be rounded to the appropriate precision. Input values outside the minimum and
maximum range specified will produce an error.

Table 24: Minimum and Maximum Limits for Technology Improvement Values in Vocational Input File

Modeling Parameter

Units

# of Decimals

Minimum

Maximum

Vehicle Speed Limiter

MPHorNA

1

54.0

65.0

Weight Reduction

lb

0

0

10,000

Delta PTO

g/ton-mi

3

0.000

3.000

Neutral Idle, Automatic Transmissions Only

Y/N

-

-

-

Start-Stop

Y/N

-

-

-

Automatic Engine Shutdown

Y/N

-

-

-

Accessory Load

%

1

1.0

10.0

Tire Pressure System

%

1

1.0

10.0

Other

%

1

1.0

50.0

26

-------
IV.C. Trailer Input Files

The next lines of the trailer input file contain the model inputs. In the first two columns, shown in
Trailer Sample 1, the user provides a Run ID and the Regulatory Subcategory for each run. For
trailers, there are four regulatory subcategories modeled in GEM and Table 25 shows the naming
convention. While the trailer program does include reduced standards for some trailer types, trailer
manufacturers do not use GEM for compliance, and we did not configure GEM with additional
subcategories for those trailers. These four vehicles are sufficient to create the GEM-based equation
used in trailer compliance.

Trailer Sample 1: Input File Run ID and Regulatory Subcategory Inputs

Run ID

Regulatory Subcategory

Unique Identifier

(e.g. LDV)

LDV 1

LDV

LRV 1

LRV

SDV 1

SDV

SRV_1

SRV

Table 25: Trailer Input File Naming Convention for Trailer Regulatory Subcategory Inputs

GEM Input Name

Regulatory Subcategory Description

LDV

Long Dry Van

LRV

Long Refrigerated Van

SDV

Short Dry Van

SRV

Short Refrigerated Van

The next few columns contain the trailer performance parameters and technology improvement
options that are user-defined in GEM shown in Trailer Sample 2. A description of these parameters
has been provided in prior sections and additional information is available in the preamble to the
Phase 2 rulemaking. Note the aerodynamic improvement for trailers is measured as a delta CdA and
not the absolute CdA value used in the tractor program.

Trailer Sample 2: Input File Performance Parameters and User-Defined Vehicle Characteristics

Aerodynamic
Improvement (Delta)

Trailer Tire

Aerodynamic Drag
Area (CdA)

Rolling
Resistance Level

mA2

N/kN

Technology
Improvement

Weight Reduction

lbs

Technology
Improvement

Tire Pressure System

There are limits associated with the user-defined and technology improvement input values as
shown in Table 26. GEM will produce an error if the fields are blank or if any of these values are
out of the acceptable range. Users that wish to model their vehicle with weight reduction can enter
the cumulative weight reduction value (0 to 5,000 lbs or "NA"). The tire pressure system is a set
percent value for use of an automatic tire inflation system or a tire pressure monitoring system. See

-------
40 CFR 1037.515, noting the ATIS and TPMS values listed in the regulation are in decimal format
because trailers will use a GEM-based equation for compliance.

Table 26: Minimum and Maximum Limits of Technology Improvements in Trailer Input File

User-Defined Parameter

Units

Minimum Value

Maximum Value

Aerodynamic Drag Area (CdA)

Tire Rolling Resistance Level (TRRL)

N/kN

Weight Reduction

lbs

5000

Tire Pressure System

IV.D. Supplemental Input Files

Supplemental input files are required to provide the necessary component data to GEM. The GEM
installation package contains sample input files including four folders of supplemental input files as
follows:

• Axles: 1 example axle definition file

• Engines: 30 example steady-state and cycle average engine definition files

• Powertrains: 2 example powertrain definition files

• Transmissions: 11 example transmission definition files

Tractor and vocational vehicle manufacturers are required to generate separate engine and
transmission input files or a single powertrain input file for GEM. A vehicle manufacturer would
generate a separate engine and transmission file for each engine and transmission used in its
vehicles, or separate powertrain files for each engine and transmission combination tested. The axle
input file is optional for manufacturers that would like to include more axle loss information. As
discussed in the following section, each of the different types of supplemental input files consist of
various sections (tables) within each file. Examples from the sample input files are included as well.
Some of these tables and columns are required while others are optional. These files must be in .csv
format to be properly read by GEM. Each of the supplemental files consist of tables which must be
separated by an empty row to be processed correctly. Manufacturers are recommended to choose a
consistent naming convention that provides unique file names for each of these supplemental input
files. The GEM trailer model does not use these supplemental input files and instead relies on
default values built into the trailer model.

IV.D.l. Engine Input File for Tractor and Vocational Vehicles

There are two basic formats for the Engine component input file for GEM. The standard format uses
a complete fuel map of the engine which is interpolated during simulation. Due to the potential
inaccuracy of interpolating a steady state fuel map to simulate more transient operation, as occurs
during the ARB transient cycle, an additional cycle average map is required which contains results
from a set of transient tests. More details on the formatting of these tables and on the cycle average
method are presented below and described generally in section VI.B. The second file format uses
three different cycle average fuel maps, one for each test cycle. The main fuel map is reduced to
only include the low speed points which allows adjustments to be made for differences in the idle
speed of the engine tested and the vehicle being simulated.

The first line of the GEM engine input file reports the GEM version for which the input is intended.
The first section consists of various engine parameters such as Manufacturer Name, Combustion
Type, Fuel Type, Family Name, Calibration ID, and Displacement, as shown in Engine Sample 1.

-------
Combustion Type and Fuel Type have specific options for a manufacturer to choose. The user can
choose any name for the other header fields, but the names should be consistent with other
regulatory documents from the manufacturer.

Engine Sample 1: Input File Header Information

GEM P2v4.0 Engine Definition

Manufacturer
Name

Combustion Type

Fuel Type

Family Name

Calibration
ID

Displacement

(e.g. Cummins)

(Compression Ignition /
Spark Ignition)

(Diesel / Gasoline
/LNG/CNG)

(e.g. abcl2345)

(e.g. 123abc)

liters

EPA

Compression Ignition

Diesel

GENERIC

The next three sections in the engine input file specify the operating range of the engine. An
example with specific numbers removed, is provided in Engine Sample 2. The first section, titled
"Engine Full Load Torque Curve", specifies the maximum torque curve for the engine, and is used
to define the upper bound on engine torque in the simulation. Similarly, the third section, titles
"Engine Motoring Torque Curve" defines the lower bound from the motoring torque curve. The
second section provides the maximum torque curve of a parent engine, which is used to estimate the
transient torque response of the engine during simulation. The test procedures for each of these
inputs including the number of speed and torque points to measure can be found in 40 CFR
1065.510.

Engine Sample 2: Operating Range of the Engine

Engine Full Load Torque Curve

Speed

Torque

RPM

###

Parent Engine Full Load Torque Curve

Speed

Torque

RPM

###

Engine Motoring Torque Curve

Speed

Torque

RPM

###

The next two sections in the engine input file contain information on the fuel consumption as shown
in Engine Sample 3 and Engine Sample 4. The data is detailed in three columns including engine
speed in RPM, torque in Nm and fuel consumption in grams per second.

The "Engine Idle Fuel Map" is only required for vocational vehicles and provides data on fuel
consumption during extended idling, corresponding with the parked idle cycle weighting for

-------
vocational vehicles in Table 6, Table 7 and Table 8. The fuel map data should consist of the four
test points specified in 40 CFR 1036.535. The table will be interpolated based on the idle speed
entered for the tractor or vocational vehicle and torque estimated from the simulation.

Engine Sample 3: Engine Idle Fuel Map

Engine Idle Fuel Map

Speed

Torque

Fuel Rate

RPM

grams/ sec

###

The "Engine Fuel Map" can exist in two forms, either as a full engine map or a limited low speed
map if cycle average data is provided for each test cycle. The procedures to select the test points and
run the test can be found in 40 CFR 1036.535.

Engine Sample 4: Engine Fuel Map

Engine Fuel Map

Speed

Torque

Fuel Rate

RPM

grams / sec

###

The remaining sections in the engine input file include the cycle average fuel maps titled "Transient
Cycle Average Fuel Map" and "Cruise Cycle Average Fuel Map" shown in Engine Sample 5. The
section for the ARB Transient cycle is required while the others are optional if a complete engine
fuel map is provided. Each section consists of columns for engine cycle work in kWh, the ratio of
engine speed to vehicle speed (more detail on this in Section VLB), and fuel consumed over the
cycle in grams. The section for the ARB Transient cycle contains additional columns for engine
speed and load at idle. This information, along with the idle speed from the tractor of vocational
input file is used to adjust the fuel consumption calculation for differing idle conditions.

GEM features limited backward compatibility for older engine input files. Cruise cycle average fuel
maps formatted for GEM P2v3.5 which feature separate 55 mph and 65 mph cruise maps can be
used instead of the combined map. While GEM may be able to load these older maps be sure to
verify their usage for certification is allowable under the current applicable regulations.

-------
Engine Sample 5: Cycle Average Fuel Maps

Transient Cycle Average Fuel Map

Engine Cycle Work

Simulation N/V

Fuel Mass

Idle Speed

Idle Torque

kWh

rev / m

grams

RPM

###

Cruise Cycle Average Fuel Map

Engine Cycle Work

Average Speed

Average Torque

Fuel Mass

kWh

RPM

grams

###

IV.D.2. Transmission Input File for Tractor and Vocational Vehicles

The first line of the GEM transmission input file reports the GEM version number for which the file
is intended. The first section, as shown in Transmission Sample 1 contains basic information on the
transmission such as Manufacturer Name, Type, and Model Name. The vehicle manufacturer can
choose any Manufacturer Name and Model Name for the transmission, but it should remain
consistent with other regulatory documents. The transmission field Type has specific options for the
manufacturer to choose.

Transmission Sample 1: Input File Header Information

GEM P2v4.0 Transmission Definition

Manufacturer Name

Type

Model Name

(e.g. Eaton)

(AMT / MT/AT/ DCT)

(e.g. 7100)

EPA

AMT

HHD

If the transmission is an automatic transmission where the torque converter clutch lockup can occur
in a gear lower than the GEM default of 3rd gear, an additional column may be added to specify the
minimum lockup gear. The Minimum Lockup Gear input does not apply for AMT and manual
transmissions. The optional additional transmission header information for automatic transmissions
is shown in Transmission Sample 2.

Transmission Sample 2: Optional Input File Header Information for Automatic Transmissions

GEM P2v4.0 Transmission Definition

Manufacturer Name

Type

Model Name

Minimum Lockup Gear

(e.g. Eaton)

(AMT/MT/AT/DCT)

(e.g. 7100)

Number / NA

EPA

HHD

-------
The next section in the GEM transmission input file contains information on the transmission gear
ratios as shown in the sample shown in Transmission Sample 3. If engine torque is limited when
operating in a specific gear, the limit amount should be entered in the appropriate column. This will
constrain simulated engine torque when operating in the specified gear and alter the shift strategy
accordingly.

Transmission Sample 3: Transmission Gear Ratio Information

Transmission Gears

Gear Number

Gear Ratio

Input Torque Limit

The next section, titled "Transmission Power Loss" is optional and can be used to include data on
transmission losses. Instructions for obtaining this information are found in 40 CFR 1037.565. The
format of the power loss table is shown in Transmission Sample 4. When providing power loss
information not all gears need to be included. Neutral, signified by gear number zero, is optional. If
loss information is provided for a given gear, all higher gears must also be included. For example,
supplying data for only 8th gear on a 10-speed transmission would not be valid, data would need to
be provided for 8th, 9th and 10th gears. The data points for this section are transmission input speed
and torque as well as total power loss across the transmission, in kW, at each point. If power loss
data is not provided, GEM uses default loss maps which are scaled based on the regulatory
subcategory of the vehicle and type of transmission.

Transmission Sample 4: Optional Transmission Power Loss Information

Transmission Power Loss

Gear Number

Input Speed

Input Torque

Power Loss

RPM

The final section in the transmission input file, which is also optional and only available for
automatic transmissions contains data on the torque converter characteristics as shown in
Transmission Sample 5. The test procedure to generate the information can be found in 40 CFR
1037.570. This section contains columns for speed ratio, torque ratio and the calculated k-factor in
RPM per square root Nm. If this data is not provided, GEM generates torque converter properties
based on the engine torque curve.

Transmission Sample 5: Torque Converter Characteristics

Torque Converter Properties

Speed Ratio

Torque Ratio

K Factor

RPM/sqrt(Nm)

-------
IV.D.3. Optional Powertrain Input File for Tractor and Vocational Vehicles

In lieu of providing separate engine and transmission performance data, a single powertrain test
input can be used for GEM simulation. Procedures for generating the data to populate a powertrain
input file can be found in 40 CFR 1037.550. This section summarizes the input file format.

The first line of the GEM input file for powertrain contains the GEM version for which it is
intended. Next are three header sections, shown in Powertrain Sample 1, each contain a single row
of data summarizing the engine, transmission and combined powertrain tested. The engine
parameters include engine rated power which is used to scale the default powertrain during
simulation to match the engine. The powertrain test configuration contains information on whether
the measurements provided in the following sections were conducted at the transmission output or
at the wheel hubs.

Powertrain Sample 1: Input File Header Information
GEM P2v4.0 Powertrain Definition

Engine Manufacturer
Name

Combustion Type

Fuel Type

Family Name

Calibration
ID

Rated
Power

(e.g. Cummins)

(Compression Ignition /
Spark Ignition)

(Diesel / Gasoline
/ LNG / CNG)

(e.g. abcl2345)

(e.g. 123abc)

EPA

Compression Ignition

Diesel

abc

CAL1

340

Transmission
Manufacturer Name

Type

Gears

Model Name

(e.g. Eaton)

(AMT / MT / AT / DCT)

Number

(e.g. 7100)

EPA

AMT

abcl

Powertrain Family
Name

Calibration ID

Powertrain Test
Configuration

(e.g. abcdl2345efg)

(e.g. 123abc)

(1: Trans. Output,
2: Wheel Hubs

EPA

CAL1

The next section in the powertrain input file contains data on engine fuel consumption at idle as
shown in Powertrain Sample 2. For powertrains with a singular fixed idle speed this section would
contain a single row while those with a calibratable idle speed must feature multiple entries
spanning the range of options. Each entry must contain the idle speed setpoint and fuel consumption
rates for idling in gear and when parked, the latter of which is only required for vocational vehicles.
This data is used by GEM for vocational idle fuel consumption and to adjust for differences
between tested and simulated idle speeds. The procedures for collecting the necessary data can be
found in 40 CFR 1037.550(o).

Powertrain Sample 2: Engine Idle Fuel Consumption

Idle Fuel Rate

Idle Speed

Drive

Parked

RPM

grams / sec

-------
The final three sections contain the powertrain cycle fuel maps and are somewhat analogous to the
cycle average fuel maps of the engine input file. Each contain columns for work, N/V and the
measurement of fuel consumed. The data for work and N/V are relative to the point of
measurement, which was noted as Powertrain Test Configuration in the file header. The map for the
ARB transient features an additional column to report the calibrated engine idle speed used in the
test. This data is used to adjust for differences in idle speed between the tested powertrain and the
vehicle simulation.

Powertrain Sample 3: Additional Performance Information

Transient

Powertrain Cycle Work

N/V

Fuel Mass

Idle Speed

kWh

rev / meter

grams

RPM

Cruise

Powertrain Cycle Work

Average Speed

Average Torque

Fuel Mass

kWh

RPM

grams

GEM features limited backward compatibility for powertrain input files. Powertrain fuel maps
formatted for GEM P2v3.5 which feature separate 55 mph and 65 mph cruise fuel maps similar in
configuration to the transient map above can be used instead of the combined map. While GEM
may be able to load these older maps be sure to verify their usage for certification is allowable
under the current applicable regulations.

IV.D.4. Optional Axle Input File for Tractor and Vocational Vehicles

The first row of the GEM input file for axles provides the GEM version and the subsequent rows
contain user-specified details including Manufacturer Name, Family Name, and Type, as shown in
Axle Sample 1. The user can type any text in the first two fields, but the names should be consistent
with other regulatory documents from the manufacturer. The valid options for axle type are
"SINGLE", "TANDEM" and "TANDEM WITH DISCONNECT".

Axle Sample 1: Input File Header Information

GEM P2v4.0 Axle Definition

Manufacturer Name

Family Name

Type

(e.g. Dana)

(e.g. abcdl2345efg)

EPA

TANDEM WITH DISCONNECT

The next rows contain two comma separated tables for Axle Loss and Disconnect Axle Loss, as
shown in Axle Sample 2. Each section includes output speed in RPM, output torque in Nm, and
power loss in kW. See 40 CFR 1037.560 for the test procedure to map axle efficiency and
determine appropriate values for this input file's tables. The disconnect axle losses are only
included by axles of type "TANDEM WITH DISCONNECT".

-------
Axle Sample 2: Axle Loss Table in Axle Input File

Axle Loss

Output Speed

Output Torque

Power Loss

RPM

Disconnect Axle Loss

Output Speed

Output Torque

Power Loss

RPM

V. GEM Output File Structure

An output file will be generated when the vehicle simulation is complete and will automatically
save to the same location as the input file. When users run GEM from the Start Menu or command
prompt, the output file name will be given the same name as the input file with a " result.csv"
appended. During the simulation an error log file will be generated, even if there are no errors, with
the same base name as the input file with an "_errors.txt" extension. Identical output files are
generated for each of the methods.

V.A. Standard GEM Outputs for Compliance

Each output file is aligned to the input file for the simulated vehicle, with additional columns if the
.csv file populated with the model's results. Sample 2 shows the standard results of an example
simulation. These results fields are the same for each vehicle type, but their exact column location
in the file varies. The first result column indicates the date and time when the simulation was
performed. The next two columns are the raw GEM CO2 emissions and fuel consumption values.
The two columns after that contain the Family Emissions Limit (FEL) values EPA and NHTSA use
for compliance. For trailers and vocational vehicles, EPA's FEL CO2 results are reported as integer
values and for tractors, the results are reported with a single decimal place precision. NHTSA's FEL
fuel consumption results are calculated using the EPA's FEL CO2 values and are reported with four
decimal place precision for vocational vehicles and trailers, and five decimal place precision for
tractors to ensure consistency between the agencies' results. The vocational vehicles have
additional GEM and FEL C02 Emissions and Consumption result columns, highlighted below in
yellow, for each of the duty cycles, Regional, Multipurpose and Urban.

-------
Sample 2: Example Results Columns in GEM Output File

Date/Time of
Run

GEM C02
Emissions

GEM
Consumption

Default FEL

C02
Emissions

Default FEL
Consumption

Subfamily

Subfamily
FEL

Subfamily
Volume

YYYY-MM-DD
HH:MM:SS

g C02 /
ton-mile

gal / 1000
ton-mile

g C02 /
ton-mile

gal / 1000
ton-mile

Name

g C02 /
ton-mile

YYYY-MM-DD
HH:MM:SS

###

##(.#)

##(.####)

YYYY-MM-DD
HH:MM:SS

###

##(.#)

##(.####)

The final three columns of the output file are left blank by the model. Manufacturers would
manually add the appropriate subfamily name, target FEL for the given subfamily, and the volume
of vehicles that will use the resulting FEL value prior to submitting their output file for compliance.

VI. Running GEM

There are two options for running GEM. The first option directly accesses the program's executable
file via one of the Start Menu entries or a GEM icon on the desktop. This initiates a GEM pop-up
window to select an input file and run the model. The second option uses the command prompt,
which will similarly initiate GEM, but also has several options to generate additional output
information. Each of these options is described in the following sections.

VI.A. Preparing for GEM Simulation

Prior to running GEM, an important consideration is to locate the necessary input files. Many
computers will produce errors if users try to alter files in a folder where they do not have write
permissions such as the default installation folder (C:\Program FilesYUS EPAVPhase 2 GEM). See
the previous sections for a description of the input files and their formats.

VLB. Running GEM from the Start Menu Icons

Users can access Phase 2 GEM from a collection of Start Menu icons found within the EPA Phase 2
GEM folder. The "Phase 2 GEM" shortcut runs the certification version of GEM while "Phase 2
GEM Check Inputs" and "Phase 2 GEM Cycle Creation" provide other functionality discussed in
the following sections.

A pop-up window will ask the user to select an input file. By default, the program will first look in
the installation folder. As mentioned previously, all the input files should be moved to a separate
location to avoid permission warnings. In this example, the files were moved to the desktop and the
folder was renamed GEM P2 Sample Input Files. Navigate to the input files folder and select the
appropriate input file to begin the simulation.

Once the input file is selected, the program will begin to run. A status window will step through
each configuration as it runs, as shown in Figure 5. At any point in the simulation, users can stop
the model by clicking the red "X" at the top right corner of the status window. A new window will
appear that asks, "Cancel current simulation?" If the user chooses "Yes", the execution will stop
once the current simulation has completed (which may take a few seconds) and produce an output
file with only the configurations that completed.

-------
S3

1 <=> s Ms-ll

(s A A

% sSfei

YW 1 ~J/

US Environmental Protection Agency
GEM P2v3.0

Processing "C:\UsersMlser111\Desktop\GEM P2 Sample Input Files\GEM_tractor_sample_inputs.csv

Processing Run ID Sample 1 Complete!

Processing Run ID Sample 2

< I ~

15%

Figure 5: Sample Status Window Showing Progress of GEM Simulations

The status window will indicate when the simulations are complete with messages like those shown
in Figure 6. The resulting output file will be saved in the same location as the input file and given
the same name as the input file with a " result.csv" extension added to the end. When the
simulation is complete, as indicated by the green "Complete!" message, users can close the status
window by clicking the red "X" in the top right corner of the window. If an error occurs during any
simulation, the window will indicate which configurations failed and the model will continue to the
next simulation. See Section VI.C of this Guide for examples of errors and a description of the
_errors.txt file produced.

H 1^11

(i O \

Processing Run ID Sample 5

Processing Run ID Sample 6

Processing Run ID Sample 7

Processing Run ID Sample 8

.Complete!
.Complete!
.Complete!
.Complete!

Processing Run ID PT Samplel

— Batch Simulation Complete! —

Complete!

\ | 1"

< i

rrr

COMPLETE!

Figure 6: Sample Status Window Showing Complete GEM Simulations

VI.C. Testing Input Files for Errors

Prior to running a GEM simulation with a large number of runs, users are recommended to test their
input files for errors. GEM will ensure the appropriate data is included in the required input file
fields, the file headers and data are in the proper format, and supplemental input files exist with the
representative information. Note that some GEM inputs are case-sensitive and will produce an error
if inputs do not exactly match.

To test an input file, initiate the "Phase 2 GEM Check Inputs" shortcut and select an input file as
before. GEM will run through each line of the input file and search for errors. Figure 7 shows the
status window that appears when checking for errors. The status bar will progress as GEM checks
each Run ID and indicates "FAIL!" or "Input Valid!" Users can close this window and check the
output files for more detailed information.

-------
H CsD b ^

UJIIICIILaUIIMI L_H.USVJLUI\£U IU.UT. I J J.U ^IC\Jfllll^lC III^Ul I IBSML

FAIL!

FAIL! —

FAIL!

FAIL! E

FAIL!

...Input Valid!

Hf ~| ~

65% WITH 5 ERRORS!

Figure 7: Status Window Checking a GEM Input File for Errors

Once an input test ends (by completing the check or after being terminated by the user), the two
standard result.csv and _errors.txt files are created. The results file will simply display if the inputs
were valid or produced an error, as shown in Sample 3. Users are instructed to see the error file for
details.

Sample 3: Example Results.csv Output File Generated when Testing GEM Inputs

Regulatory Category
Manufacturer Name
Model Year
GEM Version
Run ID

Unique Identifier

Sample_l -- ERROR: see error file for details!

Sample_2 - ERROR: see error file for details!

Sample_3 -- ERROR: see error file for details!

Sample_4 - ERROR: see error file for details!

Sample_5 -- Input Validated
Sample_6 - ERROR: see error file for details!

Sample_7 -- ERROR: see error file for details!

Sample_8 - ERROR: see error file for details!

PT_Samplel -- ERROR: see error file for details!

Figure 8 shows the error output file from this example which provides details about each error.
Error messages describe the powertrain area with the error, file location, line item and text
description.

Processing Run ID Sample_1 ..
Processing Run ID Sample_2 ..
Processing Run ID Sample_3 ..
Processing Run ID Sample_4 ..
Processing Run ID Sample_5 ..
Processing Run ID Sample_6 ..
Processinq Run ID Sample 7 .
'

-------
US EPA Phase 2 GEM P2v4.0 Error Log

Sample_l -- Invalid Engine Definition File EPA_2018_D_GENERIC_350_trans_cyc_avg.csv Line 2131 Incorrect Transient....
Sample_2 -- Invalid Engine Definition File EPA_2018_D_GENERIC_350_trans_cyc_avg.csv Line 2131 Incorrect Transient....
Sample_3 -- Invalid Technology Improvement Automatic Engine Shutdown Value "F" not recognized must be "Y" or "N"
Sample_3 -- Invalid Axle Definition File EPA_AxleB.csv- Unable to open file
Sample_4- Unknown Regulatory Sub Category "LHQ" see Documentation for Valid Options

Sample_6 -- Invalid Engine Definition File EPA_2018_G_GENERIC_300hp_trans_cyc_avg.csv Line 2 Incorrect Data Field ....

Sample_6 -- Invalid Engine Definition File EPA_2018_G_GENERIC_300hp_trans_cyc_avg.csv Line 2403 Incorrect Data....

Sample_6 -- Invalid Engine Definition File \EPA_2018_G_GENERIC_300hp_trans_cyc_avg.csv Incorrect number of Engine....

Sample_7 -- Invalid Technology Improvement Start-Stop Value "T" not recognized must be "Y" or "N"

Sample_8 - Invalid Drive Axle 2 Tire Rolling Resistance Level should be NA for Axle configuration 4x2

PT_Samplel -- Invalid Powertrain Definition File EPA_Sample_Powertrainl.csv Line 22 Incorrect Transient Data Field....

Figure 8: Example _Errors.txt Output File Generated when Testing GEM Inputs

Users should correct any errors in their files and rerun the test until they receive a confirmation that
all the inputs are valid. Once completed, the input files are now ready to be used in a GEM
simulation.

VI.D. Cycle Average Engine Fuel Map for Tractor and Vocational Vehicles

Tractor and vocational vehicle manufacturers must include a cycle average fuel map for the
transient cycle within the engine input file and can optionally apply cycle average fuel maps to the
55 and 65 mph cruise cycles. The cycle average method involves using GEM to generate a
collection of engine dynamometer test cycles based upon a collection of vehicle simulations. These
simulated vehicles are intended to span the range of potential vehicle applications for the engine.
The engine is run over each of the cycles on the dynamometer and summarized in the cycle average
map by work, the average speed (represented via N/V) and fuel consumed. Testing over the ARB
transient also requires average speed and load at idle to be recorded. As discussed in Section IV.C. 1
the cycle average map is interpolated to determine fuel consumer over the test cycle.

Included in this GEM installation package are two sample vehicle input files that serve as an
example of the inputs needed to generate the dynamometer cycles the cycle average map:

1. "GEMtractorEnginePrepinputs.csv" for tractor vehicles, and

2. "GEMvocationalEnginePrepinputs.csv" for vocational vehicles

These input files rely on different engine input files located in the sample subfolder "Engines" and
transmission input files located in the subfolder "Transmissions". Note the sample axle file is not
used when creating the cycle average map.

To generate cycles for the cycle average map, six, eight, or nine vehicle configurations are
simulated in GEM depending on the vehicle class(es) that may use the engine as defined in40 CFR
1036.540. Engines used in heavy-haul tractors are evaluated over six specific heavy-haul and
sleeper cab configurations. Engines installed in vocational vehicles qualifying as Light HDV or
Medium HDV are evaluated over eight light- and medium-heavy duty configurations. Engines
installed in vocational vehicles qualifying as Heavy HDV and for tractors that are not heavy-haul
tractors are evaluated over nine vehicle configurations. The example EnginePrep files show some of
the default values that may be applied. The method of selecting appropriate axle ratios, tire sizes
and other parameters to construct the cycle generation configurations intended to cover the expected

-------
range of potential vehicles where the engine will be used is defined in 40 CFR 1036.540. Note the
configurations may vary by test cycle therefore separate simulations may be required.

As mentioned in Section IV.C.l the engine input file exists in two formats, a full steady-state engine
map and cycle average map for the ARB transient or a cycle average map for all three test cycles.
The engine input file for cycle generation similarly varies depending on the expected format. If
creating a final engine input using the steady-state map, the same steady-state map should also be
included for cycle generation. If the final engine input uses the three-cycle average map, the steady-
state map should be omitted.

The EnginePrep input files are used with the GEM executable (i.e., the "Phase 2 GEM Cycle
Creation" executable from the Start Menu). Each simulation line in the input file will generate three
engine cycle files in .csv format corresponding to the transient, 55-mph, and 65-mph drive cycles.
These output files are named with the text of the unique identifier of the simulation followed by the
drive cycle (e.g., "EngineXYZ_cyclel_transient_cycle.csv"). The .csv files include the average
vehicle speed over the drive cycle and three columns defining the engine cycle: Time (sec), Engine
Speed (RPM) and Engine Torque (Nm). A sample of a generated cycle file is shown in Sample 4.

Sample 4: Generated Test Cycle File

GEM P2v4.0 Engine Test Cycle

Configuration ID:2018_Engine350_all_cyc_prep_cyclel

Simulation Average Vehicle Speed: #.### m/sec

Simulation Crankshaft Work: ##.### kWh

Time

Engine Speed

Engine Torque

Vehicle Moving

sec

RPM

BOOL

The engine cycle files are then used to test the actual engine according to 40 CFR 1065 to generate
the following results that will be added to the end of the engine input file with the appropriate
headers as described previously in Section IV.C.l:

1) N/V Ratio (rev/meter) calculated from time average engine speed during engine test divided by
average vehicle speed from the generated cycle. Note the average vehicle speed is determined
by GEM and is stored in the GEM output file containing the engine cycle.

2) Positive cycle work (kWh) calculated from the engine test.

3) Total fuel mass for the test corrected for mass-specific net energy content of the fuel, according
to 40CFR 1036.540.

Two additional parameters are included for the ARB transient to enable correction of fuel
consumption at idle, which could vary based upon different transmissions or idle speed calibrations.

4) Average speed at idle where idle is identified by the flag in the generated cycle output file.

5) Average torque at idle where idle is identified by the flag in the generated cycle output file.

VI.E. Running GEM from the Command Prompt and Advanced Options

Users can also operate GEM using the command prompt, calling the executable directly where
additional functionality is available beyond what is provided in the start menu shortcuts. The default
installation directory for the GEM executable is "C:\Program Files\USEPA\Phase 2 GEM4.0.

-------
Users can then initiate the same "Select an Input File" pop-up window as when using the start menu
links by executing GEM.exe with no additional arguments.

In this example, the executable is stored in the default Program Files location and the input files are
located in a folder on the desktop. The program will directly initiate a status window similar to the
one shown in Figure 9.

SS I CJ II B llrf^l

Figure 9: Command Prompt to Initiate GEM and Directly Apply Input Files (no File Selection Window)

The different GEM operating modes are accessible by specifying the mode as the first argument
(e.g. GEM.exe cycle-generation). The modes are listed along with a summary of their operation in
Table 27. The 'cycle-generation' mode is used to generate the cycle average maps (Section VLB)
and the 'check-inputs' mode is used to test the input files (Section VI.C).

Table 27: GEM Operating Modes

Mode Command

Description

certification

Simulate vehicle parameters for certification [DEFAULT]

cycle-generation

Export the engine speed and load data from each simulation of each phase to a separate
file for use in cycle average testing (see Section VLB)

check-inputs

Test the input file for errors such as syntax and out of range parameters, but do not
simulate (see Section VI. C)

simulink-parameters

Export the model parameters for later running Simulink version (see Appendix B)

hil-parameters

Export model parameters for the hardware in the loop Simulink model (see GEM HIL
Model)

As mentioned previously, the command prompt method of running GEM has several options that
can alter the output format or generate additional data. These options are included after the optional
mode specifier as single-letter flags with a leading dash and users may include more than one flag in
a single command (e.g., GEM.exe -d-s or GEM.exe cycle-generation -c). The available flags are
shown in Table 28.

icrosoft Windows [Uersion 6.1.7601]

:\Users\UserName>cd C:\Program FilesSUS EPfi\Phase 2 GEM

I:\Program FilesSUS EPfiSPhase 2 GEM>GEM.exe "C:\UsersSUserName\DesktopSGEM P2
iple Input Files\GEM_tractor_sanple_inputs.csu"

i:\Program FilesSUS EPfiSPhase 2 GEM>

-------
Table 28: Command Line Options for GEM

Flag

Option Name

Description

-vN

Model output
verbosity

N = 0 (default) logs only the data required for certification
N = 1 turns on the model's energy auditing
N = 2 or 3 add additional signals of interest for debugging

-d

Detailed output

Generates additional cycle-specific data columns in the results.csv file

-P

Preserve files

Preserve the simulation raw output as a .mat file for later examination

-n

No tech improvements

Turns off all technology improvements and result modifiers

-c

Console Only

Disables the status display window and runs in a console mode

-s

Stringency mode

Bypasses requirement of cycle average map for ARB transient cycle and
rounding of input parameters is disabled

B

icrosoft Windows [Uersion 6.1.7601]
opyriglit (c) 2009 Microsoft Corporation. All rights reserved.

C:SUsers\UserNane>cd C:NProgram FilesNUS EPANPhase 2 GEM

C:\Progran FilesNUS EPANPhase 2 GEM>GEM.exe "C:\Users\UserNane\DesktopSGEM P2 Sa
tuple Input FilesNGEM_tractor_sample_inputs.csu" —c

C:\Progran FilesNUS EPANPhase 2 GEM>US Environmental Protection Agency
GEM P2u3.0

Processing "C:NUsersNUserNameNDesktopNGEM P2 Sample Input FilesNGEM_tractor_sanp
le_inputs.csy"

rocessing Run ID Sample_l 			Complete?

rocessing Run ID Sample_2 	Complete?

rocessing Run ID Sample_3 	Complete?

rocessing Run ID Sample_4 	

Figure 10: Command Prompt Display When Using the Console Only Option

To specify a GEM input file, which bypasses the file selection dialog, specify the path to the file as
the last argument (e.g. GEM.exe "pathJoMnpiitJile.csv" or GEM.exe -v 2

"path to\anotherJile.csv"). If the file or path contains spaces, be sure to enclose the file in quotes.
VII. Final Notes

Users are encouraged to look through the additional information provided in the Documentation
folder included with the GEM installation. For more information on the Phase 2 rule or the
Technical Amendments, please see Docket EPA-HQ-OAR-2014-0827 available at
www.regulations.gov.

42

-------
Appendix A. GEM HIL Model

Included with GEM is a Simulink model which can be used in conjunction with the procedures in
1037.550 for simulation during powertrain testing. There are three powertrain testing scenarios, and
their usage with GEM is summarized in Table 29 below.

Table 29: Powertrain Testing Scenarios

Components Tested

Resulting Input Format

Model Used

Engine & Transmission

Powertrain (see Section IV.C.3)

Equations in 1037.550 or
GEM HIL Driveline & Vehicle

Engine, Transmission & Axle

Engine with Hybrid System

Engine (see Section IV. C.l)

GEM HIL Transmission and
Equations in 1037.550 or
GEM HIL Driveline & Vehicle

Model Description

The GEM HIL model, shown in Figure 11 below, consists of multiple sub-models that are used to
meet these different scenarios. The model as provided, targets the most complicated scenario of a
hybrid engine. Testing a hybrid engine requires the use of the transmission sub-model (marked in
blue) along with the driveline and vehicle sub-models (marked in green), which can be substituted
with the equations found in 1037.550. The remaining sub-models can be altered or replaced as
necessary to implement the testing procedures. A short explanation of the content in each sub-model
is provided in Table 30.

GEM Model for HIL









-*





i r

9

Figure 11: GEM HIL Model

43

-------
Table 30: GEM HIL Sub-model Descriptions

Sub-model

Description

transmission

Physical model and gear selection logic for automatic and AMT transmissions. Physical
model mirrors GEM certification model, but the shift routine is updated to enable an
improved interface for real engine hardware versus what is simulated in GEM.

driveline

Physical model for axle(s), brakes and tires, including axle losses and rolling resistance
losses. Model mirrors GEM certification model. When used for HIL testing the wheel
slip model used in GEM certification is disabled.

vehicle

Physical model for vehicle mass and aerodynamic drag. Model mirrors GEM
certification model. Also includes integrator for vehicle speed.

HIL engine

Interface to convert engine dynamometer measurements to transmission model inputs
including calculating engine statistics used within the model. Also includes logic for
handling dynamometer operating in torque mode, which drives modeled engine speed to
match measured speed. The values in the hil data structure can be adjusted to tune this
behavior.

HIL accessories

Lookup tables versus time to adjust engine load as necessary. For the default
implementation these values are set to zero.

HIL ambient

Information on ambient conditions available for use by the physical models.

HIL cycle

Lookup tables to interpolate the drive cycle including current phase, grade and whether
the transmission should be in gear. Also includes the distance compensation logic, which
can be disabled. The definition for the drive cycle is contained in the dnive_cycle data
structure.

HIL driver

A base PID control with additional logic to drive vehicle speed to match selected cycle.
Also includes logic to split output between accelerator pedal (to engine) and brake pedal
(to brakes model). The calibration for the driver can be tuned via the d r i ve r data
structure to adjust behavior as needed.

HIL signal routing
&

HIL signal routing 2

Translations to convert model signals to / from a more comprehensible format for
interface with test cell systems. Includes conditions to enable adjustments to the load
applied at idle, which can be activated via the hil data structure.

HIL interface

Primary point to interface a test cell system to the model. A more detailed description of
the various signals is available in Tables Table 3 ITable 32.

The HIL Interface block is intended to provide a common point where signals can be sent to and
received from the model. Contained in Table 31 is a list of signals provided from the model. The
signals are divided into three groups, powertrain model signals (those that correspond to
dynamometer commands), drive cycle signals and driver signals. The latter two correspond to
functionality that can be implemented separately if desired. Table 32 contains the list of signals
input to the model such as dynamometer measurements and driver inputs (which may be passed
through if using the provided drive cycle and driver sub-models).

Table 31: GEM HIL Interface Model Outputs

Signal

Description

-------
dynoidleloadNm

Load to apply during idle conditions when the dynamometer is operating in a
torque control mode.

dynotgtspeedrpm

Dynamometer target speed during normal running operation.

transdrivelineengagedbool

Status flag showing whether the simulated driveline is engaged. The signal will
be false when the transmission is in neutral during shifts.

transcurrentgear

Status signal showing the current transmission gear.

trans_eng_load_limit_Nm

Torque limit signal from transmission logic to engine. When running this limit
may correspond to a limit on torque for the current gear. During shifts
(particularly for AMTs) the limit is used to cut engine torque enabling the shift
sequence to begin.

transengspeedcmdrpm

Speed control signal from transmission logic to the engine which is used by the
AMT transmission to coordinate shifts. Downshifts require the engine speed
achieve this target in order to complete the shift.

trans_eng_torque_cmd_Nm

Alternate implementation of the transmission speed target which uses a PID
within the model to generate a supplemental torque target to be included in the
command to the engine.

vehiclespeedmps

Status signal showing the current modeled vehicle speed.

cycle_pos_secs

Drive cycle position signal used as the input to the various interpolation blocks.
When distance compensation is disabled should be equivalent to time.

cyclekeyon

Flag from drive cycle lookup for ignition to be on, included for completeness.

cycleingear

Flag from drive cycle lookup to instruct transmission to be put into drive.

cyclespdmps

Current drive cycle target vehicle speed.

cyclelookaheadspdmps

Drive cycle target speed at a point a few seconds into the future used to
anticipate speed changes.

cycle_lookahead_accel_mps2

Average acceleration anticipated over the next few seconds of the drive cycle,
which is useful to estimate the power required for vehicle acceleration.

cycle_grade_pct

Current drive cycle interpolated grade, which can be either based on distance
(integrated vehicle speed) or time.

gemdriveraccelnorm

Accelerator pedal command from driver model for command to engine.

gemdriverbrakenorm

Brake pedal command from driver model to command brakes within model.

Table 32: GEM HIL Interface Model Inputs

Signal

Description

dynomode

Signal of the currently commanded dyno mode:

0	indicates the dynamometer is being commanded to a given torque

1	indicates the dynamometer is being controlled to match a target speed.

45

-------

More detail on how the model operates in these different modes is in the
subsequent paragraphs.

dynotorqueNm

Dynamometer measured torque signal.

dynospeedrpm

Dynamometer measured speed signal.

engloadestNm

Signal from the engine controller indicating the current engine crankshaft torque
output. This is used by the AMT shift logic to delay the shift sequence until the
torque has been reduced to near zero.

driverkeyon

Boolean input comparable to engine on; could use provided drive cycle lookups,
external signal or always true.

driveringear

Boolean input to indicate the driver is shifting the transmission into drive, could
use provided drive cycle lookups or external implementation.

driveraccelanticipatebool

Boolean input to indicate a stopped drive cycle phase is about to end, triggering
transmission gear engagement if neutral idle is active or beginning the clutch
engagement process for an AMT.

driveraccelnorm

Accelerator pedal signal used to estimate driver demand within gear selection
and torque converter lockup logic.

driverbrakenorm

Brake pedal signal used to command brakes within vehicle model in addition to
clutch logic.

cycle_grade_pct

Grade to implement within the vehicle model, could be from provided cycle
lookup or externally.

The dynamometer mode signal is important for keeping the GEM HIL model operating properly
when running at idle conditions with the dyno controlling torque. Normal model operation receives
torque measurements which drive changes in vehicle speed and thus engine speed (dynamometer
speed setpoint). At idle conditions this process flow needs to be altered. The model provides an
output for the torque to apply, however this could result in a deviation between the dynamometer
speed and the target speed generated by the model as there is no longer feedback coordinating the
speeds. To correct for this, a PID is included within the HIL engine sub-model to drive the modeled
speed to match the dyno measurement. The dynomode flag set to false activates this functionality.
The calibration for the PID can be adjusted to ensure proper behavior.

-------
Using the GEM HIL Model

The first step in making use of the GEMHIL model for powertrain or hybrid engine testing
involves integrating the provided model or selected sub-models with the test control hardware. As
outlined in 1037.550 the test cell dynamometer must be configured to receive control setpoints from
the simulation and the appropriate measurements and signals provided to the model. The
descriptions in Table 31 and Table 32 provide information on these signals and their use. Given the
variations in different testing hardware and associated systems a detailed description of such
integration is outside the scope of this document.

The next step is generating the parameters for use by the model, which is done via the GEM
executable in hil-parameters mode as briefly discussed in Section VI.E.. The command line syntax
for hil-parameters mode with an input tractor or vocational input file named inputfile.csv would
be:

gem.exe hil-parameters input file.csv

In hil-parameters mode, each row in the input file is processed and a MATLAB mat file containing
the HIL model parameters is created. The data within the mat file is similar to the parameters used
for the certification GEM simulation with adjustments to match the GEM HIL model
parametrization and values altered from their GEM defaults to match the applicable test procedures.
Examples of these differences would be revising the transmission losses or disabling the tire slip
model. One parameter of the model not included within the generated mat file is the drive cycle. A
separate mat file 'GEM hil drive cycle.mat' containing a sample drive cycle structure is included
with the GEM package for reference.

As noted earlier, there is room for customizing the integration of these models within the test
environment. A specific example could be alternate implementations of the driver. Similarly, given
the variety of test automation systems and vendors it is difficult to provide a full example of
implementation.

-------
Appendix B. GEM Simulink Model

The GEM software used for certification or cycle average test generation is built from precompiled
models which simulate vehicle operation. To better document the content of these models the
original Simulink source code is bundled with the GEM install package and a simulink-parameters
mode is included to generate the necessary model inputs. Usage of the Simulink model requires
MATLAB/Simulink and the stateflow package.

Model Description

This document is intended to provide a rough overview of the GEM model and its components. The
goal is to provide documentation on where to look for certain functionality but not necessarily be a
complete explanation of the model or its subsystems. The top level of the GEM model is shown in
Figure 12 which illustrates the sub-models for the powertrain, vehicle (chassis), driver, ambient
conditions and routing of signal buss helping the various parts of the model communicate. Within
the powertrain model, as shown in Figure 13, are the component models for various powertrain
elements explained in more detail later in this section. The vehicle sub-model contains items
corresponding to the physical chassis of the vehicle, notably the mass of the vehicle and the
aerodynamic drag losses. The driver model contains both the drive cycle to simulate and logic to
operate the virtual pedals. The ambient sub-model provides information about external parameters
such as temperature and grade for use during simulation.

GEM Vehicle Model

Figure 12: GEM Vehicle Model Top Level Diagram

-------
GEM_CVM Conventional Vehicle Powertrain

Figure 13: GEM Vehicle Model Powertrain Sub-Model Diagram

The powertrain sub-model provides all the functionality converting the driver request via the pedals
to force output at the wheels. The CVM control block houses the start-stop logic to shut off the
engine when appropriate in simulation. The GEM engine and accessories sub-models take the driver
requested accelerator signals and provides an engine torque output in addition to fuel consumption.
There is some modeling for transient behavior along with specific logic for idle control and
coordination during transmission shifting. Some mechanical constant power accessories are also
included; the load associated with these is determined by regulatory subcategory.

The electric model contains additional accessory losses intended to replicate the alternator and
charging system. A starter is also present to enable simulation of start-stop operation.

The transmission model contains two sub-models, one for an automatic transmission with a torque
converter and one for an AMT, which requires coordination with the engine for proper shift
behavior. Within GEM, manual transmission vehicles are simulated using the AMT model but are
adjusted post-process to slightly reduce their efficiency. Both transmission architectures use the
same gear selection logic ALPHAshift which seeks to minimize fuel consumption by selecting an
appropriate gear.

The driveline block contains the components such as the driveshaft, axles, brakes and tires, with
their associated inertias and losses. To simplify simulation and improve calculation speed, the
different axles and tires are lumped into one single equivalent axle.

-------
Using GEM Simulink Mode

To generate the GEM Simulink model input parameters a GEM tractor or vocational input file is
passed into the GEM executable (or MATLAB source code if preferred). The -s (stringency mode)
flag is desirable to avoid the need for cycle average maps in engine input files.

Calling from the command line:

GEM.exe simulink-parameters -s "sample_inputs\GEM_tractor_inputs.csv"

Calling from MATLAB with GEM source code:

» GEM_main('simulink-parameters', '-s', 'sample_inputs\GEM_tractor_inputs.csv');

One of the resulting mat files can be opened in MATLAB:

»load sample_inputs\run_*******_inputs.mat

When in the Simulink folder provided with the GEM release, the sim_gem_phase2 script can run
the simulation and postprocessing:

» sim_gem_phase2

This will result in a variety of variables created within the workspace such as audit, datalog, result
and postproc result which contain continuous and phase data from the simulation for analysis.
Contained within the script is the loading of the drive cycle, if desiring to simulate different cycles
the drive cycle variable can be altered or replaced. Similarly, various component model or
parameters can be replaced or adjusted for educational purposes. A slightly more detailed version of
this information can be found in the README_RunningGEM.txt document bundled with the
source code.

50

-------
Appendix C. Changelog

Summary of Changes for 3.4 from 3.0

•	Fixed typo in Class 7 default tractor transmission torque limits

•	Revised engine idle control

o Reduced gains for improved stability
o Added feed forward for a smoother transition to idle

•	Improved handling of larger input tables

o Engine torque curves previously could exceed size allowed by model executable
o Tables are now pre-interpolated to fit within allowable dimensions

•	Revised engine cycle generation outputs

o Corrected engine cycle generation torque output from model based on simulated

inertia and rate limited speed target
o Added output of simulated crankshaft work to cycle header

•	Added scaling of powertrain simulation default engine & transmission maps based on new
rated power input

•	Recalibrated driver over-speed allowance on cruise cycles from 3 mph to 2.5 mph

•	Increased weight reduction input range for tractors up to 25k lbs

•	Changed fuel map interpolation to be consistent between simulation and post processing

•	Moved idle speed parameter from engine file to the main (vehicle) file

•	Added adjustment to cycle average fuel for differences between tested and simulated idle

o Added inputs in transient cycle average map for speed and load when stopped
o Regression lookup of cycle average map uses only portions with vehicle moving
o Post process adjustment to fuel consumption based on simulated idle speed

•	Fixed bug where parked fuel map was used for ARB transient idle correction

o Parked fuel map is now only required for vocational vehicles

•	Adjusted accessory work correction for powertrain inputs to exclude work that is provided
by the inertia of the vehicle

•	Added ability to print detailed output (-d) when using engine cycle generation (-e)

•	Detailed output columns rearranged

•	ARB Transient detailed output columns for engine work, average speed and load only use
portions with vehicle moving only

•	Added vocational output columns with result of each cycle weighting (where applicable)

•	Added detailed output columns for idle speed and load for transient simulation

•	Added torque converter input table featuring speed ratio, torque ratio and k-factor to
transmission input file

•	Added check that engine fuel map contains a sufficient number of points

Changes for 3.5

•	Increased tractor weight reduction limit to 45000 lbs

•	Modified stringency mode (-s) to disable input range checking

•	Improved component input file parsing for better error handling

Changes for 3.6

•	Fixed issue with parsing torque converter table when transmission power loss not present

•	Added ability to turn off wheel slip in compiled model

51

-------
•	Fixed issue with default engine map for vocational vehicles using 3 cycle average maps

•	Fixed incorrect cycle weighting factor for HHD vocational vehicles

Changes for 3.7

•	Fixed idle fuel adjustment to properly account for time at idle

•	Adjusted method for blending idle map and default engine with 3 cycle average maps

•	Changed gravitational acceleration to match CFR

•	Updated model and subsystems to enable HIL model operation

o Revised command line interface with different modes
o Revised model, cleaning up libraries to be common with HIL model
o Added output mode where simulation input workspace is tweaked for HIL or

Simulink mode and dumped to a MAT file
o Significantly updated AMT shift routine (only HIL version of model for now)

•	Added Intelligent controls as a vocational input

•	Fix very old bug where so HHD, C8 & C7 AMT transmissions use clutched upshifts
Changes for 3.8

•	Include GEM HIL model for release

•	Revised HIL pre-process scripts to match model revisions and test procedures
Changes for 3.9

•	Migrated runtime to MATLAB MCR 9.8 (2020a)

•	Added adjustment factors into GEM software to normalize results back to GEM P2 v3.0

o Based on regulatory subcategory, model year, combustion type and transmission
o Outputs labeled GEM are unadjusted, FEL are adjusted and rounded

•	Added 5% allowance for idle speed vs idle fuel map speed range

•	Modified input file version checking to allow component input files for prior GEM versions
if they contain the necessary inputs

•	Removed fuel map blending for cycle average engine inputs, use default map for simulation
and idle map for post processing

•	Update engine idle speed control logic for improved response

•	Remove accessory work from powertrain post processing as it is now expected to be applied
during testing

•	Revised carbon mass fraction for E85 to match 40 CFR 1036.530

•	Modified conversion from C02 to fuel gallons to be based on engine combustion type

Changes for 4.0

•	Correct Adjustment factor for C8 SC HR with automatic transmission

•	Fix bug with printing of header in cycle generation outputs

•	Enable backward compatibility with P2v3.5 style cruise cycle average engine fuel maps

•	Enable backward compatibility with P2v3.5 style cruise fuel maps for powertrain inputs

•	Added return value to monitor for failed simulation runs

•	HIL model updates to improve robustness and ability to tune AMT shift behavior

52

-------