Leonardo Electronic Journal of Practices and Technologies Issue 15, July-December 2009

ISSN 1583-1078 p. 19-30

Semi-Empiric Algorithm for Assessment of the Vehicle Mobility

Ticusor CIOBOTARU

Military Technical Academy, 81-83 George Cosbuc, Sector 5, Bucharest, Romania.

E-mail: cticusor2004@yahoo.com

Abstract
The mobility of military vehicles plays a key role in operation. The ability to
reach the desired area in war theatre represents the most important condition
for a successful accomplishment of the mission for military vehicles. The off-
road vehicles face a broad spectrum of terrains to cross. These terrains differ
by geometry and the soil characteristics.
NATO References Mobility Model (NRMM) software is based on empirical
relationship between the terrain characteristics, running conditions and
vehicles design. The paper presents the main results of a comparative mobility
analysis for M1 and HMMWV vehicles obtained using NRMM.
Keywords
Mobility; Soft Soils; Traction; Vehicle.

Introduction

There are numerous scientific attempts to describe the traction of vehicles on soft soils
in direct correlation with the soils characteristics.
A large amount of papers have developed analytical approaches based on Bekker’s
equations [1]. The last developments of the analytical models deals with the detailed study of
the design influence on traction characteristics [2, 3].
The complexity of the phenomena occurring within the terrain - vehicle interface
leaved enough room for the development of empirical methods for evaluation of the vehicle

19
http://lejpt.academicdirect.org
 Semi-empiric algorithm for assessment of the vehicle mobility
Ticusor CIOBOTARU

mobility. The first successful attempts includes the work of US Army specialized laboratories
[4]. Considerable efforts were made in order to set up empirical relations who cover the
current design diversity of the vehicle; finally, the mathematical relations were implemented
into NATO Reference Mobility Modelling (NRMM) software [5, 6].
NRMM is a large scale digital simulation, which predicts the on-road and cross-
country performance of a vehicle in a global sense. The measure of performance is speed-
made-good and percent of an area denied due to immobilization.
NRMM is used in the development of requirements, cost-performance analyses, as an
evaluation tool, in support of contract selections and to support other mobility studies.
NRMM is an engineering model (constructive type) which marries the engineering
characterization of the vehicle (power train, running gear, suspension, mass and inertial
properties and geometry) with the engineering characterization of the terrain (slopes, soil
strength, obstacle geometry, surface roughness, and vegetation density). It is an analytical
simulation model which predicts the vehicle's maximum attainable speed cross-country and
on-road for a specified geographic region and environmental condition (wet season, rainfall,
snow), and defines within the geographic region areas of immobilization and the reasons for
such situation. Meanwhile, this model is based on empiric relations experimentally validated
for al large variety of vehicles in terms of size, weight and running gear type.

The Mobility Index of the vehicle

The Mobility Index, (MI), represents a parameter that is related to the Vehicle Cone
Index (VCI) performance of the vehicles running on fine-grained soils. It is composed of
many traction parameters influencing vehicle characteristics. [6].
The calculation of Mobility Index (MI) is performed using design data specific to each
vehicle by the following formulae [5]:

⎛ CPF ⋅ WF ⎞ (1)
MI = ⎜ + WLF + CF ⎟ ⋅ EF ⋅ TF
⎝ TEF ⋅ GF ⎠
where: CPF – contact pressure factor:
• for tracked assemblies:

w (2)
CPF =
n ⋅l⋅b
• for wheeled assemblies:

20
 Leonardo Electronic Journal of Practices and Technologies Issue 15, July-December 2009
ISSN 1583-1078 p. 19-30

w (2)
CPF =
0.5 ⋅ n ⋅ d ⋅ b
• for wheeled assemblies:

w (3)
CPF =
0.5 ⋅ n ⋅ d ⋅ b
TEF = traction element factor done by:
• for tracked assemblies:

b (4)
TEF =
100
• for wheeled assemblies:

100 + b (5)
TEF =
100
WLF – wheel load factor:
• for tracked assemblies:

w (6)
WLF =
10 ⋅ Nr ⋅ As
• for wheeled assemblies:

w (7)
WLF =
2000
WF – weight factor, GF – grouser factor, EF – engine factor, TF – transmission factor,
adopted from tables [5].
For the above relations, the following nomenclature applies: w = weight beneath
assembly i, lbf; n = total number of elements (tracks) on assembly i; or total number of
elements (tires) on assembly i; l = length of single element (track) in contact with ground, in.;
b= width of single element (track) in contact with ground, in.; or single tire section width
(undeflected and highway inflation pressure), in.; g = grouser height, in.; Nr = total number of
road wheels on assembly i in contact with ground; As = area of one track shoe, in.; hc =
minimum ground clearance of assembly i, in.; PWR = engine gross power to weight ratio,
hp/ton; d = single tire outside diameter (undeflected and highway inflation pressure), in.; h =
single tire section height (undeflected and highway inflation pressure), in.; δ = single tire hard
surface deflection for scenario j, in.;

21
 Semi-empiric algorithm for assessment of the vehicle mobility
Ticusor CIOBOTARU

Soil characterization using the Cone Index

The Cone Index, (CI) characterises the in situ shear strength of a soil. It is obtained by
the measurements performed using a specific device, the cone penetrometer. In fact, the Cone
Index represents the force applied to a cone penetrometer with 0.5 sq in base area and 30 deg
angle to penetrate 150 mm into the upper layer of the soil. The cone penetrometer realized by
Military Technical Academy uses a strain gauge sensor for force measurement, the
displacement being measured with a string potentiometer, as it is shown in Figure.

Figure 1. The cone penetrometer

Usually, several wheel assemblies act successively on the same soil area, so the
repetitive loading has to be taken into consideration. The behaviour of the soil under
repetitive loadings is described by the Rated Cone Index (RCI) which is measured using the
same cone penetrometer but the soil is compacted by applying a defined number of blows.
The value of the rated cone index may be estimated using the following formula:
RCI = ea+bln(u) (8)
where a and b represents coefficients extracted from tables, and depend of the type of soils
and its humidity u, in %.

22
 Leonardo Electronic Journal of Practices and Technologies Issue 15, July-December 2009
ISSN 1583-1078 p. 19-30

For the soils described in the Unified Soil Classification System [6], the variation of
the Rated Cone Index with the humidity was computed, the results being presented in Figure
2. The curves indicate a rapid decrease of the Rated Cone Index with the increase of the
humidity for all types of soils.
RCI
700

600 SM, SC, SM-

SC
CL
500
ML
400
CL, ML
300 CH

200 MH

100

0
10 20 30 40 50 60 70 80 90
humidity (%)
Figure 2. Variation of the Rated Cone Index versus soil humidity

Vehicle Cone Index

The Vehicle Cone Index, (VCI) represents the minimum soil strength in the critical
layer, in terms of rating cone index for fine-grained soils or in cone index for coarse-grained
soils and muskeg, required for a specific number of passes of a vehicle, such as one pass
(VCI1), or (usually) 50 passes (VCI50).
The calculation of the Vehicle Cone Index for a single pass (VCI1) is done by the
following empirical formulae:
• for tracked vehicles:

⎛ 39.2 ⎞ (9)
VCI1 = ⎜ 7.0 + 0.2MI − ⎟
⎝ MI + 5.6 ⎠
• for wheeled assemblies with MI<100 psi:

⎛ 39.2 ⎞ ⎛ 0.15h ⎞
0.25 (10)
VCI1 = ⎜11.48 + 0.2MI − ⎟⋅⎜ ⎟
⎝ MI + 3.74 ⎠ ⎝ δ ⎠

23
 Semi-empiric algorithm for assessment of the vehicle mobility
Ticusor CIOBOTARU

• for powered wheeled assemblies with MI>100 psi:

⎛ 0.15h ⎞
0.25 (11)
VCI1 = 4.1MI ⋅⎜
0.446
⎟
⎝ δ ⎠
For assessment of mobility of vehicles used according with the scenario which
provides large units displacement, the trafficability is better described by the Vehicle Cone
Index for 50 passes, which takes into consideration the degradation of the upper layer of the
soil do to the passes of the previous vehicles. The formulae used in this situation are the
following 1:
• for tracked vehicles:

⎛ 125.79 ⎞ (12)
VCI50 = ⎜19.27 + 0.43MI − ⎟
⎝ MI + 7.08 ⎠
• for wheeled assemblies with MI<100 psi:

⎛ 92.67 ⎞ ⎛ 0.15h ⎞
0.25 (13)
VCI1 ⎜ 28.23 + 0.43MI − ⎟⋅⎜ ⎟
⎝ MI + 3.74 ⎠ ⎝ δ ⎠
• for powered wheeled assemblies with MI>100 psi:

⎛ 0.15h ⎞ (14)
VCI1 = 9MI 0.446 ⋅ ⎜ ⎟
⎝ δ ⎠
The results obtained using the above formulae for M1 tank and HMMWV wheeled
vehicle are summarized in Figure 3.

120
98.59 VCI1 VCI50 MI
100 88.92

80
65.47
60.48
60

40 28.84
26.34
20

0
M1 HMMWV
Figure 3. Vehicle Cone Index and Mobility Index for M1 tank and HMMWV wheeled
vehicle

The data presented in Figure lead to the following conclusions:

24
 Leonardo Electronic Journal of Practices and Technologies Issue 15, July-December 2009
ISSN 1583-1078 p. 19-30

1. The Mobility Index predicts the ability of the vehicle to cross obstacles and
difficult terrains, in this case the tank being superior to the wheeled vehicle;
2. The grater values of VCI1 and VCI50 respectively indicate the better behaviour of
the wheeled vehicles for one pass and for multiple passes on soft soils.
The last conclusion is obvious if the total weight of the vehicles compared is
considered. Nevertheless, the difference is not so great, and the tank keeps a high mobility
despite the large effect of damage performed by the track on the upper surface of the soil.
The excess rating cone index, (RCIx ) is related to a vehicle's traction performance on
fine-grained soils and is computed as:
RCIx = RCI - VCI (15)
Greater the excess cone index, better mobility characteristics the vehicle has. To
increase the value of the excess rating cone index, for the same soil (same RCI value) it is
mandatory to decrease the vehicle cone index (VCI) value by specific design.

Traction performances
Computing traction performances relied to the drawbar at nominal slip, SLIP at
"Maximum" Soil Strength, and Motion Resistance calculation using empirical relationship
derived from experimental research for various types of soils according to USCS, slippery and
non-slippery surfaces and different degrees of humidity.
As an example, in
Table there are presented the relationships for the drawbar coefficient at nominal slip
for fine-grained soils. Similar structure of relationships exists in the case of motion resistance
too.
The main variable of the above prediction equations is the Rated Cone Index of the
soil. For a homogenous soil, the results of the computation are compared with the vehicle’s
capabilities in terms of maximum traction force. If the traction force is high enough to
overcome the motion resistance, a maximum speed for given soil conditions is computed
taking into consideration the engine power. The final results may be:
ƒ NO GO situation when the drawbar pull is bellow to traction effort developed by the
vehicle,
ƒ otherwise, maximum speed for given soil conditions.

25
 Semi-empiric algorithm for assessment of the vehicle mobility
Ticusor CIOBOTARU

For the real terrains, the soil characteristics are not homogenous. The method used for
mobility assessment consists of the spilling the area of the terrain into rectangular areas, small
enough that the Rated Cone Index is constant. Then the computations are performed for each
rectangular small area.

Table 1. Prediction equations for the drawbar coefficient at nominal slip for fine-
grained soils
Relationship description
Prediction Equation
Criteria Soil
D RCI 4.5585
CPF>4 psi SM = 0.5200142 − + 0.374746007
WNom RCI x + 8.117059
w/out Tire RCI
Chains SC, ML, D = 0.6152356 − 6.183363
+ 0.05261765
CL, CH WNom RCI x + 9.258565
Wheeled CPF< 4 psi
D RCI 6.183363
w/out Tire CH = 0.6152356 − + 0.05261765
Chains W Nom RCI x + 9.258565
D RCI 5.036329
w/Tire Chains = 0.6452267 − + 0.03994437
WNom RCI x + 7.35047
ML, D RCI 4.90683
= 0.6612633 − + 0.02224646
CL,SM WNom RCI x + 7.285463
CPF>4 psi
D RCI 5.131209
Tracked SC, CH = 0.6969994 − + 0.03483978
WNom RCI x + 6.99228
D RCI 4.838035
CPF< 4 psi = 0.7241738 − + 0.0435981
WNom RCI x + 6.301396

The final results of the mobility assessment may be presented in following manners:
ƒ as graphics presenting the maximum achievable speed on the percentage of the total
terrain considered (Figure4);
ƒ as maps for which the colours code indicates the maximum achievable speed.

26
 Leonardo Electronic Journal of Practices and Technologies Issue 15, July-December 2009
ISSN 1583-1078 p. 19-30

Figure 4. Mobility assessment output in terms of NO GO percentage of the total area

Comparing the same two vehicles, from Figure4, the main conclusions are detailed
below.
Within the 1st zone, the speed of the wheeled vehicles is grater with about 10 km/hour
than the speed of the tracked vehicle; this zone represents about 45% of the total area
considered, and consists of firm soils for which wheeled vehicles develop grater speed and,
consequently, they have grater mobility.
The 2nd zone, which represents about 40% of the total area, considered by this
scenario, the speed of the two vehicles is almost the same. This zone corresponds to the
typical soft soils of the usual battlefields.
For the 20% of the terrain, the wheeled vehicle is not capable to develop enough
traction effort to keep moving, and get stuck. This zone corresponds to the difficult terrain
within the area considered.
This representation relied to a statistical approach which cannot offer adequate
answers for mobility assessment for a given itinerary of the vehicle’s displacement.
A deeper analysis could be performed taking into consideration the specific soil
characteristics in correlation with the area in which they occur. This type of analyse involves
a mapping approach, illustrated in the figures below [7].
For the dry terrain, a terrain scenario was adopted for comparative simulation of the
mobility. The results obtained are presented in Figure5 to Figure7.
The comparative analysis reveals the area of the terrain for which one vehicle may
achieve better mobility characteristics with respect to speed or GO/NO GO situation.

27
 Semi-empiric algorithm for assessment of the vehicle mobility
Ticusor CIOBOTARU

Figure 5. Mapping of HMMWV maximum speed on dry terrain

Figure 6. Mapping of HMMWV maximum speed on dry terrain

Figure 7. Comparative analysis of mobility assessment o dry soil

In Figure5 an itinerary was marked from point A to point B. The same itinerary was
marked on the map presented in Figure. If this itinerary represents to most probable course of
action, the instantaneous speed of each type of vehicles may be evaluated. Consequently, the
mission plan may be adjusted with respect to mobility factors, for instance limiting the speed
of the hybrid column of vehicles or planning separate columns of vehicles in correlation with
their mobility.

28
 Leonardo Electronic Journal of Practices and Technologies Issue 15, July-December 2009
ISSN 1583-1078 p. 19-30

Conclusions

The mobility of vehicles running off road relies to the characteristics of soils to be
crossed. These characteristics vary significantly with the type of soil (granulation, cohesive or
sandy particles content, etc.) and with the humidity.
The characterisation of the soil with the respect to the vehicle’s traction is done by a
single parameter, the Cone Index. The measurement in situ of the Cone Index supposes
simple procedures and hardware but, usually, the results of the measurements have a large
variance due to the non-homogeneous soils. Consequently, this method relays mainly to a
statistical approach than to a deterministic algorithm.
The relationship for Drawbar Pull coefficient at nominal slip, and the Motion
Resistance Coefficient are simple in essence but are specific for each type of running gear
(tracked or wheeled), and for each type of soil. More than that, the relationships are specific to
the nature of the running surface (slippery or non-slippery). All the above mentioned aspects
introduce severe difficulties in adopting the right formula to be applied due to the fact that the
correct characterisation of the soil type supposes the existence of a high expertise.
Nevertheless, the validation of the relationships used by NRMM through a large amount of
experiments carried out on a large diversity of vehicles increase significantly the confidence
on the final results.
The computing amount of work is low, so efficient mathematical algorithms may be
used for a shorter time of computing. Consequently, this model may be implemented for real-
time modelling in war game software.
The comparative analysis is useful for general mobility assessment, but few
differences results between similar vehicles competing in acquisition programmes; so, a
limited applicability of NRMM is expected for comparing two or more similar vehicles.

References

1. Bekker, M. G., Introduction to Terrain-Vehicle Systems, The University of Michigan

Press, Ann Arbor, 1969

29
 Semi-empiric algorithm for assessment of the vehicle mobility
Ticusor CIOBOTARU

2. Wong, J.Y.: Terramechanics and Off-Road Vehicles, Amsterdam, Netherlands,Elsevier

Science Publishers, 1989

3. Kitano, M. and Kuma, M., An Analysis of Horizontal Plane Motion of Tracked Vehicles,
J. of Terramechanics, Vol. 44, No. 4, 1977, pp 211-225.

4. Jurkat, M.P., Nuttal C.J., Haley P.W., The AMC '74 Mobility Model, Tech. Report No.
11921(LL-149), May 1975, Waterways Experiment Station, Corps of Engineers, Dept.
of the Army

5. Ahlvin, R.B., Haley, P.W., NRMM II Users Guide, Vol 1,2, Ed. 2, Army Corps of
Engineers, Waterways Experiment Station, Dec. 1992

6. Jody D. Priddy, Stochastic Vehicle Mobility Forecasts using NATO Reference Mobility
Model, US Army Corps of Engineers, Waterways Experiment Station, Technical Report
GL-95-8, June 1995

7. Ciobotaru T., Jones R. A., Comparative Mobility Assessment Using NRMM, International
Conference Off-Road Vehicles, Bucharest, 22-24 September 2008

30