dding=7 cellspacing=0 width=100% bgcolor=ccccff> « back to results for ""
Below is a cache of http://www.emlsymposium.org/13th_papers/docs/A-EML115.pdf. It's a snapshot of the page taken as our search engine crawled the Web.
The web site itself may have changed. You can check the current page or check for previous versions at the Internet Archive. Yahoo! is not affiliated with the authors of this page or responsible for its content.
Research on the Cooling System of a 4QT Prototype Machine Used for HEV Research on the Cooling System of a
4QT Prototype Machine Used for HEV

Ping Zheng, Senior Member, IEEE, Ranran Liu, Peter Thelin, Member, IEEE, Erik
Nordlund, Member, IEEE, and Chandur Sadarangani


AbstractA four-quadrant transducer (4QT) is a hybrid electric
vehicle (HEV) power train concept. It can enable the internal
combustion engine (ICE) operate at optimum efficiency during
all driving conditions with both torque and speed control. The
thermal property is a vital problem due to its special
configuration. In order to find the hot spot and investigate the
cooling system of the 4QT, the thermal field is calculated with
finite-element method (FEM). The simulation is validated by the
comparisons with the experiment. The influences of the forced
air cooling, water cooling, and the distribution of the cooling
channels on the cooling performance are discussed, respectively.
The stator air ducts are removed and 12 water channels evenly
distributed in the housing are employed. The design principle of
the cooling system is presented. The speed impacts of air flow
and water flow are discussed and provide a parameter reference
for choosing the forced air cooling fan and water pump properly.
Index Termsfour-quadrant transducer (4QT), thermal property,
cooling system, design principle
I. I
NTRODUCTION

A combustion engine in a conventional vehicle operates rarely at optimum
efficiency. The optimum efficiency is about 30-40% and can only be obtained in a small
speed-torque region. The mean efficiency is much lower because of the small maximum
efficiency region. To operate a combustion engine at optimum efficiency during all
driving conditions, a full control of the torque and speed is required [1]-[4]. A four-
quadrant transducer (4QT) concept, which can fulfill both torque and speed control,
gives possibilities to operate the internal combustion engine (ICE) totally independent
of the road load [5].





Project 50577011 supported by National Natural Science Foundation of China.

Paper: #115 (a) (b) (c)
Fig. 1. Main component of the manufactured
radial-radial flux 4QT prototype machine: The
wound inner rotor (a), the canned outer rotor with
permanent magnets on both sides (b), and the
stator and housing (c) .
As a combination of two electric radial flux permanent-magnet machines, the
torque production capability depends on the temperature of the magnetic system. The
accurate knowledge of the temperature allows prevention of insulation breakdown and
permanent-magnet demagnetization effects in the design stage of the machine. In
addition, compared to the conventional configuration of machines, the slots of the 4QT
inner rotor are much deeper, leading to the increased current and copper loss especially
when operating above the base frequency, which is much higher than the industrial one
[6]. Considering the increased copper loss in the slots of the inner rotor, more
ventilation will be needed for the 4QT.
Thermal modeling of electrical machines is often performed using lumped-
parameter networks considering thermal resistances due to contact transitions and thin
insulation layers at various places in the electrical machine [7]. By using a proper
thermal network, the mean temperatures of different parts of a machine can be predicted.
However, in order to find the hot spot of the 4QT and to achieve the temperature values
of different points such as ones in the slots of the stator and inner rotor with a greater
accuracy, the thermal finite-element (FE) analysis should be considered [8].
II. T
HE
4QT

P
ROTOTYPE
M
ACHINE

The 4QT is an electric machine made up of two combined radial flux permanent-
magnet synchronous machines. The 4QT comprises three parts: a stator with windings,
an outer rotor connected to the load, and an inner rotor with windings connected to the
ICE. The stator and outer rotor operate as one machine called the stator machine (SM),
and the outer rotor and inner rotor operate as another machine called the double rotor
machine (DRM). The SM either increases or decreases the torque produced by the ICE
to the torque required at the final gear,
while the DRM changes the speed of
the final gear at the same time allowing
the ICE to operate at its optimal speed
point. Thus the operation at optimum
efficiency of the ICE during all driving
conditions can be realized. A 4QT
prototype machine was designed and
constructed to verify the simulation
results, as shown in Fig. 1.
III. T
HERMAL
A
NALYSIS
O
F
T
HE
4QT

P
ROTOTYPE
M
ACHINE

FE Thermal Model
Two-dimensional (2D) FE thermal model was performed in the commercial
software Ansys. Because of the geometrical symmetry, only a quarter region of the 4QT
needs to be considered in the thermal analysis, as shown in Fig. 2. The proposed model
is to compute the temperature distribution for the steady-state condition, so the thermal
capacitances are neglected. The analysis is implemented with the following hypothese: Fig. 2. 2D FE thermal model.
1. The influence of the uneven temperature distribution in the axial direction caused by
the external fan mounted at one side of the
machine is ignored.
2. Only the heat flux in the radial direction is
considered.
3. Each cylinder is thermally symmetrical in
the radial direction.
4. The heat sources are uniformly distributed
on the corresponding regions.
Thermal Parameters Calculation
Heat Source
The power losses of the 4QT prototype machine are input as the heat source for
temperature calculation and properly allotted to the corresponding nodes in the FE
thermal model.
The copper losses are computed from the currents and the measured ohmic
resistances, which can be expressed as:

2
3
Cu
ph
ph ph
P
k R I
=

(1)
where
Cu
P is the copper loss,
ph
k is the increased loss factor (1.1 for the stator copper
loss and 1.5 for the inner rotor copper loss ) considering the circulating currents in the
windings,
ph
R is the dc resistance of one phase and
ph
I is the one-phase current. To
simplify calculation, the varnished copper winding and the insulation can be considered
as a homogenous solid in the plane of the FE model with the equivalent thermal
conductivity [9]. The copper loss is uniformly allotted to the elements of the solid.
The iron losses are computed for the magnetic core region with knowing some
parameters obtained from the data sheet of the magnetic steel lamination [10]. The
calculated data are obtained by rotating the inner rotor when calculating the inner rotor
iron loss and rotating the inner and outer rotors synchronously when calculating the
stator iron loss. The computed iron losses of the stator and inner rotor teeth and backs
are respectively assigned to the elements of the corresponding regions.
Material Thermal Conductivity
The thermal conductivities of different materials with the isotropic thermal
properties are chosen from the textbooks and the data sheet.
Convection Heat Transfer Coefficient and Bulk Temperature
Radiation in motor thermal analysis is always ignored when considering the
convection heat transfer. The heat transfer coefficients in the inner rotor axial air holes,
stator air ducts, and water cooling channels are analytically calculated. When the water Fig. 3. The temperature distribution of the FE
thermal model at full load. Temperatures are
given in degrees Centigrade.
or forced air flow is laminar, the Nusselt number is constant, which gives a constant
convection heat transfer coefficient. The convection heat transfer coefficients of the
turbulent flow are calculated according to the modern formulations developed by B. S.
Petukhov as a vastly improved description of forced convection heat transfer in pipes
[11]. The inner and outer air gaps are considered according to [9].
The inner air, ambient, and water temperatures are taken from the measurements
made in the comparing tests, and the measured temperatures are used as the bulk
temperatures in the simulations.
Comparing Tests
Several temperature sensors were mounted in the 4QT prototype machine, one in
a stator slot, one in the inner rotor winding, and two respectively in the air of the inlet
and outlet side. The water inlet and outlet temperatures and the water flow speed were
measur