Environmental Engineering Reference
In-Depth Information
Ice layer
Blade
Fig. 8.19 ANSYS model of accumulated ice with a uniform thickness of 3 mm with initial
volume V
ice
¼
1
:
17
10
4
m
3
, covering the upper surface of the blade from the leading edge to
about 40 % of the blade chord length
conductivity. The heater material was assumed to be ''structural steel'' from the
ANSYS material library in creating the models.
8.9.1 De-icing Performance Metric
In order to quantitatively evaluate the de-icing performance of distributed thermal
actuation, a performance metric is needed. This performance cost function should
consider the de-icing time [t
di
(sec)], maximum global temperature to the blade
structure during de-icing [T
max
b
(C)], the volume of the blade experiencing higher
than a certain temperature [30 C here as V
T[30
(m
3
)], and thermal power con-
sumption. Both T
max
b
and V
T[30
give indications of the level of thermal stress on
the blade due to the localized heating. A more advanced metric may also consider
additional parameters such as setting priority for de-icing in those regions on the
blade with higher contributions to the aerodynamic torque production (regions
closer to the leading edge and the blade tip), cost of the heaters, etc. In general, the
total cost of a heater network is a function of wind turbine size, blade dimension,
and total number of heaters on a full-scale blade. This financial cost has not been
included in the performance cost function in this chapter and is an area of future
work.
In order to compare de-icing performance between different heater layouts,
assuming the same amount of thermal power consumption, a quadratic perfor-
mance cost function J is defined as
J
¼
X
T
W
m
X
ð
8
:
4
Þ
T
where X
¼
t
di
;
V
T [ 30
;
T
max
b
½
is the de-icing performance state vector, and W
m
is a diagonal weighting matrix
