Physics Procedia 5 (2010) 735–744

www.elsevier.com/locate/procedia

LANE 2010

Potential of thermally conductive polymers for the cooling of

mechatronic parts
Heinle,C.*,
Heinle, C.*,Drummer
DrummerD.
D.
Lehrstuhl für Kunststofftechnik, Am Weichselgarten 9, 91054 Erlangen, Germany

Abstract

Adding thermally conductive fillers to polymers the thermal conductivity can be raised significantly. Thermal
conductive polymers (TC-plastics) open up a vast range of options to set up novel concepts of polymer
technological system solutions in the area of mechatronics.
Heating experiment of cooling ribs show the potential in thermal management of mechatronic parts with TC-
polymers in comparison with widely used reference materials copper and aluminum.
The results demonstrate that especially for certain thermal boundary conditions comparable performance between
these two material grades can be measured.
⃝c 2010 Published by Elsevier B.V. Open access under CC BY-NC-ND license.
Keywords: Thermal conductive polymers, cooling ribs; thermal management; LED;

1. Introduction

The thermal management with injection moulded functional parts of thermal conductive polymers is an
economic approach in creating highly innovative cooling concepts for mechatronic systems. For instance, the three-
dimensional structure of a mechatronic assembly’s housing can be integrated entirely into thermal management [1,
2]. Furthermore, plastics materials modified with ceramic fillers may be employed for mechatronic components,
such as inductance coil housings. It is therefore possible to achieve thermal, isolatory as well as mechanical
functions in a single step by overmolding the electronic component /3/.
Metal or ceramic fillers such as aluminum [4], graphite [5] or aluminum- and boronnitride [6, 7] increase heat
conductivities of plastics from approximately 0.15-0.5 Wm-1K-1 up to 20 Wm-1K-1 by using 60 vol.-% filler content
[8].
Nowadays more than 85 percent of all bord-based electronic circuits are realized in surface-mount-technology
(SMT) [9].
No matter if for rear lamps, headlights, or interior lighting: more and more LEDs are used – mounted to circuit
boards by means of the IMS-Technique (Insulated Metallic Substrate) [10].

* Corresponding author. Tel.: +49-9131-85-29-722.

E-mail address: [email protected].

1875-3892 ⃝c 2010 Published by Elsevier B.V. Open access under CC BY-NC-ND license.
doi:10.1016/j.phpro.2010.08.106
736 C. Heinle, D.name
Author Drummer / Physics
/ Physics ProcediaProcedia
00 (2010) 5000–000
(2010) 735–744

Fig. 1. Design and functionality of cooling concepts in mechatronic applications

left: air gap between BGA and cooling structure; frequent failure cause [Lehn]
middle: LED-assembly with aluminum heat sink; Fa. Diehl – airbus-interior lighting
right: thermo camera image of the LED-assembly in use

To dissipate the heat generated in the LEDs, designers increasingly apply cooling elements from aluminium
produced by extrusion or, sometimes, die casting; see Fig. 1 centre/right.
To provide for insulation between metal cooling element and circuit board, heat conducting films from plastics
are usually applied. This however does not only imply elaborate assembly work, but also a low coefficient of heat
transfer between the individual components. This hybrid, multi-component structure means an obstacle in terms of
thermal management between e.g. component and circuit board, circuit board and cooling element, etc. The quantity
of heat that can be dissipated is thus reduced considerably, making it often difficult to utilize the cooling element’s
high cooling potential entirely, or sometimes even leading to failure of the assembly group [11].
Moreover, conditions prevailing while using these components are frequently very thermally restrictive, with no
such components as fans etc. available to intensify the cooling effect. This is another reason why the metal cooling
element’s high degree of thermal conductivity, i.e. cooling potential, is not utilized entirely, because cooling takes
place with natural, or sometimes minor forced convection.
This paper is due to show that, starting from the thermal conditions described, a cooling element from a
thermally conductive polymer is well able to achieve a thermal output that is similar to that of a metal element.
TC polymers can be processed by injection moulding [3], Fused Deposition Modelling (FDM) [12] or selective
laser sintering (SLS) [13]. While design is extraordinarily free in terms of component geometries here, the cooling
potential can further be increased. There are thus materials available with a high potential for application in thermal
management of mechatronic assembly groups. Thermally conductive polymers can thus be employed for a wide
range of applications, to temper mechatronic assembly groups, and replace conventional hybrid systems.

2. Basics

In thermodynamics the heat transfer in technical processes is described by defining heat as a form of energy that
is transferred across system borders only because of a temperature difference [14]. Heat transfer is possible by three
different mechanisms:
x convection (free and forced),
x radiation, and
x conduction
 C. Heinle,Author
D. Drummer / Physics
name / Physics Procedia
Procedia 5 (2010)
00 (2010) 735–744
000–000 737

Conduction is the transfer of heat by direct contact of particles of matter. Steady state conduction is the form of
conduction which happens when the temperature difference driving the conduction is constant so that after an
equilibration time, the temperature in the conducting object does not change any further.
In a stationary process, increasing the thermal conductivity leads to a linear improvement in heat transport [15]:
dQ dT
O ˜ A ˜ (1)
dt dx
where
Q: heat input or heat lost [J] over a time t [s]
: thermal conductivity [Wm-1K-1]
A: heat transfer surface area [m2]
dT/dx: difference in temperature along a certain path x [K]

Convective heat transfer take place through diffusion, the random Brownian motion, and by advection, in which
heat is transported by the larger-scale motion of currents in the fluid or gas.
There are two major types of heat convection:
1. heat is carried by a fluid motion which is driven by external forces. This heat transfer process is often
termed forced convection.
2. heat itself causes the fluid motion (via expansion and buoyancy force). This process is called free or natural
convection.
The Quality of the heat transfer between a solid surface and a fluid or gas is described by the heat transfer
coefficient (HTC). This is the proportionality coefficient between the heat flux, Q/(At), and the thermodynamic
driving force for the flow of heat (i.e., the temperature difference, T).
dQ
D ˜ A ˜ 'T (2)
dt
where
Q: heat input or heat lost [J]
 : heat transfer coefficient [Wm-2K-1]
A: heat transfer surface area [m2]
T: difference in temperature between the solid surface and surrounding fluid [K]
t: time period [s]

For the calculation of technical applications mean (or) effective heat transfer coefficients are often used. These
are defined for a specific set of boundary conditions, such as part geometry, fluid and wall temperature. In the case
of forced convection the mean HTC  can be calculated using the following equation [15]:
O
D ˜ Nu(Re, Pr) (3)
L
where
L: characteristic length [m]
: thermal conductivity of the fluid [Wm-1K-1]
Re: Reynolds-number
Nu: Nusselt-number
Pr: Prandtl-number

For a specific geometry the Nusselt-number only depends on the dimensionless Reynolds-number and the
prandtl-number. These are calculated with equations 4 and 5:
v˜L˜U K ˜ cp
Re (4); Pr (5)
K O
where
v: characteristic flow velocity [m/s]
: viscosity [Pas]
cp: heat capacity [J/gK]
: density at mean temperature of the fluid [g/cm³]
738 C. Heinle, D.name
Author Drummer / Physics
/ Physics ProcediaProcedia
00 (2010) 5000–000
(2010) 735–744

Approximately the HTC for free convection can be calculated using the following equation [16]:

D 2  12 ˜ v (6)

3. Experimental: Materials and Specimen

Compound preparation and production of specimen:

On a ZSE 27 HP-40 twin-screw extruder by Leistritz compounds were prepared. As a matrix material, PA6
(Ultramid B3 by BASF) was employed, and 10 and 40 vol.-% of copper powder (Cubrotec 5000, Schlenk
Metallpulver GmbH) were integrated into the compound. The copper filler has a platelet-shaped, i.e. clearly non-
isometric geometry, as shown in figure 2 left. The platelets’ average diameter D50 is approx. 35 μm. Microscopic
analyses revealed that the platelet particles are approx. 1 μm thick.
On a Milacron K110S2-F standard injection molding machine by Ferromatik, the produced compounds were
processed to plate specimens with rib structure on the top, as can be seen in figure 2 right.
The main process parameter such as injection velocity (vin = 12.7 cm³), mould temperature (TMould = 80 °C) and
melt temperature (Tm = 260 °C) have been kept constant using standard values.

Fig. 2. left: SEM-image of the platelet shaped copper fillers used for the copper compounds
middle: geometry of the injection moulded sample specimen „rib structure“
right: pepared specimen “heat sink” for heating experiments

The ribs structure was eventually cut out of the test specimens and prepared; see Fig. 2 right. The reference
aluminium and copper specimen were machined.

Thermal conductivity measurement:

To determine the thermal conductivity of the injection moulded test specimens, a so-called Hot-Disk device (by
C3-Analysetechnik GmbH) was used. The sensor introduces the heat, which then dissipates in all three directions in
space. The characteristic value obtained is thus named integral thermal conductivity Oint.
 C. Heinle,Author
D. Drummer / Physics
name / Physics Procedia
Procedia 5 (2010)
00 (2010) 735–744
000–000 739

Heating experiments:
The heat sink presented in Fig. 2 right, was examined as to its cooling performance in a heating test. While
testing, thermal input was varied, as were cooling conditions. Via a heating plate of A dimensions of 8.96 cm²,
heating input P of between 1 and approx. 20 watts is possible. The setup is described in Fig. 3.

Fig. 3. Set-up for the heating experiments, heating source without and with mounted heat sink and schematic description

The heating plate is insulated, and temperatures are measured entirely by thermocouple inside the insulation. As
this determines the lost heat, actual heat input can be set precisely. At the interface between heating plate and
bottom of the heat sink, temperature TC was measured by means of thermal elements.
The efficiency of heat sinks is analytically calculated in electrical engineering by the difference between heat
sink surface temperature and junction temperature [20]. In our case TC corresponds to the junction temperature.
With decreasing TC, an increasing cooling efficiency of the heat sink can be assumed. Therefore comparable flow
conditions and surface properties are necessary.
The trials were undertaken at standard environment, and geometries remained unchanged. This is why
approximations and comparative investigations could be conducted to calculate an effective heat transfer coefficient.
Merely flow velocity, i.e. the effective heat transfer coefficient, was thus varied during the tests. Heating-up trials
were conducted with natural convection, i.e. little cooling, as well as forced convection, i.e. increased cooling. To
achieve different degrees of defined cooling input with forced convection, a fan was employed to vary air flow
velocities; see Fig. 4.

Fig. 4. Set-up for variation of cooling conditions free convection (box) and forced convection (flow channel)
740 C. Heinle, D.name
Author Drummer / Physics
/ Physics ProcediaProcedia
00 (2010) 5000–000
(2010) 735–744

The set flow velocities were measured by a hydrometric vane (G25 Höntzsch GmbH type) and are presented in
table 1. This table also shows the approximated effective heat transfer coefficients  obtained at different air flow
velocities, which were calculated under simplified assumptions (standard environment and no local flow effect) on
the basis of Equations 3 – 6.

Table 1. Approximated effective HTC  for the specific flow velocities of the air cooling

Velocity air-flow vair [m/s] free convection 1 3 5

HTC:  [Wm-2K-1] 7 12 22 29

4. Results and Discussion

The thermal response obtained at the heat sink, i.e. the cooling element, can be divided into two periods. During
the instationary period (heating process), the heat flows resulting - and thus the system temperatures inside the
component – are still subject to temporal changes. When heat input into the component and heat output due to
convective cooling have come to a thermodynamic equilibrium, then a so-called stationary state is achieved.
Temperatures inside the component are thus constant. Fig. 5 presents the contact temperatures TC for cooling
elements of copper and of a Cu-compound (10 Vol.-% filler content)

100
90
80
temperature TC [°C]

70
60
50
PA6 + 10 Vol.-% copper
40 PA6 + 40 Vol.-% copper
30 Aluminium
Copper
20
0
0 10 20 30 40 50
time t [min]

Fig. 5. Development of the time dependent contact temperature TC; heating power P=7W, flow velocity: v=5 m/s
 C. Heinle,Author
D. Drummer / Physics
name / Physics Procedia
Procedia 5 (2010)
00 (2010) 735–744
000–000 741

During the investigations, scientists determined the period of time required until a stationary state, i.e. constant
temperature, was reached. In case average temperature deviation is below 0.1 K/min, a stationary state can be
assumed. With heat input set to 7 watts, the copper cooling element reached the stationary state after approx. 24.8
min, while this took approx. 31.5 min for the copper compound.
However, the major differences showed in contact temperature TC. It merely reached 60 °C with the copper
cooling element, while the PA6 + 10 vol.-% copper cooling element’s temperature rose to 100 °C. This can be
contributed to the fact that TC polymers’ thermal conductivity is clearly below that of metal materials. The heat
conductivities of the used copper compounds determined by Hot-Disk measuring can be seen in table 2 below.

Table 2. Thermal conductivity of the cooling-rib materials: TC-polymer (Hot-disc), aluminum [17] and copper [18]

material PA6 + 10 Vol.-% PA6 + 40 Vol.-% aluminium copper

copper copper
thermal conductivity 1.1 W/mK 4.5 W/mK ~ 220 W/mK ~ 400 W/mK

heat sink

To compare heat outputs taking place from the heat input interface, the temperatures TC determined for the
stationary state will be used in the following.
In order to enable comparison of cooling performances, variations were tested in
- Heat input values of P = 1, 3 and 7 watts, and
- convective cooling values of approx.  = free convection, 7, 12, 21 and 29 Wm-2K-1
Fig. 6 shows a photo made by a thermographic camera. It presents the stationary-state temperature distribution at
the surface of the rib cooling element for copper-compounds with 10 to 40 vol.-% filler content.

Fig. 6. Thermocamera scan: surface temperature of the cooling rib; TC-polymer; heat input: P = 3 W, free convection;
left: PA6 + 10 vol.-% copper
right: PA6 + 40 vol.-% copper
742 C. Heinle, D.name
Author Drummer / Physics
/ Physics ProcediaProcedia
00 (2010) 5000–000
(2010) 735–744

With low filler content, temperatures at the cooling element’s basic surface are relatively high, which can be
contributed to the low cooling effect of the ribs. As filler content is stepped up from 10 to 40 vol.-% of copper,
temperatures decrease from 80 °C to 60 °C.
The temperature TC at the interface between cooling element and heating plate shows the same behaviour, see
Fig. 7.

90
PA6+10 vol.-% copper
PA6+40 vol.-% copper
80 copper
temperature TC [°C]

aluminum
70

60

50

40

0
0 5 10 15 20 25 30 35
-2 -1
heat transfer coefficient D [Wm K ]

Fig. 7. Contact temperature TC dependant on the HTC  on the rib surface, heat input P = 3 W
 C. Heinle,Author
D. Drummer / Physics
name / Physics Procedia
Procedia 5 (2010)
00 (2010) 735–744
000–000 743

As convection is natural, increasing the filler content from 10 to 40 vol.-% of copper will reduce TC by approx..
10 K. It is interesting to note here, that deviation in temperatures is only minor, if compared to pure copper or
aluminium. It amounts to just a few °Kelvin. As the cooling effect is stepped up due to intense flow, i.e. forced
convection, the difference between cooling elements from metal and TC polymers gets larger. However, even a flow
velocity of 5 m/s will lead to only approx. 15 K. This cooling performance is roughly similar to that of a commercial
16 V PC fan operated at maximum speed.
The cooling elements’ cooling performance, is determined not only by convective cooling, but also significantly
by heat input. The obtained contact temperature TC is presented in Fig. 8 for cooling elements of copper as well as
PA6+40 vol-% copper, in relation to the different values of heat input.
With natural convection, the differences in temperatures between both materials are only slight. If flow velocities
are stepped up to increase cooling, then, at approx. 7 watts heat input, temperature deviates by approx. 15 K.

free convection
150
PA6+40 vol.-% copper
135 copper
120
temperature TC [°C]

105 forced convection (5 m/s)

PA6+40 vol.-% copper
90
copper
75
60
45
0
2 3 4 5 6 7 8
heat input P [W]

Fig. 8. Contact temperature TC dependent on the heat input p; Cooling: free and forced (vair=5 m/s) convection, material: copper and PA6+40
vol.-% copper

However, reducing the heat input that acts upon the interfaces also reduces the temperature difference between
the two material systems. As a result, with low cooling, input and forced convection, the cooling effect of the TC-
polymer is close to that of metals. Heat output is determined by environmental cooling conditions to a growing
extent.
With the existing experimental conditions, in case of poor cooling, as is the case with natural convection, the
differences between the two materials, metal and TC-polymer, can be neglected.

5. Conclusion

The investigations have shown that, under specific thermal boundary conditions, thermally conductive plastics
have a high potential for cooling mechatronic systems. In the area of low to average dissipated energy, if linked to
minor degrees of convective cooling, polymers’ cooling performance is almost identical to that of metal materials
(aluminium or copper). It is thus possible to remove the energy usually dissipated by LEDs, i.e. approx. 0.5 to up to
approx. 5 watts [19], by means of TC-polymer cooling elements.
Unlike conventional hybrid concepts, the techniques of polymer processing usually offer a high level of freedom
in geometrical design. New designs are thus possible for cooling elements, and new concepts of thermal
management can be implemented, which are comparable or even superior to the known concepts.
Additional potential may be expected from using ceramic fillers. These materials’ heat conductivities are similar
to those of the TC-polymers presented; they are electrically insulating, though. They might thus enable the designer
744 C. Heinle, D.name
Author Drummer / Physics
/ Physics ProcediaProcedia
00 (2010) 5000–000
(2010) 735–744

to omit intermediate layers, e.g. insulating films, and eliminate heat transfer at the same time. By directly on- or
overmolding circuit boards with thermally conductive polymers, cooling structures for thermal management can be
integrated. This will further raise the level of integration, in terms of construction as well as production.

Acknowledgements

The authors would like to thank the German Research Foundation (DFG, SFB 694) for funding the work.

References
1. Heinl, D.: Hochgefüllte Duroplaste mit definierten Eigenschaften für Elektronik und Elektrotechnik; In: Hochgefüllte Kunststoffe, Springer-
VDI-Verlag, Düsseldorf, 2002
2. Miller, J.; McCullough, K.: Thermally conducting Moulded Polymers; Advanced Packaging, 1999
3. Heinle, C.; Drummer, D.: Thermisches Management mechatronischer Systeme mit wärmeleitenden Kunststoffen; AmE 2010 - Automotive
meets Electronics, GMM-Fachtagung, 15.-16. 04.2010, Dortmund
4. Wong, C. P.; Bollampally, R. S.: Thermal conductivity, elastic modulus, and coefficient of thermal expansion of polymer composites filled
with ceramic particles for electronic packaging, J. Appl. Polym. Sci.,1999; 74
5. King, J. A. ; et. al.: Electrically and thermally conductive nylon 6.6; Polym. Compos., 1999, 20(5):643–54
6. Xu, Y.; Chung, D. D. L.; Mroz; C.: Thermally conducting aluminum nitride polymer-matrix composites; Composites: Part A, 2001;
32:1749–57
7. S. Yu, P. Hing, X. Hu; Thermal conductivity of polystyrene-aluminum nitride composite; Composites: Part A, 2002; 33:289–92
8. Amesöder, S.; Wärmeleitende Kunststoffe für den Spritzguß; Dissertation, Lehrstuhl für Kunststofftechnik, Universität Erlangen-Nürnberg,
2009
9. Noelle, L.: Elektronische Baugruppen gut gekühlt; E&E Kompendium 2009/2010, S.117 ff.
10. Loutschonig, F.: Schaltungsträger mit hoher Wärmeleitfähigkeit für LED-Applikationen; Elektronikpraxis, 02.05.2008
11. Lehnberger, C.: Kühlkonzepte für die Leiterplatte: Engpässe im thermischen Pfad einer Baugruppe erkennen und beheben;
Elektronikpraxis; 02.01.2007
12. Wendel, B.; Feulner, R. W.; Schmachtenberg, E.: Wärmeleitende Funktionsbauteile aus Kunststoffdrähten
Kunststoffe 5 (2008), S. 34-37
13. Wendel, B.; Dallner, C.; Schmachtenberg, E.: New developments in Selective lasersintering of polymers
Proceedings of the 5th LANE, Erlangen, 2007, S. 323-331
14. Incropera, F.P.; De Witt, D.P.: Fundamentals of Heat and Mass Transfer (3rd ed.). John Wiley & Sons., 1990
15. Leipertz, A.: Wärme- und Stoffübertragung; 2. korr. und erw. Aufl. ed. 2008, Erlangen
16: N. N.: VDI Wärmeatlas - Berechnungsblätter für den Wärmeübergang; VDI-Verlag; Düsseldorf 1991
17. N.N.: DIN EN 485-2 Bleche und Platten
18 N.N.: Cu-ETP.pdf ; Deutsches Kupferinstitut e.V., Düsseldorf, 2005
19 N. N.: Produktinformationen „High-Power-LEDs“; Fa. Lumitronix, 2010
20 Adamczewski, F.: Berechnung eines berippten Kühlkörpers; Fa. Thermoconsult, FAd/2002