Success Story

Keywords: Thermal Analysis, Heat Transfer, Optimization

Company ProfileDSM – Bright Science. Brighter Living.™

Royal DSM is a global science-based

company active in health, nutrition and

materials. By connecting its unique

competences in Life Sciences and Materials

Sciences DSM is driving economic

prosperity, environmental progress and

social advances to create sustainable value

for all stakeholders. DSM delivers innovative

solutions that nourish, protect and improve

performance in global markets such as

food and dietary supplements, personal

care, feed, pharmaceuticals, medical

devices, automotive, paints, electrical and

electronics, life protection, alternative

energy and bio-based materials. DSM’s

23,500 employees deliver annual net sales

of about 9 billion Euros. The company is

listed on NYSE Euronext. More information

can be found at www.dsm.com.

The ChallengeDSM Engineering Plastics is a global

provider of high performance plastics used

in a variety of engineering applications.

One such application is the heat sink used

in LED lighting applications with the main

function of dissipating heat, generated

by LEDs, to the environment. The first

generations of LED lighting applications

use a heat sink made of aluminum. Being

a very good heat conductor, aluminum,

however, also has its limitations for this

kind of applications. Major limitations of

aluminum are the costs related to the

further need to machine die-cast parts,

limited design freedom, recyclability and

weight. DSM Engineering Plastics offers a

highly thermally conductive polymer which

can overcome these issues aluminum

heat sinks show. Combining its material

expertise as material supplier with its CAE

Thermal analysis of LED lamps using AcuSolve

IndustryHigh Performance Plastics

ChallengeNeed for accurate simulation results and prediction of heat dissipation in LED lighting applications at DSM.

Altair SolutionUse of AcuSolve from the HyperWorks CAE Suite for modeling a variety of heat and flow problems in the development of heat sinks made of plastics.

Benefits• Accurate prediction of the thermal performance• Possibility to optimize the heat sink geometry

for a given application• Good correlation with measured results• Shorter and more streamlined development

cycles leading to a better usage of the DSM high performance plastics

Key Highlights

design capabilities DSM is able to provide

a plastic-based solution for the heat sinks

needed in the lighting industry. In order

to support its customers in their product

development, the DSM design department

uses the state-of-the-art CAE tools provided

by global software suppliers. To find an

optimal design of the heat sink made

of polymer material and to predict the

thermal performance, DSM has to model

the natural convection and the radiation

cooling as these are the mechanisms by

which LED heat sinks dissipate heat to the

environment.

The SolutionRecently, DSM implemented the software

package AcuSolve, a product of Altair

Engineering in its development processes.

AcuSolve is a general-purpose CFD based

software capable of modeling a variety

of heat and flow problems. Currently, DSM

uses the capability of AcuSolve to model

natural convection and radiation cooling.

The necessity to cool LEDs originates from

the requirement that these electronic

components must maintain their maximum

temperature below a certain point in

order to reach the life time specified

by the manufacturers of LED lamps. In

order to evaluate the cooling capacity of

a polymer heat sink, AcuSolve is used

along other valuable CAE tools (i.e. CAD-

based software). In the following use

case overview, the predicting capabilities

of AcuSolve regarding the evaluation of

thermal performance of a plastic heat sink

are described. The simulation results are

compared to temperature measurements

showing a good correlation.

The Use Case – Modeling heat dissipation of LED heatsinksIn order to model heat dissipation from a

heat sink to the environment it is necessary

to take into account the relevant heat

transfer mechanisms to the environment.

Obviously heat loss by convection occurs.

However, it is noted that in many natural

convection problems the heat loss to the

environment by radiation is not negligible

as opposed to forced convection problems.

In the problem considered here, heat

loss by radiation is approximately 1/3 of

the total heat loss to the environment.

To model the heat loss of a heat sink to

the environment accurately, both natural

convection and radiation need to be

included in the analysis.

To model heat loss by natural convection a

tool is required that couples the heat loss

to the air flow around the heat sink which

results due to density differences of the

air in the proximity of a heat sink surface

which obviously has a higher temperature

than the air. Also, air flow stagnation

due to boundary layer effects need to be

accounted for, particularly when a heat

sink is designed with a too small fin-to-fin

spacing. AcuSolve includes the capability

to model these boundary layer effects in

three dimensions and it can prevent the

designer from making errors regarding the

introduction of fins that are too closely

spaced.

On the base of the heat sink a PCB (green)

is located which contains three LED’s (red).

Each LED acts as a heat source and generates

a heat of 1.4 W. Between the PCB and the

heat sink a thin layer with the thermal

conductivity of thermal grease is placed in

order to model the thermal contact resistance

that exists between PCB and heat sink.

Around the heat sink an air volume (light blue)

is modeled with dimensions of about 8D and

8H, with D = heat sink diameter = 46 mm

and H = heat sink height = 39 mm

The right picture shows the temperature

distribution of the heat sink and surrounding

air.

Figure 1: AcuSolve model of the heat

sink with a PCB and three LED’s.

Modeling heat dissipation of LED heatsinks

Another issue in cooling analysis of heat

sinks is related to modeling the radiation

heat loss. A certain fin design might

involve radiation between two neighboring

fins instead radiation from a fin to the

environment. In the first case, the net

heat loss by radiation to the environment

will be limited which is not desirable in

cooling applications. Obviously, the second

case is the preferred one as the goal is to

maximize the heat loss (by radiation) to

the surroundings. In order to take care of

this, view factors for different parts of the

heat sink need to be computed. AcuSolve

takes into account the view factors using

hemicube algorithm. One advantage of

using this view factor approach is the

speed of computation for radiation is

not drastically increased compared to

non-radiation cases. In this way, one

can optimize the heat sink design and

maximize the radiative heat loss to the

environment. AcuSolve is able to solve

for all the heat transfer phenomena

(conduction, convection and radiation)

encountered in this system.

In order to test the capability of

AcuSolve to model heat dissipation to

the environment via natural convection

and radiation, a heat sink was designed

and manufactured. The plastic heat sink

is shown in figure 1. On the base of the

heat sink a PCB (green) is located which

contains three LED’s (red). Each LED acts

as a heat source and generates a heat of

1.4 W. Between the PCB and the heat sink

a thin layer with the thermal conductivity

of thermal grease is placed in order to

model the thermal contact resistance that

exists between PCB and heat sink. Around

the heat sink an air volume (light blue) is

modeled with dimensions of about 8D and

8H, with D = heat sink diameter = 46 mm

and H = heat sink height = 39 mm. In the

model the environment temperature of 25

°C is assumed. The heat input is modeled

as a volumetric heat source applied to each

of the volumes that represent the LED’s.

The steady-state temperature distribution is

computed at these conditions. The analysis

includes the coupling of the heat flow with

an air flow around the heat sink in order to

capture the natural convection effects. Also

view factors are computed with AcuSolve

and taken into account as they are needed

for an accurate modeling of the radiative

heat loss. The steady-state temperature is

shown in Figure 2. The local temperature

as well as some characteristic locations is

listed together with the experimentally

observed values.

The experimental setup used for the

measurement of the temperature

distribution is shown in figure 3 together

with the positions where various

thermocouples were attached. Also shown

is the observed temperature evolution over

time. After roughly 1 hour the steady-state

temperature distribution was reached.

The environment temperature during the

experiment was 25 °C.

www.altair.com

Steady-state temperature distribution as

predicted by AcuSolve. At specific locations

the predicted temperature is predicted with

measured temperature.

Figure 2: Prediction of temperature

distribution

TInterface = 62 °C

TLED = 69.9 °C

TFin = 55 °C

Benefits and Conclusions“Thanks to AcuSolve we could accurately

predict the thermal performance of heat

sinks, used in LED lighting applications. We

were very satisfied with the results we

received and could – thanks to the good

results our analysis showed – move our

development efforts significantly from real

testing to virtual development methods,

leading to fewer prototypes and shorter

development cycles, resulting in reduced

development costs. In this way we are

helping our customers to make better use

of our materials offerings in the area of

high performance plastics” said Adnan

Hasanovic, Research Scientist/Design

Engineer at DSM Ahead B.V. AcuSolve

proves to be a valuable tool in predicting

the thermal performance of heat sinks

used in LED lighting applications. Not

only a qualitative

prediction of results

is obtained, the

absolute values of

the temperatures at

different positions

are also predicted

quite accurately.

In this way, it is

possible to design

and optimize the heat sink geometry for

a given application with high accuracy.

AcuSolve’s preprocessor AcuConsole enables

the use of templates in a manner such

that generating meshes for new cases is

fast and easy. At DSM, a similar analysis

was also performed for a number of other

heat sink geometries and the predicted

results were compared to the experiments.

A good correlation

to measured

results was found

in all these cases.

With this method

the development

focus could be

shifted to virtual

prototyping, leading

to a much more

streamlined development process in terms

of less physical prototypes and reduced

development costs and finally to a better

usage of DSM’s high performance plastics

at their customers.

Experimental setup of the heat sink with

PCB and LEDs, incl. the position of various

thermocouples (left). Measured temperature

evolution over time (right).

Figure 3

We were very satisfied with the results we

received and could – thanks to the good

results our analysis showed – move our

development efforts significantly from real

testing to virtual development methods,

leading to fewer prototypes and shorter

development cycles, resulting in reduced

development costs.

Altair®, HyperWorks®, RADIOSS™, HyperMesh®, BatchMesher™, HyperView®, HyperCrash™, HyperGraph®, HyperGraph®3D, HyperView Player®, OptiStruct®, HyperStudy®,

HyperStudy®DSS, MotionView®, MotionSolve™, Altair Data Manager™, HyperWorks Process Manager™, HyperForm®, HyperXtrude®, GridWorks™,

PBS Professional®, and e-Compute™ are trademarks of Altair Engineering, Inc. All other trademarks or servicemarks are the property of their respective owners.

Altair Engineering, Inc., World Headquarters: 1820 E. Big Beaver Rd., Troy, MI 48083-2031 USAPhone: +1.248.614.2400 • Fax: +1.248.614.2411 • www.altair.com • info@altair.com

LEDFinHS/MCPCP