Using CFD to Optimise a LED Luminaire Design
  • Using CFD to Optimise a LED Luminaire Design

    Posted on Nov Mon, 2015 by seacad_admin

    For today’s post we would like to show how CFD analysis is used to help optimise the design of a LED Luminaire product.

    Whenever we do a simulation study, as an engineer we need to know what are:

    1. The Engineering goals – what type of results are we looking for? Is it a specific operating temperature, natural frequency, etc
    2. Variables that we could change in the design – is the material used, size, etc?

    Based on this, the simulation study needed becomes much more focused and easier to perform. Now back to our LED design.

    Every electric lighting products produces heat which is unwanted. As an engineer, the goal is usual to:

    1. Reduce the heat produced when the LED is turned on,
    2. or a more efficient way of transferring the heat away from the LED.

    Usually the latter options is much more probable to change as it can be change by many factors such as design geometry, cooling channels and many more, where else the first one is just based on the material.

    The key to successful LED system design is to transfer the active device’s heat efficiently from its own PN junction to the ambient. The path involves both the printed circuit board, which mounts the LED, and the enclosure. The engineer must know that the enclosure must effectively transfer heat away from the bulb and any heat sensitive parts.  Not to mention, also to make sure that the aesthetics of the design is top-notch and that its design is intended to retrofit into existing fixture, as it is essential to match the thermal and cooling characteristics of the original equipment.

    The designing process always starts with an initial design of the product. (such as below). The depicted system includes an integral connector (shown in yellow) attached to a lamp housing whose fins act as a heat sink. The light source is a power LED mounted on a metal-core PCB. The lamp’s lens is omitted to expose a more detailed perspective on the LED itself.

    Now that the model has been done, its time to test it out. SOLIDWORKS Flow Simulation automatically checks the volume to recognise fluid volumes and solid volumes. Not only does it detect internal volumes but also external volumes to take into consideration heat loss due to convection. Once this is done, the mesh is also automatically done which looks something like below:

    As you can see the mesh seems to be most refined in areas with smaller channels and gaps. This is automatically done so that better results are obtainable where it matters. Next the engineer must define boundary conditions (what was decided before the simulation began), that is the operating parameters and limits that will be used in the calculations. Values for the external air temperature and the LED device heating power must be specified, and can be incremented over multiple iterations of the CFD analysis.

    The image below displays the outcome of the CFD operation on the cross-sectional slice. It not only shows the heat distribution across the physical elements of the lamp, but also the vectors for air flow due to convection over the exterior of the lamp.

    In this view, the color spectrum goes from red (hottest) to blue (coolest). Of course, the object of this exercise is to ensure that a proposed physical design will transport heat away from the LED source and conduct it safely toward the ambient environment.

    The image below provides an answer to this important question. In this view, weightless particles trace the air flow path. Here the color spectrum reveals the heat distribution on the enclosure as well as the heat increase of the air particles as it travels up. The higher the heat change of the air particles as it travels, the more efficient the cooling system is. As an engineer, we would need to re-design until we get the best results:

    Convection carries the warmed air up and away from the lamp. Is this sufficient dispersion for the lamp itself and any other housings that will be part of the final design? It is a question only the engineer can answer, but SOLIDWORKS Flow Simulation has provided the data necessary to support an informed judgment.

    Finally for safety reason, we may want to know what is the temperature of the housing so that users will not get burnt if they were to touch it. We can plot the surface temperature (which is the contact temperature) to find out:

    Once the results comply with the standards that is set, it is now ready for production. Clearly we can see how is it that SOLIDWORKS Flow Simulation can help you:

    1. Optimise your product design
    2. Reduce cost of physical prototyping and testing
    3. Reduce the design time needed to achieve faster Go-To-Market

    If you would like to know how SOLIDWORKS Flow Simulation can help you save time and cut down on cost, contact our highly skilled engineers at +65 6372 1416 or email

Leave a Reply