Thermal dissipation of an LED PCB 2

[kevc70]

I am looking to do simulate thermal dissipation from an LED PCB with a large heat sink with fins. I was wondering if there was software (freeware or trial) that would allow me to accomplish this without having to upgrade to Solidworks Premium.

 

[kellnerp]

Are you looking for a simple thermal FEA model or a CFD analysis?

TOP
CSWP, BSSE

http://www.engtran.com

http://www.niswug.org

"Node news is good news."

 

[rfus]

you could simulate it for free on a piece of notebook paper, but your going to need a heat transfer book, understanding of the equation Q=hA(deltaT), a Reynolds number, and a Prandlt number.

 

[kevc70]

specifically, I am looking to do a thermal analysis on a heat sink. I am attaching some LEDs to an LEB board, which will be mounted to a square block of aluminum. I will be adding fins on the bottom of this aluminum block to aid with heat dissipation. This block (heat sink) needs to have a thermal resistance of 15 degrees C/watt in order to ensure good life for the LEDs. The lower the thermal resistance the better. So I would like to simulate how large of a heat sink I would need to design in order to achieve this specified thermal resistance i.e. surface area of the fins, how many fins needed, how large of a block needed, etc. The LEDs give out about 10W of heat. Ideally, the temperature of the heat sink should not exceed 50C. So, an intensive CFD would probably be required to simulate/approximate the temperature of the heat sink, and how large of a heat sink is needed to achieve the desired theraml resistance.

 

[rstupplebeen]

Ask your VAR for an evaluation trial of premium.

Rob Stupplebeen

 

[lumenharold]

Perhaps you could check into this software:

http://www.frigprim.com/index.html

Harold
SW2009 SP4.0 OPW2009 SP2 Win XP Pro 2002 SP3
Dell 690, Xeon 5160 @3.00GHz, 3.25GB RAM
nVidia Quadro FX4600

http://www.lumenflow.com

 

[hendersdc]

Does your aluminum block serve another purpose or is it just a heatsink? There are all kinds of stock heatsinks out there with given thermal resistances.

 

[kevc70]

the aluminum block would strictly be used as a heat sink to dissipate the heat from the LEDs. Actually, after further computation, I need the heat sink to have a thermal resistance of no more than 3 degrees C/W. The block w/ fins measures 2.375" (x), 2.83" , with total depth of 1". Depth of the non-fin area is 1/2". Does anyone know of a heat sink manufacturer that can produce a heat sink with these dimensions and this particular thermal resistance? I also plan on mounting a fan near the heat sink to further aid in the cooling process.

 

[GRF]

You do not want SW premium. You want either Solidworks simulation professional or Solidworks Flow Simulaiton.

You need to know what your environment around the heat sink will be. Are you going to be inside another enclosure, such as a light fixture? Is your heat sink exposed to open air.

 

[kellnerp]

It is quite acceptable to do this kind of initial design with a hand calculation. It is a whole lot cheaper and will get you closer to where you want to be quicker. After you have gotten yourself into the ballpark with that then you can verify it with more sophisticated tools. Neither FEA or CFD can point you to the most efficient solution as easily as a hand calc. They can give you insight into what is going on once you have arrived in the neighborhood of a good solution.

I would also not look just at the SW solutions, but look at 3rd party solutions to this. There are a number of good ones that will be just as cost effective if not more so than Simulation.

TOP
CSWP, BSSE

http://www.engtran.com

http://www.niswug.org

"Node news is good news."

 

[IanLougheed]

I'm going to agree with kellnerp and the others and suggest hand calculations. Finned arrays are not difficult to do relatively accurate calculations for. Even if you find a great computational solution to this and future design questions, you should always validate one result with at least one other using a different method. There are hand calculations tailored specifically to determining the optimum length for a fin or pin of a given cross section. You would have to iteratively run many FEA simulations to achieve similar optimization.

You will need to know some information about the environment, since the temperature and convection coefficient of the air flow affects the thermal resistance of your finned array. You may find that passive cooling is enough (ie. you may not need a fan to cool this component - this is worth finding out!).

In addition to the above comments, I always find it a good rule to "design it like you're going to do it again." I usually end up with handy Excel references or MATLAB scripts that I can reuse to solve similar problems later. This turns out to be a valuable time saver more often than not, even if it takes incrementally more time to create at first.

Also, it sounds like the dimensions you have given are relatively firm. In that case, you know that you have 1/2" long fins, and have only to select the number and cross section of these fins or pins. At this point, it's up to you to optimize the fin or pin design to suit the design requirement (thermal resistance), and the capabilities of your manufacturer.