SOLIDWORKS Thermal from Beginning to End (Simulation Webinar)

[@brocklydoodle]

Something like this should absolutely be a part of Engineering course taught at colleges. I would have loved to have done this as part of my heat transfer course. Getting to use hand calcs paired with software to solve a real-world problem would have been so helpful.

[@ShubhamKumar-id9sm]

Thanks for this. Simulation webinar is must needed. Please continue the series.

[@goengineer]

We are glad to help.

[@romanmateo2566]

i realize I am kinda off topic but does anyone know of a good website to watch newly released series online ?

[@ericluis2290]

@Roman Mateo Lately I have been using FlixZone. Just google for it :)

[@desmondsimon6261]

@Roman Mateo try Flixzone. Just search on google for it :)

[@stbny4444]

Shaun makes the best videos.

[@muchammadjamil8688]

Thanks a lot teacher. . ... very lucky see your tutorial

[@goengineer]

You are welcome!

[@atulsnair6687]

Very cool webinar I see you have assumed thermal resistance of the heat pipe is it safe to assume thermal resistance? If not how can we calculate through solidworks

[@aleXelaMec]

Big thank great webinar

[@goengineer]

Most welcome!

[@AnmolHasmukhbhai]

Highly informative, is it possible to share the excel file of calculations ?

[@vaibhavshelar6176]

thanks

[@spsi2217]

Thanks for posting this tutorial Question, 23:33 does your thermal paste have a thermal resistance of 0.007K/W? Does that mean -273C/W?

[@goengineer]

1 K = 1 C° Notice I am using "C°" instead of "°C". A subtle distinction which is discussed a bit here: www.answers.com/Q/Difference_between_degrees_Celsius_and_Celsius_degrees So 0.007 K/W is equal to 0.007 C°/W.

[@spsi2217]

@@goengineer Thanks for clarifying!

[@jeremytanweihern1229]

Dear teacher, I've gone through the hand calculation in the spreadsheet you provided and I have a few questions. 1. For the board contact-board part, why do you take the thickness/ length and divide them by 3 to get the effective thickness/ length? 2. For the board contact-board part again, we should be using board thickness of 1mm to get the effective thickness, right? I'm just unsure of why do you use 1.8mm instead of 1mm to get the effective thickness.

[@goengineer]

Very observant and excellent questions! Honestly didn't think anyone cared to comb through the spreadsheet. I think you'll have to bare with me, though since this spreadsheet is over a year old and my memory of why I made certain assumptions might be fading a bit. 1.) Perhaps an obvious disclaimer to start: this shouldn't really be the way we model it (perhaps the right way is to use finite elements with detail like we do in the actual study) but if we do hand calculation, we have to start with some simplifying assumption. In this case, we are simplifying the board as heat flow THROUGH the board and then ACROSS the board, but with an additional question: How far does the heat have to flow THROUGH the board before it begins flowing ACROSS it ON AVERAGE. Assuming a triangular distribution of heat flow at the beginning of the ACROSS journey, ON AVERAGE I'm assuming the heat makes it 1/3 the way THROUGH the thickness before going ACROSS. Sort of a "centroid of a right triangle" assumption. I use the same assumption for ACROSS the board to the AIR. One main point in this discussion is that it can be difficult to hand calculate without some type of simplification/approximation. 2) The actual board geometry used in the study is 1.8mm thick. If you open the tutorial model that this is based off of, you should be able to take the measurement to confirm. You may then ask the next reasonable question: why didn't you use the 1.8mm the total thickness and for the PCB definitions? Unfortunately, I followed the tutorial document for some of the setup and in the document there is a typo where they use 1mm (see page E1-12) which I didn't question at the time. Only after digging into some of the weeds and taking actual measurements did I discover a discrepancy, but after I had captured and written a large portion of the content. At that point, I decided to just make this assumption consistent across all the studies so it should still give me a fair apples-to-apples comparison for all three studies. I hope this helps to clarify a couple of your points. Please let me know if you have further questions and I hope you got something out of this!

[@jeremytanweihern1229]

@@goengineer Hi teacher, thanks for the prompt reply! 1. Understood. I am trying to understand more on the hand calculations as I am planning to use hand calculations to validate/ verify my thermal simulations/ flow simulation results. Hence, this leads to another question, is there any other way to validate our simulation results? 2. Alright, thank you for the clarification. As you can see, I got a little confused on why is there a discrepancy in the thickness value used.

[@goengineer]

@@jeremytanweihern1229 This is a big question difficult to answer in a KZbin comment but I hope my brief answer can inspire some ideas: Simulation can be viewed as a tool to help automate long hand calculations. Each cell in my mesh is like one small hand calculation connected to its neighboring calculations. So part of this question may be "how does one validate hand calculations?" Though simple experiments, physical tests, intuition, logic and reason. These same approaches for validations can be pursued to help validate simulation results. Comparison to known hand calculations is convenient when it is available for some simple problems.

[@jeremytanweihern1229]

@@goengineer Thanks a lot for the insightful explanation, I understand that hand calculation can only obtain temperature at one point only. How do you compare the simulation result with the hand calculation then? For instance the heat sink. Do you probe the temperature at a certain location on the heatsink? If you do, may I know which exact location do you probe, let's say for the heat sink?

[@goengineer]

@@jeremytanweihern1229 I would use the AVERAGE temperature of the bodies for many comparisons to the hand calculations in my example. This approach works best for highly conductive bodies (internal resistance of body is close to zero) such as the heat sink where the min/max/average are all very close to each other. For poorly conductive bodies such as the board it is not trivial and requires some assumptions. The internal resistance of the board should be accounted for, so I break it up depending on the heat flow path I'm considering which is why you see a "BOARD CONTACT" temperature (average temperature on the junction surface contacting the board surface) and "BOARD" temperature (average temperature of the board surface which would contact that air).

[@norbsclips7091]

Thanks for this tutorial. In the SW Flow Sim two-resistor model is the Theta(JC) always the top of case package or it can be bottom of package? In case for example I will be using component without heatsink on top(most SMDs) but with epad soldered on board,(PCB as heatsink) is the Theta(JC) still referring to the top of case package in the two-resistor model or th e bottom face of the package needs to be selected as the top face? Also If in case datasheet of SMD component only provide Theta(JC) and Theta(JA) is given is there a way to derive Theta(J-Board) which is needed as input for two-resistor model in SW flow sim. Can you share your hand calculation spread sheet if that is okay.

[@goengineer]

This is a great question and could probably use some expansion on what I'm about to write, but here's the short version. We are just assigning the resistance you apply to θjC to whatever face you mark as the "Top Face" (see 35:11) so in principle you can select the bottom if you wanted. Take a quick look at the Flow Simulation Technical Reference (found under Help, SOLIDWORKS Simulation) to make sure you understand fundamentally what we are doing with this 2R Component tool before you get too creative, though. There's a little diagram in there that should clear a few things up for you on roundabout page 90 (check the Table of Contents). To estimate θjB (junction to board resistance), use the following relation: θjA= θjB + θjC Where θjA is the total thermal resistance from junction to ambient θjB is the junction to board thermal resistance θjC is the junction to case thermal resistance Finally, I don't know the spreadsheet will be as handy as you might think. The thermal circuit in it was designed for this specific example. But if you insist, I can see what we can do to host it somewhere.

[@goengineer]

I posted the spreadsheet now since I've had another request for it. See the description in the video.

[@norbsclips7091]

@@goengineer Thanks for this.

[@AminRahimi-1]

greattttt,thanks

[@husseinsharafi7060]

Thanks for the grate content. Could you plase explain how you came up with the thermal resistances of the PCB after running thermal and flow simulation to import it back to the excel sheet for comparison?

[@goengineer]

Please see my discussion with "Jeremy Tan Wei Hern" in this comments section to see more information about how the PCB was modeled. The spreadsheet modeling was done intentionally WITHOUT the use of Flow Simulation and Thermal analysis in order to illustrate the capability and limitations of an "only spreadsheet" approach. Then the Flow Simulation and Thermal analysis were performed afterward and I compared to this results.

[@yuliaroughneck1474]

I don´t quite understand how you came up with the 2 Values for heat convection in Thermal. According to the SolidWorks help page you linked in your older Video, the heat convection depends on the surrounding material (Stagnant/Flowing Air, Water, Oil) not on the part itself. Why does the Heatsink have a different coefficient than the rest? 5 - 25 is natural convection, so without any airflow. Shouldn´t the values be the same across the board?

[@chintanjoshi6804]

Please make a video of SOLIDWORKS HVAC module.

[@Yo_dream.]

How to get a PDF tutorial with related 3d models?

[@erikedstrom971]

Does Solidworks treat a void in a model as being filled with air by default? e.g. I make a hollow sphere using revolve. Are the internals atmospheric air by default? Thanks. Great video.

[@goengineer]

SimPro Thermal: by default gaps with no thermal conditions applied will behave like they are perfectly insulated and will not participate in any heat transfer (neither air convection or radiation). Flow Simulation: Gaps will be filled with the default project fluid (normally air). Exceptions/considerations: if gap is an enclosed cavity without boundary conditions, it will be excluded from the analysis by default unless you disable this option in the wizard. Also some mesh refinement in these gap areas may be needed to properly resolve them if its an important phenomena to capture in your study.

[@amirrizman8428]

Hye, sir did you know why my solidworks only have general simulation and didnt have the advanced simulation like thermal study? How can i add the simulation?

[@goengineer]

Who did you purchase your license from? You will likely need to contact them and ask them about getting a license of SOLIDWORKS Simulation Professional or SOLIDWORKS Flow Simulation depending on which tool will be suitable to you.

[@buzzymclaughlin]

Can you share the material properties you used to approximate the heat pipe in the simulation model? Is there a means of assigning thermal resistance?

[@goengineer]

The one at 21:29 or the one at 37:00? The properties used at 21:29 in the Thermal analysis were set such that the resistance of the body itself would be close to zero (I likely used a conductivity of 10000 W/m.K). Instead of relying on the simplified CAD body, I want the thermal resistance of this heat pipe pathway to be handled by the contact I create at 24:09. Here you can see that thermal resistance I am assigning is 0.3 K/W and this is intended to be the heat pipe resistance. If I know contact resistance between the pipe and the heat sink as well, I should actually add this to the 0.3, but I simply assumed zero on that term for this study. Another nice to approach this would be to split the heat pipe into halves (a cpu side and a heat sink side) and assign a contact resistance between the two halves to mimic the thermal resistance of the pipe. If you are interested in the heat capacity characteristics as well, please let me know. At 37:00 I can be more direct in just assigning the thermal resistance to the heat pipe component itself, but Flow Simulation uses a fairly similar method behind the scenes to what I do at 21:29. I hope this helps to answer your question, but please let me know if part of this is still unclear.

[@simonfeghali5214]

How did you add the time graph and did that simuation with time

[@goengineer]

For which part? In the thermal analysis, you must go to the STUDY PROPERTIES and make it a TRANSIENT study. Then you can probe a node and request a RESPONSE GRAPH. For the Flow Study, you can make it time-dependent in the Wizard/General Settings.

[@simonfeghali5214]

​@@goengineer for the flow after making it time dependent how do i see how the flow changed as animation?

[@goengineer]

@@simonfeghali5214 We used the TRANSIENT EXPLORER. This can be enabled in the Calculation Control Options and then viewed after running the study by Right-clicking on the RESULTS folder

[@dlamz92]

If the enclosure has no fan but just vents, can we use thermal simulation?

[@goengineer]

You can use Thermal Simulation whether you have fans or not, but the usefulness will depend on how good you are at estimating convection coefficients. Either way, you will want to use conservative coefficients and make sure your design passes which will inevitably lead to larger degree of over-design compared to Flow Simulation.

[@timtruo]

Thanks a lot for the video. I have an SW premium 2016. Is new simulation software compatible with 2016 SW?

[@goengineer]

Thank you for your question. Officially per KB S-019449: “Starting with version 2009, you must have the same service pack for SOLIDWORKS Simulation and SOLIDWORKS. In version 2010 and up, there is no separate installation, and therefore this requirement is automatically fulfilled.

[@muchammadjamil8688]

waiting for next tutorials. .... .

[@muchammadjamil8688]

Dear teacher/ Tutor. hello teacher, I see you are very magnificent in this area. I ask for help with your tutorial related to time dependent flow simulation which is showing result for every part with the graphic (graphic Temperature versus time in second). Thanks a lot for your help. Best regard jamil

[@goengineer]

In Flow Simulation, you can use the Transient Explorer to animate Surface Plots. This is a great way to get an overview of the thermal performance on important bodies involved.

[@jatinjethwa4874]

Can you please send me cad file, excel sheet.

[@goengineer]

The CAD file is in your Flow Simulation installation folder (usually "C:\Program Files\SOLIDWORKS Corp\SOLIDWORKS Flow Simulation\Examples\E1 - Electronics"). You'll have to do a little tweaking (split the CPU body for the thermal study and suppress some components) to it to make it look precisely like mine. This isn't the first request I've gotten for the spreadsheet so I'll also look at hosting that as well. I'll try to remember to post an update to this comment when its up.

[@goengineer]

Ok, it should be up now. Try it out and let me know if you have any questions.

[@josebafernandezguinea6248]

Dear teacher, Thanks for posting this tutorial. I have a doubt: I am running a transient analysis, with 2 bodies in contact. The first body has a heat generation and transfers heat to the second body. I would like the simulation to stop when the second body hits a temperature of 190°C (for instance). I have tried many different ways, non of them worked. Any idea? My original idea was to put a Volume Goal of temperature in the second body, and then go to SOLVE - CALCULATION CONTROL OPTIONS - Goals criteria - manual value - 190°C. It does not work, solver simply keeps calculating after going beyond the set threshold. I tried different combinations, none of them worked. Here´s a little challenge for you, thanks in advance! Cheers!

[@goengineer]

Instead of "manual" set it to ">Absolute Value" and also make sure "Goal Convergence" is the only finish condition you have enabled. Setting it to manual just sets the allowable amplitude excursion for the goal rather than a threshold for termination. See the Flow Simulation Help for more information on this. If you still have a hard time getting it to work, please post your question to the SOLIDWORKS Forum and attach the model for me to look at. Then link it here.