Fluids

Fluids

Topics related to Fluent, CFX, Turbogrid and more.

Transient Analysis

    • M Usman
      Subscriber

      Hey Everyone,  I am working on a problem in which a hot body having temperature 200-300 degree celcius is getting cooled by the air which is flowing over the body  having 0.5 mach speed. I want to know the time it takes for the body to attain normal temperature or alternatively for specific time period likle 2s, 3s or 5s, what is the temperature drop in body. I have tried normal steady state approach and Conjugate Heat transdfer approach using stead state and transient mode as well. But results are same as the temperature contours which initially developed on the body remain same and the iterations on further converge those results like wven in trainsient, I am unable to control the time for which the air is flowing over the body. And when I do the FSI, the analysis time it shows is 1 second which I am assuming is flow time of air over the body. Transient analysis gives the same results only soure time changes if I try to change the flow time.

      Help needed from the experts.

    • Rob
      Forum Moderator

      What does the FSI bit do? 

      If you're not seeing any cool down of the solid, how are you defining the surface between the fluid & solid?

    • M Usman
      Subscriber

      I am seeing solid cool down but the analysis time is 1 second which I am unable to change.I want to see how much body gets cooled down at diferent times like in 3 seconds, 5 seconds etc. And for that I want to know which approach should I use.

      • NickFL
        Subscriber

        It sounds like you are using ANSYS mechanical for a thermal analysis and not Fluent or CFX. This is the fluid board, and you may want to post over there to get a better answer. These fluid packages could solve for the thermal temperatures in the body and the flow field. Obviously, if you just need the temperature in the body this is a bit of an overkill. What are the boundary condition you actually have?

        There is a nice Mechanical thermal analysis example in one of the Innovation courses (it involves radiation, but the model building is the same). I recommended you look at that. Keep in mind that if you are modeling the air domain in your model, you won't actually see any fluid moving.

        But to answer your question directly, click on Analysis Settings in the tree. In the details panel change the end step time. You can also specify the interval you want to solve at.

    • M Usman
      Subscriber

      I am actually using CFD fluent to do the analysis (Conjugate Heat Transfer Analysis) and only using FSI to see the temperature changes on the solid body. The problem I am talking about is to set different flow times in ansys fluent to see the temperature changes in the body. For example in the image below the body temperature changes from 300 degree to lower temperature when air with 27 degree celcius temperature flows with 0.5 mach speed. Now here analysis time is 1 second written which I beleive is air flow time ovewr the body. I want to variate this time to see when does the body reaches a normal temperature. But even in transient analysis, when I change the flow time and import the results to the body using FSI, only the source time changes while analysis time remains the same.

      • NickFL
        Subscriber

        Based upon what I see, you are importing the body temperature. Then you really are giving Mechanical nothing to solve as you are telling it what the solution is.

        Think back to your heat transfer class. What type of heat transfer do we have here? And when solving convection problems, what was the one term we were constantly having to find using correlations? That is the quantity you want to be taking from Fluent and mapping onto the surface of your Mechanical model. Then you can set up this transient analysis with an initial temperature of 300° in the body and then the software will calculate the temperature field in the body.

        Keep in mind with the heat transfer coefficient you also need to define the free stream air temperature.

    • M Usman
      Subscriber

      I am only using mechanical to see body temperatures as fluent shows static and total temperatures which are different from body actual temperatures. My all working or approach relies on ansys fluent. I just want to vary analysis time or air flow time to see body temperatures after time periods of 2s , 3s or 5s etc. Ansys fluent automatically gives me the cooling of body to certain temperatures as shown above after time period of 1 second. That is my concern here.

    • Rob
      Forum Moderator

      Please can you explain what you mean by body temperature? 

    • M Usman
      Subscriber

      Actually there is a hot solid body having 300 degree celcius temperature and is being cooled by air flowing over it with 0.5 mach speed and 27 degrees temperature. I want to see how much it cools the body over the period of time which can vary or I can control. For that I did Conjugate Heat Transfer analysis with air as fluid and rod as solid body. Body gets cooled from 300 degree celcius as shown above but it is for specific period of time. I want to cool it to room temperature but I am unable to do that as I am unable to change the time for which air flows over the body. The more is the air flow time, the cooler the body will be.

      • NickFL
        Subscriber

        You want to be mapping the heat transfer coefficient on the outer surface not the temperature of the body.

    • Rob
      Forum Moderator

      You can set the number of time steps and time step size in Fluent to simulate a certain amount of time. 

    • M Usman
      Subscriber

      Yes I have done that on Ansys Fluent assuming that it would work but the temperature contours developed on the body remains the same irrespective of the number of seconds I set as flow time. And when I import those temperatures on the body in mechanical, only the source times changes while amnalysis time remains the same. This is the reason I am stuck on this simple problem from many days. Below is the first image of steady state analysis, and second image is of transient analysis in which I set the flow time to 2 second. You can see the temperatures on the body are same.

    • Rob
      Forum Moderator

      How many time steps did you use to get to 2s? 

    • M Usman
      Subscriber

      200 time steps, each time step is of 0.01s.

      • NickFL
        Subscriber
        1. Are you solving for the temperature distribution within the solid body in Fluent, correct?
        2. You are mapping this temperature onto the body in ANSYS mechanical, correct? If so, what time step are you using?
        3. Again, the approach I would take is to be mapping the heat transfer coefficient from the steady state model onto the outer surface of a transient thermal mechanical model. Then you can simply run for whatever length of time until the body reaches the temperature you want. 
    • Rob
      Forum Moderator

      And did each time step converge?

    • M Usman
      Subscriber

      In actual I want to simulate how the cooling of Fighter Jet Gun firing at 3400-3600 rounds per minute and Jet moving at 0.5 mach speed through air gets cooled by the convective heat transfer due to air flow. The concered area is gun barrel in which heat is generated and its wall reaches 200-300 degree celcius temperature. So to first just set the right approach, I am using a solid rod having a temperature of 300 degree celcius gets cooled by air. So yes i am solving for the temperature distribution on the solid body and I am using mechanical only to import temperature loads on the body to see the temperatures on the body. Not running ay solution in Mechanical. Now the only difficulty I am facing is to vary the air flow time and get diferent temperature contours.

    • Rob
      Forum Moderator

      To echo Nick's comments. There's no need for Mechanical here. 

      Flow around the cylinder is done in Fluent, and have a look at the NACA wing for some ideas about compressible flow. The cooling will also come from that, but you will need to decide on your initial conditions. Ie do you cool from a stationary point or assume it's at Mach 0.5 and then cools?

    • M Usman
      Subscriber

      Boundary Conditions: At inlet, the fluid (air) is flowing at 0.5 mach speed with 27 degree Celcius as its temperature. On rod boundary walls, the temperature is 300 degree celcius. The meshing is done in such a way that Conjugate Heat Transfer Occurs and body gets cooled downn by air flow. But results become constant 1 second air flow time and no matter how much time I set as air flow time in transient analysis, the results remain the same.

    • Rob
      Forum Moderator

      Please post a contour from Fluent showing the temperature field of the air flow & solid: a plane down the centre is sufficient. On the same surface also plot velocity and Mach Number, post those too. 

    • M Usman
      Subscriber

      Here are the contours:

      • NickFL
        Subscriber

        Are you running this simulation with constant properties? If so, you can save yourself a bundle of time by removing the flow equations from the transient solve. Here would be a possible approach:

        1. Conduct a steady-state flow analysis. (& save!)
        2. Set-up the model as a transient solve.
        3. Patch the temperature of 300° onto the solid body. (Patch is under the Initialization panel in the Tree. Be sure to patch ONLY the temperature as if you do anything else you will destroy the flow field you already solved for and are continuing to use.)
        4. Turn off the flow and turbulence equations and solve only for the energy. This can be set in the Tree under Solution->Controls->Equations
        5. Set up an Monitor points on the body, .cxa animations or other quantities before you run.
        6. Save the case file to a new name (indicate that it is transient)
        7. Run, and enjoy a good cup of coffee while the computer does its thing!

        I would also recommend that you do a free-slip wall condition on the upper and lower surfaces of your tunnel. It won't change the results much but be more aesthetically pleasing.

    • Rob
      Forum Moderator

      OK, and if you run on for some time steps?  How did you set the temperature in the solid?

    • M Usman
      Subscriber

      I set the temperature in same way. By checking the energy equation option and applying the temperature of 573 K on solid walls.

      • NickFL
        Subscriber

         

        On closer look (sorry, I am editing what I wrote earlier) you have a problem with the interface. What you have in the above panel is say there is a fixed temperature at the surface of the body. Keep this set as coupled that allows the heat transfer thru the interface from solid to fluid. This is why there is no cooling to your body because you are setting the wall temperature as 300. You only want to set the initial temperature to be 300. The easiest way is using the patch command in intialization tab.

        My recommendation is to go back and look at the example of conjugate heat transfer in tutorials.

         

    • M Usman
      Subscriber

      My simulation is working now correctly. Thanks alot for your help.

Viewing 19 reply threads
  • The topic ‘Transient Analysis’ is closed to new replies.