CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   CFD assistance (http://www.cfd-online.com/Forums/cfx/73094-cfd-assistance.html)

enr_venkat February 26, 2010 08:37

CFD assistance
 
Hi Everyone,

I'm quite new to ANSYS -CFX and i don't have wide experience in CFD analysis. However, i'm exposed to fluid mechanics, heat transfer and numerical methods and i've read CFD fundamentals. I still find difficult to understand certain concepts. I'm a fresh postgraduate and i used CFD for my thesis. I'm practising CFD now. I'm currently working on an automotive exhaust component such as EGR valve to determine the temperature proven capability. I'm not sure about the boundary conditions that i use. To be precise, the valve pushes down inside the EGR assembly so that flue gases enter through the gaps and leaves out. Valve timing would be 150 ms operation. I calculated the mass flow rate of gas mixture (standard data taken) and i know the inlet pressure. Generally the total pressure is the sum of static pressure / gauge pressure and operating pressure. I also know the differential pressure across the valve (5Kpa) and inlet temperature (750 C). My aim is to calculate the temperature distribution on the solid plunger that moves up and down i.e I'm looking at temperature distribution on plunger due to fluid flow. I don't get proper convergence trends. My initial conditions are static pressure and temperature. Even i'm not sure about these values. I had guessed and put these values. How do we judge initial conditions and how do we interpret results in CFD. When i ran the simulation, the temperature distribution on valve seems to be almost equal i.e. i see slight variation. I need to convince my boss that CFD analysis helps to accelerate product development. Please help me.

rikio March 3, 2010 11:22

What type of problem are you running?
Maybe a transient analysis is needed for different temperature distribution on the plunger. Because the T will go to a almost the same value for the whole part in a steady state analysis, just as you have got now. Am I right?

enr_venkat March 4, 2010 02:13

That sounds right ! But steady state simulation is performed with the available data. Do you suggest me to run a transient simulation and check the results. I didn't use thermal boundary conditions for the product. I kept the Heat fux option to be adiabatic which is not right in real cases. Atleast i could show my boss about the flow patterns. How do we calculate physical timescale for fluid and solid domain ? I also get messages during iteration that "some portion of fluid flow is trying to enter the outlet domain, If you persist this condition, switch over to opening type boundary conition" Why should i do that ? I hope opening boundary conditions depend on the type of problem handled. I was trying to do a conjugate heat transfer and visualise the hot spots in the fluid and solid domain. Please pitch on physical timescale and boundary conditions


All times are GMT -4. The time now is 08:11.