Discussion Groups

Simulation Mechanical and Multiphysics

New Member
Posts: 2
Registered: ‎11-15-2012

2d heat sources?

157 Views, 3 Replies
11-15-2012 04:54 PM

Hey All,

Suppose I want to simulate a physical system where I have a disc-shaped heater that is producing uniform heat evenly over its area. I believe an axisymmetric simulation is the fastest way to simulate this. Is the correct way to set up this simulation to designate the heater (a line in 2D) a heat source with a constant magnitude? In reality, the heat generated varies radially (because of the changing arc length), is this taken into account by the software?

Best,

Keith

Valued Mentor
Posts: 510
Registered: ‎08-30-2012

Re: 2d heat sources?

11-15-2012 07:15 PM in reply to: brownka

Hi Keith. Welcome.

You are correct on both points.

• The heat source is applied to the edge of the mesh, or a "line" in a 2D model. You indicate this in the software by putting the lines of the mesh on a unigue surface number. (If you use the automatic 2D mesher, each construction line creates a different surface number in the mesh.
• I think the input in a 2D model is energy per area (such as W/m^2). But even if there is an option to enter the total energy (such as W), the processor calculates the equivalent energy/area, and multiplied by the area of the revolved mesh gives the total energy input.

John Holtz, PE
Mechanical Engineer
Pittsburgh, PA

16 years experience with Simulation Mechanical
New Member
Posts: 2
Registered: ‎11-15-2012

Re: 2d heat sources?

11-15-2012 11:25 PM in reply to: AstroJohnPE

Thank you for the fast reply.

Just to be clear, if I make a surface that is parallel to the y-axis a heat source in an axisymmetric simulation, that is equivilent to having a disc heat source in which the heat is distributed evenly on the surface of the disc?

Thanks!

Keith

Valued Mentor
Posts: 510
Registered: ‎08-30-2012

Re: 2d heat sources?

11-16-2012 07:57 AM in reply to: brownka

That is correct.

Anything associated with a 2D axisymmetric element (2DAE) is treated as if it is revolved around the Z axis. The elements are revolved. A "nodal" force applied to a 2DAE becomes a circular line force. A constraint at a node supports the entire circumference of the model at that radius. If you were to connect a truss element to a 2DAE, the truss element is providing a force at the node where it connects, so it is treating the truss as if it were revolved.

One thing to aware of. The analysis is usually based on the mesh being revolved 1 radian (57.296 degrees), not a full revolution. This effects any thing where the "total load" is applied to the model. Per unit area loads are not affected. For example, a force of 100 lb applied to a node is equivalent to a total of 628 lbs in the real world. (= 100 lb/radian * 2 * pi radian/revolution). A pressure of 5 psi or a heat flux of 1.25 J/s/m^2 is applied to an area which is independent of whether the area is based on 1 radian or 2pi radians.

John Holtz, PE
Mechanical Engineer
Pittsburgh, PA

16 years experience with Simulation Mechanical