I am analyzing a small furnace, and to start, I am just trying to get a simulation which gives similar results as the test lab. I can't post pictures, but it can be best described as a shell and tube heat exchanger (6 tube passes with complex orientation). The fluid in the shell is air, and the fluid in the tube is an air/combusting propane mix. I am modelling both fluid regions as air, and in the combustion chamber ("tube"), I have included a solid body (turned "off" in component control) which generates 35,000 BTU per hour.

Boundary Conditions: I am specifying volumetric flow rates at the inlets of both the shell and tube. I am specifying ambient air pressure at all outlets. It is important to note that there is one outlet for the tube/combustion chamber, which I will refer to as the exhaust, and that there are FOUR outlets for the shell.

Goals: For my goals, I am measuring the exhaust (tube outlet) temperature, as well as the average fluid temperatures and volumetric flow rates out each of the 4 shell outlets.

Problem: My resulting exhaust temperature (tube outlet) is too high, and my shell outlet temperatures are too low, indicating that heat exchange is not happening at a realistic rate. (I am comparing these to real-world lab results). More heat should be transferring than the model shows. Due to the thin wall nature of the heat exhanger, I had assumed that it is a convection issue. I have adjusted material properties the best I can, and I have accounted for some radiation. I have also specified turbulence parameters at inlets and outlets that seem reasonable. ARE THERE ANY PARTICULAR SETUP OPTIONS I CAN LOOK INTO TO GET MORE EFFECTIVE HEAT EXCHANGE? WHAT WOULD BE THE MOST LIKELY THINGS TO TROUBLESHOOT?

It is pretty important that I use the boundary conditions I have specified. If I start putting boundary conditions on the heat exchange surfaces directly, the model will become too specific to particular scenarios and not be useful for future investigation. I do not want want to specify surface heat flux or convective heat transfer coefficients - these need to be driven by the simulation.

Thank you for the help.