Hi. Some aspect of your question is not clear. It was better to define your software as a solver and simulation device. The other thing that I have to say is about it is not clear how's your opinion about unclear results.
Having said that, some influential and important aspects that you have to consider,
First of all you have to pay attention your simulation and geometry exactly.
Secondly, you can valid your numerical simulation based on results of an experimental research or similar numerical study.
Thirdly, you should simulate one geometry and evaluate various flows based on different parameters such as Reynolds number, heat fluxes and so on.
Finally, you have to result in these various consequences based on changing influential parameters such as Reynolds number. By observing trends of accomplished results, you can achieve the rational results that you want.
Dear Osakpolor really the subject have been treated profusely in the literature. there are a lot of paper about numerical simulation of heat exchangers and enhancement of heat transfer. check the papers about that subject published by C.C. Wang or the works done by gent university in belgium. You need to build your geometry as realistic as you could, and check every boundary condition used. Velocity inlet, pressure outlet, wall, simetry and periodic are the usually applied. Be carefull with the unsteady condition created by periodic flow behind the tubes. For the higher Reynolds number, even inside the laminar regime you could find that problem.