xy plot help -- large variation in results
 

I'm fairly new to FLUENT. I have been simulating the convective flow of air above a heated plate.

I'm particularly interested in looking at the velocity behavior of the centerline (to then compare it to what theory predicts).

So far, my simulations have converged (looking at residuals and monitoring points of interest), and the contour plots look like I would expect them to.

However, when I do an XY plot of the centerline velocity of my buoyant plume*, while the values definitely follow a trend, their variance is considerably large.

*XY plot, axial velocity vs distance, following a pathline that originates at the center of the hot plate.

I've uploaded an image here:
http://picasaweb.google.com/roommate...46918201836722

It doesn't seem normal to me that the velocity along a pathline can change abruptly from one point to the other, considering that what I am doing is tracking a single particle along its path. (I particularly can't understand thos three outlier points below the curve.)

My question to the experts is: does this seem like a normal-looking plot? Is it ok to have such variations? Are my simulations not really converging yet? Am I doing something worng while plotting?

Turbulence model:
I am using an RNG turbulence model. I have used laminar flow to model a free jet entering a room, and the variation of my results is similar. (I mention this because I initially tought that the variation may be due to the presence of turbulent eddies).

Mesh:
I have used several meshes, from fine to finest, and the results are the same.
The plotted results correspond to the following mesh:
Element size ~.001 with growth rate of 1.05. The hot plate has a diameter of 0.2m, and the room dimesions are 2x2x2.5m.
http://picasaweb.google.com/roommate...49574699460306

Simulation:
3ddp, second order upwind calculations. Boussineq calculation of density.