FASV task report

Name of associated manuals : Planned / Effective working schedule. Back to index SESSION 7

Dabireau (mfn03) and Massart (mfn08)

Flow Around a Solar Vehicle.

Objective:

This task aims to make the three dimensional mesh of a solar vehicle developed for the next World Solar Challenge in Australia

TASKS:

Sphere

.1: Construction of the mesh of the fluid domain around a sphere using PREBFC

.2: Determination of the suitable boundary conditions of the problem.

Solar Vehicule

.1: Construction of the mesh of the fluid domain around the real solar vehicle using PREBFC

.2: Determination of the suitable boundary conditions of the problem.

.3: Computation of the flow due to the motion of the solar vehicle using FLUENT.

.4:Determination of the subsequent drag coefficient.

MANUEL MAILLAGE 3D

SPHERE

In order to improve our knowledge about Fluent, and to calibrate our computations, we decided to simulate the flow around a sphere. The results would be easy to compare with the experiments.

Back to index

1:Construction of the Mesh

We decided to put our sphere in a cylinder in the aim of compare our numericals results with experimental results given by Mr. Schimt.

Here is the geometry:

Back to index

2:Determination of the boundary conditions

We wanted to know how the choise of the boundaries condition can play on the results.

In fact we had two possibility for the output :

• pressure-outlet
• outflow

We could also choose the boundary conditions between :

• wall
• symetry

Back to index

SOLAR VEHICULE

1:Construction of the Mesh

As we explained it in the previous session, we didn't obtain the real geometry of the solar vehicule... We weren't more lucky this time. In fact, we only have a drawing of the vehicule with some dimensions. Nevertheless we decided to built the geometry with such poor informations. It needed a lot of time, but we were able to simulate something! This is the reason why our results are not acceptable.

In order to improve our knowledge about Fluent, and to calibrate our computations, we decided to simulate the flow around a sphere. The results would be easy to compare with the experiments. (cf. to the second part of our task report)

Here is the document we used:

This is a Fax-copy of a drawing of the solar vehicule. Of course, the scale is precised. This is the only document we obtained. In fact, the people of the ENSICA worked on a link between their software and PreBFC , but they didn't succeeded before the dead-line of our defence.

With a lot of PATIENCE, we choosed many points to built the structure of the vehicule. The geometry remains very precise refering to the datas we had.

Here is the geometry:

The mesh:

Back to index

2:Determination of the boundary conditions

The outlet boundary condition used is Outflow.

The ground and the walls have the same boundary condition of speed ( here 20 m/s).

Back to index

3: Computation of the flow

We had a lot of problems to simulate this case...

In fact, it was impossible to reduce the node length under 0,3 m!

The capacity of the computer seems to be the reason of this failure. Indeed, the quantity of memory was not enough.

But we think the problem could be solved by a better construction of the geometry. It seems obvious that a perfect geometry, we mean a geometry directly obtained of a drawing software , would be better.

So the results below are given to illustrate our work and are not good.

Back to index

4:Drag coefficient

0.015 is an evaluation of the drag coefficient Cx (in the conditions of the simulation we described before).

Because of the conditions of the simulation, this value mustn't be considered seriously. Furthermore, this value is very small and appear incorrect.

Back to index