[dommartyn]

For my Masters Dissertation in Mechanical engineering I designed a new windscreen wiper system.

It a nutshell it is a single wiper arm/blade system that is propelled in a linear motion using the Maglev principle. To prove this single wiper system would create a better vehicle fuel economy I used CFD.

I did 2 months worth of running CFD using ANSYS Fluent 17. I simulated the normal style twin oscillating wiper blade system in three positions, Parked sweep, Half sweep and full sweep. I then simulated a control ( No wiper blades or arms ) Finally I simulated a single linear wiper arm/blade in the middle position of the car windscreen. I drew all the CAD exact from a donor VW Polo 2010

The results for the pressure on the frontal area were obtained and then converted into Drag Co-efficiency numbers.

I truly expected the Half sweep wiper blades to have the highest drag co-efficiency as they were in the full force of air flow (Set at 70mph).

However, all the results came out backwards. What I expected to be the worst was the best. My Single Blade system came out worst.
I checked and double checked with no change.

After research I discovered that Bernoulis theorem states that an increase in the speed of a fluid occurs simultaneously with a decrease in pressure or a decrease in the fluid's potential energy. If the highest turbulence results gave the lowest drag co-efficiency it is possible the reason why.

However, I still cannot get my head round the idea that using a single, linear wiper arm in the perpendicular position in the centre of the windscreen would in essence cost more money in vehicle fuel to run than the standard twin oscillating system commercially used, due to the Drag co-efficient result.

I am stuck on writing my conclusion. After 163 pages and 24000 words I cannot manage to write that I have proven my own idea to be totally wrong. Yet, if I can find a way to justify it, I will be happy.

Any advice I would be greatfull.

[FMDenaro]

I don't know the details of your flow problem and if yor numerical results make sense but the Bernoulli integral requires the fluid to be incompressible, irrotational and without any source of energy dissipation. That is not the case of turbulence.

[dommartyn]

Thank you for your reply.
Not too sure what you would like to know to clarify the flow problem.
To explain a little more.
The standard twin wiper system was modelled on a vehicle profile. The vehicle was placed in a fluid domain enclosure (Rectangle) 31.3 mps was set as the air flow speed. Air density was set as normal. Ansys fluent did 10,000 iterations and gave a pressure reading on the windscreen as an integral of the pressure balance front to rear multiplied by the frontal surface area.
5 models were set as advised.
The results were
Setting Drag Coefficient
Half Sweep 0.642610665
Full Sweep 0.672939959
Parked Sweep 0.675295802
Maglev 0.678422453
Control 0.681166852

The fluid is incompressible, irrotational and without any energy dissipation.

However, there is obviously turbulence.

With the wiper blades set with the full force of 31.3 m/s air flow the turbulence caused a drop in pressure, hence getting a lower drag co-efficient number.

My only reasoning for the single blade having a higher pressure, less turbulence, hence higher drag co-efficiency was the Bernoulli theorem.

Please feel free to enlighten me.

[FMDenaro]

Originally Posted by dommartyn

Thank you for your reply.

The fluid is incompressible, irrotational and without any energy dissipation.

However, there is obviously turbulence.

I strongly advice to study the fluid mechanics basis before doing any usage of CFD tools...what you wrote as student would drive me as professor to stop any master exam...

[flotus1]

Apart from that, the differences in the drag coefficient seem to be quite small. You have to make sure that the contribution of all other sources of error is smaller than the difference you expect to see.
In other words: you have to perform a proper error analysis http://www.cfd-online.com/Wiki/Ansys..._inaccurate.3F