Hi.

My thread "cavitation" has a follow-up question. Flow corrosion is due to shear stress.

Q1: Is there a level of the shear stress where corrosion occurs?

Q2: If yes, can this critical level be linked to the aggressiveness of the fluid

A1: Corrosion is a chemical reaction. Rate of corrosion depends on the amount of corrosive agent available at the wall. Due to the similarity of viscous and diffusive boundary layers corrosion is high in the region of high stress.

A2: Once the surface is corroded there is another factor that can speed up the process. It is erosion which is due to mechanical interaction between fluid and surface. Cavitation can play a major role in erosion.

Angen

Hi Angen

Thanks for your answer. I have a few more:

Q(A1): Can something be said about what "high stress" is? 10Pa? 1e3Pa?

Q(A2): Do you know how to evaluate or estimate erosion from a CFD calculation? Can the level be linked to e.g. "k"? (and if yes, is there a critical level?)

/Christian

A1: In fact, corrosion rate is not a function of stress, but it is a function of concentration gradient. Similarity between viscous and concentration boundary layers imply that thickness of concentration boundary layer is proportional to the thickness of viscous boundary layer. So I do not think there is any magic number that can be assign as a high stress (or concentration gradient). Generally, the greater is stress the greater is corrosion rate.

A2: I do not know, but I am sure that typing a few keywords in a web search engine will provide you with plenty of sources.

Angen

Christian,

Years ago I did some work to try to predict flow induced corrosion / errosion in valves. The key was to have data on "material removal rate" from testing for the specific material and fluid involved, and then use this with the CFD results for velocity and turbulence. I'm fairly certain that this work was not published, but there may be similar information available.

Alton

Hi again

Is it possible to say what a "high" level stress is in numbers?

And the same question for "low" level stress

"High" or "low" are totally arbitrary in this context. Even at zero flow rate (no stress) there will be a corrosion due to a molecular diffusion of corrosive agents. The only way I can think about assigning some meaning to "high" or "low" is to define what is an acceptable level of corrosion and then to correlate it to stress.

Andy