Material Selection for Increased Cavitation Erosion Resistance 2

[sry110]

Hello All -
I am chasing an issue with a torque converter that we use in our mechanical package and I could use some outside advice. The blades in the turbine wheel are exhibiting advanced erosion due to cavitation in the working fluid (ISO VG 32 light turbine oil). The root cause of the accelerated wear lies in the operating conditions which we are discussing in a parallel path with our customer; meanwhile, I am also tasked with revising the turbine wheel design to incorporate a different material for the blades which will result in higher durability under the current operating conditions.

The current turbine wheel design incorporates blades made from cast 17-4PH H1150 stainless steel. My understanding is that 17-4PH stainless steel is theoretically a good choice for this application; however, I am wondering if we could achieve an appreciable increase in erosion resistance (and correspondingly increased time interval between failures) by changing to blades machined from 17-4PH forging. My premise for this thought process is based on the forging having a finer grain structure and smoother surface finish than the casting, resulting in increased erosion resistance. That said, my area of expertise is in designing and assembling these systems, but not in the metallurgical makeup of the internal parts, so I am admittedly out of my depth here and could use some help. Thank you in advance for any advice!

 

[Heaviside1925]

OP,
I appreciate your situation. A customer comes in with a failed part, "all hands on deck" "throw any ides against the wall and see what sticks". The parallel path you are describing may not be the most pragmatic approach. The first question should be, Is cavitation a normal or abnormal process condition? If its abnormal, the solution would be to redesign the system to prevent this. After all engineering approaches to remove the cavitation in the system are exhausted or are deemed unfeasible do you approach cavitation as being a normal operating condition. If cavitation is determined to be a normal operating condition, then model all the areas the cavitation cold be affecting. I say this because you may design the turbine to handle the cavitation only for the customer to came back because the casing or valving or other component is now failing. The dynamics of cavitation involve extreme stress to a very localized area. You need a material that can either deform to absorb this energy or is extremely hard. Since cavitation is an interface event, consider a ceramic coating. There are many hard metals that can be sed as well but will lead to a reduction in strength. If only the turbine is to be hardened, first consider options from pump turbine manufactures as well as valve manufactures for cavitation resistant materials of construction. Then evaluate the forgeability and machinability along with any other fabrication costs to see if these materials are even an economically viable candidate. Once materials have been selected, model these materials in the design to ensure their properties won't cause unintended consequences, such as reduced cyclic loading as a result of reduced material strength. This can be a very arduous process, which is why, if cavitation can be removed in the system, then this should be the initial solution.

 

[EdStainless]

How large are these? What is the thickest section?
You could start by polishing the surfaces of the cast impeller.
Are these being solution annealed prior to aging?
They should be.
In fact we used cast, anneal, overage (H1150M, 1400+1150), then we did whatever work was needed to finish the part.
Then we re-annealed and aged.
And speaking of that you could easily go to an 1100F aging to gain a little hardness without losing significant ductility.
Or maybe even 1050F, depending on where you think fatigue will become an issue.
In the past I have used Nitronic 40 (21-6-4) for cavitation resistance because it work hardens greatly (ultimately reaching 200ksi).
The only bulk alloys that excel at cavitation resistance are Co alloys.
I have seen coating used that we cobalt alloys and not carbides, but ductile alloys.

There is journal called Tribology that publishes a lot of papers on this subject.

https://www.researchgate.net/public...n_Resistance_Influence_of_Material_Properties

https://pubs.acs.org/doi/10.1021/acsami.2c19210

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[sry110]

@Heaviside1925 - Thank you for your reply! I am very confident that the root cause of this material failure is the customer's insistence on operating our torque converter in its reduced pressure mode for prolonged periods, whereas it was intended to operate in the reduced pressure mode for only a short transient (less than 1 minute) during each run. I believe that the reduced pressure in the torque converter traction circuit exacerbates the oil's tendency to cavitate at the turbine blade surface due to the localized pressure in that area dropping below the oil vapor pressure. We have other customers using the same system while following our recommendation to operate in full pressure mode, and their time between service intervals is on the order of ~5x that of this particular customer who insists on running in the reduced pressure mode.

So, yes, I agree that the correct approach to solve the problem is for the customer to change their operating profile to match our recommendation, but unfortunately they are not committing to doing so while asking us to develop this Band-aid approach to help get them a bit more time between failures. I've told them that any material we throw at it will degrade rapidly unless they modify their operating profile to match our recommendation, but they aren't interested in doing so. Tough situation for us.

 

[sry110]

@EdStainless - The turbine blade is 2.25" wide, approx. 1.40" long, and approx. 0.44" max section thickness (at the entering end of the blade). Polishing the blades is an interesting idea, and something we could accomplish without too much additional work. Do you have any detailed recommendation as far as polishing method, surface finish to achieve, etc?

 

[EdStainless]

I would think that a small gelt sander (finger sander) would work.
Start fairly rough (50), I would do the first and then 80 with solid paper backed belts.
Provided at this point you have smooth surfaces I would move to cushioned belts (non-woven, like ScotchBrite) at 120/180/240 to get a nice smooth surface.
It doesn't need to be prefect, but any improvement will help.
If the user won't change then just charge them to polish and maybe even coat.
This will likely double the cost of the impeller.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed