Failure of diaphragms in reciprocating compressor 1

[qualitypro]

We are experiencing frequent failures in the 1st stage diaphragm of compressor since being put into operation last year. The material spec states that it is made of Inconel 718 material. The in situ hardness measured was HRC 34 as against HRC 36. Can a marginally low hardness of the diaphragm really impact the fatigue strength to this degree? The life of this component was quite short. Please share your thoughts as to what could be the potential failure causes for this and remedial actions that could be taken to stop recurrence of this issue. Are there other investigations that need to be done to get to the root cause of this failure?

 

[3DDave]

From all I've seen and read, hardness doesn't have great effect on fatigue life.

I would contact the pump maker and whoever is the supplier of the diaphragm(s). You could also retain a stress analyst.

Edit :34 vs 36 is very different from 21 vs 36.

 

[Tmoose]

For sure check with the mfr first.

Seems like the diaphragms would be being replaced under warrantee.
What are you compressing? Is it on the list of acceptable environments for the compressor?
Are inlet and outlet pressure and temperature conditions OK per mfr specifications?
How about drive motor rpm, and amperage, RE: OEM specs.

What do the failures look like? Please provide detailed pictures of the failed diaphragms, and the installation.

Sometimes long runs of inlet piping can cause a degree of "supercharge" with surprising results.

"The material spec states that it is made of Inconel 718 material"
Is this the mfr's spec, or results from a PMI tester?

 

[metengr]

First, did you have a proper failure analysis conducted by a metallurgical lab to determine the actual failure mechanism? Fracture initiation site, material defect? Failed material properties..... This all has to be performed to determine a particular failure mechanism(s). Once this step is completed, you evaluate causal factors related to design and service conditions until you find one or more root causes, and decide on corrective actions. This is not going to be accomplished on a forum.

 

[EdStainless]

and 718 has multiple heat treatments (they optimize strength, creep, or EAC resistance).
Work with the manufacturer on this.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[mrfailure]

I certainly agree with metengr's approach, and that you should not just look at one anomaly that you can measure .

That said, to your original question: Yes, there is a relationship between hardness, yield/tensile strength, and fatigue strength. You said hardness was measured in-situ, presumably with a portable tester. You will have better results when hardness is measured in the lab.

 

[qualitypro]

Thanks to all for your replies.
The failure analysis is underway. The failed diaphragm material has been sent to a lab for investigation and the results are pending. We were told that the hardness seen was HRC 21 as against the requirement of HRC 36 minimum. The OEM states that per their BOM the material is Inconel 718. We have asked for the CMTR to verify the material traceability and to see if the proper heat treatment was indeed performed.

 

[metengr]

Thanks for the follow-up. Make sure you are involved, ask questions and work with the lab. You are on the right path to determining the cause and prevention.

 

[MagBen]

HRC 21 is even low for annealed 718! Material in use was not aged?! If the strength is critical to your application, it is not supposed to use this material at annealed condition. Use it as solution annealed + aged, ideally double aged.

 

[EdStainless]

Don't go by the hardness, the difference between the two aging treatments is not much. RC 36 vs 38, and you want the lower one.
Fully annealed this stuff will be RB 90, which is below RC20.

Solution anneal at 1700-1850°F followed by rapid cooling, usually in water, plus precipitation hardening at 1325°F for 8
hours, furnace cool to 1150°F, hold at 1150°F for a total aging time of 18 hours, followed by air cooling.

While this is the lower strength of the two common aging treatments it will yield better fatigue resistance.

http://www.specialmetals.com/assets/smc/documents/inconel_alloy_718.pdf

I find 718 an odd choice. There are alloys that you can cold work to similar strengths, which have better corrosion and fatigue properties. 718 is a great option if you are operating at 1200F, but otherwise I am not so sure.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[MagBen]

Ed, I got thousands of hardness data points when annealed at 955C/1 hr, which were never lower than 20RC. 20-21RC is the very lower end. HRB 90 is about 9HRC if it is convertible. surprised to see RB 90 from Special Metal data sheet!

 

[EdStainless]

We see 90-95 RB on tubing all of the time. But our in process anneals are 1950F.
We use 1850F for a final anneal, but the values are still mid-90's.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube