forged shaft flaws

[Tmoose]

What are folks' experiences with hot rolled 4140 shafting6 to 10 inch diameter being free of internal flaws?

I worked for a precision spindle company that made lots of steel shafts, usually from forged 8620, sometimes 4140 etc.
The only material flaw I recall was a seam or crack uncovered when a section of a fairly large (over 18 inch?) billet (Hot rolled?) was whittled down to a toothpick, relatively speaking.

My current employer has 6 foot long machine shafts made typically from annealed 4140, and includes an ultrasonic inspection, but would like to start skipping that step. I'm most concerned with surface flaws in radiuses, etc, so believe the finished shaft needs to be magnafluxed.

thanks

Dan T

 

[arunmrao]

Tmoose,it depends on the source of your round stock . If you can trace back the manufacturing process,perhaps you might feel confident in dropping the UT . Else ,due diligence must be exercised for such large parts.

I have not failed. I've just found 10,000 ways that won't work." — Thomas Edison
_____________________________________

 

[metengr]

I would not skip ultrasonic examination of shafts for the sake of cost or a false sense of security. IF this is for rotating equipment, it is worth the incremental cost of a volumetric exam to avoid shaft failure in service.

 

[mfgenggear]

Are the shaft rough machined?

if so I would recommend Mag & Demag at this point.
verify parts are good from the forging house.

a final mag of finished parts is mandatory.

 

[steelmtllrgst]

If final parts are fully tested, which I assume they are, a simple cost analysis will give you the answer.

is intermediate NDT more expensive than additional costs associted with failure of NDT in final part (assuming same failure rate as you find now intermediately)???

 

[metengr]

I believe there is some confusion here based on the post by steelmtllrgst. A volumetric examination (like UT) is a one time exam to evaluate the soundness of the forging either before or after final machining. There should be no intermediate volumetric exam and a final volumetric exam.

Surface NDT as a minimum should be after all fabrication is completed. The soundness of the forging should be performed before machining in the event the forging must be rejected.

 

[mfgenggear]

metengr

you are one of the most helpful on this board.
we all appreciate your Imput.

however in this case I will have to disagree in one point.
My experience both with mag & UT it is better to rough machine forgings then NDT test.

The rough machining removes all surface imperfections.
better results and no false indications.

In my line of work 99.9% of all parts for aircraft
are NDT after final machining?

 

[metengr]

mfgenggear;
Perhaps I was confused. I agree with you in that both volumetric examination and surface exams should be performed for critical components. I thought that was what I had stated in my second post, not the first post. Maybe what I should have added in the second post is that the sequencing of these exams is based on how your Level II or Level III examiner would like the surface prepared.

The point to all of the above is that eliminating volumetric examination needs to be carefully considered for rotating shafts. Simply relying on past vendor experience is NOT good enough. I always like the adage, trust but verify.

 

[mfgenggear]

Metengr

Thanks for the reply :>)

I agree.

 

[tgmcg]

We are presently dealing with two steam turbine rotors which have just failed NDT inspection after finish machining. There is a large number of linear indications on both smaller diameter ends of the rotor...hundreds of indications.

We are puzzled as to why this was not detected before via UT. UT was performed at the foundry.

No way would I EVER accept a vendors assurances that rotors should be defect free and therefore skip volumetric examionation. Some vendors will ship you swiss cheese if they can talk you into accepting it.

That said, I'm wondering why volumetric examination failed us in this instance. It's looking like our delivery schedule is blown.

Best regards,

Tom McGuinness, PE
Turbosystems Engineering

http://www.turbosynthesis.com

 

[metengr]

tgmcg;
I can give you one possibility having gone through a recent technical audit because we are purchasing a forging, not just a cast ingot, for a large turbine generator. The volumetric nondestructive testing for a large forging is usually conducted after forging and rough machining and is performed in zones.

For most reputable mills that supply rotor forgings for turbine or generator rotors, the UT is normally done using an automatic process. Some still use manual scans. The zones I referred to above are evaluated based on how critical the locations are relative to stress levels in service. My point to this is that there may have been different acceptance criteria depending on location in the rotor. I would review the procedure used by the foundry and their acceptance criteria for the entire rotor.

 

[tgmcg]

The turbine vendor is a well known name. The forge is located in Canada. Unfortunately the turbine vendor does not provide much source inspection anymore. We need to investigate the whole process but this will wait till after we formulate a recovery plan.

I've been reading of late about phased array UT and eddy current techniques. Do you have any suggestions as to the best NDE method for detecting core defects on rotor forgings? In this case the raw A470 Class 4 forging had a diameter of 9" in the bearing and seal areas and a finished diameter of 4". Today we are performing FPI to assess how many of the defects penetrate the surface.

Is there a good reference for acceptance criteria for rotor forgings? ASTM A470 limits linear defects to 0.125". In our case a few indications exceed this, but the entire shaft end is peppered with smaller defects. A photographic guide would be helpful.

Thank you for your suggestions.

 

[redpicker]

It sounds as if you are finding non-metallic inclusions near the center of the original ingot. If they are, it is not suprising that they were not revealed during an ultrasonic inspection, particularly if the inspection was performed prior to heat treatment. Generally, these inspections are looking for internal bursts and voids and they would not be expected to detect inclusions. These are best handled by steel cleanliness requirments.

rp

 

[metengr]

tgmcg;
The best inspection method for rotor forgings is an automated UT system using phased array transducers. Nothing beats this system provided you have a qualified procedure, reference standard and certified examiners. For our rotor forging, we carefully evaluated the vendors NDT program and reviewed their acceptance criteria. The vendor also qualified their phased array procedure on actual portion of a rotor forging that contained defects. The acceptance criteria for critical zones in the rotor forging (after forging, heat treatment and rough machining) was smaller than 0.125". For the non-critical zones, 0.125" was used.

 

[mfgenggear]

TGMCG
Please advise What type of steel & heat treat?
Parts where ground?

 

[TVP]

My thoughts are similar to those of redpicker: linear indications that deep in a billet/bar/forging are most frequently non-metallic inclusions, and therefore not necessarily detected using UT. I also agree that these are best controlled by clearly specifying the requirements for chemical composition, inclusion assesment, and working with the steel mill on a suitable manufacturing method (melt chemistry, ladle additions, slag type, use of vacuum, etc.). Any chance you can provide the composition, especially O, N, S, & P? Any inclusion ratings (ASTM E 45 or similar)?

 

[tgmcg]

Thank you all for the great suggestions.

According to our inspector who witnessed the FPI inspection of the two rotors, "the shaft journal and seal areas looked like the Milky Way". In several areas, the FPI results lined up with the MPI results. Some surface indications exceed 0.500" in length. Despite this, the vendor continues to advocate a fracture mechanics analysis of the shaft as a possible means for determining shaft suitability for service...wishful thinking I suspect. The vendor would love to ship these rotors, and the customer would very much like to take posession of the machine prior to the scheduled shutdown.

In view of the unusually large number of confirmed surface defects, I have to wonder at what concentration of subcritical defects does fracture mechanics analysis cease to be a valid method for determining shaft integrity? I am far from convinced that fracture mechanics analysis is an appropriate tool for this situation. Any thoughts/comments?



Best regards,

Tom McGuinness, PE
Turbosystems Engineering

http://www.turbosynthesis.com

 

[tgmcg]

Thank you all for the great suggestions.

According to our inspector who witnessed the FPI inspection of the two rotors, "the shaft journal and seal areas looked like the Milky Way". In several areas, the FPI results lined up with the MPI results. Some surface indications exceed 0.500" in length. Despite this, the vendor continues to advocate a fracture mechanics analysis of the shaft as a means for determining shaft suitability for service...wishful thinking I suspect. The vendor is desperate to ship, and the customer is desperate to take posession of the machine prior to the scheduled shutdown.

In view of the unusually large number of confirmed surface defects, I have to wonder at what concentration of subcritical defects does fracture mechanics analysis cease to be a valid method for determining shaft integrity? I am far from convinced that FMA is an appropriate tool for this situation. Any thoughts/comments?

 

[metengr]

tgmcg;
I agree. I would not recommend you accept the new forgings under stated NDT results. I deal with large steam turbine rotors and having to disposition a new forging with this density of defects including your statement

Some surface indications exceed 0.500" in length.

is unacceptable. If I were the Purchaser, I would not accept the rotors. My questions is this, was there an acceptance criteria for the new forging? I have done numerous audits and for the large turbine OEM’s, most have very conservative acceptance criterion for new forgings with no disposition of flaws using fracture mechanics. If the forgings are found to contain rejectable defects by volumetric UT, they are rejected.

 

[redpicker]

I am not sure I understand your abbreviations

FPI-Fluorescent Particle Inspection?
MPI-Magnetic Particle Inspection-OK this one I get
FMA-Fracture Mechanics Analysis?

Understand that a mag indication is not a defect unless it exceeds the acceptance criteria. What is the accpetance criteria for the shafts you are inspecting? If the indications exceeds those, they are defects and must be removed or the shaft must be rejected.

It sure sounds as if what you are seeing are non-metallic inclusion stringers. It is rare that these are detected with dry powder MPI, but with wet flourescent MPI, they can be seen. Even with wet MPI, it is rare that such indications would exceed 0.125 inch, but I've seen it. To detect such imperfections with a volumetric UT would be very rare, unless they are much larger than you are describing.

If these are non-metallic inclusions, and they are oriented parallel to the direction of principal stress, they may not be all that detrimental. I have accepted such imperfections in motor output shafts (4-6 inch diameter) with no problems reported. I'll admit that the reasons I accpeted them included that I had no justifiable reason to reject them; the manufacturing specification was not written very well and required wet mag to inspect for forging defects, quench cracks, etc..., which these definately were not. I will add that after seeing them, and talking to the mill that produced the steel, I started requiring AMS 2300 cleanliness testing on the bar stock used (aka "Aircraft Quality") because I wasn't entirely comfortable with accepting them. You might consider a dye penetrant inspection to convince yourself there is not any depth to these indications. This may or may not help you with your decision.

I don't think I'd put much weight behind a fracture mechanics analysis, since the big concern these would cause me would be initiation points for fatigue cracks, which fracture mechanics would not consider.

The big question that will need to be answered, of course, is who is going to pay for them? Unless the vendor can be shown why they don't conform to the specs he was given, he is going to want to get paid. Of course, the customer would be better off to scrap them (eating the cost) if they are going to fail in service, but he won't be happy about it.

rp