Shaft defect, crack or weld 1

[thiso]

Hi All,

This is my first post so thought i would say hi and fun this past some people that may have seen something similar in the past.

Situation is that we had some blokes cleaning a shaft (300mm dia x 2100mm) getting ready to assemble and they came across a defect which has raised some concern.

This shaft is supposedly brand new (never seen dynamic loading), it was assembled with bearings and shrink disks as per specification and then the job was put on hold for 18 months so the assembly was disassembled and prepared for storage. Wrapped in denzo tape, crated and stored in a shipping container.

The defect is located in between a shrink disk, and a bearing, around 200mm apart this is approximately in the middle.

To me it looks to be the toe of a weld, which could be from a manufacturing defect. As the defect has no sharp end and has a deviation towards the top I do not think it is a crack. The red line shows the path of the defect.

What is this suggesting to you?

I would love to provide more detail but i cannot due to the sensitive nature surrounding this shaft.

Any input would be appreciated.

Cheers

Thiso

 

[metengr]

thiso;
First what I would do is contact a nondestructive testing (NDT) company and have them perform a shear wave ultrasonic examination to evaluate if the surface indication has any radial depth.

If the shear wave UT results indicate no depth into the shaft, I would use emery cloth and polish the surface followed by wet fluorescent MT or Liquid Penetrant testing.

It may be a surface blemish or line of corrosion pits or a crack, hard to say from one photograph.

 

[metengr]

Note: if the shaft material is nonmagnetic, a Liquid Penetrant examination must be performed in lieu of wet fluorescent MT.

 

[thiso]

The shaft has been locally cleaned with emery tape. NDT has been conducted shortly after to determine the defect.

Mag part showed no cracks, nor did die penetration testing. This is strange due to the shaft having defect having a lip which you can feel when you run your finger nail over the crack. However the testing did not show.

As for UT, the test conducted was Pulse Echo is this similar to what you described? This too picked up no result.

You may be correct in that it is just a blemish, however the fact there is a physical edge is deceiving me.

Would an etch test show if the area has been welded? Is it possible to carry out such a test on site normally?

I have attached some more images taken from the NDT report, clearly show there are several defects in the vecinity.

Quite strange, i have never come across this before. Could these defects be a result of forging? As the shaft was forged then turned to achieve the final design.

 

[metengr]

It could be a minor weld repair. What is the shaft material?

 

[thiso]

From my understanding the material is 4340 forged and turned.

Some info extracted from material cert;
*Tensile*
HBW - 269
0.2% Proof Stress - 720MPa
UTS - 890MPa
Gauge Length - 50mm
Elongation - 20%
Reduction Area - 50%

I am unsure as to why it would require welding as it is a new shaft and there is no component seat at this location. Unless, it was a machining error. for example they under cut the shaft and needed to build up to adjacent dia.

 

[Andries]

thiso,

You may be right about the repair. Years ago I had a similar problem and after various tests we found that the shaft had actually been repaired (weld build-up) and then machined. We thought it was new. So look into this possibility.

Andries

 

[metengr]

I would polish, and have a final surface NDT. If you are concerned about weld repair for this shaft material you can have portable hardness testing performed to evaluate hardness variation from possible weld repair. Otherwise, leave it alone.

 

[JJPellin]

We experienced a shaft failure of a 5.25 inch diameter shaft in a critical liquid ring vacuum machine. During examination of the failed shaft, we found that the vendor had performed a weld repair along the edge of a keyway. They considered that acceptable. Eventually, we found that the shaft failure was not primarily caused by the presence of the weld. There were a number of other issues including a stress concentration at the bottom of the keyway and rotor resonance. But, some manufacturers consider it acceptable to weld repair new shafting. We have modified our specifications to disallow it in critical applications. We were able to confirm the existence of the weld repair by metallurgical testing following the failure. I am not sure what methods could be used non-destructively.

Johnny Pellin

 

[MiketheEngineer]

Why not bring in the mfg and ask him to explain what is going on??

 

[Cockroach]

Could do the old black light with sodium dye penetration, look for the crack. Maybe if something more sophisticated is required, shoot it (X Ray) and look for a potential crack to light up.

If is is a weld repair, it could be a thermal crack where the root failed to gain penetration into the alloy steel matrix. On the other hand, you could be right in noting it as the toe of the weld and not being an issue.

Probably sodium dye NPT would be the quickest, less expensive and comprehensive exam. Good luck with it.

Kenneth J Hueston, PEng
Principal
Sturni-Hueston Engineering Inc
Edmonton, Alberta Canada

 

[MintJulep]

Possibly the inclusion of some contaminate during the forging.

That could explain why you don't get an indication with DP or MP.

 

[1gibson]

Visual with magnifying glass? Could be simple as a deep scratch that didn't polish out with the emery paper. If it's a non-critical area, do a little more sanding?

The zig-zag looks suspiciously like a component got stuck on the shaft and was wiggled to get it past that point.

 

[MrMyers]

Can you take a metallurgical replica to investigate the grain structure at high magnification.

 

[RolMec]

Hi thiso,

this could be a line of slag, coming from impurities @ semifinish manufacturing process. This may be superficial and thus will not give so much when doing NDT. However, i'd have the whole shaft checked, where there's one bug there maybe others.
Concerning welding: For 4340 i'd be most careful, esp. after heat treatment. How one will correct all those changes in microstrucure? c/o website of weldingdesigndotcom
If the zone is not that important, check whether you can loose sthg of the dia. and have that place lathed away.
For a shaft, i'd not accept a surface fault like that. That's a shower to everyone who likes to discuss anything afterwards, and a crack just loves to come from such a place.

Regards

R.

 

[rob768]

etch the surface with nital. If there is a weld, i will become visible

 

[thiso]

Hi All,

Thank you for your input and you suggestions.

We sent the shaft off site for more testing, mag particle and die pen as previously done. These reviled no cracks or identified areas of concern.

The TOF (time of flight) UT, which did not pick up anything outside the bounds of the standards.

The shaft was etch tested to determine if it was a weld. Results show that it was not a weld. The Micro structure visual inspection completed along with the etch testing determined that the "imperfection" was in fact corrosion. So with a little more wire wheel action the imperfection soon vanished into oblivion.

Thanks again for your assistance, if you guys want some more information let me know and ill follow up.

Thiso

 

[MintJulep]

Star for metengr for identifying the defect in the very first response to the OP.