repair welding of a cracked compressor disk 1

[NJavan]

Hi to All,

We are going to repair welding of a cracked compressor disk. The cracks are formed due to low cycle thermal fatigue. The disk material is ASTM A471 class 12 (Fe-0.45C-0.8Mn-1Cr-0.5Mo). We should qualify the welding procedure before welding on the disk. However, A471 is not available to make a test plate for welding qualification. Is it possible to use another material for making a test plate for qualification purpose? Are 4140 or 1045 steel being suitable choice? Would you please comment on this matter?

Many thanks,

Nima Javan

 

[EdStainless]

You need to match chemistry (well at least similar) and thermo-mechanical history.
So do you know the exact HT on your disc?

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy

 

I concur with EdS; this is a critical part and you need to develop a procedure on the same material, not just the same category of material (e.g., as with P-number categories in ASME IX).
Distortion and residual stresses are potential concerns that a weld qualification test gives no information about.


"Everyone is entitled to their own opinions, but they are not entitled to their own facts."

 

[NJavan]

Thanks EDS and ironic metallurgist. Yes, the part is critical. But, the exact materials is not available. So, we have to use a similar material. 4140 has similar composition to that of A471-class 12. But, as you suggested, before welding, I think we should heat treat 4140 (Quench and tempering) to produce a tempered martensite microstructure in the steel. Thanks.

 

SAE 4140 is a reasonable facsimile.

"Everyone is entitled to their own opinions, but they are not entitled to their own facts."

 

[EdStainless]

yes, that is how to do it. HT the sample material to give you the same strength.
Then you can work with pre-heat, welding, and PWHT.
Since this is a rotating componant you are really going to need to look at toughness and NDT.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy

 

[Tmoose]

Do you have pictures showing the cracks in the disc?
What operating temperatures are involved?


Is the disc to be repaired in situ? (So no PWHT as mentioned by others)
Will the operating conditions that caused the cracks continue?
If so, I'd be plenty scared that weld repairs will be quite short lived and develop cracks of their own.

Have you determined how deep the cracks are?
Are they in a location of high and significant hoop stress ?

Sometimes if grinding parts thinner by 10% or less, with large gentle radiuses and with excellent craftsmanship removes the indications it can be a satisfactory repair.
I'd follow up the grinding repair with shot peening.

 

[NJavan]

Thanks EDS and ironic metallurgist for your comments.

 

[NJavan]

Tmoose,

Thanks for your comments.

The picture is attached. The cracks initiated from the balance holes on the surface of the disk. Some cracks are through thickness (i.e., the depth of some cracks is about 4 cm!). The cracks are probably formed due to thermal fatigue as a result of start up and shout down thermal stresses. The operating temperature is about 300 oC. The life of the disk is about 150,000 h. The hardness of the disk is about 240 HB with a tempered martensite microstructure.

The simulation results showed that the stress in the cracked region is about 350 MPa.

We should weld with sufficient pre-heat and then immediately PWHT using proper conditions.
Would you please give me your idea about GTAW filler metal selection? Two options are plausible. The first, is the use of match filler metal (e.g., ER110 or ER120). The second is the use of an undermatch filler metal (e.g. ER80) to reduce the susceptibility of cold cracking. What is your opinion?

 

What is your opinion?

Scrap and replace.

"Everyone is entitled to their own opinions, but they are not entitled to their own facts."

 

[NJavan]

Dear ironic metallurgist, Thanks for your comment. But, what is your reason for this comment? The length of the crack?

 

[kingnero]

This is obvious a critical repair (cfr. application, stresses, material, ...), meaning this must be done to the highest standards. Are you confident you(r welders) will be able to weld this without any internal defects?
Keep in mind, any significant defect can (will) cause cracking to reappear.

While I regularly advise on difficult repairs, I do not feel confident on the (practical) feasibility of this one...

 

While I regularly advise on difficult repairs, I do not feel confident on the (practical) feasibility of this one...

Precisely the basis for my advice.

BTW I have to wonder when I see balancing holes 180° opposite each other.

"Everyone is entitled to their own opinions, but they are not entitled to their own facts."

 

[NJavan]

Dear kingnero,
Thanks for your advise.

 

[NJavan]

Dear ironic metallurgist,

Thanks for your advise. You mentioned that " I have to wonder when I see balancing holes 180° opposite each other ". Would you please explain a little bit more? Thanks.

 

[Tmoose]

How many cracks ?

Did your simulation include disc rotation as well as thermal goings-on ?

If just one or 2 I'd be mighty tempted to machine out the cracks with a ball end mill and call it good. Machine a similar detail opposite in an effort to maintain balance.
Will the rotor be re-balanced with the blades installed? Is there an OEM procedure how to seat the blades ( rev it to XXXX rpm )
if the part is to be balanced as an assembly.

I'd Also machine a radius on all the holes to improve the stress distribution.
Whacking/pressing a large ball bearing into each hole entry to create a chamfer loaded with compressive stresses might be of more benefit.

 

[weldstan]

The best welding repair will not be equivalent to a new machined part in terms of fatigue cracking resistance. Any such repair should be considered temporary - sometimes very temporary.

 

[Tmoose]

If only one balance procedure has been performed in the component's life so far, and the balance set up was good, and the operator was capable, then the drilling should only have been in one region ( the "heavy" spot ) on thin disk shaped object.

"Significant" holes opposite each other suggests a severe over-correction, or the set up was bad, and then fixed, or the operator mistakingly drilled on the light spot ( doh !!), or something.

 

[Compositepro]

Lining up the balance holes at the same radius as the bolt holes seems to be a design error which is causing a significant stress increase in the area between holes.