In turbomachinery rotor repairs, are we trying to match the original hardness?

A substack reader (thanks Alex) recently asked me two questions that I think deserve a proper answer, because they come up more often than you'd think in the field.

Questions:

1. When repairing a rotor, are you trying to match the original hardness of the material?

2. Are there situations where you would intentionally coat a section of the rotor with a harder material to prevent future damage?

I will start by providing a very short answer to your questions before diving into deeper principles.

Short Answers:

1.      It would be ideal if we could. But it also depends on what you are trying to do.

2.      Yes. Since the processes used to restore shafts have a side effect of producing hard materials, we sort of lean into it and apply materials that provide improved erosion or damage resistance.

But the real answer to both requires understanding why matching hardness is harder than it sounds 😉

Deep dive

We are talking in the context of steam turbine or compressor shaft repairs.

Let’s start by making an important distinction between these two types of turbomachinery.

Most centrifugal compressor rotors consist of an assembly of impellers mounted onto a shaft.

If a compressor rotor suffers shaft damage, there is always an option to remove the impellers, repair the shaft, and then re-stack the rotor.

It is also convenient or affordable, if the damage to the shaft is extensive, to manufacture a new shaft, and re-use the existing impellers.

Now, some low-power and low-speed steam turbines are also made from assembled disks on a shaft.

But most API 612 “Special Purpose” steam turbines in critical services are designed as “integral rotors”, meaning that the disks are one piece with the rotor. These rotors are manufactured this way as forgings.

And guess what? Manufacturing forgings takes a long time, and the places that can provide the materials of the required quality are few.

To put things in perspective, it can take 8 to 12 months to manufacture a forging for a modest mechanical drive steam turbine.

In contrast, it can take 4 or 5 weeks to procure a bar forging to manufacture a compressor shaft.

So, these long wait times for steam turbine forgings is what has driven the repair industry to find clever ways to restore metal surfaces.

Let’s now talk about the types of shaft/rotor damage you may observe when inspecting a steam turbine or compressor that has been in service for a good 10+ years.

Light Scratches

Exposed shaft areas are common to receive light scratches on the seal areas, journal areas. These can be the product of dirt in the lube oil system, damage from installation or removal of seals, or bearings, etc…

Nicks and scratches are normal. These scratches may only be 0.001” to 0.002” deep sometimes and can be buffed or polished out.

Figure 1- Shaft with light scratched on a journal surface.

Corrosion Pitting

If a shaft area is not well preserved and is in contact with moisture, it can develop corrosion pitting. Here in this picture a hydraulic coupling surface has been compromised by pitting.

The depth of these pits may not be more than 0.002” deep.

If we install a new coupling, the contact between the hub and the coupling will not be optimal, so these areas need to be restored.

You can grind the surface clean of pits, but that will change the location of the coupling and can create a safety hazard.

So, it is best to restore by adding material and maintaining the interference and location of the coupling the same as the original.

Figure 2 - Coupling fit with corrosion pits.

Grooving

In seal areas it is common to see grooving, from stationary seals lightly rubbing the shaft during operation.This is very common in compressors, in interstage seal areas between impellers, or process seal areas.

These grooves will be sometimes 0.005” to 0.010”deep.In order to prevent leakage these areas need to be restored.

Figure 3 - Compressor shaft seal area with grooving from rubbing against stationary seals

Figure 4 - Steam turbine seal area with grooving from rubs and corrosion pits.

Now that we have a picture of what the common damages look like, let’s get back to the original question.

When repairing a rotor, are you trying to match the original hardness of the material?

If you have a damaged rotor, and you wanted to restore its geometry by adding material, you would love to match the original hardness. Afterall the original material specs were chosen to provide the necessary properties for the rotor to be in form and fit for service.

Unfortunately, we do not have an easy way to add material to a rotor made from a low alloy steel and end up with the same original properties, without altering the original base material, and requiring some sort of post process heat treatment.

What if we restore a shaft surface with weld?

When we weld, let’s say a submerged arc or a TIG or MIG welding process, we melt both the substrate (base) metal and the filler metal, and you end up with an unintended heat treatment of that melted zone (Where both metals mix).

We call that the heat affected zone (HAZ). The HAZ ends up with a different microstructure than the original material, and by consequence, ends up having different hardness, different toughness, and even may have residual stresses from the weld operation itself.

In order to restore and heal this HAZ, we have to apply a Post Weld Heat Treatment (PWHT).But to be effective, we have to subject the entire affected areas, we need to make sure we get to 100% of the HAZ. And based on the geometry of the rotor, whether the rotor is in an assembled state, if it has blades installed, etc.… this may not always be feasible.

Most target temperatures for a PWHT on an Cr-Mo low alloy steel are in the neighborhood of 1,100 to 1,200 F.

That much heat, also increases the risk of causing distortion, to any area of that rotor that is already at final dimensions (i.e. a journal, a coupling) may distort.

So in short, the risk of collateral damage during a weld repair of a rotor is high, and that is why we use it when we must make major reconstructions.

For example, rebuilding a complete disk on an integral steam turbine rotor, or re-establishing the surface of atmospheric seals, journals and couplings.

What if we coat with a thermal spray?

Now let’s say the damage is not a “major reconstruction”, we are talking about restoring a seal area that has some corrosion pits, or a journal with some scratches, or a coupling with that has galling damage from removing a coupling.

If the amount of restoration, is in a thickness no more than 0.050” of thickness, we would consider a thermal spray like High Velocity Oxygen fuel a candidate.

We often use Tungsten Carbide, often denoted WC (for the first letters of their atomic symbol).

HVOF torches or guns are literally supersonic flamethrowers where we inject WC powder into the fiery stream. The temperature melts the metal powder, and the velocity ensures that you get a very dense coating that adheres to the base metal.

The tradeoff is that you end up with a very hard surface.

So we use this carefully on journals (making sure not to make it too thick so it does not crack).

We use it on seal areas, where we improve the surfaces resistance to wear.

We use it in impeller bores, we use it in applications where we need to restore a thin amount while ensuring the coating won’t crack or come off.

Why not spray a powder made from a low alloy steel? Won’t we get better results?

Well, the same thing that happens when we weld happens here as well.

When the melted low alloy steel powder lands on the base metal, it will cool really fast, and the whole coating turns into untampered martensite, and martensite is very brittle. So, it is like you are “glueing” in little pieces of a metal that has the correct chemistry, but it very brittle that don’t bond well to each other.

To make matters worse, you cannot PWHT such a thin layer of coating, the coating will crack and come off during the process.

So by design, HVOF and WC produce a hard face coating.We use it to restore, and we like to think that we use it to also enhance those same surfaces making them more resistance to wear.

The one warning is to make sure you understand the process, so you don’t end up with a coating that can crack in service.

Lastly, what if we use lasers!?

 Now let’s imagine the best of both worlds. A welding process that produces a metallurgical bond had a baby with a thermal spray.

This is what “laser metal deposition” (LMD) or laser welding, or laser cladding brings us.

And it is important to make a quick distinction because the topic of the applicability and validity of laser welding for rotor repairs can be argued to death by engineers and metallurgists. So much so, that API 687 for instance does not consider laser welding a valid repair.

API 687 is explicit, they say laser welding is "not recommended for shaft repairs due to insufficient experience and procedures."

I don’t agree with this position. Laser welding has been around for a very long time, and it has been implemented and validated by many manufacturers and repair shops.

Despite that, all original equipment manufacturers that I know apply it, all the independent service providers that I know, apply it as well.

We all use laser welding, not necessarily to provide structural welds, as if we were joining high load bearing components together. We are mostly using it to restore thickness over damaged material. Or to install a very hard face shield or cladding, onto wear surfaces.

Laser welding, like the other welding methods, involves heating both the substrate metal and the weld powder, so you still end up with a Heat Affected Zone. The difference is how small it is. In my experience, and I have measured this, the HAZ from a laser restoration runs about 10% the thickness of what you would get from submerged arc welding.

In a turbomachinery repair shop, we will employ a combination of methods to repair rotors.

To understand which method to apply, you need to understand the damage, the root cause, the implication of a repair, and have a validated process.

Sometimes the first repair option is to “do nothing”. Depending on the condition, sometimes scratches can be superficial or look a lot worse than they are.That is why a proper condition assessment, performed by a qualified inspector using non-destructive examination methods, will tell you which.

A second option sometimes is to remove the damage and re-establish a clean surface. This is called “under sizing” or “bastardizing”.  IF the amount of damage is a few thousands of an inch you can grind a new surface.

Most equipment owners don’t like this because you are not restoring the shaft to its original dimension, and it has an impact on having to compensate the new shaft sizes on bearings and seals.

But in an emergency, under sizing is a valid option.

I’ve written about shaft restoration methods before:

https://www.allaboutturbo.com/post/api-rp-687-chapter-0-annex-g-restoration-methods

And lastly, if we must restore a surface, we can pick one of the three methods I described.

The ranges of applications that I have experience with are illustrated in the image below.