Why do outside tubes of an exchanger suffer fretting the most

[MaterialsDude]

Well I am doing another anlysis on why some Type 304 stainless steel tubes failed by fretting in one of our exchangers. One of the things I noticed with the tubes located on the outrside perimeter of the tube bundle is that they tend to fail mainly by fretting. I know fretting in exhcnager tubes is usually a result of vibrations by high velocities, but can't understand why the outside tubes are the most vulnerable.

 

[IRstuff]

I'm unclear on your context of "fretting" since that's a term used to describe wear damage

TTFN

FAQ731-376

 

[btrueblood]

IF it truly is fretting, then you might ask yourself:

Are the outer tubes located physically closer to one another? Does the shell-side flow path cause high-velocity streams to impinge in crossflow or swirling flow on the tube bundle (outer tubes may bear the brunt of such if it is so)? Is the fretting occuring at baffles, and are the baffles secured against motion?

 

[metengr]

In my experience with heat exchanger tube failures over the years, the outer tubes are located near nozzles where process fluids enter. If there is no impingement plate or diffuser because of failure or improper design to re-direct fluid flow, you can cause fretting damage between the tube OD surface and the tube support plate holes because of local vibration.

Second comment - I have seen on occasion where support plate holes are not sized properly in one part of the tube support plate(s), and it is possible this could be the case.

 

[EdStainless]

another possibility is that the outer tubes always see flow in a primary direction, so their vibration mode is fairly constant. The inner tubes see very confused flow pattern coming across the other tubes. this more random flow may result is more distributed wear.

If you ever see this the check list is short.
1. is there any possibility of corrosion of the support plates? Make them out of the correct material.
2. are the flow diffusers adequate?
3. Were the vibration calculations rigorous enough?

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[rmw]

If the Hx is in a condensing application, all the mass flow entering the section of tubes in question (between two support plates for example) passes across the outer row of tubes, while the flow past each subsequent inner row is reduced by the amount of condensation that occurs on each previous row in the flow path. The velocity is constantly decreasing.

If it is a non-condensing application, then the velocity is ever decreasing as heat is transferred from the gas stream as it penetrates the bundle, result being that the velocity is highest in the outer rows.

rmw

 

[tbuelna]

MaterialsDude,

If your heat exchanger tube has liquid flows, then the surface corrosion you are seeing is more likely due to cavitation and not fretting. The cavitation pitting occurs when the local surface fluid's temperature and pressure are such that boiling occurs. As the vapor bubbles rapidly expand and collapse on the tube surface, very high pressures can result. If the pressure are high enough and/or occur enough times, the tube surface will eventually fail in compressive fatigue. The result of which is the surface pitting (or more correctly "spalling") that you see.

This same cavitation erosion is commonly seen on the liquid cooled surfaces of diesel engine cylinder liners. It is understood that the fluid cavitation is mainly due to the cylinder liner wall vibrating from internal pressure spikes each time that cylinder fires.

Hope that helps. I know a bit about engines, but not much about heat exchangers.
Terry

 

[cloa]

Materials dude needs to clarify it is liquid side or solid side of Heat Exchanger that the failure is radiating out of. Fretting is a solid phenomenon. Cavitation is liquid phenomenon.

 

[whammett]

If there is no diffuser pate to drop velocity of the fluid and change direction, turbulence can result as the procss fluid enters the shell and causes greater delta T and pressure on the outside tube of the bundle. However, this would only account for fretting at the inlet of the shell. Does said Hx have transversed baffles?

"Scientists dream about doing great things. Engineers do them." -James Michener

 

[MaterialsDude]

Thanks guys for the answers.
The shells side is a naptha/steam vapor mix, and the tube side is water. The purpose of the exchanger is to warm up water. There is an impingiment plate at the vapor entrance but it only covers a certain area. I've been told by one of the other engineers that the vapor enters at a rate of 300 ft/s and so given that the failure occurs mainly on the peripheral tubes, I believe that this high velocity is playing a role on the fretting wear observed (not corrosion just wear)

 

[EdStainless]

Get out your copy of HEI and work on some vibration calculations. I think that you will find that this unit is way under designed for these flows.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[rmw]

I have seen impingement plates that did no more than deflect the incoming moisture droplets onto the Hx inner wall where it (1) eroded the wall and (2) bounced back and damaged the tubes below the impingement plates. I have on a couple of occasions modified the impingement plate design to wrap it around the upper rows of tubes so that the splash back couldn't get to the outer rows just below the impingement plate.

Very few (if any) impingement plates that I have seen on the original design were adequate for the service.

rmw