Fracture location in large cast iron part

[ebarba]

Hi,

We’re going to attempt to locate a suspected fracture in a large cast iron machine part. The part is a shallow, hollow cylinder with vertical axis, about 400 mm tall and 1500 mm in diameter. To access the inner part of the cylinder there are a number of inspection doors about 120 mm día in the outer part of the circular wall (not very easy to get in and use nondestructive fluids). The bottom is accessible from a outside, as well as the circular outer wall.

We have an ultrasound thickness measurement instrument, will it locate the discontinuity of the crack, if the mic passes over it? My guess is that there will be a noticeable variation on the measured thickness were the crack is. Is this reasonable to think?

Thanks!

 

[racookpe1978]

The answer is a definitive "maybe".

Depends on the "tightness" of the way you hold the sensor against the outside surface: How good is the "connection" of the sounding sensor against the surface of the steel, what liquid coupling agent is present (might be good, and what angle of the sensor's "cone" of noise out from the surface towards the interior? For example, assume the outside of your PV is very rough, and the outside is curved so you can't "aim" the signal very well towards where you assume the flaw is inside the metal. Makes sense that your echo back at the surface would be weak, distorted, and not very clear.

Most important, what is the size and depth and orientation of the flaw against the surface and the sound wave "cone" aimed from the outside inward towards the flaw? See, if the flaw is parallel to the "cone" of noise, it may reflect very little energy. If the flaw "canyon" is perpendicular against the cone of noise going into the steel, the echo (which is the signal received back at the surface) might be very clear.

 

[MintJulep]

No.

 

[metengr]

The UT will not be effective because the graphite flakes will attenuate the sound waves through the casting. I would use a good fiber optic borescope with an articulated probe head and magnification for a visual examination.

 

[arunmrao]

I would go with metengr's choice. How do you propose to salvage the casting , or are you planning to scrap it ( I don't ever like the idea of scrapping).

Good Luck in your defect hunt.

"Even,if you are a minority of one, truth is the truth."

Mahatma Gandhi.

 

[ebarba]

The thickness of the wall is 10, 15 and 20 mm, depending on the specific part of this cylinder-like structure.

Internally it's very clean and the surface is raw casting (no machining). Externally we expect it to have some deposits after thoroughly cleaning it and unmachined in most of structure. Both internally and externally a paint coating is present (thickness < 0.5 mm).

We plan in using the usual ultrasound gel to improve the contact.

The measuring device has a 5 MHz probe and is capable of both T-E and E-E operation. Since water is getting inside the cylinder, the suspected crack should go through the whole thickness of the wall. About the orientation: hard to be 100% sure, but since it has to be through the wall, it will somehow be perpendicular to the surface (and unfortunately rather parallel to the ultrasound cone).

My guess (hope?) is that the discontinuity will manifest in a sudden and rather noticeable difference in the thickness reading, in a localized spot or line across the part, but yes: I am aware that the flaw being parallel to the sound beam will make it difficult.

We also have a boroscope, so we'll also try using non destructive fluids to make the crack appear. It's cumbersome, though because of the reduced access diameter, large inner diameter and large surface. I was hoping to be able to do it all from outside.

Thanks!

 

[Compositepro]

If you have leakage though the wall, why not apply slight pressure and locate the leak with soapy water?

 

[ebarba]

It’s impossible to pressurize, the top part is large and practically impossible to make airtight

 

[Tmoose]

If Compositepro's excellent suggestion is for some reason impractical I'd brush "paint" the outside with LPT penetrant. Probably Fluorescent penetrant. can

I'd expect in less than 30 minutes to see "indications" on the inside.

I picture this casting is pretty valuable. Loctite 290 ( wicking) is quite effective at sealing porosity etc, introduced from whichever side is higher pressure in service. Although the LP Penetrant might ruin its effectiveness.

 

[ebarba]

The cast in itself is not very valuable, but a lot of other stuff is attached to it. If it's really cracked it will mean stopping a 24/7 line for at least one week to replace it.

There is no effective pressure on the wet nor dry side. The outer part is constantly sprayed with process water and we suspect that the crack is so dismal that only under specific load conditions allow for the water to seep in by capillarity. To have an idea, it takes months for about 100 liters of water to get into the cylinder, so the rate is really low.

We'll give it a try with the penetrant and see what happens.

Thank you all for the excellent suggestions.

 

[EdStainless]

We used to leak test large parts that we couldn't pressurize by florescent dye penetrant.
Add some dye to the water (just a hint, if you have additives in the water now such as corrosion inhibitors it already fluoresce, or if there is any oil in it) and then search using fiberscope blacklight.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[mrfailure]

As metengr said, you really can't use ultrasound for this application because graphite will hide the indication, no matter what the probe. I also am skeptical that this vessel can effectively be weld repaired, especially if it happens to be gray iron.