Leak Testing

How about magnetic particle testing? MP or DP will detect cracks, but they may not necessarily be leaks. To test for a leak, the casting or weldment must have a pressure differential between surfaces. One method that is used in shipyards, is to build a box around the item to be tested. The box is then sealed against the item and a vacuum drawn. Any leak will cause the pressure gage to rise.

For castings, the shape is irregular. A number of times, we get seepages. Is there any alternate material to Kerosene ? Can water be used ?

For very large surfaces, doing LPI for very small pin holes is very time consuming and not always fool proof (In sqc we always try to reduce the 100% part due to monotony effect). And it only tells me about the surface, not whether this has passed through.

The problem with water is as I see the much higher surface tension and viscosity, thereby it is much less likely to premeate through very small and complicated paths.

"In fact in certain cases we have heard of DP( dye penetrant) used but may not be feasible for this size."

The size or basic material of the product is no matter for carrying out Dye Penetrant procedure. Ensure degreased cleanliness of the area and carry out test. This is most ideal and inexpensive inspection check for cracks followed extensively in aero space programmes

http://www.ardroxengineering.com/will help you understand the procedure or Google ARDROX DYE PENETRANT CHECK for wealth of information.

The "feasibility" was mentioned due to two reasons

a) In sand castings a lot of false indications are expected especially if the proper surface cleaning is not done

b) since the surface area will be quite a few tens of sqm per item, it will be very easy to miss a few small indications

c) due to surface itself, false indications are expected if penetrant is not carefully wiped off.

d) We have seen quite a few seepages due to microporosity on surface, and over cleaning may mask these.

e) Considering the amount of area and volume of work (in terms of number of castings) it is highly likely that the inspector will be not too careful.

f) We are interested in surface cracks, but also in through porosities. A bit of surface porosity is acceptable, provided they are not through.

Dear sb,

I do appreciate your points graphically explained. I agree Adrox dye penetrant method may not exactly serve your needs.

Best of luck!

Hi sb,

If I have understand well your problem, you have to "leak test" (not just testing for surface opened discontinuities but for "through wall thickness ones".

Maybe one alternative, quick and relatively cheap is to make leak test using the "bubble technique". You need pressurize internally the casting with air (just one bar overpressure is enough, so no high risk due to gas) then while under internal pressure, brush a solution of water and any suitable tensoactive (soap). If leaks are present, bubbles will form in the location which leaks.

For smaller pieces (you talk about some cubic meters) the process is even quicker as you just need to submerge the pressurized component in water and the bubble presence indicates you the leaks.

Another possibility is (once pressurized the part) to check it from the outside part with an ultrasonic leak detector. When pressurized gas goes through a leak generates turbulent flow which in turn generates high frequency sound waves over the audible range but easily detected by ultrasonic detectors specifically constructed for this service. I've used it in the past when I had to detect leaks in many rubber seated doors. The problem was how to pressurize just the door and not the complete room. I did it with adhesive tape and plastic film, creating a "chamber" along the four door borders at one side, closing the door, injecting air in the chamber and "sniffing" with the ultrasonic detector from the opposite side. We detected all leaks.

There are many manufacturers of such equipment and you can find one not expensive for your specific needs. Just google "ultrasonic leak detector" and you'll find a lot.

Kind regards

More problems think of a flat castings - bottom cover for gear housings, as well as gear hsg itself - having depression and ribs. These in casting stage itself are currently tested by kerosene by forming some dams/ filling the depressions and the space between ribs.

The problem is the availability of kerosene in this huge qty as well as HSE hazards.

In those cases you can form, either on a flat surface, or a depression between ribs, a "chamber" with plastic film and adhesive tape and inject air and try any of the methods I proposed (bubble formation or ultrasonic). Of course I realize the problem with kerosene in huge qty: costs and health and safety issues.

I don't know the manufacturing lot sizes, if there are dozens or hundreds of identical dimensions gearboxes or just tailored ones. Depending on the lot size which impose you the frequency and quantity of leak tests, it may be cheaper to prepare some auxiliary devices to form the "chamber".

Kind regards

Sherlock leak detection, under air pressure, or.....

A talc mixed with alcohol sprayed on the surface while pressurized with kerosene will reveal any leakage after the alcohol dissapates leaving the talc.