Non-destructive strength measurement

[CaliSolarMechE]

I have a small population of completed welds whose strength is suspect. We're not prepared to rework everything, and we have too many to perform destructive strength testing. I need a non-destructive way of estimating the strength of the joint.

In words, the joint is supposed to be a complete joint penetration groove weld all-around, 3mm size, used to splice two ASTM A500 4" square tubes (3mm thickness) together end-to-end.

Can you think of a method I can use for this non-destructive examination need? This isn't a regulatory requirement. We are thinking that we will take another look at our applied loads and see if "what we have" is good enough.

Thanks!

Elliot

 

[Guest102023]

Radiography or ultrasonic. A boroscope is simple and cheap if you can reach the weld root with it.

 

[CaliSolarMechE]

Brim, thanks for your response.

We have performed radiography with the disappointing result that the welds fail the standard, but to an unknown degree. If the X-ray image could tell thickness with any certainty, we would be doing fine!

I found a method of UT called "crack tip diffraction" (

http://www.ndt-ed.org/EducationResources/CommunityCollege/Ultrasonics/MeasurementTech/cracktip.htm)

that could determine depth of penetration. Is there any other likely method?

 

[MintJulep]

Even if you could measure strength non-destructively (and I don't think you can with and degree of real confidence) static strength may not be the whole story.

If the joint is supposed to be CJP and it's not then it's fatigue performance will be compromised.

Just bite the bullet and rework things now. It will be faster and cheaper for you in the long run.

 

[dhengr]

UT would normally do this, tip depth indication, but on 3mm thick mat’rl. I’m not real sure it would be accurate enough to tell you the percentage pen. What are you using as a backing strip, or what is the quality of the weld at the inside surface under normal conditions, or are you fooling yourself about ever really getting a good full pen. weld? Who speced. a full pen. weld, and is it really needed? ME’s are notorious for calling out full pen. welds for everything. Is this primarily a static structure or is it fatigue critical? How is it loaded and at what stress levels? Section a few of the welds to see what you really have on average, maybe you can live with that. How did this problem come to light, failures in the field?

 

[Guest102023]

Outside NDE will be expensive and ultimately semi-quantitative at best, so I concur with previous advice given to bite the bullet and rework.

 

[Duwe6]

With a 3mm thickness, any repair will involve cutting out the entire weld on that side of the tube. If you have 2 or more adjacent sides that need repair, cut the tube at the weld, reprep, and reweld. Shop welds are less than half the cost of an X-ray or UT. On something this thin, not repairing it is very expensive and timeconsuming.

If the defect was lack-of-penetration at the root of a 1/2" 12mm thick tube, and there were more than a couple of tubes, UT testing and engineering recalculation might be worthwhils. The UT tech has to be able to perform Crack Sizing, because reliable depth measurements are critical.

 

[Mr168]

How about cutting out what appears to be the worst on film, and destructively testing one or a few?

 

[Tmoose]

Could you add some external reinforcement?

http://www.fgg.uni-lj.si/kmk/ESDEP/media/wg11/f0430003.jpg

http://www.renehersebicycles.com/headtube lug upper.JPG

 

[Ron]

NDE and NDT are primarily available to locate and characterize flaws in the weld or base metal. They are not intended to determine strength, and agreeing with MintJulep that there is not likely a reliable method to do so, your best opportunity to determine weld strength is a proof test. Secondarily, determining the geometry of the weld and calculating the strength based on assumed average material properties can be done, but with caution.

Again, piling on with MJ, strength is not the only consideration. Some weld flaws, particularly root discontinuities, can be detrimental for long term crack potential, even in mildly dynamic situations such as moderate temperature fluctuations. As MJ noted, if the conditions are susceptible to fatigue or vibration, then why take on the liability and risk of known defects in the weld?

Failing tests or observations have to be resolved or your liability can increase exponentially.

Why did it get so far with bad procedures? You might want to re-visit your QA/QC program.