Not logged inAmerican Welding Society Forum
Looking for an NDT method that would best show depth of penetration of an EB weld. The part is a cylinder, approx 1/8" dia x 1/4" long w/ a wall thickness of of .050 to .060.
The cylinder starts out as two semi-curricular halves, a flat (.013 thick) is placed between them and the cylinder is reassembled. (forming two separate chambers that run the length of the cylinder)
The EB weld runs along the seam fusing the left & right half with the flat. Since this is not a full pen weld, we would like to try and find a non-destructive method of measuring the depth of penetration.
Any suggestion would be appreciated.
I saw this post a while back and was waiting for someone to reply. I'm curious what methods of NDT you might have at your disposal. I'm not as familiar with RT as I would like, but I would think that if you could position the shot to get a clear view of the welds, it might give you a volumetric picture of the depth of penetration. These guys that are more familiar with RT can probably give you some guidance on what to use to compare the weld with to determine distance in relation to the depth of penetration. This is a really small part compared to the work I'm used to. I don't know if you have plenty of these to spare, but it would be easier to saw one in half, polish, and etch the weld zone to see exactly what you have. Sorry, I didn't give you much to go on.
I'm inclined to agree with you. It would seem that one could qualify the parameters needed to produce the required depth of penetration by initially destructive testing some samples, then perhaps pull a sample periodically for cross-sectioning to verify the parameters are still producing the required depth of penetration.
If I am picturing your joint right, it is like welding a T-joint from the back side, except the plate you are welding has a square butt joint at the same point the "T" meets the outer cyclider. If this is correct, I am wondering why you say it is not a full pen weld. You would have to achieve a full pen weld through the outer cylinder in order to fuse with the inner partition. Are you really wanting to know the width of fusion at the point the weld enters the partition? (Which would have to be less than .013")
I would also have to agree with the previous post that on such a small part, that unless they are prohibitively expensive, that sectioning a small percentage for quailty assurance seems to be the way to go.
You are close on the joint design, but the .013 flat extends to the surface. I referred to the weld as only a partial penetration because it doesn't go through to the ID.
We have ultrasound, mag part, and dye penn in house, but the size is too small for us to handle.
As far as the destructive testing, that has worked for us in the past but isn't an option this time. A new application requires us to NDT the entire seam.
Is there any way to do a pressure test? It sounds like you might have a number of parts to test and I would think you could do that type of testing fairly quickly. It would not demonstrate weld penetration but it could prove that the parts are serviceable.
I am inclined to agree with Chet. I cannot think of an NDT technique that would give reliable and repeatable results in this instance without doing some experimentation. I would suggest pressure testing. Although, it would be a good reaearch project for your in-house Level III.
Does anybody know if Eddy Current would work in this application?
Also, wouldn't machining these parts from solid stock be much simpler?
I was thinking of Eddy current myself. Depends on the material type. Didnt see it listed. Must be a ferromagnetic material. If it is you would still have to do some destructive testing to prove out your setup. Would take alot of work but I think it could be done. I used Eddy in the past to somewhat successfully predict case depth in induction hardened shafts.
What is EB welding?
I'm not sure about this, but I thought that any metal could be tested with eddy current testing, not just ferromagnetic materials?
Also, will it work on parts as small as was described?
Yep I think you are correct Chet....does not have to be ferromagnetic. I was thinking of mag partical. Getting old...has been a number of years since I dabbled with Eddy Current. As far as size of part I wouldnt think that would be a problem...they can custom make probes of various sizes.
Is this going to be a long term production item or are you just running a few of them? If you do find soemthing that will work are you going to buy the equipment and if so what is you budget.
I still believe Eddy current might be the route to take. It would however take alot of development so it would not be something that could be done over night. As I stated in earlier post I used it to predict case depth in axle shafts....wokd pretty nice at time. Could get it right within about .010" when it was maintained. Problems we had was when you would change heats of materials. To get by that problem we would have to heat treat sample and enter the data into system then all was find. But that required destructively testing part as well. Wasnt an issue for us due to the large volume of parts ran. Unit I used was produced by a company in Michigan by the name of K. J. Law....dont know if they are still in business or not....do a search and see if you can find them. If they are contact them and see if they can help. The unit I used work on the principle of linear regression. Well enough babbling....
I would think that UT could only tell you if you have voids or not. Unless they have advanced since I messed with them.
Something else you might consider is thermal imaging? Havent worked with it personally but maybe someone else can guide you hear. Just throwing out some ideas.
The material is 420 stainless.
EB = Electron Beam
The part is very complex, so as not to confuse everyone my description was broken down into its basic geometry relevant to the weld. Many hours of machining go into just the sub-components of this part, a 1 piece unit is beyond current technology.
Pressure testing would not be an option because the customer requires weld depth penetration.
Eddy Current may be a possibility, although everyone I have spoke to say it won’t work for our application. Have been getting mixed info on x-ray too. Our staff says no NDT method will work but I did find a company that believes they can UT the parts, hopefully I will know something in a week or so.
Seems the biggest problem is the size of the part and we want to measure depth of penetration, not just full or incomplete.
I appreciate all the input. Still looking so anyone with x-ray or Eddy Current experience wants to offer an opinion it would be greatly appreciated.
Would a bore scope be small enough to check inside?
Brian J. Maas
Thats what I was going to suggest. A borescope to at least see the penetraton. Of course measuring it would be next to impossible.
By the way, Eddy Current will work on both ferro and non-ferromagnetic materials, but works better on Non-Ferromagnetic materials. That is why it is so widely used to detect cracks on aluminum aircraft skins.
UT can give you depth and thickness measurements, however, in thin material like this, it would not be possible to focus the sound beam to a point where you could measure LOP in such thin material.
If this company says they can UT this part, I would verify their results with destructive testing and continue to do so on a random basis. Don't let them know which parts are being cut. If they can tell you fairly accurately, please tell us how they do it, because I would like to know. I would also like to know what they are using for a calibration block.
I think maybe your client needs to provide some assistane in this matter.
A borescope may work (I'm not too familiar with the available sizes), we have been using microscope's for assembly and inspection.
Planning on confirming the results via destructive testing.
Found another firm that is suggesting x-ray, but I'm not so sure they are going to be able to measure prtial penetration along the seam.
Will keep you posted.
I think I would rethink the design of this thing and go to an extrusion if possible. Problem gone.
Brian J. Maas
If you could give me little more information about the end use of the item I could possible ask around . I work at a reasearch facility that performs and developes some very specialized NDE techniques that are not currently available in the general industry. You could e-mail me at
How's the weather in Los Alamos?
I can't remember the website address but, Lawrence Livamore National Labs has a great website also with respect to some sophisticated, advanced NDE methods not yet available to general industry either... http://www.ndt-ed.org/index_flash.htm is a website associated with Iowa State University. These folks really know their NDE/NDT! They also may provide some assistance to PaulyZ's concerns.
SSBN727 Run Silent... Run Deep!!!
Brian, extrusion won't work for our application and design change is out to the question. (for now)
Iceman, this link may shed a little light on some of the applications for our product.
SS, thanks for the link.
I would try Neutron radiography. Excellent imaging and sensitivity. Micro-applications are common and you will have a pictorial record of your tests.
Powered by mwForum 2.29.2 © 1999-2013 Markus Wichitill