Please Help! TIG weld 17-4H1025 bending failure

admin

Dear everyone,I am Michael, I am a mechanical engineer,at the moment, I need to do TIG weld 17-4H1025 (150 KSI Yield Strenght, 160 UT strength, 39HRC).Weld sample is tube, OD 3.6" , wall thickness .2" ( OD 91mm, wall thickness 5mm)I'm using Rod ER630, 1.6mm and Argon 99.99% , Flow rate 16l/m , Amps 150We have welded 2 times and sent the samples to Third Party Laboratory Test in accordance with  ASME Section IX (2010)But all samples  failured at bending test. All samples broken.I realy don't know how to solve this problem now.Could someone please advise me something in order that I can pass the test. (e.g : amps , volts, Flow rate, .... or anything )I am looking forwards to your help!Thank you very much,Best regards,Michael
Reply:If you are saying that your personal weld test coupons were sent to a testing facility, and broke during a destructive test, then it's going to come down to the fact that you need more practice. Passing a weld test isn't something that you can just set your machine to "magic numbers" and viola! you pass. I know that wasn't much help at all, but there really isn't much else to it.
Reply:Thank you, dcoffmanir !I have tried and sent sample to Third Party Laboratory Test 2 times. Sure, I will practise more. And I will wait for some advises more futher.Thank you very much,Best regards,Michael
Reply:im not sure what all that fance mombo jumbo was but is it a root/face bend?? nick break??can you put up picts of partial welds?
Reply:Hello ksmrf,Thank you very much for your reply,They are root & face bend,Here I have some picts for reference ... Tomorrow I will send picts of bend test sampleRegards,Michael
Reply:With that geometry, root bend test will fail almost certainly. You've got a HUUUGE initiation on the outer tensile side. The thinned down wall from the "inside" pipe won't do anything as no force will go through there, so you've basically need "to do it" with half the wall thickness.Face bend test fails also? Where? detailed pics please. This should work, even though it's a not-optimal design.What is the pipe for anyway? when you're using PH, one can assume it's for corrosive medium, and you'll end up with a crevice inside that you cannot get rid of with this geometry.
Reply:Hi Kingnero,Thank you very much for your reply,I agree with you about "The thinned down wall from the "inside" pipe won't do anything as no force will go through there, so you've basically need "to do it" with half the wall thickness."So actualy the Lab removed "The thinned down wall from the "inside" pipe" while did bend test.Yes, the pipe for corrosion resistance application.A nother thing I would like you to notice. Usually 17-4 at H1025 heat treat condition has 145 ksi yield (ASTM A564). But our customer require 17-4H1025 with 150ksi yield, so we purchased it.     Weld metal is as welded, no PWHT. The weld is fusion completely.Thank you,Best regards,Michael
Reply:I am sorry, I'm not familiar with ASME as I'm from Europe, and almost exclusively work with european standards.If I understand correctly, the inside is machined away, so you've completely removed the crack initiation? If tested like that, there should be no reason for the test to fail for a metallurgically sound weld.You are aware that you are dealing with martensitic steel, which is very prone to forming both martensite and precipitation of all sorts of carbides? My guess is that, with 5mm wall thickness, you're cooling too rapid.I'd preheat a bit (150-200 °C) with the filler you're using, certainly no PWHT, but I mainly don't understand why you are using overmatching filler in a joint geometry that only uses half of the "possible" strength of the tube.Together with the application, this doesn't make sense to me.
Reply:You are breaking right through the root? Partial penetration?Weld like a "WELDOR", not a wel-"DERR" MillerDynasty700DX,Dynasty350DX4ea,Dynasty200DX,Li  ncolnSW200-2ea.,MillerMatic350P,MillerMatic200w/spoolgun,MKCobraMig260,Lincoln SP-170T,PlasmaCam/Hypertherm1250,HFProTig2ea,MigMax1ea.
Reply:@ 150 amps? forget about it... not that it matters anyway, the joint won't see any gains whatsoever from a full pen weld.
Reply:Dear everyone,Here are some picts of weld test sample,Hope I will have some comments more further.Best regards,Michael Attached ImagesLast edited by Michaelpham83; 12-06-2013 at 11:31 PM.
Reply:Dear everyone,Here are some picts of weld test sample,Hope I will have some comments more further.Best regards,Michael
Reply:Is that cold finished bar flat? Has it been normalized?I agree with Kingnero, 150 amps is too light.Weld like a "WELDOR", not a wel-"DERR" MillerDynasty700DX,Dynasty350DX4ea,Dynasty200DX,Li  ncolnSW200-2ea.,MillerMatic350P,MillerMatic200w/spoolgun,MKCobraMig260,Lincoln SP-170T,PlasmaCam/Hypertherm1250,HFProTig2ea,MigMax1ea.
Reply:Do I see a crack in the macro emanating from the root? That is a fail right there. Why even bend?Weld like a "WELDOR", not a wel-"DERR" MillerDynasty700DX,Dynasty350DX4ea,Dynasty200DX,Li  ncolnSW200-2ea.,MillerMatic350P,MillerMatic200w/spoolgun,MKCobraMig260,Lincoln SP-170T,PlasmaCam/Hypertherm1250,HFProTig2ea,MigMax1ea.
Reply:Originally Posted by shovelonIs that cold finished bar flat? Has it been normalized?I agree with Kingnero, 150 amps is too light.
Reply:Based on the information given in this thread, I see one major ( i ) and two minor ( ii and iii ) fails.i : joint geometry. for a pipe that contains a medium, nothing short of a full penetration weld is acceptable. Also, the 0.03 standing below the root, what's the purpose of that ? And is concentricity that important for the application? Or are you trying to avoid using backing gas?ii : weld execution. But that can also be because the (p)WPS is not good enough, and the weldor is strictly following that.iii : choice of filler metal.The weld cracks in your filler metal according to the face bend test. That shouldn't be the case. Either by weld defects, most likely because of the hard microstructure. Ask for hardness values in base metal, haz and filler metal.That your root bend test fails is obvious when you look at the macro, like shovelon already mentioned.Only increasing welding current isn't going to solve much here.A 15 minute talk with a welding engineer, even at $250/hr, will pay for itself by eliminating much more tests...But, that's only my $0.02.Last edited by kingnero; 12-07-2013 at 04:20 AM.
Reply:Looking again at your pictures, what is the diameter of the middle roller with which the bend test is performed? It does look like a rather severe test, compared to our test method. To which standard is the bend test performed?
Reply:I agree with kingeroo. Why is it bent so sharp, instead of with a round die. Even mild steel shows cracks through the mill scale and stress in the steel with a quick bend like that. I think a big part of the problem is the way they are performing the bend test!I hate being bi-polar it's awsomeMy Heroes Have Always Been Cowboys
Reply:Originally Posted by weldermikeI agree with kingeroo. Why is it bent so sharp, instead of with a round die. Even mild steel shows cracks through the mill scale and stress in the steel with a quick bend like that. I think a big part of the problem is the way they are performing the bend test!
Reply:Ditto the comments on:- pre/post/parent HT, HAZ conditions-over matched filler-sharp bend testing with sharp guide-OP's  presumption that anybody can produce ASME code welds once the machine parameters are established-customer selection of this type of joint for corrosive useMacro showing circumferential root crack? …..As Shovelon points out--this sure looks like one. I'd suspect that if more macros were taken, this cracking will be seen in most, if not all segments around the joint.This joint design allows the inner turned down section to fully bottom in the counterbore of the mating tube. As the groove weld forms and cools-contracting as it's being made, it has to have a tensile strength high enough to overcome this solid, 'no give', stepped counterbore mating of the tubes. It obviously does not have that strength--(the selection of the over matched filler may be an effort to create a stronger weld)….but this cracking is occurring immediately as the weld solidifies with the tube joint only partially welded.An example of this cracking effect of fully restrained joints can occur in socket welds, where there's been no spacing to allow for weld contraction.Additionally, the .030/side gap creates a crack initiation point, as this macro image implies. Purposely establishing a contraction gap-prior to welding might be a stop gap solution, in addition to a deeper root pass that sees fusion below this .030 open space…..sorta……. maybe….roll the dice…..But isn't the intent of creating a procedure to ensure reliability and repeatablity?Blackbird
Reply:Dear all,I would like to say Thank you very much for your help,I will try to do again but now the groove will have a gap (.02") for deeper penetration.And also ask the Third party Lab about the radius of die/plunger of jig bend test. May be its radius is too small.Best regards,Michael
Reply:Dear all,I will study your advices carefully.Regards,
Reply:Any updates?
Reply:Dear everyone,Dear Kingnero,Now I am studying ASME Section IX again, may be the radius of die/plunger of jig bend test is about 32 7/8 x T ( ~ 32.875 x T, with T is wall thickness ) Because My English is not very good so I have difficulties in studying ASME Section IX.  After I test the sample again, then I will inform you.I would like to say thank you very much for your help,I will come back after I have some news.With best regards,Michael
Reply:I know that, If someone has ASME Section IX in hand and he is good in English, He can read and advise me.If someone has ASME Section IX, please see QW-466.1, you can understand which is the radius of die/plunger of jig bend test(my weld sample has 10% elongation).Best regards,Michael... but I know, I should study ASME IX by myself ...With Regards,Michael
Reply:Dear All,It is so long to see you. Sorry for my late update.Today I have performed Bend Test.  The result is a little bit better now. The face bend is OK, there is no crack or broken .But the Root bend is broken completely. Root bend still has problem.Yesterday I made The Bend Jig by myself after so many days to study ASME IX. Because my English is NOT very good so it took me so many days to study.Because the elongation of material 17-4H1025 is 19%, elongation of TIG Rod is 12%.  So I dicided I should follow elongation 19% of material. So I used Radius of Plunge is R.394" (R10mm). Thickness of sample is .216" (5.5mm)   (Note b as per QW-466.1, ASME IX.)Now I repeat  my trouble again, hope you will listen to:I am welding 17-4 H1025, 150 KSI Yield Strength, 38HRC. Elongation is 19%The Tig Rod ER630 has 12% elongation.Last time we send samples to Third Party Lab. Both Face & Root Bend were boken.Now I think that last time, Third Party Lab  luckily  used proper/correct Radius of Plunge. They follow calculation R=2t (they told me). But regarding to  Elongation19%, It should follow Note b as per QW-466.1, ASME IX.Note: Our customer do NOT require PWHT. That is my problem.Can you please review and advise me further.Thank you very much,With Regards,Michael
Reply:
Reply:
Reply:
Reply:Wow I'm getting a head ache......  Miller Dynasty 350Twenty Six HammersThree Crow BarsBig Rock
Reply:I'm trying again: why going through all that trouble, while the joint geometry isn't what it should be? Why that particular preparation?
Reply:Originally Posted by kingneroI'm trying again: why going through all that trouble, while the joint geometry isn't what it should be? Why that particular preparation?
Reply:No matter what preheat, root bend will continue to break as long as the vertical cut is there (the small part below the V-bevel).I'm willing to bet the higher dilution plays a part in it as well.
Reply:Dear Kingnero,My experience of welding is just a little bit ...Can you please explain me further about the reason why root bend will continue to break.Thank you very much,With Regards,Michael
Reply:The first page of this thread contains many, many clues... The main problem is even circled in red. It doesn't get any more obvious.I would say, if your customer requires this joint geometry, I'd have them propose a valid solution.As a welding engineer, I'd bluntly refuse to go through with this.