Star Shape Cracks on Push-Welding Electrode Make Star Shape Weld

Disclaimer: I'm not an expert at any of these topics, but I did stay at a Holiday Inn Express last night.

The radial cracks on the lower electrode may be caused by stress corrosion cracking. The mechanical stresses on the lower electrode may cause a higher rate of electrode erosion at those points of higher stress.

As for the weld failures, I would offer that the 'uneven' weld may result in discontinuities leading to stress risers that result in crack propagation and ultimately failure of the weld.

These are the two things that come immediately to mind. Others may percolate up later.

Yeah, radial cracks. I think I said axial. Yes, they may be due to mechanical stress, but they also may just have had too much current run through them for too long, and have surpassed their normal life expectancy (by far). Or, a combination of both. I'm not sure how much pressure is supposed to be applied by the top electrode, which pushes down on the fitting (hence, the term "push" welding). It may be too hard. Trying to find a way to confirm that one way or other. Their manufacturers support has been poor. They seem more motivated by selling us the most expensive parts, which is the electronics and transformers, without any real guarantees that that will fix the problem. But that was a few years ago when they last came in.

Hoping someone that has experience with this particular type of welding, who can offer me some basics that I should check and rules to follow, will respond with some things to try.

There aren't really any failures of the weld out in the field. We catch them before they go out. But scrap rate is way too high. Good product. Bad efficiency.

And by the way... jeeze, you two old bastards have been on here a long time. Been talking to you both for 10-11 years! Don't you have anything better to do? I was a sign engineer when I first found this place of wayward, discarded engineers and pseudo-engineers. Don't even recall how many contracts, companies, states and jobs I've been through since then.

I will say that a few of the most belligerent cr4 members seem to have faded away. The cabals and cliques are no longer here. It's a more mellow bunch these days. That's a good thing. I had to take a break from here for several years. It was getting too nasty. It seems like it was worth the wait.

Except for lyn, He's still crotchety.

Can't speak for Brave Sir Robin, but I'm retired and repair BMX bikes to give to school kids whose parents can't afford them.

So, no. Not much. Cutting the grass and cleaning the pool doesn't keep me occupied all week and I finally have a "job" where I can work from 8:00 AM to noon, with two hours for lunch and have beer any time I want to.

Eat your heart out.

I admit... yes I am.

At one point, I thought I still have another decade. Then life happened, and it seems like it might be another 15 before I can hang a hammock between two palm trees. By then, hopefully I won't be breaking bones when i fall out of that hammock.

But for now, I'm still enjoying my gigs.

You have grass and a pool! If only we were that lucky, I was cutting our grass the other day & it fell off the window sill.

Yes, we are lucky.

@Out of Box Experience

Looking at the inside tank view, are the radial features raised, indicating metal flowed, or etched away? I'm not doubting Greg's expert opinion, just curious.

Ah, good point. They are raised. In between the ridges it is smooth and flat, like it should be all around. What I find interesting is how regularly spaced they all are. Very symmetrical. But yes, in this case they aren't being etched, but rather almost flowing into the cracks, which is quite extraordinary, considering the very short time frame we're talking about.

Then I wonder, is your scrap really scrap? Or is it just a hidden blemish with no ill affect on performance?

It's mostly scrap when it doesn't pass the pressurization test, just before going into the powdercoating booth. There are several welds during the fabrication process... all potential leak points. Or sometimes there's a burn through during one of the girth welds and its obvious. That's one of the issues I'll be getting to eventually. There's a very high output through this plant. So even a small scrap percentage can mean quite a few tanks.

But even at the welder we're talking about here, some welds of the fittings just don't take for various reasons. Especially the tanks made of the thicker metal, as I mentioned.

You know best.

This is the only social media worth my time.

I enjoy it, primarily because it helps keep the mind 'exercised' with like-minded people. I like that we can have some fun with things and still get serious about real engineering challenges.

I don't spend a ton of time on the site, as you may note I rarely post anything on weekends. That's time for family, friends, and projects around the home.

I'm not retired yet. Ten years more, plus or minus a few depending on investments/health/family/etc. I'll be 56 later this year, so I may be one of the 'younger' ones.

I don't even remember how I found this site back in '07. I was probably looking at industry directories or something looking for some piece of equipment or components and saw this weird CR4 URL and clicked the link in the search engine results.

Yeah, I'm not going to be too hard on them. They have a lot of issues, but that's why they hired me; to make all these things right. They actually have a very good QA & QC program. Their product is very good. But their scrap rate is unacceptable. There are many reasons. I have to track them down, and this is one of them.

But yes, I agree that I need to buy a new set of electrodes, to start with. Even not being an expert on resistance welding, this seems to be obvious in any case. They have been recycling and alternating just two electrodes for each size (of which there are 4 sizes), and both are so old that their attempts at refurbishing and alternating the two, are pretty ineffective at this point. Replacing junk with junk over and over. And the sandpaper isn't really doing the trick in any case. That's why I re-faced them on the lathe and discovered this radial crack pattern. But no experts are around for this particular welding machine. And support from manufacturers has been poor, for whatever reason. I don't know all the history. I just started 6 weeks ago.

So I'm reaching out to see if anyone is familiar with this type of "Push Welding" which is similar to flash welding. Working one problem at a time.

Maybe you could solder a copper washer on there...

I would need to find a copper washer with the exact inner diameter to fit. I don't know that that's a standard size. And I don't know if solder would hold and not melt. It would help if I knew for sure the cause of the cracks. Too much current? Too much pressure? Age?

The issue looks to be uneven pressure on one side causing the metal to form to the breaks in the outer electrode. They do not look like arc tracks I have seen in the past due to the lack of discoloration.

Are the cracks always on the same side? It would be worth while to have them mark a location on the welded piece to show you orientation.

No, actually the star pattern is usually all around. This one I took a photo of was a reject in the scrap heap. Not sure why it was rejected.But when they are all around, still I it makes sense they are conforming to the electrode cracks.

The first thing I want to confirm with one of you is that the ridges in the weld make for a weld that is sub-standard. It should be flat and smooth. I don't know, but I suspect that the ridges are an issue that should be eliminated.

Thank you. That's good information. I haven't gotten to the coolant issue yet. I assume it's necessary. To be honest I haven't yet checked if coolant is flowing at all. I'll do that Monday.

Yes, I told them that sanding surface is not a good idea. It does conform to shape as there is a flat at top of tank.

Can you explain exactly what heat checking is, as it applies to this situation? So I can explain it to them.

Heat checking is crack development caused by differential thermal expansion whenever there is an extreme temperature gradient between the surface and interior of the solid body. Once a crack develops, it becomes a stress concentration point and will continue to grow deeper into the body.

I see. Okay. So... do you suspect this radial crack pattern has more to do with heat checking, than the mechanical pressure of the top electrode pushing down on the steel fitting? I'm thinking that these electrodes have a certain lifetime, a certain number of cycles, before they are not usable anymore, despite the fact that there is still a couple inches of copper. I think these guys believe that the electrode should still be usable until the time that they are too short from refacing (or in this case, sanding). This might be a situation similar to a car tire that is old and dry-rotting, but is still being used because the tread depth is still within limits. Is that a suitable analogy? I can use that. Analogies are good.

Life time of the contacts are dependent on how you setup the spot welder.

Pressure during the weld causes the contact to mushroom. This can be removed with a deburring tool, what the operators call refacing. There by shorting the contact.

The cracks are from high temperatures. Think of how the pressure is applied you should see wrinkles or mushrooming. The direction of the cracks are wrong for it to be from pressure. The cracks would radiate from the outside in, if the cracks where from pressure.

How you determine, if the contacts can still be used is by thickness. There will be a point where the thicks is too thin and you will start to see the tips will glow. The minimal thickness can be found in your manual or by contacting the manufacture.

I'm not exactly sure what you mean by "contacts". There is just the upper electrode pushing down on top of the steel fitting, pressing it against the sheet metal. And on the other side of the sheet metal, is the lower electrode. Both upper and lower are cylinder shape. The upper is about 3" diameter and the lower is about 2" diameter. Those two flat circular surfaces are all that touch the two steel parts. It's not spot welding. It's pushing the two steels together from top and bottom. You can see them both in the top photo. The center steel pin on the lower electrode just lines everything up by sliding through the holes in both steel parts.

Not sure if you are aware that spot welding and resistance welding are the same.

Change contacts with electrodes in what I said. If electricity is not involved then the process is called friction welding. Your statement "It's pushing the two steels together" leads be to believe no electricity is involved.

If this is true, then everything I said will not work.

Edit:

I have a question for you. Are you in the maintenance department or are you an engineer?

I know that spot welding is a type of resistance welding - RSW. I'm an engineer. But all resistance welding is not spot welding. The type of resistance welding this machine is doing is the push welding or upset welding type of projection welding, which isn't exactly standard spot welding. If it was standard spot welding, welding two similar thickness sheets together, I would be more familiar with it. Or if it was studs. But the added requirement of needing a weld here that will pass a pressurization test adds aspects that I haven't had to address before.

Also, I'm a design engineer. So the actual welder itself is not something that I've needed to know the finest details, nuances, and ins and outs of before.

Will the new smooth electrode eliminate that arcing?

I'll be memorizing the talking points of this into bullets, and use all of this as I lead a meeting next week regarding my conclusions, and the way forward.

I have worked with resistance welders a bit, we have production that utilizes these machines. I agree with Gringogreg 100%. Electrodes are trash, cooling, etc.

I may have missed it, and if so I'm sorry to be repeating it, but your electrodes must line up. 3/8" is way too much misalignment. I would guess that the cracks in your part line up with the main mass of your offset. I would even guess that the power was increased to try to allow for the open side offset and contours not sealing well.

New electrodes made to whatever contour is needed, check cooling, center the electrodes, check the settings (power, dwell, pressure). Sanding an electrode may work a few times, but to correctly renew the surface it must be cut to prevent arcing, as gringogreg said.

Yes, that pretty much sums it up. I haven't even mentioned the crudely folded over copper shim stock that was sandwiched between the upper heat-sink block and the part that the upper shunt connects to. It seems that the upper electrode assembly was not perpendicular to the ground, so they folded copper shim about 4 times and slipped it between those two blocks on the front edge to put it at 90°. So there is a slender wedge of empty air between those two parts. I can imagine the arcing that goes on in there, not to mention that the metal to metal contact is reduced to the surface area of the shims, and the back edge where there is no shim. I assume that that is a correct assessment of the situation?

Yeah, so the misalignment of maybe 3°-4° probably has something to do with that, at least in part. I've got to figure out how to get the upper assembly perpendicular, the proper way, instead of using shims to tilt it into position.

Hypothetically speaking... if coolant was not running, or not running well, what would be the symptoms? I'll have to look into that on Monday.

The symptoms are exactly what you observe in your electrode!

Right. As I suspected.

Okay, good. This is all great information. So I will be first buying new electrodes... having at least one spare to alternate... throwing these old ones into the recycle bin... and fully checking out the cooling situation. They may be cooling. I don't know. But the last week of discussion, I haven't heard the word cooling mentioned even once.

I also need to center the upper electrode with the lower. I don't know precisely what the misalignment might be doing, but aligning things is just common sense basic shit... whatever you're working on.

I still don't know how to gauge the pressure we're applying with the downstroke of the upper electrode assembly.

There is still an entirely other issue I haven't even touched on here. i'll be dealing with that once I get the welder operation correct. And that would be the design of the edge of the steel fitting, that is coming into contact with the sheet metal tank that it is being welded to. It isn't just flat, with the entire wall thickness of the fitting coming into contact with the sheet metal. There is a specific contour that allows some collapsing of that edge. That is a whole science in itself, and I don't think they have that at an ideal design either. But I may tackle that in a future question all its own. First I have to get the welder working at its optimal level, and design a proper PM checklist and schedule for these guys.

Before you go out and buy new electrodes, I would suggest looking into their design. From what I see, I think they are relying on the fixture base as a heat sink. I don't think this is enough. I would want to see coolant channels within the electrode itself.

Ah, interesting. Yes that fixture base is a huge hunka copper square bar. That may be heat sink; in which case, they may not be using any coolant at all. But you think the welder itself has the capacity to use coolant, but is not set up that way for this particular job? And is using a heat sink instead? That can't possibly be as effective.

It's time to throw in another factor. I didn't want to mention too many variables at once, to not muddy the water. But we are not getting enough current. We are running one set of tanks that is thicker material. That is where most of our failures are. With that tank, we are at the top edge of the manufacturers recommended thickness. But it is not performing that way. In reality, it is under-powered to properly make that weld on the thicker tanks. It should be sufficient, but it isn't quite there.

What you've been suggesting is that we are leaking power in various places and in various ways. This is where the vendor of the welder and of the electronics keep wanting us to buy new expensive parts, but not guaranteeing improvement. It appears that the answer is in finding all the small leaks, that in combination are robbing me of current where it needs to be, thereby not realizing the full capabilities of the welder's potential. This is showing itself most clearly, when welding the thicker tank material.

This is all coming full circle now, and making sense. I have to plug all these holes.

Should the top electrode also have coolant? Behind the top electrode is a big copper shunt that is made up of dozens of layers of copper sheeting. This stack of copper sheets is folded into a curve... like you took a ream of copier paper and bent it into a right angle. Proportionately it's about the same the same look as a new package of paper. About half as wide though. The outside-most layers are broken.. cracked at the outside bend. This also must have some effect on all of this. More loss of current that is not finding its way to the welding surfaces. This shunt may be the upper heat sink.

Normally, the top electrode should have what is called a "fountain", a tube leading to the tip of the electrode, and a co-axial channel to provide return flow of the coolant. The bottom electrode should have one too, but the placement of the pilot pin prevents this. I would recommend redesign of this electrode to provide coolant flow directly through the contact area.

The "shunts" should also be water cooled in this application. Shunts with a water cooled jacket are available from most resistance welding suppliers.

Also I would consider connecting the bottom electrode directly to the current return shunt, thus eliminating another high resistance joint between the electrode and the base.

Good stuff. A lot of that is ambiguous with very large ranges for several different welding styles, but it does confirm some of my assumption regarding having several electrodes on hand, facing them off on a lathe is important.

Also "electrode should be cool to the touch"... that, I know is not the case. Leads me to believe that proper cooling may be one of the problems.

Mushrooming is also visible, but not the electrodes. Rather the solid copper square 4x4 column that the electrode assembly is sitting on. It's not shown in the illustration above.

Good. Gives me more things to check Oh, and there is no operators manual for this machine. That certainly would have helped.

Yes there are many critical parameters that can be tweaked....pressure of clamp, time elapsed of weld, level of amperage, contact surface conformity, the temperature and surfacing more for longevity....Being able to measure all these parameters and calculating the proper settings for material and type of weld also very helpful, rather than trial and error method....although in the end, pretty much what it comes down to...Helps to know what over and under on each of these settings looks like on the piece...not just for you but the machine operators as well...Good luck with that...

Yes, ideally, they would put out an ad looking to hire someone who truly understood the system. As is, what we have are pure "operators" who move the parts on and off, and recognize the scrap parts, but that's about it. But I've gained some good information here. Enough to now know what I'm looking for, and to put together at least some kind of PM procedure. That's a good start.

Excellent explanation. Thank you very much. That's more or less what I was thinking, but without the specific details of the process.

The same sort of thing happens in die casting molds and in plastic injection molds.

Good description of heat checking.

Yes, between the two of you, I'm going to sound like I know what I'm talking about when I explain all this to my peers.

Throw in a few phrases about "intermetallic dislocation growth" and "thermal heat flux propagation" and they might really swallow that University California Pacifica diploma as the real thing.

It's actually University California Pacifica, LLC diploma

Spoken like a proud owner.

Just remember, you must judge the level of technical competence and willingness to listen to, and accept, what you present to them of your audience.

Be prepared to answer their questions in a confident way.

Know your subject and you will do well.

Good luck!

Well, to be honest... I already seemed to suss out more about these electrodes in 15 minutes of cleaning and refacing them, than they seemed willing to figure out in 5 years of using them. Anything I learn here is above and beyond.

I just can't believe they had no PM schedule for it, whatsoever. No PM, period. Ah well... it gives me a job to do. No complaints here.

Yeah, I had my talk, and they were all in awe like I was speaking the words of God. One said... "It took us going through three engineers to finally come up with good information".

So... I'll take that as a good sign. All were happy except for the guy in charge of the Maintenance dept., who were supposed to be doing regular PM on it since the 4 years since the welder rep was last here. They totally dropped the ball. Don't even have the PM list any more.

So... I also now found out that those electrodes are four years old, minimum. Possibly a half million welds with the same electrode. It needs to be melted down and made into a copper bust or something, for having not shattered into pieces by now.

That's good to hear.

Maintenance needs to understand that they have a vested interest in PM. It helps the bottom line. Bring him, or her, around to your way of thinking.

This has been a very worthwhile thread!

They are a good maintenance team... with everything else. I believe they were just intimidated by the complexity, as everything else is simpler machinery without a lot of electronics or complex assemblies. Its mostly shears, presses, conveyors, hand welders... sheet metal working machines. Some pneumatics. Not even a CNC. So they may have gotten confused and had no support system that allowed them to ask for help after the first set-up, and things just went slowly downhill. I'll make sure they get better training and not be scared of it.

Yes, this thread has helped me tremendously.

Get it cast into a belt buckle.

With the amount and size of the cracks, depending on how high the 'fountain' or 'bubbler' tube channel goes, it may likely be too thin once the cracks are visibly removed anyway. Copper also has a tendency to move during machining and may hide the cracks that continue deeper. Facing the electrodes should be a regular operation when it gets discolored. This will remove any transfer of the steel to the copper, (just did this to one today). This will also create the need for more amperage and may make it seem like it needs more dwell.

Okay, now that it's Monday, I can update some things.

Yes we are using a coolant of some sort. Might be just water. Not certain. The upper electrode that pushes down on the NPT fitting remains cool to the touch. The lower electrode is warm. Very warm, but not too hot to touch. The copper block, square column, that the electrode sits on has the coolant running through it. It just acts as a heat sink to the bottom surface of the actual electrode, which is held down against it by clamps as you can see in the photos. It seems to be less than optimally effective in keeping the electrode cool.

Does this have anything to do with the radial cracks? Or are those cracks purely from cycling current for too long?

When I install new electrode, is there anything I can do differently to prevent the cracking, or is that just a symptom of using the same electrode for too long and through too many cycles?

I mean the electrode and the cooled square column below it are direct from the manufacturer of the welder. They are not home-made. Is this system designed wrong, for the purpose we're applying it to?

Meanwhile, I'm trying to find a way to check the flow of coolant, in case it isn't flowing properly. But it is flowing. I just don't know how much. Is checking this a reasonable thing to do?

To me those "cracks" appear to be too regular and even to be accidental or the result of damage from over use.

I wonder if you are looking at a mechanical support structure or even the coolant paths, which were originally inside the electrode, the original electrode material having been eroded away long ago!!!!

Perhaps your surfacing of it on a lathe even removed the last remnants of the electrode material....

Interesting conjecture, but no... I have other electrodes in smaller sizes that don't have the cracks. They are just solid hubs of copper with a hole in the middle. I had earlier commented on the equal spacing. It's quite fascinating.

I'm curious if the copper is some special copper alloy. They are kind of expensive to replace the whole set with two of each size. I'm thinking of buying one set and making additional replacement sets to swap out, on a lathe... if I knew that it was just regular copper round bar.

Thanks for that. Actually, here is the electrical info:

KVA: 560

Volts: 440

Cycles: 60

Secondary volts: Min 7.28 Max 16.25

I just checked the coolant temperature. I read that suggested temp should be between 68° and 78°F. Ours was running at 80°. I've got it down to 70° now. The lower heat sinks... those tall square columns that the electrode sits on... were definitely not cool to the touch. The columns have the coolant paths running in them. I'm curious to see what 10° cooler water will do.

I agree with your suggestion for the electrode design. Having the coolant flow directly through the electrode, rather than depending on the base as a heat sink, should provide a significant improvement to electrode life.

That's helpful. Thanx.

You have several things on your plate, but I suggest that your major concern is that the design of your weld is such that you are attempting to weld a rather large cross sectional area by resistance welding. I submit that you should convert your weld geometry to a projection weld configuration, This would involve forming a weld projection (much like the projections formed into stud welds) into your welded nipple. You could then weld with considerably lower weld current, which would provide a strong weld, reduce water cooling requirements, and provide several additional benefits. We do this kind of stuff for a living. If you provide drawings of your parts, I will modify the weld interface dimensions and provide you with a recommended component, An alternative is to insert a washer between the two parts you are welding. This is a simple approach but has disadvantages.

Thank you. We actually do have a projection profile around the steel fitting, that collapses, but I don't like it. I don't believe it was well designed and I was going to get to that once I handled all the actual overdue maintenance and adjustments to the welder itself.

I'd love to show you a cross section of the current profile. I'll put it here tomorrow.

I've been trying to research who actually designed this profile, to determine their qualifications for being able to do it properly. Evidently it's been many years, and I can't really verify what his qualifications were. In any case, I know it can be a better design. It's still too heavy and thick.

I have the company rep coming out next Wednesday, from the welder company. Problem is... he's the same guy that came out 4 years ago and told them it was okay to put several layers of shim stock between the upper heat sink (that the upper electrode clamps to) and the mounting block above it, in order to make the electrode perpendicular to the welding surface. The problem with that is, they put the shim along one edge, between the blocks, which now has a narrow wedge shaped air gap between the copper block and the mounting block. It reduced the metal to metal contact between the two, considerably. It also set up a situation where arcing is now probably going on between the two surfaces. I believe that is one of the problems we are having in getting strong welds when we switch to slightly thicker sheet metal. It appears to not have enough current available, although, by all rights, it should. I'm just leaking current in so many places, and so many other things are wrong, that it's all adding up.

So, I don't have a lot of faith in what he may suggest.

But in his defense, I don't believe our maintenance team properly followed whatever PM schedule he left for them 4 years ago... which seems to have disappeared in any case.

I believe a new shunt for the upper electrode assembly will also make considerable difference. The current one (it's made up of many layers of thin copper sheet, and then bent to a right angle) has a majority of the layers broken, or ripped. More than half of the stack, is broken at the curve. So, I've got that ordered as well.

A lot of things going on, as you can see. But I see I'm also on the right track in identifying what needs attention, thanks to all the opinions I've received here over the last few days.

Here you go. That's the projection we have on the bottom outer circumference of the adapter. I believe it's too thick and too broad an angle. The flat that goes around the inner circumference is the "stop", when the projection edge melts down. It doesn't weld. Also, I'm not sure we need that trough that's in between the inner and outer diameters. And lastly, I'm wondering if the projection should be symmetrical. Basically, the entire design is suspect, and I don't believe it is ideal for this part.

The photo right above here, it is of a different size adapter, and has a hex on the outside, but the projection rim is exactly the same design as the drawing.

I have two concerns about your parts. The first is that your steel fitting is much longer than the thickness of your seel shell. A fundamental rule of resistance welding is that the two parts you are welding should be about the same thickness because the the maximum temperature will occur midway between the two parts. This rule can be modified by projections and other features, but your design is an obvious violation. The result means that the highest weld temperature is in your fitting, which has to be overheated to achieve proper weld temperature at the weld interface. You could fix this problem by machining a flange on the OD of your weld nipple and putting the weld projection on the OD of that flange. The design would then resemble a stud welded component. This design would use considerably less weld power because you wouldn't be heating the whole volume of the fitting. The 70 degree included angle of your projection is very traditional, and I suggest you retain it.

I appreciate the suggestion of the flange, but that is probably beyond my ability to authorize. This company is a large national company with branches all over the place. They all weld similar parts, albeit with slightly modified details. My mission is to get the welder at peak performance, because it has finally been admitted that those who were previously responsible for it, dropped the ball. And making some improvements to the projection would help to. But a major change like adding a flange is probably not going to be approved. And it would also cost considerably more per fitting, as a flange would need much larger bar stock and much longer and more expensive machining. Unless I welded the fitting to a separate flange (washer) then welded the flange to the sheet steel. But then that's two welds. So none of that would really be an option.

While I understand your reluctance to stick your neck out, you should understand that we do this for a living and know what we are talking about. However, I have a suggestion that may be easier for you to deal with , and may help you better understand the parameters. Try taking one of your weld nipples and cutting it in about 1/2. Finish the weld end square and smooth and throw the other end away. Then make a weld to the shell using this cut-off end. You should reduce the weld current because there will be less material to heat. You should find considerably improved weld performance, and assuming you do, we will be happy to show you various methods for achieving the performance you are looking for. I should also tell you that we weld hundreds of similar parts/day with no electrode wear, no water cooling, no cracks, and all to tight military specifications. Best regards

Thanks, but I can't do this. As I said, my mission is to make the welder work better. It's not reluctance. I'm one person in a 10,000 employee international company. I don't have anywhere near the authority to change their product line in this drastic way. There is far more to it than just making it shorter. That would affect a dozen things down the line. I have to work with what I've got. Millions of tanks are being made each year with this length of fitting, in this way. I'm just looking to improve the projection profile. And compensate for no PM having been done on the welder for at least 4 years, and set up a PM program. So, I appreciate the suggestion but that isn't an option.

Thanks, but I can't do this. As I said, my mission is to make the welder work better. It's not reluctance. I'm one person in a 10,000 employee international company. I don't have anywhere near the authority to change their product line in this drastic way. There is far more to it than just making it shorter. That would affect a dozen things down the line. I have to work with what I've got. Millions of tanks are being made each year with this length of fitting, in this way. I'm just looking to improve the projection profile. And compensate for no PM having been done on the welder for at least 4 years, and set up a PM program. So, I appreciate the suggestion but that isn't an option.

IMHO the problems you are experiencing are largely related to a very antiquated nipple design. This is a very simple thing to fix and I'd be happy to show you how to do it. You would need to slightly modify the nipple design, the upper electrode features, and your welding machine settings. These modifications are simple,can be done in a day and should solve all your problems. If you become disenchanted with the way the efforts are progressing, the whole effort can be reversed. Please feel free to contact me. I may be able to travel to show you how to do this in person, or recommend someone closer to your facility. Everything we do is reversible, so if you are not pleased, the modifications can be eliminated. I suspect that when people in your other locations see what you have accomplished, they will want to participate in the modifications and you will become their hero.

Not sure I agree with Welderman and don't know who developed the original projection geometry. I can see that the outer rim forms the seal weld an that the inner ring acts more as a stop. Is there an extruded bead around the outside of the fitting when the weld is complete? In bandsaw blade welds, there is usually and extruded fin of material that basically makes sure the final welded joint has been purged of contaminated material. With bandsaw blades, the extruded fin is ground back to clean material and a smooth surface.

Do you have a cross section cut from one of the completed welds? It would be interesting to see where the metal is melted and where it is compressed to a closed, but not welded, joint. I would assume that the weld is qualified by hydraulic test, either for leak tightness or for process qualification by test to failure.

There is a bit of an extrusion around the outside. I can see the line between sheet and fitting. All the way across. It does fade away at the outside edge, right at the extruded bead, actually.

Looks about the way I expected. Approx the outer 25% melted and comingled and a crevice for the inner 75%. I'm surprised that the groove in the fitting face totally disappears, but then I suspect if you check the volumes, the groove matches the volume of upset from the outside ring. I was thinking that the crevice might be a fatigue crack source, but with the fitting base as backing, not much flexure can occur. Is the tank expected to see much alternating pressure over it's life, or is it mainly at a static pressure, as in a propane cylinder? The static case should stand up well and alternating pressure/vacuum wouldn't be much worse.

It's purpose is alternating pressure. Not real often. Just about 20,000 x a year. Roughly 2 1/2 times per hour, 24/7. The swings are about 20 psi up or down.

Each tank gets extreme pressure tested before leaving, to 100 psi. But normal operating range is within 40 to 60.

Still probably OK if it's +40 psi to +60 and back. Pump discharge accumulator?

Yes, we don't use that term but you could call them pump discharge accumulators.

At my last job we had 3 spot welders I had to upkeep and fix.

My PM list included:

• Removing slag from the face of the machine
• Replacing any hoses or filters as needed.
• Verifying pressure, water, air, and clamping between contacts. They make a tool for this.
• Ohm\ diode check the contacts, to check the transformer is still good
• Checking all fuses and breakers
• replace contacts as needed

The first thing I would do is contact the manufacturer for a manual and PM schedule\ list if one can not be found at your place of work.

If the spot welder has not been serviced since it was brought all hoses, filters, and regulators should be replaced.

Everything that carries current needs to be taken apart and cleaned of oxidation. By everything I mean everything from the contacts where you weld all the way to the transformer. This will allow you to run at lower voltages and therefore have lower temps at your contacts. High voltage allows heat to build in you contacts, shortening their life and greater potential for sparks.

After you have cleaned the oxidation make sure there is no continuity between your supply and the chassis or return path to transformer. Only one contact should have continuity to the return of the transformer. Otherwise, part of your current will divided and you will have to turn your voltages up.

The spot welders we used had, used replaceable tips for the contacts for the weld. Might want to look into replace\ make new contacts.

As for the star patter, that is from the heat. The inner part heats and expands and after you are done inner part cools, but cool faster than the rest of the metal. You need to replace that before it explodes.

Finally, Check all voltages in the brain box are what they should be and not what the operator wants. High temperature make brittle welds, but look better than low temp welds. Also look at the welds it self make sure there is not a hump or a well. Humps are produced by poor fit up and gases getting trap. Wells are produced by too much pressure.

One comment about your fit up for the two pieces of metal. Spot welds work best when the metal over laps. From the pictures, it looks like the fitting is pressed fit and then welded into places. If this is true, then you have sheer stress at the weld. Change the fitting to have a flange which can be welded as well, will cause tinsel stress, better for welds.

"The first thing I would do is contact the manufacturer for a manual and PM schedule\ list if one can not be found at your place of work"

Yes, I did that yesterday. And I ordered the coolant intake hose replaced, as it had collapsed at a bend and was kinked to about 50% flow.

I also had the coolant temp lowered from 80° to 70°

But before I go further, this is not spot welding, it's upset welding, or push welding.

The fitting that is being welded does not fit into the hole in the sheet metal tank shell, the hole is smaller than the fitting and the fitting is welded on top, centered on the hole.

"Verifying pressure, water, air, and clamping between contacts. They make a tool for this."

Which tool is this? Does it have a name? Is it specifically for welders, or just a general tool?

I've already ordered new electrodes.

The rest, I will add to my to-do list. Thanks. That's all great info.