Not logged inAmerican Welding Society Forum
Forum AWS Website Help Search Login
Up Topic Welding Industry / Technical Discussions / Automatic welders want to change gas
- - By MrDisibility (*) Date 11-28-2007 16:40
Hello there people. I have been lurking and find your topics very interesting and informative! So when this question came up at work I thought I would ask here for some more info.

I work for a manufacturer utilizing robotic welding cells to weld up stainless steel exhaust components. The current gas being used is a He tri-mix. This costs almost 3x as much as 97/3 Co2, which my purchasing guy claims we could use. And we have 95/5 O2 plumbed throughout the building, so that would be the ideal gas to use for cost and convenience.

But what I am hearing from our production staff is that with the 95/5 O2 we get burn through and dark welds. It just gets too hot. Might 97/3 be a viable alternative?

I really want to switch the gas on these things to save cost but I am being told the combination of He tri-mix and Lincoln 308L .035" wire is the only way we can get a good weld.

Seems to me there should be something we can do here. Any input? Forgive me if I am light on technical knowledge. I am a process engineer, not certified welder. So I am here to learn more!

Marcus
Parent - By Lawrence (*****) Date 11-28-2007 18:03
Ahhh, but the problem is a process control issue not a welding issue.

So we have high expectations from you right from the get go  :)

Welcome to the forum!

Ed Craig  www.weldreality.com is a big proponant of 98/2  Argon/C02 for spray and pulsed spray on austinetic stainless steels.  Ed also has a couple of textbooks he sells that will provide excellent robotic parameters that will make your work  look like the work of a process control engineer!.......... Oh wait... your an engineer already... Get the book!  :)   :)

If you are short circuiting than the tri-mix may be the best choice.
Parent - - By chall (***) Date 11-28-2007 18:05
In addition to Lawrence's advice, remember that if you change the gas composition, you have to requalify the weld procedure.

Charles
Parent - By MrDisibility (*) Date 11-28-2007 18:43
Yes, all the more reason to make sure we do some good homework before getting started! We've got somewhere north of 500 programs right now. All in the name of saving a few bucks. The important thing though is that as our arc time goes up with increased efficiency (ahem, due in no small part to my efforts, presumably), so does our gas consumption!
Parent - - By Sean (**) Date 11-28-2007 19:08
In my experience the 98 / 2 that has been mentioned works quite well for most austenitic stainless steel applications. 

When you change your gas you will likely need to adjust your parameters and perhaps adjust some of your work & travel angles on your robot programs.  The large amount helium in your trimix will promote a large droplet globular transfer and your cross sectional penetration is elliptical.  With the argon bimix you will get more finger like penetration and it will promote spray transfer.  Is it possible that the "too hot" observation may be a function of the penetration profile caused by the shielding gases?  The dark welds are likely due to the amount of O2.

If you are looking for cost savings you may want to remember that 70% to 85% of welding costs are labour related.  Perhaps you could get the same cost savings by changing your deposition rate.  

BTW it may be useful if you told us what parameters you are currently using joint type(s), material thicknesses, and fit-up tolerances and I'm sure that someone reading the forum will have more feedback for you.

Sean
Parent - By Kix (****) Date 11-28-2007 20:00
98/2 Argon/Co2 with a 304 base and a 308L wire is the BIZ!  Not that it wouldn't be good with any other base. I too am phasing out the trimix for 98/2 and the website Lawrence gave you is very VERY helpful.  You can use 97/3 Argon/Co2 as well, but 98/2 is the way to save a buck.;-)  I've used both 97/3 and 98/2 and I would go with the 98/2.  What types of power sources are you running your bots of of?  Do they have pulse cabability?
Parent - - By MrDisibility (*) Date 11-28-2007 20:38 Edited 11-28-2007 20:41
I will do some more homework on the parameters being used now. I have only just begun looking at this process so I have not fully investigated yet. We are concerned with the exhaust system assembly, which involves welding brackets, flanges, hangers, mufflers and tips to 16g tubing. We are doing fillets and laps in rotational paths, and also in linear paths. Our tolerances could use some tightening; sometimes there are air gaps there.

The penetration profile may have something to do with the burn though as well. This is very possible. I am wondering if this is the case then are we truly limited in our choices?

I am reviewing weldreality.com now. Lots of propaganda there. Trying to sift through it before pulling the trigger. Although it seems like some of his resources there will get my lingo and knowledge up to speed...
Parent - - By MDG Custom Weld (***) Date 11-28-2007 21:15
You are totally right about Ed Craig, lots of propaganda, but somewhere in there is a wealth of information.
98/2 is right on, however we run Ar/O2 and not the Co2 as mentioned above.  We make the same type of parts that you do, 304, 309, 316, 409, 439, 441, 442, all used in hot end exhaust parts and manifolds.  Thicknesses range from .035 wall up to .072, and the fit up is always a challenge on thin wall bent tube (especially mandrel bent with no wrinkling).  We're using Motoman, ABB, and Fanuc robots, but mostly prefer the Motoman.  We use Miller Auto Axcess, and Fronius TPS 4000, mostly spray, but some short circuit.  Weld speed ranges from 35-65 ipm are normal depending on the joint.
Parent - - By Lawrence (*****) Date 11-28-2007 21:31
MDG

I'm assuming your pulsing with your Auto Access.... Why else spend 11 grand eh?   Anyhow, do you find that 0.045 pulses the thin material better than the 0.035 wire?

All of my experience here is hand held manual, but it seems like for whatever reason that the 0.045 wire can make fillets even on pretty thin stuff without burning through and the .0.35 is a bit harder to control.

What are your thoughts????
Parent - By MDG Custom Weld (***) Date 11-28-2007 21:52
Lawrence,
  Thanks, I did forget to mention that all of the welding is pulse.  We use 409, and 439 wires in the .045 dia, but they are both metal core (not to be confused with flux core), not solid wire.  That pulse program is very consistent and smooth on thin material, and with a minor arc length change, the depth to width ratio can be moved to the required amount.  We also have the wave writer software and with some DOE work we came up with modifications to the factory wave the made spatter and cold starts disappear, even with welding speeds in the low to mid 60ipm range. 
Now anything in the 300 series we use solid wire.  This is where the pulse programs really aid in going fast and maintaining a consistent fusion at the toe of the weld. 
We don't see much of a difference in .045 verses .035 from a burn thru control standpoint, except when we have poor fit up.  The .045 can bridge and maintain the arc with better control, but I think that comes from the metal core wire more than the wire diameter.  But then again, it's really easy to control the arc when the robot is doing the same thing every time.
Mark
Parent - By DaveBoyer (*****) Date 11-29-2007 05:10
From What I recall on Ed Craig's site He is a big fan of .040" dia. wire. I didn't know such a thing was available [in the US] but as that is about 1 mm I guess it would be a common size in the metric[the REST of the] world. Do any of You use .040 wire?
Parent - - By MrDisibility (*) Date 11-29-2007 00:09
Well, I looked a little closer and we are using the Blue Max 308LSi wire with tri-mix. I am still not getting the need for using this gas after spending the majority of the day on Ed's site. Even Lincoln recommends an Ar- 1% O2 for pulse welding (which we are doing btw). I tried getting some more informations on our feeds and volts/amps...it varies widely from program to program. I don't really understand why though...we weld 304 to 409 to 416 to 430 in all different combinations. Shouldn't we be able to get a ball park setting for everything?

Anyway, I am starting to think I need this book, if for nothing other than my own interest.

Marcus
Parent - - By eekpod (****) Date 11-29-2007 00:52
We also use the Blue max 308LSi w/ a tri mix. in pulse mode w/ program #44 for pulse GMAW on 304 grade material.  I have inquired about switching to a dual mix gas versus a tri mix to save money, but it appears no one here is willing to stick there neck out and experinment w/ it.  There thought is the welding machine rep/ salesman says this is the way to go and that's that.  I personally took a step back and questioned this mentality but was brushed off, so I didn't pursue it.  I guess if it's not broke don't fix it, even if it appers it could be made to run better and cheaper.  Chris
Parent - - By ssbn727 (*****) Date 11-29-2007 05:41
SOMEBODY'S GETTING KICKBACK!!!!!
Parent - By MDG Custom Weld (***) Date 11-29-2007 16:02
All of our solid 300 series wire is Sandvik and we run the .039 dia.  It is drawn in Scranton PA, and sold by most suppliers.  We have spent countless hours testing other manufactures, and found that nothing runs as good as the Sandvik product.  It is about 10% more than some of the offshore products that are out there, but the consistency of the wire makes up for it in the lack of welding/ feeding problems that normally contribute to robotic welding downtime.  Everything is 98/2 ArO2 and pulse spray robotically welded.
Parent - By MrDisibility (*) Date 11-29-2007 16:03
Something being cheaper and something being 67% less are two different ballparks though. There seems to be little resistance for me here, but before I get in there and delay production and ruin product I better have a degree of confidence in place. Especially given that our efficiency is going to be increasing and more and more gas will be used in the future.
Parent - - By Sean (**) Date 11-29-2007 16:20
Having worked for a gas supplier I can tell you that the majority of the sales people will promote products that get them the best commission.  Personally, I'm not a big fan of trimixes as in most steel and stainless steel applications it helps the supplier more than the manufacturer/fabricator.  Keep in mind that trimixes tend to be patented and it can be harder to switch gases so the sales guy keeps the price higher (especially with Helium which is typically more expensive) to get the better commission... I'm not saying there isn't an application for trimixes (eg duplex), but that its typically better for the supplier than the user.  

When it comes to the penetration profile that can be changed by adjusting work, travel angles and your welding parameters.  If you materials are consistent thicknesses and joint configurations you should be able to keep roughly the same ball park parameters, although you may have tweak some of the weld starts and stops.  IMHO I would experiment take my bread and butter joint & materials, set the machine to my ball park productivity range, change my gas and start with checking the work & travel angles of the program.  It may be as simply as increasing your travel angle and travel speed.  If that didn't work I start playing with my parameters (one at a time).  Don't forget to write down what you changed and what happened...
Parent - By Kix (****) Date 11-30-2007 16:16
Do you tend to keep your bot in more of a push angle or more of a drag angle?  I prefer a slight drag angle for better penn and less spatter.  If your pulseing i'd say you could do pretty much what ever you want regarding the drag or push to cut down on spatter.  If your setup good on pulse it will slick her in there regardless.  You'd probably have better consistency in the push for your particular application it being thin sheet metal with gaps in it sometimes.  Is there anyway you can get a bottle of 98/2 and do some testing before you jump right in on production parts?  Like people are saying on here " don't allways listen to your suppliers sales guys because they are trying to make a buck".  Airgas here in the valley has everyone that is running stainless mig using the 90% helium tri-mix.  Miller Electric doesn't even use that high of a % helium when writing stainless pulse programs.
Parent - - By MrDisibility (*) Date 12-03-2007 19:21
Thanks for all your posts guys. I am pulled some Technical Papers from SME.org (I am a member of SME, pat me on the back) and am beefing up on some tech background on welding and related topics in an effort to make some more educated hypotheses.

Marcus
Parent - - By Lawrence (*****) Date 12-03-2007 20:16 Edited 12-03-2007 20:19
Marcus,

Try to keep one thing in mind while your doing your research.

Just because your working with robots things don't have to be terribly complicated.

Process control
Correct, consistant, fitup makes robots work. Not multiple gun angles, or shield gas or pulsation.

With proper fixturing and fitup... a common gun angle, wire feed speed, travel speed, voltage, run-in, crater setting can be used a surprisingly large variety of fillet and lap joints.

Buying pulsers and changing parameters to allieviate burn through and inconsistancy due to poor fitup and bridging gaps is a loosing cause and a giant snowball growing ever larger as is cascades down your mountain.

Investing in excellent fitup and fixturing up front will allow you the flexability to choose less expensive gasses and power supplies. 

Don't get me wrong.  While it is not a silver bullet; Pulsation has its place in high speed automation.... But so do traditional CV's
Parent - - By MDG Custom Weld (***) Date 12-04-2007 14:44
Lawrence hit it on the head.  Robots are only as good as the people that program them, parts they weld, and the fixtures that hold them.  It sometimes cost 2 or 3 times the more when buying robotic welding fixtures when compared to their manual welded counterparts.  Also, part tolerances must also be considered when robots are involved.  If the robot is repeating to +/- .005 (as most industrial robots are capable of), but the part is +/-.030, then the wire is bound to track off then joint and make scrap or rework.  Sure some gas mixtures and pulse power supplies will help the process be more tolerant, but it will NOT solve a lack of focus on the above listed items.
Mark
Parent - By MrDisibility (*) Date 12-04-2007 16:11
Yes this concern has been in my mind as well. Generally our fitups are good, but there does leave something to be desired.
Parent - By DaveBoyer (*****) Date 12-05-2007 05:36
Inconsistant part fitup was the major problem with robotic welding in the auto frame plant. The solution there was to set the fixturing to make a tight joint on 1  side of the part and weld that roboticly, and let a person deal with the inconsistancy on the other side. NOT a great solution.
Parent - - By ctacker (****) Date 12-07-2007 03:26
If i remember right, you can get software that helps your robot track the joint! we just finished about 1200 mud pumps we had welded with a robot, the operator had to finetune each and every one because of the inconsistancies in the casts, took about 30 minutes of dry running it before welding it, That job still paid for our robot and got us a bonus every month for the last year and half!
Parent - - By DaveBoyer (*****) Date 12-07-2007 04:34
I am sure great strides have been made since My time at the frame plant, I left there in '91.
Parent - - By ctacker (****) Date 12-07-2007 05:52
I imagine they have, i couldnt understand why my company invested 200k into the robotic setup and didnt invest 3k into the software, would have shaved at least 30 minutes off a part, and would have been a wise investment. (from my point of view)
but i'm not a number cruncher and when it comes to making decisions of that sort, no ones ever listened to me so i quit trying and am learning to go with the flow!
Parent - By MrDisibility (*) Date 12-07-2007 22:24
We can do seam tracking but I was told the time to perform the weld is considerably longer to perform. Besides, I am not sure how well it will work with our welds traveling around concave and convex surfaces. Personally I'd like to see the gaps get more consistent. But that's not really the point of the thread.
Parent - - By MrDisibility (*) Date 12-12-2007 23:19
New info found out:

We are actually SC welding not Pulse welding. (hmmm, some of our other cells that are welding pulse are on our Bulk Ar 95/5 O2...the welds are dark as in they have buildup on the outside, which we deem unacceptable for the cells I want to change. But they are doing it.).

Also, our general practice is to orient torch angle to push the puddle. This is to avoid too much heat buildup in the part.

I found the screens that hold our weld parameters also. We have 4 or 5 standard ones we use for welding different materials together. I will report those soon, maybe tomorrow.
Parent - - By Lawrence (*****) Date 12-13-2007 11:26
That is a horse of a different color.

If you are short circuiting..... The tri-mix is the right choice

I think most would agree that 5% 02 is a bit high for most common stainless applications whether its spray or pulsed spray.
Parent - By MDG Custom Weld (***) Date 12-13-2007 14:11
I agree, 5% is too high.  We did some testing a while back (year or two) and found that anything over 3% caused excessive spatter and did not help the weld in any way.  Still, I think tri-mix is a waste of money regurdless of transfer mode.  The added cost far out weights the minumial gains in speed.
Parent - - By MrDisibility (*) Date 12-13-2007 16:26 Edited 12-13-2007 16:31
So my question becomes, why short circuit then? If we can change our parameters and consumables to accommodate pulse (we can set them up for MIG or MAG pulse welding) then maybe that is something to consider.

I am sure we can change the mix on the bulk as well. The equipment using the 5% is deemed to produce inferior welds (aesthetically speaking). Maybe this would help across the board.
Parent - - By Lawrence (*****) Date 12-13-2007 17:56
You mentioned your Short Circuit welds were made with a push gun angle...  Push angles and short circuit are often an indicator for thin materials with the possibility of burn thru issues.

If you can create procedures with GMAW-P that don't burn thru than you will probably have an overall better chance of superior sidewall fusion in fillets, which means lower reject rates with the very real benefit of increased travel speeds with the GMAW-P.

Listen to MDG about those gas selections... His experience in the trenches goes way beyond my understanding of the subject.

Seems to me like you've been asking some very good questions.
Parent - - By MDG Custom Weld (***) Date 12-13-2007 19:37
MrDisibility, that is truly a good question...Then why Short Circuit???  The answer (and I'm sure Ed Craig might disagree) from my prospective is that it's not needed at all in thin section SS or even CS when using automation.  Early pulse machines (pre 2002) were very unstable at low voltage/ wire settings, rendering them inconsistent on penetration and weld profile.  The "new generation" of pulse machines have much better control even on lower settings, and have the versatility to refine the pulse wave for your specific application.  Miller Auto Axcess and Fronius TPS4000 or 5000 have very controlled arc length settings that are very stable on thin material.  We have one application were we weld a heat shield that's .035 thick onto a 409ss pipe that's .055 thick, and then weld seams on the heat shield to itself.  This is done using .045 409 metal core wire.  We had 70% rework on this part using older pulse machines until 2002 when we put the Fronius system in place.  Our consistency jumped to 95% good parts almost overnight.  Since then we have used the Miller and Fronius power supplies on all new production equipment.
We weld pulse on almost everything and control the level of penetration with the parameters and torch angle.  This started out all robotic, but in the last 2 years, even our manual welding department has switched over to pulse and 98/2 ArO2 for all ss applications.  Pulse has given us better spatter control, and faster welding speeds.  We also push everything 5-20 degree as needed to control the puddle.
Hope this helps,
Mark
Parent - - By MrDisibility (*) Date 12-13-2007 21:33
Thanks for the support here. I like to think my job is all about asking the questions "Why (or why not,m as the case may be)."

Mark, in reference to Ed Craig (http://www.weldreality.com/short_circuit_under100.htm):

    If welding stainless or carbon steel gage applications from 20 gage to 12 gage,  short circuit is an excellent weld transfer mode. If using specific > 2005 pulsed equipment and the correct pulsed settings, the pulsed mode which in the past has been unstable in the low current range can now also produce consistent weld quality with no weld spatter.

    A Short Circuit and Pulsed Transfer Weld Consideration.

    Short circuit weld transfer available from a low cost, constant voltage, traditional MIG power source with an 0.035 (1mm) wire, is ideally suited for manual steel applications 20 to 12 gage.

    Weld Fact: A benefit of the pulsed mode for thin gage applications is a lower cost, easier to feed, 0.045
    (1.2 mm) carbon steel or stainless wire may be used instead of the 0.035 (1 mm) wire.


This indicates that a switch from SC to Pulse would not only allow the use of Argon, but also a switch to .045 wire. I have a request in with purchasing for potential savings (man, having resources sure is sweet, coming from self-employed life not long ago).

I am wondering about the aesthetic quality of your SS welds. Are they discolored at all, and do you consider the corrosion resistance to be deteriorated after switching from He?

Marcus
Parent - - By MDG Custom Weld (***) Date 12-14-2007 14:47
Marcus,
We work in the same field, robotic welding ss exhaust components.  I work on the "hot side" parts which are usually from the catalyst up stream to and including the manifold.  The usage of stainless for our components on the "hot side" is not so much for corrosion resistance as it is for it's ability to stand up to the exhaust gas temps.  Even 300 series manifolds that are running exhaust temps of 1250F + will show surface rust after only a few months in service.  That being the case, we don't see failures of the welds due to corrosion.  Sometimes during dyno thermal testing of a new design we do see weld failures but it's usually due to unpredicted thermal expansion and contraction of the part that causes work hardening and failure of the heat affected zone at the weld toe.

Good quotes from Ed Craig on this.  My 2002 dates were from beta testing we did with Miller on the Auto Axcess during the product development, and that's also the year Fronius started their North America division.  Our plant is only 10-15 minutes from the Fronius NA office.  We were the first place to use their product in the states.

Back to the gas again.  He mixtures were very beneficial on the older pulse machines, it helps the puddle wet out and adds some stability at lower settings.  But as technology has advanced in the past few years, the power supplies have taken control and He gas mixes are not so critical.  Like I said, we use Ar/O2 for nearly all production welding.  It's much more cost effective to mix your bulk gas and have 1 main welding gas supply the entire shop.  As for appearance, the GMAW welds have a dull silvery look, not to be confused with the grey tint that you get when shielding gas is not adequate.

Good discussion you have prompted here!!

Mark
Parent - By MrDisibility (*) Date 12-14-2007 23:38
You know, the discussion would be nothing without educated and helpful posters. I come from a world where posting in a forum takes somewhat of a leap of faith which only half the time results in an intelligent discussion. This is very gratifying for me. I am considering joining the AWS because of it, and really taking hold of this subject as a professional pursuit. If I can get this shop on one gas (instead of 3-4) it will through no small part of you guys' help.

Anyway, we are working upstream as well, mostly catalyst sections on that side. We also produce mufflers and cats themselves from base raw materials. But the focus is really on our exhaust systems. Visual quality is a priority. To that end, failure due to corrosion is not our concern. It is primarily on the overall aesthetics.

Our mufflers welding with 5% O2 do not have great "finish good" visual quality welds. They are nice beads, but have a sort of greyish/brownish scaly buildup on them. I was told it was "glass" as a result of silicon buildup, or something of that nature. That was one of those moments when I scratched my head, as someone I know has history in the company and with the product was giving me honest to goodness 3rd or 4th hand information. We could scrape the buildup off and reveal a lustrous weld, but the labor at that point is too intensive. As such, mufflers with welds having a brownish buildup are "acceptable."

This appearance is not acceptable on the cat back systems though. With the current configuration on the cat-back robots we have low spatter (meaning, some spatter on start arc here and there, but not a major issue), satin grey welds. There is minimal discoloration on the base material, and no buildup whatsoever. This looks better next to our polished tips and mufflers, and satin stainless pipes than the "grey/browninsh" color ones. But it is not silvery in my initial mental image. All in all, the welds look good from a structural standpoint, but not evoking a "wow, look at those welds!" reaction. When we internally say the weld look good, it might only be in comparison to our mufflers.

Anyway, I pulled some parameters from the teach pendant. Ed called me today and an naturally suggested an AR/CO2 mix. Bump the feed by about 20% and the travel about 10%, drop the volts 'til we're happy. We might get some black or dark coloring on the surface, but we might be able to live with it. I guess it is time to give it a shot!!!

Marcus
 
Parent - - By MrDisibility (*) Date 01-05-2008 00:28
OK! Reporting back. We did some testing with a 97 Ar 3 CO2 today.

Flat stainless sections, just making 6-10" lines.

Welds look great. We are on a knife-edge though. We either start sinking into the material a bit, or get sparatic spattering. Nothing major, but the least we can get the better.

We used Ed's verbal advice and it got us in the right ballpark. After about 50 lines we got a few parameter settings that were really nice looking.

Thanks for your help and encouragement guys!!!

Marcus
Parent - - By MDG Custom Weld (***) Date 01-07-2008 14:31
Now that you are seeing better results, try using more of a push angle on the weld.  That should help you control the the puddle and keep it from sinking into the material.  Gun angle and travel speed have more to do with it than parameters alone from my experience.
Parent - - By MrDisibility (*) Date 01-07-2008 15:56
We are already around a 45 degree push, travel speed is around 235 in/min. Our best ones are at 185 amps -95 volts. We have a setting for "welding speed" which we are pretty sure is the wire feed rate. That likes to be around 35-36.
Parent - - By MDG Custom Weld (***) Date 01-07-2008 17:15
I think you might have the two numbers mixed up.
35-36 sounds like IPM of welding travel speed, and 235 sounds like the wire feed speed.

235 inches per minute of travel speed is like 3 seconds per foot, and that's faster than any mig welding I have seen! (other than laser hybrid, and I think that was in the 120ipm range)

What robot are you using?
Parent - By MrDisibility (*) Date 01-07-2008 17:31
You know that makes a lot of sense....but when we set the welding conditions it seems the opposite. I am a little confused by these setting actually, and I am relying on my programmer (whose  got limited welding knowledge, he's up there with me). I say "welding speed" is the travel speed, but he says it is the feed rate.

235 in/min is more like 3.9 in/sec. But that is still really fast. I tend to think this setting is more likely the wire feed as well, but we don't set this parameter in the AS command, we set it as a line movement.

We are on Daihen OTC robots.
Up Topic Welding Industry / Technical Discussions / Automatic welders want to change gas

Powered by mwForum 2.29.2 © 1999-2013 Markus Wichitill