Outside Corner Joints- Grinding

By dwantz Date 01-25-2012 21:02 Edited 01-27-2012 12:46

First time poster here. We are having somwhat of a disagreement here as to whether grinding of weld material that extends past the parent material is allowed by code. (in this case D17.1, but it could also be D1.1 , D1.2, D1.9 etc.) I am refering to, is the area in the toe of the weld where a slight "rollover" if you will of weld (along it's longitudinal length) that typically occurs with the sheet metal thicknesses we typically weld 1/8" and under). I wish I could post a picture to try to illustrate what I am referring to, but can't figure out how to do that on this forum. To try to explain more it is a small amount "extra" weld that protrudes slightly above the parent material and beyond what the required weld size is (in this case no size specified, so a fillet weld the same size as the parent material in this case 1/16"). It has been our standard practice to just grind this material "flush" with the parent material in order to avoid potential interference problems in the down stream assembly processes. To be clear, we are not grinding away parent material and we are not compromising the minimum fillet size. The resultant weld after finishing meets the requirements had we been able to make a "perfect" weld to start with (without any "rollover"). A CWI that we have, is telling us that the weld must stay "as welded" with no grinding allowed because the weld symbol did not specify a contour or finishing symbol. Now, there is no finishing symbol that I am aware of that specifies grinding the weld in the area I am describing on an outside corner such as this, because it is NOT the weld face that is getting ground. It is my contention that so long as I meet the minimum requirements of the fillet size(without going undersize or grinding away thickness of the parent material), that I do indeed meet the code. I use the logic that even if this were a rejectable discontinuity or defect such as overlap for instance, the code allows grinding to remove the overlap condition (again, if done without detriment to the original reqirements), so surely, the same acceptance critera would apply to a weld that did not have a defect so I should certainly be "permitted" to dress this weld if I choose. Thoughts, comments?

By Superflux (****)

Date 01-26-2012 04:53

I have not had the pleasure to work with D17.1, but since you mentioned it, I have yet to find anything in D1.1 that disagrees with you.
As a Welder, I'd say to anyone on which ever side of the fence I'm on...
"Show me in WRITING where it says I CANT!"
I've lived by the moto (OK tried really hard) "Do not quote the code with out the book open to that page."

FWIW, I'm behind you.
Good luck on your argument.

By dwantz Date 01-26-2012 16:44 Edited 01-26-2012 17:29

I think I figured out how to post a picture, or at least a link to one to show what I am referrring to. The areas colored in yellow are the areas that we grind but are told that we should not be grinding. If it's not "required" to be there in the first place, why can't I remove it if I so choose?

http://s195.photobucket.com/albums/z215/DWantz/

[img

][/img]

By ozniek

Date 01-27-2012 16:13

Hi dwantz

When you grind on a fillet weld that has no "corners" from which to measure the leg length, it is impossible to know if it is correctly sized. I have made three different sketches below. The first one is the corner weld without any grinding. It is difficult to achieve this in any other way than to deposit a full sized weld. The second is a weld that had the toes ground, and the weld smoothed. While it is in fact full sized, it will tend to look very much like the third sketch, which is an undersized weld that has the unfused corners blended and weld smoothed over.

Some naughty welders will try to hide undersized welds by grinding the corners away, so inspectors often do not allow any grinding as it means they have no "reference point" from which to evaluate if a weld is full sized or not.

Regards
Niekie

By Lawrence (*****)

Date 01-27-2012 17:13

Great Illustration!

Thanks

By dwantz Date 01-31-2012 11:40

I certainly agree with this, and the "potential" for this to happen especially due to the fact that our welding is 99% light gauge aluminum. Yes we need to be very careful in this regard. We have a two step inspection process whereby the weldments get inspected in their "as welded" condition prior to any subsequent finishing of the welds. This is done mostly for dimensional verification and not necessarily for weld inspection as we have a seperate department for weld finishing. Weld finishing is not done by the welder in most cases (unless needed during the weldment assembly process). We then follow up with a final inspection by a CWI with the welds in their "final" condition (after finishing).

By waccobird (****)

Date 01-27-2012 13:02

dwantz

The one reason I give is cost.

If it is not a rejectable weld then picking up a grinder costs money.

The President here all the time goes out and walks the shop and one of his pet peeves is personnel grinding too much.

Until you pay the bills you can not fathom the waste that occurs in a shop where the employees dictate the procedures.

Just My ¢¢'s

Marshall

By dwantz Date 01-31-2012 11:54

Thanks for the replyt. I agree, I don't like doing "extra" work any more than the next guy, however when you consider all costs associated with the time it takes to clearly marking the drawings for every area that "must" be ground for form fit and function, going back and changing "old" drawings, training the design people, etc. etc. , and then factor the potential time "saved" from stardardizing a process so that the personnel doing the weld finishing do not have to decipher the drawing to determine how/ if each outside corner weld needs finished, you quickly eat away at the time you thought we were saving by not blending. I guess every company has it's unique internal processes for designing and creating drawings, and many companies have their welders do their own weld finishing. We have a seperate department for that, so there is more people involved to train etc.

The main question I have posed, is more from a "disregard cost for a second" and purely from a "Per the code, can I do it or not?" Once that is determined we (as a company) can decide what is best for our organization.

By Lawrence (*****)

Date 01-31-2012 12:05

Your answer has been clearly provided, with code quotations, based on the data you have given.

I don't understand your first paragraph at all.

"however when you consider all costs associated with the time it takes to clearly marking the drawings for every area that "must" be ground for form fit and function, going back and changing "old" drawings, training the design people, etc. etc."

If you are trying to do this constantly, I applaud you. If your trying to avoid this you will fail and be eaten by your competition, it's only a matter of time.

You would get enough "perfect" welds to eliminate your weld grinding department if you simply required your welders to blend their own work... Better quality, reduced cycle time, reduced rework, don't have to pay a weld grinding dept. I'm not guessing on this one... It is the voice of experience.

Spend the money on process control and training.

By Matthew Nemchek (*)

Date 01-26-2012 23:27

Where are the weldments going once they leave your shop? I worked for a company that used to make train components for a German company and despite the fact that we were fabing them in the USA we had to hold to the German welding standard "DIN". This standard has a few requirements that I had never experienced prior to working at said shop. From what I am told when DIN is the standard, you are not allowed to grind any welds to meet the design spec or to change the size or shape of how the weld was installed. If the weld is over sized, had reject able discontinuities, or any other reason you may want to take a grinder to a weld, the whole weld must be removed down to parent metal and reinstalled correctly.

Outside of that the only other reason may be surface finish. Depending on what you are using to grind the welds flush to the parent metal it may be leaving course grinding marks that require extensive metal finishing to take out? Or the fact that if the parts are getting painted grind marks stick out like a sore thumb even under coats of paint, and if its a part that gets visually looked at on a regular basis the customer may not be satisfied with the results. Just remember if you grind it flush and leave grind marks the metal finishing will take your grind job into the parent metal just to remove for and example, rock wheel scratches.

I only throw these out there because I have had, and heard people say just about anything to get you scared into doing it their way. It may not even have anything to do with code, they could be quoting you that as an excuse just because they think it looks better that way. When it comes to motivating people some will tell you anything.

Probably missed the boat, but I figured I may take a stab as to why. :red:

By dwantz Date 01-27-2012 11:55 Edited 01-27-2012 12:51

Matthew, thanks for the response. The components we are welding stay here in the USA , no foreign requirements. The reason to grind is mainly due to fit of the parts at the downstream assembly operations (many times they need to slide into one another). If the weld is left on they simply will not fit. Often times we (in the welding dept.) do not know how the part is being used. Unfortunately, our designers and engineers do not provide detailed enough drawings to direct the shop as to whether to grind the welds or not. They assume a "perfect" weld size will be met. We also many times have to work from customer drawings and have no engineering authority to change or enhance the drawings. Therefore many years ago we adopted an (unwritten) "standard" if you will, to just go ahead and grind these outside corner joints as I described. Yes, ....that means that there are parts that get ground that really wouldn't need to be, but in order to avoid potential issue in assembly and resultant rework, we feel it that the added cost to do this is minimal compared to the potential to have to rework components for weld interference.We use a two step process when grinding; 1st to remove the weld with a sanding disc, and then to use a D/A sander to remove the sanding marks. This allows for a fine finish that is easily covered by paint. This has been "working" for us for 35 years without a customer complaint. It has been questioned a couple times as to why we grind when we don't have too, however we have never been told (until now) that it is not allowed by the code even though no one can show me where the code states that. Remember that the weld size requirements are being met, and we are not affecting the parent material. Regardless as to any questions as to the cost and "why" we would do it when we don't have to, or why or designers are not being more specific on drawings, the only real question is this. If I submit a weldment to a CWI that has outside corners with the welds ground and finished to the way I described, and the conditions of the weld are otherwise acceptable (free from any rejectable discontinuities) Does he have to accept it or not?

By Lawrence (*****)

Date 01-27-2012 16:04 Edited 01-28-2012 19:43

Dwantz

I only have a copy of the 2001 standard at my desk today... But you should easily find the corillation in the latest ed.

5.14.2.1 (2) Welds Made Using Filler Material.
(2) An overlap, weld reinforcement, or other prodruding
indication may be removed by a metal removal method
that will not degrade the material's properties.
Corrections shall be performed in such a manner that
the weld size and base metal thickness will remail within
the drawing tolerances. The corrected weld bead shall
comply with 5.14.3

If your corner joint is autogenous your inspector may be right.

5.14.2.2 Nonmanual or Autogenous Welds.
The in-process correction listed in 5.12.2.1 is allowed only
when the engineering drawing or contract doccument allows
such correction.

D17 suggests aerospace... Most OEM's have a standard practice manual that referrs to situations like you describe. If you are supplying them (the OEM), you or your engineering staff should have access. Furthermore if it is aerospace work there are generally class 1 doccuments that provide very specific demensional requirements and include instructions and allowances for (blending) I don't like the term grinding, of reinforcement.

What you describe, in my opinion, may be termed accuratly as "reinforcement"

Edit:
You ask if the inspector "has to accept" I won't touch that with a ten foot pole. Inspection doccumentation is almost always more complex, with different data than fabrication and repair doccuments.. However, the inspector *should* be able and willing to point out the exact reason for rejection/non-conformance from a class 1 doccument. I find it hard to believe the welds are being rejected for a "time wasting" standpoint as far as blending goes. Reject processing takes more time than a blending operation.

By dwantz Date 01-31-2012 12:02

Lawrence,
Thanks for the response. That is exactly the paragraph that I used to state my position. If a weld with a defect such as overlap is permitted to be "in-process" corrected by grinding and this is acceptable per the code, then it only seem logical that a weld that didn't have overlap be permitted to be blended in this fashion. By the way all of our welds are made using filler.

By Lawrence (*****)

Date 01-31-2012 12:54

There you go.

Just keep in mind that internal doccuments for any fabricator may exceed the code while still being compliant.

Welds should not be rejected due to "opinion" There needs to be solid doccumentation to trigger a change in process flow, rework or part scheduling.