Weld contour within tolerances

[trottiey]

Do welds have to meet the bulk dimensions and tolerances of the part? Intuitively, I thought that weld beads would only have to meet if the weld symbol indicates a flat or concave contour. If the weld symbol indicates a convex contour, which is the default, I'm used to seeing big rugged weld beads that go weld beyond tolerance boundaries. These bead sizes are still limited by tolerances on the weld symbol and associated codes and welding procedures, but not by the diameters, flatness, etc, of the parts being welded together. I always assumed there was something in the standards that allowed that, but I couldn't find anything in AWS A2.4 or ASME Y14.5. Any thoughts?

 

[ctopher]

I have always created weldment drawings as assemblies. The tolerances follow the weld symbol and the assy drawing.
Any other standards are followed on the part drawings before they are added to the weldment dwg. This is how it has been done at every company I have worked at.
Others may have come up with other ways.

Chris
SolidWorks 10 SP4.0
ctopher's home
SolidWorks Legion

 

[trottiey]

I've seen that approach before, but many companies just produce weldment drawings without subordinate part drawings. See ASME Y14.24 section 4.2 and Figure 9, "Insperable Assembly Drawing." That approach has the advantage of giving the fabrication shop full freedom to adjust part lengths as needed to compensate for weld shrinkage and distortion. In my experience, this is the norm for pressure vessels and structural frames.

Figure 9 in ASME Y14.24 shows a flush weld contour, thereby avoiding the problem. But what if there was no contour symbol, so that a convex or slight fillet shape was acceptable? Would the weld bead still have to lie within the 4.57±0.25 tolerance band of the flange? My experience says no, but I've been challenged on this and I can't find standards to support this interpretation.

 

[trottiey]

For clarity, I'm not talking about following the stock dimensions of the material. Clearly, the welds can exceed those. But for example, if you have two machined cylinders welded together end to end, and your weldment drawing shows a diameter of 2.5±0.010 for the whole weldment, can the weld bead diameter exceed 2.510? To my mind, a flush contour symbol would mean yes, a convex contour would mean no.

 

[trottiey]

I got the last sentence backwards:
To my mind, a flush contour symbol would mean no, a convex contour would mean yes.

 

[ctopher]

The weld bead can exceed the part dia. The weld bead needs to not exceed the weld symbol tol.
If the weld needs to be the part dia, a separate machining dimension should be added after weld.

Chris
SolidWorks 10 SP4.0
ctopher's home
SolidWorks Legion

 

[ak762]

How you will seperate dimensions in weldment drawing which for machining after weld from some for weld assy?

 

[PeterStock]

If a surface is machined it gets a "remove material" type of finish mark, all others are as produced before welding.

Peter Stockhausen
Senior Design Analyst (Checker)
Infotech Aerospace Services

http://www.infotechpr.net

 

[nithinku]

We give Flush symbol whenever Weld not to exeed the Part dimensions.
When 2 parts welded end to end Either Butt weld or Groove weld given (for Groove weld Chamfers made in both parts) .So its easier to maintain the part dimension.

NX 6.0.2.8 MP4
Teamcenter 2007
WINDOWS XP (64 Bit)

 

[trottiey]

OK, it sounds like nithinku and ctopher basically agree with me. For further clarity, I've attached a small sketch this time. Is there any standard or textbook that says that, so I can prove it to my colleagues?

 

[PeterStock]

You have not provided enough information to determine the .062 dimension. Do the dimensions in the top section apply at the assembly or part level?

Peter Stockhausen
Senior Design Analyst (Checker)
Infotech Aerospace Services

http://www.infotechpr.net

 

[trottiey]

Hi PeterStock,

if you read my earlier posts in this thread, you will see that we are not making separate part and assembly drawings. The top section in my sketch is an example of what would appear on the weldment drawing, and all dimensions apply at the weldment level.

The 0.062 dimension is an arbitrary example. My point is that an ID bump far beyond the ±0.010 tolerance would be permitted. This is a normal feature of welds called "drop-through." On a long pipe, it's awfully hard to go in after the weld and machine down the ID.

 

[ctopher]

I don't think a welder will measure the 'bump' of a weld inside a diameter.
Add the weld info on the drawing, then a dimension for the machined ID with a note to be machined after the weld.

Chris
SolidWorks 10 SP4.0
ctopher's home
SolidWorks Legion

 

[ctopher]

It also depends on the design intent...

Chris
SolidWorks 10 SP4.0
ctopher's home
SolidWorks Legion

 

[PeterStock]

If I was reading the drawing and saw what is in the top view, I would expect that the MMC of the ID to be 2.990. Yes, welding will cause a bump inside, but if the dimension applies to the assemby, it applies after welding, unless there is a note that dimensons apply before welding or that dimensions do not include welds. If the size of the ID after welding is important, you need to specify what is acceptable after welding.

Peter Stockhausen
Senior Design Analyst (Checker)
Infotech Aerospace Services

http://www.infotechpr.net

 

[trottiey]

It seems I've received 2 votes for my interpretation and 2 against (counting my colleague) and no clarity from the codes and standards. I guess clarity must be achieved by notes. It's a pity though - this seems common enough that the standards should talk about it.

 

[SteveMartin]

My understanding is that for inseperable assemblies like the one in your sketch that all tolerance requirements apply at the weld point. So without a note of some sort, the weld cannot exceed the listed ID dimensions.