Not logged inAmerican Welding Society Forum
Forum AWS Website Help Search Login
Up Topic Welding Industry / Inspection & Qualification / Flare Groove Weld Size Question
- - By TimGary (****) Date 05-23-2017 13:18
I keep running across an issue with flare bevel and flare vee groove welds, as they're called out on drawings.
The attached sketch uses round bar to help clarify.

Please respond with your AWS D1.1 interpretation of the following:

1. A flare vee groove weld symbol with no indicated weld size calls for a CJP weld with the groove filled from radius line to radius line.

2. A flare vee groove weld symbol with a weld size detailed that is less than the groove depth details a PJP groove weld. However, to avoid an underfill defect, the weld face width is determined by the distance between lateral radius lines of the joined parts, or ensures the groove is filled flush. (The attached sketch shows that this case shows a very difficult weld situation)

3. If the engineer / detailer desires a PJP flare vee groove weld that is not filled flush, the only way to detail this is to add information in the tail of the weld symbol.

4. Welds detailed as stated in 3. above are not prequalified.

Thanks,
Tim
Parent - By 52757 (**) Date 05-24-2017 13:14
: [i]in grooves the weld face shall be full to the adjacent base metal unless otherwise specified on the drawing. Note: For flare bevel groove welds and  Flare V Groove welds where the weld symbol shows only a dimension with or without parenthesis the weld should be interpreted to mean the face of the weld can be below the surface of the base metal as long as the face of the weld is flat or slightly concave and  the throat meets the weld size. Weld size in this case measures from the root to the face of the weld as shown in dimension “E”. (This application is an approved deviation from standard symbol interpretation.)  This is a statement added to our weld standard from engineering, because of  the issue that you described. The flare bevel welds are difficult to get everybody on the same page for understanding what is required.
Parent - - By welderbrent (*****) Date 05-24-2017 13:41
Most of the time a flare bevel should be proven to be able to be achieved with the process in use by running a mockup and cutting and doing macros to prove that depth of penetration matches the required weld size.  Almost never will you get clear to the bottom of the 'V' but if the required weld size can be achieved and the weld remain lower than flush then a method of measuring the amount of decrease from flush will need to be established.  It should be a constant for the grooves in question as the radius should be consistent unless the size of material changes. 

Overall, I agree with your 3 statements and #3 and 4 were my main point as that would be a point of consideration often missed.

I seem to remember a similar conversation just recently.  Was it on the other forum? 

He Is In Control, Have a Great Day,  Brent
Parent - - By TimGary (****) Date 05-24-2017 14:34
Thanks for the replies Guys.

I posted this mainly to provide extra information for jhall773's post about "Round Bar WPS Question", without hijacking his post, and to see if I'm missing something in D1.1. Yes, we've talked about it before, but I think further discussion / interpretations would be helpful.

I like 52757's approach to documenting a deviation in the internal specification, though I'm not really comfortable with possible under-strength / non-fully fatigue resistant welds, which may also have issues with centerline cracking due to the possibility of weld depth exceeding width. I guess it depends on how critical the welds are in consideration with the weldment's loading requirements.

I think a lot of people miss that deviations from code weld joint design and acceptance criteria may be made, but only if documented by the Engineer, and if he/she is smart, backed up by PQR(s). This leads to problems during inspections, and is a recurring issue for me.

It's amazing how many people out there think that a flare bevel/groove weld size is determined by it's face width...
I've also yet to find and Engineer / Draftsman / Detailer / Cost Estimator who understands these requirements, and how varying corner radii on square tube affect the amount of weld required to adequately fill the joint.

Tim
Parent - By welderbrent (*****) Date 05-25-2017 13:07
And some of them get very large when they don't need to be. 

BB
- By pipewelder_1999 (****) Date 05-23-2017 23:54
I agree with all you have stated. I have run into instances in the field on tubing in which welds are below the surface. The welding symbol doesn't allow for it and often times the designer doesn't detail it otherwise.

If it were detailed as less than flush, I would have to see a qualified WPS for the condition and dimensions.
- - By 803056 (*****) Date 05-25-2017 14:29 Edited 05-25-2017 14:49
Flare grooves - AWS D1.1:2015 - Refer to Clause 2.4.1.4 and Figure 3.2 B-P11 (flare V) B-P10 (flare bevel).

The prequalified flare groove is filled flush the the top of the round edge and is limited (specified) per Table 2.1. The maximum size is a function of welding process, radius of the round members, and position. The maximum size is based on the groove being filled flush. A weld size less than the maximum is simple the max size minus the underfill. So, if the maximum weld size for the flare V-groove is 5/16 inch based on Table 2.1 and if the weld is specified as 1/4 inch, the inspector should be able to lay a straight edge across the round surfaces and measure a 1/16 inch (or less) space between the underside of the straight edge and the face of the weld.

The joint penetration is assumed to be the same for an undersized flare groove as it is for a full size flare groove weld. The difference is whether the groove is filled flush or not. The sketch below may be helpful.

A prequalified flare groove weld does not assume there is joint penetration to the point of tangency. Not to say it isn't possible, but it is unlikely for most situation. If one wants to take credit for penetration to the point of tangency, one simply has to qualify the process per clause 4.

The bottom line is there is a disconnect between a welding symbol calling for "complete joint penetration" relative to flare grooves, physics, and reality. If the WPS is qualified only for prequalified weld details, a CJP flare groove is not permissible. if the welding symbol specifies a flare groove weld without a weld size, I expect the WPS to be supported with an appropriate PQR for a flare bevel and a flare V-groove where the contractor has demonstrated joint penetration to the point of tangency is achieved.

Al
Parent - - By TimGary (****) Date 05-26-2017 12:44
Thanks for the input Al.

Please consider the following (concerning AWS D1.1:2015):

Agreed - There is no guarantee that a weld in a flared joint can achieve CJP through penetration meeting the point of radius tangency at the root.

There are provisions in the Design section 2.4.1.4 for determining the effective size of an underfilled flare groove weld.
- This gives the Engineer a method for calculating weld strength and an inspector a means of measuring an approximate underfilled flare groove weld size (as you efficiently noted), but no direction toward pre-qualification or how to express this desire in a weld symbol.

Acceptance criteria for groove weld profile is established in sect 5.23.3, 9.25,Table 5.8 and Fig 5.4, which depict groove weld underfill as unacceptable.
- Figure 5.4 is lacking a depiction of flare groove welds, which would be a helpful clarification, but typical groove weld profile is applicable.

Therefore, as an underfilled groove weld is visually unacceptable, such welds are not pre-qualified and are to be rejected by an Inspector unless an approved deviation is noted.
- Acceptable flare groove weld profiles must be filled at least flush to the point of tangency at the surface.

Tim
Parent - - By 803056 (*****) Date 05-26-2017 19:15 Edited 05-26-2017 19:39
I look at the situation differently. For the sake of clarity, I am referencing the New FC, D1.1:2015.

The maximum weld size of a prequalified flare groove weld is provided in Table 2.1. That is based on the flare groove being filled flush. The designer specifies the weld size to the left of the weld symbol. If the size required is something less than the maximum, it would still be specified to the left of the weld symbol and enclosed by brackets.

If the flare groove weld is something less than the maximum, the inspector simply subtracts the measured underfill from the full size weld to determine whether the design size (requirement) was met.

The sketch of the prequalified flare bevel and the flare V-groove are included in Figures 3.2; BTC-P10 and B-P11. The maximum weld size is listed in the appropriate sketch and the maximum size listed is in agreement with Table 2.1.

The bottom line is that flare bevel and flare V-grooves are prequalified provided the weld size specified does not exceed the maximums listed by Table 2.1 or the figures in 3.2.

Al
Parent - - By TimGary (****) Date 05-30-2017 12:04 Edited 05-30-2017 18:39
Agreed, there are provisions in new FC section 2 that give the Engineer guidance on determining underfilled flare groove weld strength.

Considering:
- The visual acceptance criteria for groove weld profile established in sect 5.23.3, 9.25,Table 5.8 and Fig 5.4, depicts groove weld
    underfill as unacceptable.
- Table 3.2 note l, calls for pre-qualified flare groove welds to be filled flush.
- Table 2.1 doesn't give the maximum size of a pre-qual flare groove, but the effective size of a flare groove weld filled flush.

In your scenario:
  How can a acceptable pre-qual flare groove welds exhibit underfill?
  What gives the Inspector latitude, during VT of an underfilled flare groove weld, to disregard the visual acceptance criteria?

Tim
Parent - - By 803056 (*****) Date 05-31-2017 04:28 Edited 05-31-2017 05:53
We must have different versions of FC:2015. My note say, "Weld size (E) shall be based on joints welded flush."

My understanding of the note is if the flare groove is filled flush, the weld size is E as depicted. The note does not say the flare groove must be filled flush.

Table 2.1 refers the reader to clause 2.4.1.4. As you mentioned the word used in the text is the "effective size" rather than the maximum size. However, without qualifying the groove detail per clause 4, the effective size is the maximum value the designer can can use for calculating strength. Summarizing; the clause also goes on to say that if the flare groove isn't filled flush, the size is determined by deducting the amount of underfill.

A parallel to this discussion is the lap joint. Clause 2.4.2.9 provides details on the maximum size of a fillet weld along the edge of a lap joint. That clause does not limit the designer to specifying the maximum fillet weld size in every case.

I will concede the sketches in Figure 5.4 depicts underfilled bevels and V-grooves as unacceptable. Unlike the provisions for sizing flare groove welds that are underfilled, the code makes no provisions for sizing underfilled bevel, V-, U-, or J-grooves.

Back to clause 2.4.1.3 which only addresses the minimum size of a PJP based on the thickness of the base metal. The reader is directed to clause 3.12.2.1. Clause 3.12.2.1(1) directs the reader to Table 3.5 to determine the minimum PJP based on the thickness of the base metal. The groove details selected from Figure 3.2 for flare grooves (BTC-P10 and B-P11) must produce at least a minimum size E that meets the size requirements from Table 3.5.  

In short, if the welding symbol specifies a 1/4 inch weld in a flare bevel groove consisting of a 2 inch diameter bar laying up against a flat plate, I would accept a weld that was filled to 1/16 inch less than flush. The calculation isn't difficult, the effective weld size if the flare bevel groove is filled flush is 5/16 inch based on Table 2.1. Subtract the 1/16 inch underfill and the the resulting weld size is 1/4 inch as provided by clause 2.4.1.4. 

The attached sketch shows an example for a HSS member with a 1 inch wall thickness where the minimum weld size (5/16 in.) based on Table 3.5. That is less than the effective weld size of 7/16 in. (0.468) for a radius of 1.5 inches per Table 2.1. However, consider the minimum weld size if the vertical member is reduced to 1/2 inch thick. The minimum weld size would be reduced to 3/16 instead of 5/16 while the effective weld size remains 7/16 in. (0.468) if filled flush and a low hydrogen process is used.

The smaller weld is preferred if the loads do not require a full size weld. 

Fun stuff in the FC.

Al
Parent - - By Steel5 Date 11-16-2017 22:43 Edited 11-16-2017 23:31
Ran across this thread looking for information on flare-bevel groove welds, this is a great discussion!  Sorry for posting on a somewhat older topic, hope you guys can provide me with some insights.

I’d like to ask a question regarding Al’s last paragraph starting with “the attached sketch…”.

I think I understand the general concept: flare bevel weld should be flush, but if it’s less than flush you can discount some of the weld and still calculate some strength out of it.

Where my question lies is in what the “Total Weld Size (E)” would be at the end.  I follow your calculation and sketch, but it’s written in terms of “effective weld size”, which I don’t think is the same as “Total Weld Size (E)”.

I typically work out of the AISC Manual 14th edition and have less familiarity with AWS D1.1.

I found an old edition of AWS D1.1 from 1996, and found table 3.5 (which is actually 3.4 in this version) and table 2.1, so I see where you’re grabbing the values from, but I’m concerned with the “Weld Size (E)” listed in the prequalified weld tables, which in this 1996 version is figure 3.3 and Flare-bevel-groove weld is on pg.65.

These prequalified weld tables are reproduced in AISC Manual table 8-2, which is where I most commonly refer to them.  In both the AISC manual (pg.8-61) and AWS D1.1 (pg.65 in old edition), the “Weld Size (E)” for a flare-bevel groove weld is 5*T1/8.  Interestingly, it’s not even a function of radius.  For the example you gave of a 1in thick HSS, the “Weld Size (E)” would be 5*1in/8 = 5/8in.

One thing I can initially conclude is that if the weld is flush, I have the full value of (E)=5/8in.

But what if the weld is not flush?  Say it’s just like your example, where the “effective weld size” was supposed to be 0.47in but is now actually 0.31in.

One method might be to look at the difference, which is 0.16in, and subtract that from (E) to get 5/8in – 0.16in = 0.465in.  But this seems erroneous because then if the “effective weld” were reduce to 0in, then (E) would still be 5/8in-0.47in=0.155in by that logic, which doesn’t make sense.

A better way to look at it might be to use a ratio.  0.31in/0.47in = 0.66 = 66%.  So perhaps the “Weld size (E)” value would be (E)=5/8in * 0.66 = 0.413in.  Seems to make more intuitive sense, but it’s just an idea, I’d like to know what you guys think.

Generally, I’m picturing this whole problem from an evaluating an existing condition perspective, but I also wonder what this would be like from a design perspective.  If I specify a (E)=(1/2in), how does the welder determine how much to weld?

Hope my question makes sense.  Ultimately the value that’s used in engineering design is the “Total Weld Size (E)”, so it’s very important to calculate correctly and understand.  Thanks for any help!
Parent - - By 803056 (*****) Date 11-17-2017 05:28
I'm not following your logic or your math.

If you reduce the weld size (E) to 0, the weld size is 0 regardless of the radius of the HSS.

The formula in the table are the maximum weld size the designer can assume if no testing is performed to determine what the weld size can be achieved for the proposed process and parameters. The assumption is the maximum weld size for the purpose of design requires the flare bevel to be welded flush. No extra credit is given for face reinforcement. 

The value of the radius is the assumed radius of the HSS if the designer doesn't have the means of going out and measuring the actual radius of the HSS being welded. The actual radius may differ from the calculated radius using the formula found in D1.1.

It isn't that the flare bevel should be filled flush, it is simply that the maximum weld size provided by the table assumes the bevel groove is filled flush. There is no reason the designer can't specify a smaller weld size if the loads don't justify the maximum weld size. The code provides a means for determining the actual weld size when the flare bevel groove isn't filled flush.

I believe the sketch included with the previous response shows how the actual weld size would be determined.

Al
Parent - - By Steel5 Date 11-22-2017 23:31 Edited 11-22-2017 23:33
Thanks for the response.  I agree with everything you’re saying.

Generally, when using the prequalified weld tables (AISC Manual 14th Ed. Table 8-2 and AWS D1.1-96 Fig. 3.3), the “Total Weld Size (E)” is used for calculating the strength of a weld, but it appears there are some nuances to it I didn’t realize before.

For example, for a single-bevel-groove weld (AISC pg.8-55), the “Total Weld Size (E)” = “S – 1/8”, which matches the definition of Effective Throat per Table J2.1, which is what’s needed to calculate the strength.

Another example would be a flare-bevel-groove weld (AISC pg.8-61), similar to your sketch.  The “Total Weld Size (E)” is listed as “5*T1/8”, which is a different definition from Effective Throat per Table J2.2 which is “5/16 R”.  Fortunately, for HSS members, their radius can be assumed to be "2t" per the footnote in Table J2.2, which means that these two different definitions come out to the exact same value when dealing with HSS members.

Your example through me off since you said the radius was 1.5 in. when the thickness was 1.0 in.  In this case 5*T1/8 does not equal 5/16 R.  That was the source of my initial confusion, I was convinced that the strength of the connection had to be based on 5*T1/8, but this is only true when R=2t like for HSS.  Although it might be a rare case, if the radius was truly 1.5 in. on a 1.0 in. plate like your example, you would have to calculate the Effective Throat based on 5/16 R (as you did), and disregard the 5*T1/8 equation.

Basically, your calculations are spot on, and there are limitations to using “Total Weld Size (E)” from the prequalified weld tables that I didn’t realize until now.
Parent - By 803056 (*****) Date 11-23-2017 12:15
It doesn't simplify matters when the standards that are supposed to compliment each other use different terminology or there is a lag between the two.

An example of the changes in terminology is the term "effective throat". The term is currently used when speaking of the fillet weld. It is the shortest distance from the weld root to face of the fillet weld. No credit is "given" to fusion beyond the joint root unless SAW is used. Likewise, no credit is given to the convexity of the fillet weld if my recollection is right.

The "equivalent" when working with groove welds is the joint penetration. Weld size is the joint penetration or the sum of the joint penetration in the case of a double groove weld.

No longer is the term "effective throat" used for both the fillet and the groove weld as indicated by a number of "outdated" texts and welding standards. People forget that our English language is not static unless one sticks with a dead language like Latin.

Al
Up Topic Welding Industry / Inspection & Qualification / Flare Groove Weld Size Question

Powered by mwForum 2.29.2 © 1999-2013 Markus Wichitill