allowable strength

By swnorris (****)

Date 06-21-2005 11:19

If you're referring to fillet welds made with 70 series electrodes, the maximum allowable stress on the weld would obviously depend on the weld size, and can be calculated using a .928 multiplier. For example, to calculate the maximum stress for a 3/16 fillet weld, multiply 3 x .928, which equals 2.784 kips per inch. So, a 3/16 fillet is good for 2,784# per inch. If the weld is 8" long, the weld is good for 22.272 kips, or 22,272#. A 1/4 fillet is good for 3.712 kips per inch (4 x .928), a 5/16 fillet is good for 4.64 kips per inch (5 x .928), and so on. To answer part of your question, I don't think the method of inspection would change that. The weld may be subjected to tension, compression, bending, shear, or torsion forces, which would also factor in determining the allowable stresses on the welds, and would help determine the fillet size, or whether partial or complete penetration welds would be required to support the design criteria.

By jbrotzman (*)

Date 06-21-2005 11:25

Thanks. I know the 'allowable' rule of thumb calculation and the variables that relate to loading. The question is does the allowable increase when one verifies the integrity of the weld by x ray inspection? It makes sense to limit the allowable for uninspected welds and to increase it for inspected welds. I'm trying to verify the inclusion (or exclusion) of this common sense in the code.

By swnorris (****)

Date 06-21-2005 13:05

I guess I'm missing something. Why would it need to? Why would there be a need or concern for the possibility of an allowable increase when a weld is x-rayed as opposed to a visual? Why even consider limiting the allowable stress for uninspected welds as opposed to increasing it for inspected welds, when by code, all welds are supposed to be inspected? The engineer or detailer has supposedly already calculated the weld strength and size, and if the joint is welded properly, it will perform as designed. So, why would there be a concern or need to look into the possibility of making a visually inspected joint stronger by more stringent testing? If the engineer has concerns about the integrity of the weld, he may choose to specify a method of testing.

By jbrotzman (*)

Date 06-21-2005 13:14

If, by verifying that the weld is without defect, its allowable stress is greater, then one can design with less weld metal which equals money in the pocket. And surely you do not suggest that all welds are x-ray inspected. Calculations of weld strength use only a fraction of the strength of the metal. Steels is shear yield at 57% of their tensile strength, commonly. The allowable is much less then that by code standards because of the vagaries of welding processes. And one is not 'making the joint stronger' so much as verifying its integrity and permitting a higher stress level to be employed.

By jon20013 (*****)

Date 06-21-2005 15:29

Scott; I'm interested in where the .928 multiplier comes from? Are there other multipliers for different fillers, for example stainless materials?

After 30+ years in this industry all I truly realize is how little I actually know! lol!!! What a great field we share!

By G.S.Crisi (****)

Date 06-21-2005 14:39

In my opinion, Mr. Brotzman's question is answered by ASME Code Section VIII, at least as pressure vessels are concerned.
Section VIII states that when the weld has been fully radiographed the Joint Coefficient is 1 (one), i.e., the weld is as strong as the base metal.
When the weld has been spot radiographed the Joint Coefficient is 0.85, i.e., the strength of the weld is 85% of that of the base metal; and when the weld has not been radiographed the Joint Coefficient is 0.7, i.e., the weld strength is 70% of that of the base metal.
I'll take a look at API 5L (Specification of Line Pipe), which I don't have on hand at this moment, and come back with the answer.
Giovanni S. Crisi
Sao Paulo - Brazil

By jon20013 (*****)

Date 06-21-2005 15:25

I agree with Professor Crisi. I believe the intent is not one of "strength" perse but rather one of engineering confidence.

By swnorris (****)

Date 06-21-2005 19:01

jon20013,

I got that multiplier years ago, but I can't remember where. It's referenced in the 9th Edition of the Manual of Steel Construction, ASD/LRFD Volume II Connections. If you have access to a copy, look on pages 2-29 and 2-30. Here is the formula for a 1/4" fillet shown on page 2-30:

1/4" fillet: P = 0.928D = 0.928 x 4 = 3.71 kips/in.

The "D" has been adopted to indicate the number of sixteenths of an inch in the leg size of a weld.

A more elaborate formula for a 1" fillet weld is shown on page 2-29, but the end result is the same as my simple 16 x .928 = 14.85 kips/in. The example on 2-29 is based on using E70 series, but where it shows Fv = .030 x 70 = 21.0 ksi, just replace the 70 with a 60 (for E60 series) an 80 (for E80 series), etc. I don't know about calculating for stainless.

By jon20013 (*****)

Date 06-21-2005 19:38

Thanks Scott! Good stuff!

By jwright650 (*****)

Date 06-22-2005 12:17

Jon,
Scott and I have Structural detailing backgrounds and that is an ole rule of thumb used by detailers for calculating how many inches of 70ksi weld of a given size is required. As detailer we had charts already worked out to use as a "quick glance reference" instead of having to do the math each time. There are multipliers to determine the number of HS bolts(of a given size) required also. We use these rules of thumb here at our shop to try and keep the fillet weld sizes down to just one pass where we can, by lengthening the number of inches of weld. For instance if 3/8" weld(3 passes) is required at a given length, we "try" to lengthen that weld "if possible" to where a 5/16 fillet (single pass) will suffice, such as making a plate or angle longer to accomadate the extra length of weld required.
I'm sure if you take the data for other strength fillers you can work out a multiplier that will work for those situations and create a chart for quick glance reference.
John Wright

By MBSims

Date 06-22-2005 03:37

I think if you calculate the allowable stress for a 1" long fillet using 70 ksi filler, based on leg length in 16ths of an inch, and combine all the numbers into one constant except for the numerator, you will wind up with the 0.928 constant. I'll have to run through this later to verify it though.

By swnorris (****)

Date 06-22-2005 11:05

I found a somewhat related article "Alternate Acceptance Criteria for Weld Size" in the May/June 1993 issue of the Steel Inspection News, which in part states the following:
The factor 0.30 is multiplied by the tensile strength of the weld electrode. Minimum values are given by specification, E70, as an example. The actual rod or wire will normally have a higher value. Secondly, the base metal involved in the welding also has an effect upon the strength of the weld. A36 today typically has a yield strength of 42 to 44 ksi, and the "multi-certified steels", even when sold for A36 requirements, will have a yield strength of at least 50 ksi. The resultant weld will have a strength higher than that anticipated. Overall, strength 10% above that assumed is about average.
When designing the weld, the engineer or detailer will normally "round up" to the next required size. If a fillet weld size of 0.20" is required, a 1/4" fillet is called for on the plans. Should the weld run slightly below 1/4", the weld will still be strong enough, even without consideration of the above strength factors. Only the engineer (or detailer) will know the actual weld size required.
For groove welds, the same principles apply. The required design thickness may be a 0.55" thick plate. The engineer would round up to the next standard thickness, probably a 5/8" thick plate. If the weld is underfilled, the actual thickness of material required could be used as an acceptance basis, rather than the AWS provisions. Again, only the engineer (or the person who determined the part thickness) may make this determination.

By petrogiant Date 01-09-2019 09:11

0.928=0.6*70*0.707/16/2

By Jim Hughes (***)

Date 01-11-2019 11:46

Thanks for reposting this. This is some great info.