Partial Joint Pen (PJP) Weld Strength - 6061 Alum

[ATSE]

Consider a rectangular tube vertical post - say 4"x8"x1/4" wall, welded to a 3/4" thick base plate.
I want to develop the yield strength of the post with a PJP + reinforcing fillet weld. That is, I want the post wall to yield before the weld or HAZ fractures (all of this theoretically and based on code values; not a true "capacity design" used in structural steel seismic design).

Based on ADM 2005 and D1.2, the "Code" seems to treat PJP with a reinforcing fillet weld as a fillet weld.

I have attached a one-page calc with the weld symbol and strength calculations.

I am looking for comments regarding the weld symbol construct-ability ("weldability") and the design methodology - that is, is this joint and weld appropriate for what I'm trying to do, or do I need reinforcing fins and stiffeners?

 

[DABwilldo]

Mechanical design calcs aren't my strong suit. But, did you consider that the shear strength in the HAZ of the base metal will be closer to that of the 6061-O material; about 12KSI? This is about half the shear strength shown on your calc sheet...

What's the bevel angle and what process are you planning on using to weld?

A TIG weld will benefit from a larger bevel angle, say 35-45 degrees. This will improve accessability to the bottom of the joint. Power capacity of the TIG power supply will also affect the number of passes required to complete the weld as shown. Probably a 3 pass weld, but a larger TIG power supply, like a Miller dynasty 700, could do it in 2 passes. Fewer passes = smaller HAZ and less distortion(all other things being equal)

What drives your choice of 5356 for filler metal? Not suggesting it's wrong, but I want to know your thought processes.

5356 is better for strength and ductility, but will not tolerate service temps over 150C.

4043 is friendlier in terms of cracking resistance and ease of 'weldability'. This is a heavy enough weldment that cracking might be an issue....so 4043 might be a better choice if you've a lot of these to make or a wide variety of welder skill levels employed in the fabrication...

A MIG power supply and a good spool gun will tolerate a smaller bevel angle (I feel) and get the job done faster.

Both processes will take a skilled welder to produce the highest quality welds.

David Benson
Benson's Mobile Welding & Fabrication

http://www.bensonmobilewelding.com

 

[ATSE]

The reason for 5356 over 4043 is as-welded tensile strength (35ksi vs. 24ksi for tensile, 17ksi vs. 11.5ksi for shear).
Fsuw' = welded shear strength of filler = 17 ksi per Table 7.3-1.
As much as I value ductility over strength, using 4043 makes my problem of trying to develop the pipe wall yield even more difficult.

You are correct that welded shear strength of base metal (6061-T6) is not 24 ksi, but instead 15 ksi (close to your 12 ksi estimate). Thanks for that - I will update that calc.

Service temp will not be more than 120F.

Maybe a better weld detail is in order?

 

[tbuelna]

ATSE,

I do not have access to the ADM, but your analysis approach seems OK for simple (P/A) tension or shear at the weld, assuming the weld is metallurgically sound. You did not specify the load case for the joint, but make sure your final analysis takes into account all combined loads.

It is good design practice to apply knockdown factors to your analysis for issues like stress concentrations, load reversals, or weld porosity. Stress concentrations can be due to weld issues like undercuts, or abrupt transitions from thick to thin section parts, or even from residual mechanical strains produced during welding.

Your particular case is a thicker flange attached to the end of a thinner constant section tube. There's an old saying "stiffness draws stress", and the locally increased structural stiffness occurring at the weld joint will naturally tend to concentrate stress in this area. One approach to diffuse this stress concentration condition is to add a doubler around the end portion of the tube wall that gradually tapers off in thickness away from the flange.

Hope that helps.
Terry

 

[Ron]

Your design approach is fine, it's just that the reduction in allowable stress for the base metal in the HAZ is huge for structural alloys of aluminum, so that will be your controlling factor (as DABwilldo noted).

Check your effective throat calculation. For the fillet shear condition, it should be 1/2 of what you are showing.

 

[ATSE]

Ron,
Thank you for your response.
Your last comment is a bit confusing. This is relatively important in terms of actual capacities, so please entertain my argument.

Regarding the effective throat:
For just a fillet weld (no PJP), the effective throat for a 1/4" leg fillet on a 1/4" thick plate would be (0.707)x(1/4") = 0.177".

However, the PJP alone is almost a CJP; then adding a reinforcing fillet, the effective throat must be larger than the fillet-only condition.
My throat geometry is based on Annex D (D1.2), which is the same as shown in D1.1 for reinforcing fillets.

Am I missing something?

As an aside, this calculation should be conservative, because a shear value is being used in an inclined plane that includes shear and tension, not just shear alone (I believe this is why AISC allows 1.5 increase to the shear capacity when loaded perp to the long axis of the weld).

 

[Ron]

ATSE...only the fillet portion is subjected to shear. The PJP section is in tension.

 

[ATSE]

Ron,
Thanks for the reply.
Based on solid mechanics, I agree with your last comment. However, based on ADM and D1.2, I believe that adding the design capacity of a PJP alone with the design capacity of a fillet weld alone is incorrect. That is, you need to use a single effective throat and use shear capacities, not tension capacities (unless you have a CJP). I'm not suggesting you said the opposite - just clarifying.
For the weld in question (3/16" effective) the eff throat will be about 1.414x0.1875 = 0.265" because of the reinforcing fillet. (The other limit states do not govern.) No?