WELD PEN vs. WALL THICKNESS REQUIRED FOR BUTT WELDED CYLINDER WITH INT. PRESSURE

[65CCJ]

I'm trying to determine why weld penetration values for fusion welded (TIG and LASER) tubes retaining a rupture disk don't meet min wall thickness requirements to handle internal pressure. A thin disk is compressed between two cylinders/tubes and the three parts are fused together with required weld penetration values dependent on disk thickness (See attached figure). The welds hold pressure but do not see tensile or bending loads and if the weld will see torque, the weld penetration requirement is increased from 1.25X Disk Thickness to .015 min. for diameters under 1.25". These weld penetration depths don't appear to follow standard Burst Pressure wall thickness requirements and I don't understand why they wouldn't (t[sub]min[/sub] = weld penetration). I have used Von-Mises and Barlows equation to see if the pressure rating of the disk = minimum required weld penetration but I see no correlation.

In the example shown, a .0065 thick disk is welded in place with .008 Min. Pen Requirement and this is increased to .015" if torque will be applied across the weld. If I run Von-Mises for min. wall requirement at 5000 psi and RT the wall should be .040"(.037" ASME). Doesn't this weld see full Hoop and Radial Stress? How does this joint hold 5ksi?

 

[EdStainless]

What does the assembly look like?
Are these installed inside a housing?
Is there axial compression on them?
The formulas around rupture disks are nothing like other pressure limit formulas.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[r6155]

Something is wrong.
Where is the safety valve?

Regards

 

[65CCJ]

I'm not concerned and am not asking about the disk. If there was no disk and I welded these parts together, why wouldn't the weld penetration not be equivalent to the minimum required wall to withstand internal pressure? Should weld penetration => tube wall thickness for distributing hoop and radial stress from internal pressure? My experience with these welds is the weld penetration is not close to the required wall thickness.
In General, the weld penetration is based on disk thickness for diameters < 1.25" unless you are applying torque across the weld.
Disk Thickness Up to .005” - the weld pen is 2X thickness
Disk Thickness from .006-.010 - the weld pen is 1.25 x thickness.
In my example, a .006" thick disk with 5000 psi rupture pressure (controlled by scoring, not just thickness). So weld pen for 5000 psi using .006" disk is .008". The required wall would be .040" based of Von-Mises or Lame pressure calc. That's 5x thicker wall than the required weld penetration. How does the weld joining two tubes not see the same stress of fail?

 

[r6155]

How do you plan to replace the rupture disk?.

Regards

 

[65CCJ]

You replace the assembly. Overinflation valves used on a wide variety of products are welded assemblies with AS port connections. They are throw-away assemblies that can be replaced like any other tubing fitting.

 

[65CCJ]

Here are examples of typical welded fittings with rupture disks.

These hold internal pressure until rupture and potentially after rupture. Does the weld penetration depth need to be equivalent to the calculated minimum wall thickness? It would need to handle the same stress as the surrounding wall before and after rupturing the disk. We use a factor x the disk thickness for weld pen but it doesn't give the result weld pen = min. wall thickness.