Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Drafting question - weld symbol question

  1. Oct 30, 2012 #1
    Hey all,

    Not sure if things like this are dealt with here, but I'll give it a shot:

    I have a pipe with an outer "diameter A" (small) and a flange with an inner "diameter B" (large). Between them is a conical transition piece with the following characteristics:
    • The OD of the cone at the small end is equal to the OD of the pipe;
    • The ID of the cone at the large end is equal to the ID of the flange.

    This transition piece needs to be welded to the other two pieces. The ends of the conical piece are not machined, so the groove for groove welding is already there. The locations to which the arrows point in image are the locations of the welds:

    The one at the top needs to have a smooth contour, flush with the cone below and the pipe above;
    The one on the bottom also needs to have a smooth contour, flush with the cone on top and flush with the flange on the bottom;
    Both of these welds need an extra weld on the opposite side, ground down to create a smooth transition where there would otherwise be a sharp edge (thus, both opposite-side welds should be fillet welds with a smooth contour).

    Are the welding symbols/sizes I included correct (assuming roughly 1/8" thick sheet metal)? Groove weld, ground flush (G for grinding), with a fillet weld on the opposite side, ground to a concave shape.

    Last edited: Oct 30, 2012
  2. jcsd
  3. Oct 31, 2012 #2
    I'll give you my thoughts. So you want a groove weld on one side and a concave fillet on the other. What symbol convention are you following? I usually use this one:


    So the bottom one should be the bevel groove and the rest looks fine (rotate the triangle 180 and lose the bottom).
    A summary of what that will give you. A bevel groove with a grinding finish on the arrow side and a fillet with a concave finish on the other side. Sounds right to me.
  4. Nov 14, 2012 #3
    Thanks! Did exactly as you recommended :)
  5. Nov 14, 2012 #4
    No problem. Surprised noone else answered though.
  6. Dec 15, 2012 #5
  7. Dec 16, 2012 #6
    Looks like the sum of angles phi and omega.

    I may be completely wrong, but the way I see it is this:

    You have an axis (call it "A") that is along the length of the chord member.
    Define a ray (call it "B") along the length of the brace member, starting at the intersection with Axis A.
    Define a plane containing the two (call it "Plane P").
    Define Axis A as the horizontal axis on Plane P. Also, give a positive direction for the axis.
    Make an axis on Plane P that is perpendicular to Axis A (call it "C").

    Consider the set of planes you get by rotating Plane P about Axis C. The angle between the projection of Axis A on those planes, and the projection of ray B on those planes, is angle psi. Only measure from the "positive side" of A.

    For the plane to which A is orthogonal, psi is 90°.
  8. Dec 16, 2012 #7
    Plane P, Axes A&C, Ray B

    Rotated Plane, Projected Axis A and Ray B
  9. Dec 17, 2012 #8
    Thank You Sultan. However, if you look at the bottom of the diagrammatic representation you will find a table with variable values of 'ψ' describing the very same brace but saying Part A, Part B and C referring to the diagram on the top. That is, as you start from the left towards the right, the value of that variable just changes and that as I said for the same brace. So it must be having something to do with the curvature of the brace wall or perhaps its ovality. Still vague to me !!
  10. Dec 18, 2012 #9

    Yes, as I showed, the variable changes depending on where the section cut is. The same unit, viewed from a different angle, will have a different angle between the two parts. The reason it matters is because the weld geometry is based on the angle between the two, and the angle between the two surfaces changes as you move along the curve on which they intersect.

    Regarding the actual definition - I was mistaken before. It's not the angle between the axes as seen on a section view, it's the angle between the surfaces themselves at a point along the weld.

    Here's a definition from the code:
    local dihedral angle, Ψ (tubular structures).
    The angle, measured in a plane perpendicular to the line of the weld, between tangents to the outside surfaces of the tubes being joined at the weld.

    At any point along the weld, there will be a plane tangent to the surface of the first member and coincident with the point, and a plane tangent to the the surface of the second member and coincident with the point. The angle between two planes is called the dihedral angle.

    Last edited: Dec 18, 2012
  11. Dec 18, 2012 #10
    Thank You Sultan. I appreciate your willingness to help as I am trying to model that joint on ANSYS to run FEA and obtain Hot Spot Stresses around the joint periphery.

    You are right. I think this variable refers to the angle between the brace bevel and the chord. This angle still varies within each of the segments. The question here is how can we decide that angle depending on the angle of inclination of the brace w.r.t. the chord. Blodgett refers to a table extracted from AWS that defines partial penetration dimensions but does not link it to inclination angle. Any ideas??
  12. Dec 22, 2012 #11
    A welder looking at that diagram would be lost. Aside from the messed up weld symbols, what exactly are they looking at? A notched tube TIG joint?
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook