Welding 6061-T6 Question

Discussion in 'Sheet Metal' started by lake_harley, Jan 9, 2008.

Help Support HomeBuiltAirplanes Forum by donating:

  1. Jan 9, 2008 #1

    lake_harley

    lake_harley

    lake_harley

    Well-Known Member

    Joined:
    Jan 5, 2007
    Messages:
    960
    Likes Received:
    303
    I understand that TIG welding 6061-T6 aluminum takes the "heat treatment" (strength?) out of the aluminum, but that in time it will return to about T4 strength through natural ageing. Can anyone offer information to confirm or contradict this? If time "brings it back", how much time?? The material in question is .058" wall, 6061-T6 aluminum tubing.

    Thanks, in advance!

    Lynn
     
  2. Jan 9, 2008 #2

    BBerson

    BBerson

    BBerson

    Well-Known Member HBA Supporter

    Joined:
    Dec 16, 2007
    Messages:
    12,642
    Likes Received:
    2,581
    Location:
    Port Townsend WA
    I am almost certain your statement "the strength returns through natural aging" is incorrect in the case of welding.
    The term natural aging applies to properly heat treated parts that undergo a precise process. Such as, "soak at 950F +or- 5F for 15minutes and quench in water within 1 second". If this is done the part can naturally age in a few days.

    Welding is closer to annealing (weakest and ductile heat treatment) because the proper quench does not occur unless the entire part is heated evenly.
    The strength at the weld will be lower, how much depends on the part, perhaps half strength. Steel retains the strength better because it has a property called "air hardening", this does not work with 6061. That is why most welded tube fuselages are 4130 steel even though steel is heavier.
    I did several weld tests to confirm. The weld area is very soft and breaks easily. I do not recommend welding 6061 for structures without extensive joint research.
     
    Topaz likes this.
  3. Jan 9, 2008 #3
    From Machinery’s Handbook, Twenty Fifth Edition: “ Many Alloys approach a stable condition at room temperature, but some alloys, particularly those containing magnesium and silicon (6061) or magnesium and zinc, continue to age-harden for long periods of time at room temperature. By heating for a controlled time at slightly elevated temperatures, even further strengthening is possible and properties are stabilized. This process is called artificial aging or precipitation hardening.”

    Yes, 6061 has been used in countless “structures” with great success in very demanding environments.

    Dave
     
  4. Jan 9, 2008 #4

    BBerson

    BBerson

    BBerson

    Well-Known Member HBA Supporter

    Joined:
    Dec 16, 2007
    Messages:
    12,642
    Likes Received:
    2,581
    Location:
    Port Townsend WA
    Lynn,
    The Alcoa Structural Handbook states: "the expected minimum yield strength of material adjacent to the weld in 6061-T6, in the direction parallel to the weld, is about 11ksi."

    To return the metal to T-4 would require the full solution heat treatment process in a furnace (970F about 10 minutes)and a rapid quench. Only then will it age harden to the T-4 yield of 14ksi for extruded tube and 16ksi for drawn tube. Still much below the 35ksi yield of the original 6061-T6. You can get back to the T-6 with further precipitation heat treatment (6-10 hours in the furnace at 350F)
    Are you planning to re-heat treat the part?
     
  5. Jan 9, 2008 #5

    PTAirco

    PTAirco

    PTAirco

    Well-Known Member

    Joined:
    Sep 21, 2003
    Messages:
    3,501
    Likes Received:
    1,013
    Location:
    Corona CA
    I wonder how all those Honey Bees and Pegasus airplanes are holding up with their welded 6061 square tube fuselages ? Does anyone know anybody who has one?
     
  6. Jan 10, 2008 #6

    lake_harley

    lake_harley

    lake_harley

    Well-Known Member

    Joined:
    Jan 5, 2007
    Messages:
    960
    Likes Received:
    303
    Good question. I talked with the owner of a Pegasus H-3. He didn't indicate any problems with the welds, but then again, he was trying to sell it;)!

    For the sake of discussion, the application I was hoping to use the weldment on is hinges for my ultralight's control surfaces. It would be (in one case) a 1 1/8" X .058" and a 5/16" X .058" tube welded parallel to one another in a "log" of sorts, and then cut to a width of 5/8" for each of the two sections making up a hinge 1 1/4" wide, pinned with a 3/16" clevis pin. The sections would be held to the support tube and control surface with 2 - 1/8" rivets.

    I'm not going to risk my well being on the design if it's not going to be strong enough. I wish I knew exactly what IS strong enough to hold on a aileron, elevator or rudder:ponder:. I did like the idea that the tubes would be close together to get rid of big gaps and potential added drag. I have seen other workable ideas for hinges and will go one of those routes if necessary, but what I tried to describe seemed really simple, light and tidy.

    To satisfy my curiosity I'll have to calculate what 11K PSI translates into in terms of my material dimensions.

    Thanks for the input so far!

    Lynn
     
  7. Jan 10, 2008 #7

    orion

    orion

    orion

    Well-Known Member

    Joined:
    Mar 3, 2003
    Messages:
    5,800
    Likes Received:
    135
    Location:
    Western Washington
    Personally, I was introduced to the concept of the weldment returning to a "T4" condition back at college during one of the ASME race car builds. Our space frame was welded 6061-T6 tubes and when the subject came up, the instructor and the consultant we had both indicated that the welded material will eventually age to the T4 condition.

    Since that point I've heard the same thing from one of our material suppliers, a local marine davit manufacturer who uses 6061-T6 almost as much as the marine alloys, and a retired local welder I met some time back who is certified both by the FAA and by NASA. The common theme in all the replies though is that the material returns to the T4 stage "eventually" and despite searching for a more concrete answer, I have not been able to nail down the time period any closer.

    One gentleman I talked to about three years ago indicated that he though it might be on the order of a couple of years but from the sound of it, I think he was guessing. His opinion though was that since most homebuilts take some time to build anyway, the time to age back to T4 was not so much an issue since most commonly it'll be several years before the airplane flies anyway.
     
  8. Jan 10, 2008 #8

    lake_harley

    lake_harley

    lake_harley

    Well-Known Member

    Joined:
    Jan 5, 2007
    Messages:
    960
    Likes Received:
    303
    Orion....Thanks for your post and the additional info. I hope it doesn't take 2 years for me to finish the ultralight I'm working on:).

    I did calculate the weldment strength figure based on multiplying .058" X .625" (thickness and width of welded tang/hinge tube) and multiplied that times 11,000 PSI tensile strength (6061-T6 figure reduced by welding), and came up with 390# (+/-) for the basic tensile strength of one "leg" of the side of a weld (providing I did the math right). I don't know what a "reasonable strength" is to hold on an aileron, rudder or elevator for an ultralight doing 50-ish MPH under good-weather, flying-around-for-the-fun-of-it flying. I had planned on four hinges for each aileron and the elevator, and three on the rudder. The hinge pin (3/16") shear strength would play into the picture too, as well as the two 1/8" rivets holding it on the tubes. I may have to do some searching to see what the shear rating is on a 3/16" pin. I think the rivets are rated at about 300# each in shear.

    Thanks again for all the posts by everyone! I appreciate the help.

    Lynn
     
  9. Jan 10, 2008 #9

    lake_harley

    lake_harley

    lake_harley

    Well-Known Member

    Joined:
    Jan 5, 2007
    Messages:
    960
    Likes Received:
    303
    I was thinking today (generally dangerous) about what the load would be on a control surface hinge to determine the hinge strength needed. My ultralight will have a wing loading of under 4#/Sq. Ft. but for the sake of an example let's just say 4#. I'm using a full length aileron on a 12' wing 1/2 span and even though the aileron is less than 12" chord set's say the aileron has 12 Sq. Ft. of surface area. With 4 hinges spaced evenly the two middle hinges would each "see" 4 Sq. Ft. of the aileron. At the 4# loading (providing the wing loading applies to the aileron as well) that would be 16# of load in level flight OR is there a "moment arm" that has to be taken into account? Even in a extreme manuver that would stress the airframe to 6 G's, the hinge load would only be 96# unless there's the leverage issue that I'm not seeing or thinking of. Now I realize that there could be vibrations that are more destructive than just simple, even loading but with my figures from an earlier post in this thread (hinge calculated strength of 390#) it seems that there's a big safety factor left.

    In reality I'm planning to re-build my welded hinges using 4130 of the same dimensions instead of thealuminum but would like to actually learn something in the process for later use. I'm not planning to ever design a plane from the ground-up but really want to understand what constitutes a under-built or an over-built part.

    Thanks, as always, for input and education.

    Lynn
     
  10. Jan 10, 2008 #10

    PTAirco

    PTAirco

    PTAirco

    Well-Known Member

    Joined:
    Sep 21, 2003
    Messages:
    3,501
    Likes Received:
    1,013
    Location:
    Corona CA
    The FAR 23's have a simplified table for determining acceptable control surface loadings. The tables don't go down quite as far as your wing loading of 4 psf, but 12 psf happens to be the lowest value sfyou can get for ailerons. Which seems high since the rest your wing has an average loading of on 4psf, but consider the possibilty of flutter and general rigidity of the surface. I think if you aimed at a design load of 10 psf you'll be pretty safe.
    The 3/16 hinge pins would be the least of your concern with a shear rating of over 2000 pounds, but I would use hefty rivets for attaching the hinges trying to keep them in shear as much as possible. Cherry Q-types or Avdel maybe?
     
    Last edited: Jan 11, 2008
  11. Apr 23, 2008 #11

    steveair2

    steveair2

    steveair2

    Well-Known Member

    Joined:
    Jun 2, 2007
    Messages:
    975
    Likes Received:
    684
    Location:
    Dallas Texas
    Lyn,

    My Uncles Kolb MK111 has been flying for over ten years, and has flown coast to coast a few times. It has a Rotax 912 engine.
    It uses MS20257-5 aluminum piano hinge.
    It's riveted on to the trailing edge with eighth inch stainless pop rivets.
    Look at some other aircraft that are similar to yours.
    I would not weld the hinges.

    Steve
     
  12. Dec 31, 2008 #12

    Boiler

    Boiler

    Boiler

    New Member

    Joined:
    Dec 31, 2008
    Messages:
    4
    Likes Received:
    0
    Location:
    Muncie, Indiana
    I stumbled upon this thread today doing a Google search on aging treatments for aluminum.

    It was an interesting thread, but from what I can see there is some very dangerous thinking here.

    1) I'd never base a design on the fact that someone else made a fusilage from aluminum without failures. If it is a company that builds airplanes, they surely have someone analyzing the design.

    2) "Moment Arm" should never be in quotes like it is a theoretical and somewhat phony term. Yes it is very important in your situation. If you are designing a cantilever style beam (like a wing) you should be seeing bending stresses based on the moment arm that are much higher than shear stresses.

    3) Bending stress on a beam such as a round tube that has one end welded fixed and a perpendicular load at the other end can be calculated as follows:

    Stress (PSI)= Moment (in x lbs)/ Section Modulus (in^3)

    where:

    Moment at a given point = distance (in) from the load center to that point times the load (lbs)

    Section modulus for a round tube: pi * (O.D.^4 - I.D.^4) / (32 * O.D.)

    ***Note***
    I don't pretend that this will size your equipment. This is just an illustration to how much more needs to go into typical sizing calculations. Beyond the bending, there are other stresses such as shear that need to be accounted for. Bending should be the largest stress however, with a beam like a wing that would have a length much larger than its cross section.

    Be careful guys.
     
  13. Dec 31, 2008 #13

    orion

    orion

    orion

    Well-Known Member

    Joined:
    Mar 3, 2003
    Messages:
    5,800
    Likes Received:
    135
    Location:
    Western Washington
    A couple of points here: One, your stress equation is missing one term: The distance from the neutral axis to the outer-most fiber of the beam in question.

    And two, tubes in bending cannot be analyzed in this manner since most failures are a function of local crippling, not material failure. Tubes must be analyzed through a statistical database for the type of tube, the type of load and the specific tube geometry (diameter, wall thickness, etc.)
     
  14. Dec 31, 2008 #14

    Boiler

    Boiler

    Boiler

    New Member

    Joined:
    Dec 31, 2008
    Messages:
    4
    Likes Received:
    0
    Location:
    Muncie, Indiana
    The section modulus equation is derived from inertia & that distance from the neutral axis to the skin. I plucked the section modulus equation from Marks as I don't usually deal in rounds, I deal with non standard cross sections in which I have to determine a section modulus through calculation of areas, etc.

    Please elaborate on item number two. I'm not familiar with the term "local crippling". Also if you could point me toward any online reference to this type of database or analysis, I'd be interested. I typically do beam calculations in this manner and then put a solid model of the beam into Ansys for quick comparison and checking.

    Like I said in the note, I wasn't trying to tell anyone how to calculate their stresses as much as I was pointing out that they weren't even close from what I could see. Another thing I didn't bring up was safety factor. I'm not familiar with typical safety factors used on those types of designs. Mine range from 2.75 to 3.00 in the "man-lift" business. Just throwing up a red flag more or less.
     
  15. Dec 31, 2008 #15

    Boiler

    Boiler

    Boiler

    New Member

    Joined:
    Dec 31, 2008
    Messages:
    4
    Likes Received:
    0
    Location:
    Muncie, Indiana
    To get back on topic, I'd recommend anyone new to aluminum welding & aluminum welding design to read this book. It's considered "the Bible" in our shop. It contains text on "as welded" strengths of aluminum in the Heat Affected Zone (HAZ), as well as tables showing tensile and yield strength of T6 & T4 material that has been welded, welded and aged, and welded and solution treated and aged.

    Also, there are a bunch of good articles on welding and designing aluminum weldments at lincoln electrics website. One in particular that taught me enough to be dangerous with heat treatment is linked below as well.

    e-Orders : home

    Common Mistakes Made in the Design of Aluminum Weldments | Lincoln Electric
     
  16. Dec 31, 2008 #16

    orion

    orion

    orion

    Well-Known Member

    Joined:
    Mar 3, 2003
    Messages:
    5,800
    Likes Received:
    135
    Location:
    Western Washington
    In general, when dealing with potential failure modes that are partially driven by membrane instability, I've often found that FEA systems have significant limitations in being able to predict that behavior. Yes, most do incorporate a buckling algorithm but when dealing with aircraft structures, the preferred (by industry and by FAA) solution is one that is based on hard empirical evidence rather than behavioral theory. As such, the use of classical methods is often a better approach for predicting crippling failure, although the FEA approach is better suited for predicting stress levels in the region, especially in more complex assemblies.

    A fairly good write-up of membrane instability and localized crippling, as well as a listing of applicable formulae and applicable factors, are included within the pages of Roark and Young's "Formulas for Stress and Strain".

    The failure behavior of tubes and similar constructs can be quite critical in that the crippling failure can occur substantially below the calculated strength value of the tube (as little as 30% of the value calculated by the aforementioned simple equation). Some years back I was asked to examine the possibility of increasing the gross weight of the two place Grumman aircraft - the request was based on a cursory analysis of the tubular spar used therein, which indicated that the wing might be significantly over-designed. However a detailed examination of the spar showed that although slightly conservative, the spar strength was actually only a few percent above that which is required by the Normal Category, as defined by the FARs. The increase was not justified.

    And one side note - for the most part most members herein are pretty cognizant of the need for responsible structural design, including the understanding of the applicable loading scenarios and safety factors. The analysis of a wing is not necessarily a straight forward exercise and as such, most of the discussions herein are based on the assumption that said analysis will be conducted to the level necessary for the application. But if any do try to put forth the notion that this is a simple task, you'll find that the membership will try to straighten them out in fairly quick order.

    Finally regarding welding, usually this is brought up as a potential approach for a simpler structure. It however rarely turns out to be the case and given the complexities and variables generated as a function of the localized heating, welded aluminum structures in aircraft are very rare.
     
    Topaz likes this.
  17. Dec 31, 2008 #17

    Boiler

    Boiler

    Boiler

    New Member

    Joined:
    Dec 31, 2008
    Messages:
    4
    Likes Received:
    0
    Location:
    Muncie, Indiana
    Thanks! I'm pulling 'ol Roark off the shelf right now...
     
  18. Dec 31, 2008 #18

    Midniteoyl

    Midniteoyl

    Midniteoyl

    Well-Known Member

    Joined:
    Sep 4, 2003
    Messages:
    2,406
    Likes Received:
    501
    Location:
    Indiana
    And BTW, Boiler... Welcome!!

    :D
     
  19. Mar 10, 2010 #19

    billmce

    billmce

    billmce

    New Member

    Joined:
    Mar 10, 2010
    Messages:
    1
    Likes Received:
    0
    Location:
    Aurora, Ontario, Canada
    I stumbled onto this thread looking for info on 6061-t6 and the impact of welding it.

    On analyzing A/C structure a good book to get is Analysis and design of flight vehicle structures by Bruhn. Yeah is sounds all complicated but it is a really good book, particularly section B (materials and fasteners - data) C ( strength methods) & D (fittings & connections - it covers welding briefly) which contain worked examples for many common aircraft structural calculations. It was written in a style that is as a training book for a junior engineer joining an A/C firm in the structure area. Section A is a big part of he book and covers a lot of the theory.

    Also, for the aileron loading question: I would take Vne (never exceed speed for your plane) and calculate "q" - (dynamic pressure = 1/2rho V^2) and multiply by the area of the aileron and then multiply by an appropriate Cd (see Hoerner Fluid Dynamic Drag to look Cd's or try google - flat plate drag at 90 deg to the air stream might not be a bad choice). That will give the load that the air will put on the aileron - then figure out the load at the hinge, assuming some distribution.

    I would not rely on a welded joint for the most critical hardware system on the machine.
     
  20. Mar 31, 2010 #20

    Othman

    Othman

    Othman

    Well-Known Member

    Joined:
    Aug 23, 2004
    Messages:
    355
    Likes Received:
    4
    Location:
    Ottawa, ON, Canada
    Strength of welded aluminum joints is always a topic of debate in engineering. Some engineers use them regularly in their designs (with proper analysis and testing) and some avoid them all together... much like the debate between going with sheet metal or composite construction.

    Anyway, I've attached a table from an older document, Lockheed Stress Memo 129a, Welding, (also reproduced in Airframe Structural Analysis by Michael Nui) which shows the weld efficiency factors for welded 6061 alloys. For welded 6061-T6 left as-is (no heat treatment afterwards) the table shows that the heat affected zone (HAZ) will have strength properties 60% of the original values. This value is considered to be quite optimistic in the industry outside of the big manufacturers, since it depends on how the welding was done (by machine or by hand) and by whom. 50% would be a more reasonable value.

    So to estimate the tensile and shear allowables for the weld, use 50% of the values for 6061-T6 (the base material). You will find that they are even lower than T4 values.

    The other side of designing welded joints is making sure the analysis is done right. The type of stress in a weld (tension or shear) depends on the type of weld used (butt, fillet, groove etc) and how the weld is loaded. It is not as simple as “a tensile load will result in a tensile stress in the weld”.

    Combined stresses should also be considered to determine the max shear and principle stresses. A good reference for weld analysis is Mark's Calculations for Machine Design by Thomas H. Brown, Jr.


    Ashraf
     

    Attached Files:

Share This Page

Group Builder
arrow_white