testing reality for Design of Welded Structures calculations

Hi everyone
Want some advice on how to practice the skill of designing welded joints. I imagined doing some tests and seeing how it really works.
Been talking with one of our fellow contributors here by email and he says a lot of engineers aren't at all confident in their designs - says often specify adding more weld after seeing the thing as made. So sounds like it would be doing the right thing to connect theory to practice and be confident how it really pans-out.
I've just been very enthused by the "Design of Welded Structures" module of the Welding Engineering course I'm doing
There's lovely formulae like
J_w = b^3 / b + 2d \ +     d^2 (6b + d) --- | ------ |      --- 3 \ 2b + d /      12
[R. S. Funderburk, O. W. Blodgett, C. J. Carter, M. V. Holland, L. A. Kloiber, R. M. Kotan, W. W. Sanders Jr, R. E. Shaw Jr, W. A. Thonton: "Design for Welding" in Welding Handbook, American Welding Society, (ed. C. L. Jenney, A. O'Brien) 9 th ed. Vol. 1, pp. 209-216.]
Quoting bit from the correspondence mentioned
That formula describes this cantilever off a column using a plate welded to the side
| | Force | | | | | | | | | wwwwwwww V | w/-----------------------------------------| | w| | | w| | | w| | | w| | | w\_________________________________________| | wwwwwwww | | | | | | | | | | | |
(the "w's" = weld fillet)
There's a good number of you folks used to teaching.
Any good ideas for moving on from "on-paper in class" to developing a working feel for this stuff?
Left to my own devices, I'd raid the scrap pile, do welds which would by calculation will take a person's weight at about the 2/3rds point on a beam, then actually try what it takes to snap the welded connection - how far along the beam you really have to stand to break the weld.
I'd got in mind to start with 6013 stick so weld should fail first. Need to get to 7018 and MIG along the way, but know already with 7018 the steel section can end up twisted all over trying to snap a weld even after grinding away a lot of fillet to make it easier.
So what would you recommend for someone like me wants to be a respected engineer?
???
Richard Smith
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

I'd recommend a load cell so you can measure the applied force while you test welds to destruction. This is a cheap one: http://www.harborfreight.com/big-game-scale-65613.html
You could use a hydraulic press or porta-power set with a pressure gauge added. Truckers' ratchet load binders will apply considerable tension cheaply and controllably.
I found old analog tension and compression cells cheap in a second- hand tool store. 5000 Lbs has been plenty to test the stuff f I weld up at home. I needed a lot of practice and bend testing to learn to make a strong weld.
jsw
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Hi Jim
Thanks for jumping in on this one. It's not where you want to be plowing a lone furrow!
What did you find you learned from most? Where has doing this investigation and learning made you the person of the moment?
There is the "welding school" stuff which is the first great step-up - nick-break and bend-test. You've obviously done this too. Taking same-sized slices from your welds and getting a sense of the relative properties of different welds and goodness of your weld. My question is about the next next step, isn't it? How does your complete weldment behave...???
Always think how could have put post better after posting it. Trying again:
Full-pen. butt welds are complex and expensive to do but have simple properties - they seamlessly (!) match the plate properties.
Fillet welds - easy to do - difficult to specify. I've learned how to calculate fillet size knowing plate and weld properties and joint configuration - using moments of areas - especially second moment of area and polar moment.
So being able to weld - OK, I'll duck abuse on that point ;-) - how do you build a working map between calculations on paper and real welds taking that load?
Might be able to access a big tensile tester for the next few months.
There's fillets in pure shear, where the length of the fillet is along the direction of the stress
^ F | ___ / \eg. lifting lug | O | | | | | --------w| |w--------- w| |w w|___|w wwwwwww
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

(accidentally sent first reply incomplete - sorry)
Hi Jim
Thanks for jumping in on this one. It's not where you want to be plowing a lone furrow!
What did you find you learned from most? Where has doing this investigation and learning made you the person of the moment?
There is the "welding school" stuff which is the first great step-up - nick-break and bend-test. You've obviously done this too. Taking same-sized slices from your welds and getting a sense of the relative properties of different welds and goodness of your weld. My question is about the next next step, isn't it? How does your complete weldment behave...???
Always think how could have put post better after posting it. Trying again:
Full-pen. butt welds are complex and expensive to do but have simple properties - they seamlessly (!) match the plate properties.
Fillet welds - easy to do - difficult to specify. I've learned how to calculate fillet size knowing plate and weld properties and joint configuration - using moments of areas - especially second moment of area and polar moment.
So being able to weld - OK, I'll duck abuse on that point ;-) - how do you build a working map between calculations on paper and real welds taking that load?
Might be able to access a big tensile tester for the next few months.
There's fillets in pure shear, where the length of the fillet is along the direction of the stress
^ F | ___ / \eg. lifting lug | O | | | | | --------w| |w--------- w| |w w|___|w wwwwwww
There's fillets in pure transverse load
     |     ^ F |      |     | |      |     |      |w    w w w|      |-    - - -|      |      |      |     |      |      |      |_______|      wwwwwwwww      |      |      |      |      |     |      |     | |      |      v F |
              ____________________ F <----      w| ----------------------------------------- _______________________________|w     ----> F
And fillets in a rotational case like the cantilever welded to the side-face of a column where you need the polar moment of inertia - the one I first sketched
How am I doing in thinking this through?
Richard S
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

I'm a lab technician and prototype builder so there's always an engineer eager to do the heavy math as long as I point him in the right direction. They don't always recognize problems outside their specialty, while I have a fair to good background in chemistry, physics, electronics, optics and mechanical engineering and enough slide-rule-era training that I do mental ballpark double checks that the younger simulation-dependent ones don't or can't.
For my own projects http://picasaweb.google.com/KB1DAL/HomeMadeMachines#5107452241925535618 http://picasaweb.google.com/KB1DAL/HomeMadeMachines#5107453006429714322 predicting the worst-case load was much harder than designing for it.
For the bucket loader the worst case is driving into a rock. The center of mass is above the bucket pivot so the machine jumps upwards, or the bucket bends. I originally designed it with enough strength to lift the back wheels off the ground, then checked other conditions. Since it's a home project I just doubled the calculated weld, then added more wherever convenient such as a 'cosmetic' pass over the minimum fillet.
The bandsaw mill was designed for stiffness to minimize vibration. The welds aren't loaded to anywhere near their strength. The transmission should handle twice the horsepower but then might have a short fatigue life. I can accept design tradeoffs like that which require frequent inspections, maintenance and repair on a home project, not so much on a commercial one but there a design failure is Someone Else's Problem.
Good luck jsw
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Ah yes - I recognise now. Fond recollections of polymath lab technicians. Often very strong able folks who wanted a settled family life and I gotta say I reckon they'd made a good choice.
I was sent a bit flying because in the job I had in 2009 they had "Friday afternoon in the maintenance workshop" and I'd got projects on the go. Various folk you knew in their formal capacity who were there making bits for historic cars, etc, etc. Nice feel down in the West Country (UK). Could see a similarity to backwoods USA. All fond recollections. My "capstan" and "derrick" projects really bit the dust when I was laid-off. Ah well - take it all in your stride ehh... Onwards and upwards.
Pic there of a radial aero-engine. What's that story? I like to collect stories of "machines famous for the wrong reason". There was the P47 Thunderbolt with a radial engine. Presenting a lot of front-end-drag, they compensated with a lot of power (available from a radial), but that messed up the fuel consumption and range - so outclassed as a fighter it found a niche as a ground-attack plane, Because unlike a liquid-cooled V engine (steamlined shape) where one bullet could disable it, the radial engine could have an entire cylinder blown away and run long enough to get the pilot home.
I'll see if I can blag some tensile testing machine use without frightening the course organisers.
Regards
Rich Smith
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

That is the right inboard engine of the restored B-17 "909". I was taking pix when two of the crew came up with a step ladder and started removing cowl pieces. Seeing my chance I started passing parts and had become a de facto part of the team by the time the area was roped off. The others wandered away until only the crew chief and I were working on it, replacing a cracked cylinder jug which is off in the photo. Naturally he did the critical work like setting the valves while I turned the prop etc.
jsw
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Sounds like an intersting day out.
I've seen the B17 at Duxford, nr Cambridge, UK, doing its thing.
Radials - wondered what they sounded like - especially in the American single radial engined fighter. Got an answer on that one - loud! Used to sound of in-line V-engines - meaning essentially the RR-Merlin. Spitfires and P51 Mustangs have them. Do their aerobatics over Cambridge in summer. Radial - the onslaught on ears as swept by at a few metres altitude - then hit three times more as hanger doors had just the right inclination to bounce the sound back.
Anyway, anyone out there with their experience of going from weld-specification calculations to becoming a good engineer, would be glad of guidance.
Richard S
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

I found it essentially impossible to get a degree in electrical engineering without quitting to become a full-time day student. I went as far as I could in night school.
jsw
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Sun, 12 Dec 2010 09:05:18 +0000, Richard Smith wrote:
<clip>

You need to be a good engineer to get very far with weld specification calculations for other than a few simple cases where you can look up the equations in a handbook and solve with algebra only. Only a small percentage of engineers can do this well for a wide variety of types of structures. Most specialize in a particular area; a pressure vessel designer would not likely be good at airframe or bridge design (without additional learning) and vice versa.
Pick a specialty area you like with available jobs near your location and see if you can get a job working for an engineer as a technician or CAD operator while continuing to learn on your own or at school. You need calculus first for a good understanding of mechanics of materials needed for an ability to do stress and strain calculations well ... except that finite element analysis by computer is gradually replacing traditional analysis methods, and understanding how to do parametric design using finite element analysis is probably the most useful and marketable skill a structural engineer can have.
I have been studying engineering and working with other engineers for 40 years, and IMO this is the best way to learn engineering :-).
Glen
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Probably why it's part of the P.E. certification process, then... --Glenn Lyford
PS--I have nothing productive to add, but find the topic fascinating. Thanks, guys!
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here. All logos and trade names are the property of their respective owners.