....what is it called technically....

Without images or additional context (eg, length of excessive welds relative to the interface between plates 1 and 2), could be integrated run-in/run-out tabs required for the robotic welding process. Uncertain. Or material wa$te, but suspect not. Or yet, a case where the robot was set to run a set length for many parts and then the same length was used to weld the smaller length part. Not optimized??

Where did you observe these welds and how may other locations had "welds without overrun"?

These laser and spot welding bots may know the answer, as the spot'rs will lay down about 5k separate welds on a standard automobile.

https://www.youtube.com/watch?v=muLmD_7DANw

 

Hello ManoKai,

Well I have observed them on the chassis of Kawasaki Motorcycles.

On asking around I happened to hear that it has something to do with reduction of stresses at the weld joints.

But I did'nt understand why and hence I am posting this doubt.


Basically there are gussets welded on the chassis and the welding there exceeds the gusset length by more than 30mm at places the gussets varying in lengths all over the chassis.

 

It is used alot of times to ensure proper penetration in the whole joint. Doing this allows the puddle to be hot when you get to the joint as typically there will be less penetration if you have to start at the joint. It will also reduce notch effect in the corners of the gussets.

 

Its used a lot on heavy equipment. It seems to have been started by Caterpillar, and are often referred to as "Cat Tails". I think it came from failure analysis of failed welded joints. Cracks often start along side a weld, usually either from stress concentrations or from weld shrinkage stresses. By overrunning the weld, the ends of the weld are completely on parent metal, away from the physical edged of either part.

I sometimes see them carried to extremes by designers or programmers with the Cat Tail on either end being longer than the weld itself.

I have had colleagues who attended a Lincoln weldment design seminar where the speaker felt they were unnecessary. They were occasionally of benefit to poorly designed or welded joints, but were of very questionable value in properly planned and executed joints.

 

I've seen John Deere do it on their ag equipment also

 

Contrary to the lincoln weldment design seminar. The cat tail design proves to be effective in lots of applications.
It spreads stress from vibrations away from the one corner where the piece end meets the part you welding it on to the wider parent metal area. Simple but very effective.

 

Thanks iongar!

Thanks bent!
Thanks Daniel!

 

Do you guys have any ideas on research material on this matter?
I hope to get to the bottom of this problem!

 

I used to sub for Finning "" Caterpillar "" , Brandt Tractor "" John Deere Equipment "" etc... All the upgrade and Warranty repair came with blueprint and welding procedures and run in and run off was always needed for the type of applications similar to your bike. Spreads the stress.
They have done the research.

 

I've not worked in the automotive industry, nor have I been privy to the engineering designs of various chassis components, but what you're describing is relatively common.

The most common points for a weld to fail are at the start, and at the stop. By initiating the weld ahead of the section that needs to be welded, and carrying it beyond before you terminate it, you move those weak points away from the areas of concentrated stress, thereby reducing the chance of weld failure.

 

@ kartik - the cat tail concept that 'bent' is referred is illustrative in post #5 of this related thread..

My initial assessment of "...integrated run-in/run-out tabs" was inferring the cat tail feature as a method to ensure proper preheat/fusion at the weld interface and to reduce stress concentrations. Wasn't clear in my writing. However, only personally observed these overruns on some square/rectangular elements of dirt/street bike frames, never on round/oval cross-section frames.

Clip some pictures of the weldments in question for context.

 

As a computer professional, I'd posit that it is from a failure to optimize the programming the of the robot welders.

The parts may not have to be aligned perfectly before welding, so the program was set up to cover all possible valid setups. Nobody bothered to check that they are wasting $0.002 per weld in filler and electricity, yet.

 

Hello ManoKai,

thanks for all your valuable inputs!



I have taken some pictures of the over runs observed by me on the bike chassis. Attaching them below. Let me know what you think?

 

Hello ManoKai, thanks for all your valuable inputs! I have taken some pictures of the over runs observed by me on the bike chassis. Attaching them below. Let me know what you think?

Quickly reviewed the weld lines you posted. The line appear human-produce v. robotic-laid. Stops/Starts are not consistent. If the welds were indeed robotically produced, most likely a case of a non-optimized software solution. What's the model/yr of the imaged motorcycle frame?

Kawasaki, in the creational pursuit of their $50k Ninja H2R, uses a RA1ON robotic welder. Believe Kawasaki Robitics build the rig, but uncertain. Will have to engage with the Kaw crew at FABTECH this year and find out the details.

Video of Kaw's RA1ON robotic welder in action on the Kaw Ninja H2R's trellis frame. Note, no discernible "overruns".

https://www.youtube.com/watch?v=c_kTwiUsyp4

 

I'd say that the crew that set up the robotic welder just didn't bother with "STOP WELDING" commands during the process.

Very sloppy and wasteful. No human would ever do that kind of work!

 

I did not read the thread as I don't have much interest but to me these welds are being done as if they have integrated start stop tabs. I would look at it like an insurance policy of sorts for f'd up weld operations. This robot welder crap that seems to more common than not darn near makes it foolish not to do it. Yeah codes and whatever might call for strips but anyone that actually has the nerve to identify them self welder does not need start stop strips.

 

^^ Interesting. ^^

4-axis robot welders plus the ability to rotate the part being welded.

Still doesn't excuse the programmers from stopping the weld process between nearby welds. You can see how the weld bead is sub-par between critical welds in the first 2 pics as the frame was rotated while the torch was repositioned for the next weld.

 

Hello ManoKai!

The frame belongs to a Kawasaki Z250.
I procured this in Mid 2016 so I assume the frame should have been manufactured somewhere around late 2015 - early 2016