Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

Review of testing a "robotic" welding device 3

Status
Not open for further replies.

MotoGP

Marine/Ocean
Jul 14, 2003
23
0
0
US
Greetings to the almighty Welding Engr's:

Feeling smug about my precious little knowledge of manual and semi-auto welding processes, I had the smile wiped-off of my face when I was instructed to attend a demonstration of a "newly-developed" automatic welding apparatus that will be demonstrated for the plasma-arc of Invar sheets that are used as the primary cargo containment on liquefied gas carriers.

The Korean shipyard that I am assigned to has dubbed this system as "The Spider". My limited native abilities are preventing me from seeing anything remotely similar between the machine and its organic namesake...can't seem to get anyone involved with the project to inform me as to how this name came about. Anyway, I digress...

I have tried, over the past week, to distill articles on similar welding apparatus that follow a guided path along a rigid track, have a "laser" sighting device, etc., and I just cannot seem to satisfy myself with my efforts.

I would quite genuinely appreciate any feedback that any of you can muster. My goal today is to be capable of asking intellegent questions that would specifically point at areas of concern/weakness/performance (and the like) when testing a "newly-designed" system such as this one.

As an aside, I have tried to gain access to the yard's welding lab wherein the "Spider" resides, only to be quickly escorted outside the room and being asked to wait until the machine's unveiling.

Any direct comments and/or referral to literature styled in the English language will also be to my direct benefit.

Thanks very much, in advance, for your time in anwering this enquiry.
 
Replies continue below

Recommended for you

MotoGP,
I am not a Welding Engr. but have worked with a number of systems and processes so will give some generic points to consider.
How does the system recognize the appropriate place to begin a weld? What degree of precision of fit-up is required for it to work properly? What kind of support does it require in terms of programming, maintenance, and consumables? How well does it track, i.e. keep the bead where it belongs? Does it require regular re-zeroing or just occasionally? Are there welds it cannot perform such as overhead or vertical up? How does it know when to use which type?
Hope this is what you need.
Let us hear what you see.

Griffy
 
Welding robots in general replicate the path they are taught including the angle of the gun and direction of travel weld schedules and many parameters. A welding robot can put the same weld in the same spot every time. The real problem is putting the weld joint in the same spot and with the same fitup everytime.

There have been many different methods to correct the path of the weld to the off position joint. They include location sensing using the welding wire as a sensor. Arc tracking when the weld current fluctuates making the robot adjust on the fly. Vision using a variety of methods.

Typical solutions include over-welding to compensate for postional errors, weaving to overlap out of position joints and various other techniques to correct the problem.

The problem comes back to having consistent parts. Even if the parts are wrong and they are consistently wrong then the welding will consistently weld the part. Typical problems in component parts are plate out of flat in a banana shape or twisted. Inconsistent bends in formed parts. Inconsistent part width or length. Varying size of chamfers. Varying hole size. Plate thickness variation can also be a contributing factor. A welding robot is a challenging process not only in the welding itself but in making the parts consistent.
 
Having been to HHI to wiew its automated pulsed plasma arc system, the following checks are recommended: mounting mechanisms; seam tracking capabilities & methodologies; carriage functions; carriage weight; manual overide systems; electrical/travel speed/torch angle parameter settings as the weld traverses the nodes; manpower requirements and time to reset track/welding system for a new length of weld; number and spacing of tack welds; the effect of high frequency starting with the GTAW system used for tacking on the automatic welding system (note that HHI pretacks a tank completely with GTAW using HF start prior to automatic welding; lineal feet per hour welding rate; overall cost of system; spare parts requirements; factory training costs; and warranties.

 
1. What kind of consumables are used, in house bought out.
2. MTB's, Mean Time Between Failure's..!
3. Working envolope, does it require extra axis.
4. Programming time, does it have an offline programming device, i.e ROBCAD or ACTWELD.

Spider : weld a spider can climb vertically, horizontally, overhead etc and can jump, maybe the machine does all this, you just want to hope it does jump whilst its welding..!

You better hope its not a black widow as well.
 
Status
Not open for further replies.
Back
Top