Fanuc / Power Wave Welds.

Date 23. July 2004.This report is a condensed review (changed with a little humour added) of a (Ford Chicago) MIG robot weld line used to weld the Ford 500 steel cradles and sub assemblies. I evaluated this new robot line to establish the root cause weld issues that were impacting the cradle / frame robot weld quality and production.

The Chicago location purchased approx 100 Fanuc robots with Lincoln Power Wave weld equipment. The weld wire was E70S-6, 0.045 (1.2 mm). The Fanuc robots at this plant do not have automated TCP controls or weld joint tracking equipment. which reduced the hourly production by appprox. 30%. The weld wire size selected was too large and not appropriate, for the gage parts and weld rework was > 40% and the pulsed MIG power source used is a poor / costly choice which has added > $300,0000 of unnecessary cost to the cradle robot line.

In evaluating specific robot welding issues at this facility, it's important that the unique robot weld risks be understood for the welded parts under discussion, its something manufacturing management and engineers should have an interest in yet few do.

With robot MIG welds, the primary concerns with common parts welding 2 to 4 mm is the attainment of "acceptable and consistent weld fusion", while the concerns for welding parts less than < 2 mm, is typically to avoid weld burn through.

Note: On steel parts > 2 mm, one in three or four welds in the auto industry end up with lack of weld fusion in some part of the weld.

The Ford Chicago plant provides many welds on steel parts < 2 mm. On these parts the MIG weld burn through potential is at a high level and the part dimensional issues plus robot TCP deviations were compounding the the burn through issues. As this Ford plant ramps up for it's robot weld production, it's unlikely that the part dimensions will improve (most welded part dimension issues increase with amount of stamping provided) and this combined with the lack of the automated TCP controls from the robots, will ensure that the weld issues, unless addressed will increase in proportion to the weld production attained.
To minimize the potential robot weld issues that will occur, the Ford plant management needs more focus on the implementation of best robot weld practices and to achieve this the management and engineers in this plant should be the first in line for the process control training required.

The selection of inappropriate MIG weld consumables and inferior weld process modes has for two decades created major weld issues in the Big Three plants. As the robot steel parts decrease in thickness < 2 mm, the weld quality / productivity consequences of poor weld consumable / and weld transfer mode selection really becomes evident. In your plants weld consumable selection often has more to do with a purchasing decision cost reduction decision rather than a logical weld engineering decisions, just as robot and weld equipment selection for this plant has had more to do with salesmanship and costs than it did with engineering logic.

THE IMPORTANCE OF ROBOT TCP CONTROLS . The following weld data will assist the plant in establishing best weld practices for the cradle parts < 1.8 mm.

30 MINUTES PER-HR PRODUCTION LOSS FROM NOT ORDERING THE CORRECT EQUIPMENT:

Part dimension and TCP control will always be critical when welding gage applications. The thinner the gage, the smaller the required welds and the greater the degree of weld / joint accuracy. TCP control also enables an altered program to be put back to the original program with minimal deviation of the gun to the work. In this facility, in the robot cells purchased, there are no means for accurately checking the robot TCP and you will find programmers constantly making changes to the welds, none of which have anything in common with the original program.

This plant was typically loosing 20 minutes of robot (down time / lost production from 33 robots 33 x 20 660 min/ hr from 1980 min/production/hr) due to program weld point adjustments and related contact tip issues.

Also on the parts with gaps or inconsistent dimensions, you typically have to be able to place "oversized welds" to compensate and of course over sized welds will increase the weld burn through potential, especially on the 1.5 - 2 mm parts. To avoid weld burn through with the larger welds, aggressive weld weaves are beneficial, the plant programmers made no use of the weld weaves in the robot programs. Unfortunately with weld burn through, the resulting holes were disrupting the arc causing the weld wire to end up with excess length which is a root cause for arc strike / contact tip issues.

THE LINCOLN PULSED POWER SOURCE SELECTION.

For almost two decades, the MIG equipment manufactures have been developing and promoting pulsed MIG equipment for "steel applications".

The Lincoln pulsed equipment purchased for your robot weld lines has undergone years of development yet still in 2004 provides "poor arc stability" along with arc length sensitivity. The pulsed arc length sensitivity makes the pulsed mode unsuitable for most of your high weld speed steel applications. It's ironic that the weld equipment purchased also has cost 50% more than the available superior, stable, traditional CV equipment that could have been used. The pulsed weld equipment purchased will also likely cost three times as much to repair and you will need to keep more spare equipment than normally would have been required with traditional, more durable CV equipment.

PULSED MIG AND HIGH SPEED WELD CONCERNS: When using an 0.045 (1.2mm) MIG wire and pulsed welding the cross members and similar seam lap welds, to attain the desired minimum weld travel rate of 50 ipm, on the 1.6 mm lap welds, the MIG weld wire has to virtually make contact with the weld surface. This "wire to work contact" not only causes extensive weld spatter which will impact the tip life. The weld spatter also gets on to the fixtures causing part assembly / fit issues. When establishing the pulsed trim voltage (arc length) with the high speed pulsed applications, if the pulsed weld voltage (arc length) is set to a none weld spatter condition, the weld transfer and instability at 50 ipm will cause "skip" welds (missed welds / weld blobs).

Typically with this present day power source technology, when pulsed welding we need a sufficient arc length to enable the pulsed MIG weld drops to form and transfer without making a short circuit contact with the work and wire tip. With the Lincoln pulsed MIG equipment the desired minimum arc length required for optimum pulsed weld transfers is detrimental when pulsed welding thin gage parts using 0.045 wire at the "high weld speeds".

The thin gage parts limit the allowed pulsed wire feed rate which limits the pulsed frequency utilized. The low pulsed frequency with the large wire diameters and a weld which for 50% of its time is at a background current of less than 100 amps resulted in an unstable weld transfer unsuited to high weld speeds on the Ford 500 cradle welds. The robot, high speed skip weld issues you have at the plant are the same pulsed weld issues every wheel manufacturer and torque converter manufacture has had to deal with through for the last two decade. Another ironic fact is when you use the Lincoln pulsed equipment in the traditional spray transfer modes which provides improved arc stability is the weld decision maker may not be aware that the slope influenced performance of the traditional spray mode from the pulsed MIG equipment is typically inferior to the lower cost traditional CV equipment.

Note to managers: It's a logical practice to check out weld equipment before you purchase it.

To attain high robot weld speeds with weld transfer stability, I recommend the pulsed mode should be replaced with the more stable, less arc length sensitive "spray transfer mode". While the spray mode runs hotter than the pulsed this is not a concern for the welds on the parts > 2mm as the the weld heat can be reduced through the faster weld speeds and the selection of a smaller MIG wire diameter such as an 0.040 or 0.035 wire.

Note: Remember this is 2004 pulsed equipment and although I saw the pulsed issues on numerous occasions and documented them at this site, poor performing pulsed equipment was something companies like MIller. ESAB and Lincoln did not discuss in public. There is not be the same electronic and weld transfer concerns with most of the pulsed equipment sold > 2006.

July 30-04. E Mail to Ed,

Ed: Your description of arc sensitivity with high speed pulsed welds is exactly what I am experiencing with our robots and the new Miller Accu-Pulse process / Auto Axcess. Our new auto bumpers are thin gage, 1/16 1.6mm HSLA and martensite concerns. I tried to weld above 40 IPM with the Accu-Pulse and could not do to weld skipping and arc instability. We went to an 0.035 1 mm wire and could not get the travel speeds. We changed to 0.045 (1.2mm) and had to run the pulsed arc with the arc length buried in the part, this resulted in extensive weld spatter also the part could not handle the pulsed weld heat and we would have holes all over the place. With the disappointing pulsed weld results we now use high end short circuit CV with the 045 wires and are attaining 40 IPM travel rates. I have no spatter on the part and have no arc stability problems with the short circuit. I hate to admit it but this is is another pulsed failure in my book. I could get these short circuit results CV power source for half the costs..
Regards G S.

BEST WELDING PRACTICES. WIRE DIAMETER SELECTION. As I write this report one of your competitors is utilizing 0.052 1.4 mm weld wires on Ford frames 2 to 4 mm. This company is also using the same pulsed Lincoln equipment as you. They also went through the consequences of the pulsed arc instability, and they were unable to use spray with the large weld wires. The programmers at the plant unknowingly dialled in lower "globular parameters" which are causing all types of production and quality problems, (see frame plant report at this site).The globular transfer caused over 80% weld rework.

The choice of an 0.045 wire for short circuit and spray welds that will be made on parts less than 2 mm is also an inappropriate choice.

The Chicago facility and the cradle welds will benefit from stable spray transfer and a smaller wire diameter than 0.045 (1.2mm) Lets take a moment to look at the MIG weld current compatibility with standard gage thickness used at the cradle facility.

[] With robot, single pass "butt" welds on 1.6 mm cradle parts, we typically would be in the short circuit mode, welding at 170 - 200 amps. Lap welds do allow higher weld current, however the butt weld data shows the part compatibility with the weld current utilized.

[] With the cradle 1.8 mm parts, the butt weld current would be increased to approx. 220 amps. For any application that uses a weld current over 200 amps we would want to use the high deposition, stable spray transfer mode. Butt welds on 2 mm parts could be robot welded using 240 - 260 amps.

An 0.045 wire requires a minimum of 250 - 260 amps to attain stable spray transfer. As many of the welds you produce require less than 260 amps it makes no sense to use a consumable that will simply cause weld burn through on many of your parts. The bottom line it's imperative the plant uses an 0.035 or 0.040 wire. The common 0.035 wire will go into spray around 200 amps.

BEST PRACTICES AND WELD WIRE COSTS.

I can understand why third world countries get uptight about weld consumable costs, however in many North American auto plants weld engineers should not have to worry about the $1 a pound MIG wire costs. As bigger wires cost less, in many auto plants purchasing managers rather than engineers may decide on the wire diameter selected. A typical auto cradle may today use approx. 1 to 2 lbs of weld wire per cradle. So the purchasing manager can save 20 to 30 cents a cradle by recommending an 0.045 or 0.052 wire instead of the 0.040 or 0.035 wires. In this case changing the weld wire to a smaller wire will dramatically improve both the weld productivity and quality, but hey, why worry about a daily, lousy robot production of 60% with > 40 % daily weld rework, when the purchasing manager can save 20 cents a part.

YOU ARE NOT LIKELY TO USE WELD WEAVES IF YOU DON'T UNDERSTAND THE ROBOT MIG WELD BENEFITS ATTAINED FROM WELD WEAVES:

THIN PARTS AND WELD WEAVES. The plant was not using weaves in any of the robot cells. The plant should consider when required, the use of weld weaves for specific problem welds. The weld weaves should comprise of amplitude that creates a narrow, high speed oscillation in the weld centre. This weld weave oscillation will not impact the potential weld speeds, it will however cause the fast freeze thin welds to thin out and provide slightly wider weld coverage.

Three important weld benefits are attained from the weld weaves on gage parts;

[1] Reduce weld burn through potential.
[2] Helps compensate for gaps.
[3] Helps compensate for part dimensional deviations.

If you are reading this today a few years will have passed since it was written. Do you see any similar issues in your plants? Hopefully this tongue in cheek weld article will assist some UN-blinkered auto / truck manufacturers to give a little more consideration to weld process and equipment requirements. If you want to spend less time trying to put out the weld shop fires that daily spread through too many weld shops, managers, engineers and technicians and any weld decision maker involved, all need to walk the same path to weld process optimisation, in other words for god's sake get some robot weld process control training.

 

SOME REASONS FOR THE COMMON LACK
OF BEST WELD PRACTICES AND PROCESS CONTROLS.


How unqualified individuals and
inflated egos influence welds:

KEEP SMILING, AFTER ALL IT WAS YOUR
DECISION TO BE IN THE WELDING INDUSTRY.

For the world's best self teaching and welding training
resource on Robot Process Control for welding thin parts.

Robot Mig Welding CD

Back to Ed Craigs MIG Welding Projects