It is difficult to know where to start to implement Kaizen in welding, as welding is a complex process. Moreover, in-process controls are required to ensure high quality in the weld. In other words, the finished weld might look good, showing no non-conformance, but could still contain a defect that can be revealed only through destructive testing or additional ultrasonic or radiographic testing.
To properly control welding, many in-process parameters must be verified. For example, in just semi-automatic welding processes (GMAW, FCAW and MCAW, for example), there are more than 27 parameters, including choice of welding process, choice of weld joint, plate thickness, type of base and filler materials, weld size, weld pass and layer sequence, wire diameter, wire-feed speed, welding speed, welding current and voltage, electrical polarity, shielding-gas type, gas-flow rate, welding technique and progression, preheat, post-heat, inter-pass temperature, weld backing, back gouging, electrode stick-out, deposition rate, heat input, and distortion.
To deal with this multitude of decisions, welders should have adequate training prior to performing production and, in most cases, should have access to a well-documented welding procedure that would help in choosing the correct parameters. For semiautomatic welding, what are the most important parameters, affecting quality and productivity? Kaizen should be applied to the following three parameters as a start:
To illustrate how Kaizen can be applied to welding, let us review the results of a case study. Thirteen experienced welders, working for a structural-steel bridge-manufacturing contractor, were asked to weld 12-inch-long, 5/16” fillet welds, using 1/16”-diameter FCAW wire in the 2F position. The welders were asked to execute the test plates at their own workstations, to reproduce the welding parameters used for production. These welders were experienced and were assumed to be efficient in their work. Table 1 shows the measured data and the results of the testing.
Table 1: Case history with Kaizen for FCAW structural 5/16” horizontal 2F fillet welds. After 12 months of monitoring, average welding speed increased to 14 ipm., wire-feed speed to 350 ipm, over-welding reduced to 15% and all welds passed. The welding technique was changed to a pull technique.
|
Team |
Sample number |
Required size (inches) |
Wire-feed speed (in/min) |
Technique Push/pull |
Welding speed (in/ min) |
% over welding |
Visual evaluation Acc./Rej. |
Penetration evaluation Acc./Rej. |
Global evaluation Acc./Rej. |
|
Night |
1 |
5/16 |
235 |
45º push 15º-30º |
11.8 |
-20 |
Acc. |
Acc. |
Rej. |
|
Night |
2 |
5/16 |
212 |
40º push 0º-5º |
7.9 |
10 |
Acc. |
Acc. |
Acc. |
|
Night |
3 |
5/16 |
236 |
40º push 0º-5º |
9.5 |
68 |
Acc. |
Rej. |
Rej. |
|
Night |
4 |
5/16 |
220 |
40º push 15º |
10.6 |
40 |
Acc. |
Rej. |
Rej. |
|
Night |
5 |
5/16 |
211 |
40º push 0º |
10.9 |
-10 |
Acc. |
Acc. |
Rej. |
|
Day |
6 |
5/16 |
222 |
40º push 0º-10º |
10.1 |
40 |
Acc. |
Rej. |
Rej. |
|
Day |
7 |
5/16 |
259 |
35º push 5º |
9.2 |
44 |
Acc. |
Rej. |
Rej. |
|
Day |
8 |
5/16 |
302 |
40º-45º push 5º |
9.6 |
68 |
Acc. |
Rej. |
Rej. |
|
Day |
9 |
5/16 |
219 |
45º push 15º |
7.6 |
20 |
Rej. |
Rej. |
Rej. |
|
Day |
10 |
5/16 |
241 |
45º push 0º-5º |
7.2 |
44 |
Acc. |
Rej. |
Rej. |
|
Day |
11 |
5/16 |
200 |
45º-50º push 10º |
10.0 |
-10 |
Acc. |
Acc. |
Rej. |
|
Day |
12 |
5/16 |
210 |
45º push 15º |
7.2 |
44 |
Acc. |
Rej. |
Rej. |
|
Day |
13 |
5/16 |
210 |
45º push 25º |
6.4 |
44 |
Acc. |
Rej. |
Rej. |
|
Average |
Average |
Average |
Accept |
Reject |
Reject |
||||
|
1/16"-diameter FCAW wire, 100% CO2 |
229 |
8.8 |
29.4% |
12 |
9 |
12 |
- Average wire-feed speeds were measured at 229 ipm. An average wire-feed speed of 350 ipm was attained after 12 months of monitoring.
- All 13 welders were using the wrong welding technique. Several years ago, the welders had been transferred from a GMAW welding facility (where a push technique is recommended) to a structural-steel FCAW facility (where a pull technique produces best results). The quality results were disastrous! Twelve out of 13 welders failed to produce welds to the required specification, and 9 of 13 failed to produce acceptable penetrations. Changing the technique and increasing the wire-feed speed solved the penetration problems.
- The observed travel speed was measured at 8.8 inches per minute. This value is almost half of the optimized travel speed to deposit 5/16” fillets using the flux-cored arc welding process. The average welding speed was increased to 14 ipm after 12 months of monitoring.
By Viwek Vaidya and Andy McCartney
Other articles you might like;