The Influence of Process- and Equipment-Induced Variability on Product Quality
Like the words to your favorite song playing over and over in your mind, the heat treater’s repeating lyric is producing the best quality parts by controlling both equipment- and process-induced variability. Once this lesson is learned, good things happen. Let’s learn more.
A problem that recently crossed the Doctor’s desk serves to illustrate this concept. A cast link conveyor-belt furnace running various-size bolts from multiple materials (Table 1) was shut down due to a serious quality issue: soft parts across the product line. Over time, a series of minor problems had been reported. These related to the operation of the equipment, none of which had been addressed due to the demands of production.
- Belt overloading – questionable accuracy and functionality of the weigh scale
- Sensor positioning – failure to halt loading when overloading or belt stoppage occurred
- Loader functionality – failure to consistently position parts uniformly
- Gearbox drive failures – shear-pin breakage
- Belt tracking – drive roller adjustment
- Part hang-up during transfer to quench
- Quench slot and quench chute alloy warpage and refractory wear
- Controls issues – erratic temperature
- Noise and vibration in the area of the fans
- Operational practices – intermittent versus continuous use
- Maintenance and training of operators, supervisors
When the equipment ran in the past, it produced consistently good work. The last complete reline of the furnace had occurred over 10 years ago. The furnace cast link belt – a major cost item – was approximately three years old and had two length reductions. Past experience suggests that a third belt reduction is possible prior to belt replacement. This translates to approximately one year of remaining service life left for the belt.
A thorough inspection of the entire furnace system was conducted. The overall condition of the furnace was rated as “good,” but a number of problems were revealed.
Weigh Scale, Vibratory Loader and Pan Dump System – Overall rating: “F”
- The scale failed to “zero.”
- The scale was not weighing part dumps consistently.
- The scale (overall weight) rating was lower than the actual weight being dumped per charge.
Refractory Insulation – Overall rating: “C”
- The majority of the brickwork inside the furnace was stable and undamaged.
- No significant cracking of the refractory walls or arch were evident (in undamaged areas).
Areas of missing refractory were found.
a. One course of arch brick spanning a distance of approximately 66% of the arch width was missing in the immediate vicinity of circulating fan #2.
b. The missing refractory separated at the mortar joint.
c. Fan #2 was recently replaced reportedly due to “noise” and “shaking.” Excessive fan vibration was found to be the root cause of this problem.
- The area at the rear of the furnace (near the quench chute) had damaged and missing refractory on both sidewalls.
Areas of damaged refractory were found.
a. Damage to the refractory was extensive in the areas around the radiant tubes. Tube installation techniques were suspected. Training of maintenance personnel on the techniques involved is needed.
b. The tube support ends had an extensive amount of refractory missing.
c. The burner end had fiber blanket “plugs,” some of which needed repair.
d. The areas observed will experience excessive heat loss and temperature differentials. However, the fact that these are located near a heat source will tend to mask these effects.
- The area at the rear of the furnace underneath the cast link belt showed evidence of refractory movement (i.e., bulging) outward.
- Some damage was evident at the interface between the sidewalls and the arch. Bulging was also observed.
- The arch area at the rear of the furnace had a number of insulating bricks that had moved (i.e., bulged) outward.
- A section of brick from the furnace sidewall was examined for carbon penetration. Nothing abnormal was observed for a furnace performing the carburizing process.
- Prior investigations of the furnace interior revealed parts getting stuck in the area of the refractory slot and alloy quench chute.
- A large number of parts had fallen from the belt and were observed throughout the length of the furnace under the belt. These must be removed prior to returning the furnace to service.
Cast link belt – Overall rating: “A”
a. Specific observations:
(i) No evidence of abnormal wear, distortion or other damage.
(ii) Belt edges (sides) were in excellent shape.
Radiant tubes – Overall rating: “A”
a. Minor evidence of warpage and sagging.
Thermocouple protection tubes – Overall rating: “B+”
(i) Slight amount of sagging observed on thermocouple tubes.
Fans – Overall rating: “C-“
a. Fan #1 showed evidence of metal dusting (i.e., reduction in alloy thickness) in the blade area. Replacement was recommended.
b. Fan #2 showed no evidence of damage to either the fan blades or fan shaft.
c. Fan #3 had several holes in the fan blades, which appear to be casting defects. Replacement was recommended.
Drive and idler rollers – Overall rating: none
a. The front roller could not be inspected.
b. No significant warpage or damage of the rear roll was observed (limited inspection was possible).
Quench chute – Overall rating: “B”
a. Some warpage of the chute was evident, but replacement was not warranted.
- O2 probe protection sleeve – Overall rating: “A”
Special Note: Failure of the plant air system was identified as the root cause of damage to the quench chute area (which was air cooled). An alternative method of cooling (nitrogen, water) was recommended if this issue continued to be a problem.
The soft-product issue was a direct result of the overloading of the cast link belt. The root cause was the cumulative effect of all the minor contributory factors, however, which introduced equipment-induced variability into the process. This resulted in improper heat treatment of the product being run. The parts were being overloaded and either were not reaching the proper temperature, were not at temperature for an adequate amount of time prior to quenching or were hanging up and losing temperature in the quench chute area. Together, these factors overwhelmed the process and combined to produce unacceptable metallurgical and mechanical properties in the final product. Once corrected, acceptable product (Fig. 1) was easily obtained.
This is a classic example where a series of minor equipment and ancillary-item problems resulted in a major quality issue. The old adage that there is never enough time to do it right but always enough time to do it over again is unacceptable. Take the time to correct problems as they arise, even if they interrupt production. This will help avoid major downtime and expense in the future.
- Herring. Daniel H., “The Pitfalls of Process and Equipment Induced Variability,” white paper, The HERRING GROUP, Inc., 2016
- Herring, Daniel H., Atmosphere Heat Treatment, Volumes I & II, BNP Media, 2014 and 2015
- Herring, Daniel H., Fastener Heat Treatment, in preparation (2018)