Laboratory Testing

A two-layer solid lubricant system was used on the rods. This was applied in a plant environment with conditions kept as close as possible to those used for standard production. Fifty samples were produced with a two-layer coating applied: zinc phosphate carrier and calcium stearate lubricant coating. The rods are pickled to clean them and a zinc-phosphate layer is deposited followed by a calcium-stearate layer. Another 20 samples were produced as before but with an additional coating of soap lubricant. This final coat of powdered lubricant was added to minimize the damage of the sensor's drawing die on the production wire the same way as it is done in plant.

Four different kinds of defects common in production of fasteners were introduced into the coatings (Figure 2):

1. No coating, where heat shrink-wrap was applied to rods prior to all steps in the coating process. This corresponds to missing coatings from a preprocessing stage;

2. Zinc phosphate only, where heat shrink-wrap was applied after the zinc-phosphate application. This corresponds to the missing calcium-stearate layer coating from a preprocessing stage;

3. Hammer peening of the surface of the bar. This type of error simulates defects introduced during material handling from preprocessing area to the primary processing;

Figure 2. Sample rod for the laboratory trial

Figure 2. Sample rod for the laboratory trial

4. Scratching of the coating by its removal parallel to the bar. This was introduced by filing the coating off. This type of error simulates coils being dragged across the shop floor during transfer from preprocessing area to the primary processing.

All defects were approximately 50mm in length and applied to the circumference of the rod with defects being separated by variable fully-coated lengths of rod.

The experimental test rig was used to produce the rod samples drawn with a 0.35 mm reduction from the starting diameter of 0.50 mm. The samples were drawn with an area reduction of approximately 7%, and the loads on the drawing die were monitored by strain gauges onthe rod-clamping mechanism. All defects resulted in increased drawing loads. In the case of the hammer peening, this is likely to be due to the resulting irregularity of

Figure 3. Typical nonlubricated rod sample for the laboratory trial

Figure 3. Typical nonlubricated rod sample for the laboratory trial

Figure 4. Typical lubricated rod sample for the laboratory trial

the rod diameter. In the other three cases, reduced efficiency of the coatings is due to missing lubrication components. The defect with only zinc phosphate layer resulted in the highest friction. The zinc phosphate is a soft coating and thus is likely to produce a galling or sticking effect as the rod passes through the die.

The typical force signatures for the rods with two layers of lubricants (labelled as nonlubricated samples) and for the rods with an extra layer of soap lubricant applied (labelled as the lubricated samples) are shown in Figures 3 and 4.

As can be seen, Error 4 (scratching of the coating) is visually indistinguishable from the normal condition (the one without any errors) on nonlubricated data, and only Error 3 (peening of the surface of the bar) is readily distinguished from the normal condition on the lubricated data.

The force signatures from these 70 trials were collated to create two time series, one for nonlubricated samples, and another for lubricated samples with all five possible lubricant conditions (normal condition and the four defects) appropriately prelabelled with a corresponding condition label being manually applied to each time step of the drawing force signal.

0 0

Post a comment