Wire Cross-Section Measurements
Both the round and flattened spiral loops were made from 10-gauge wire having a diameter of 0.135 in. A sample of some flattened wire was measured using an optical-comparator equipped with a computer data acquisition system. This data was used in developing a representative flattened cross section (Fig. 4) to be used in the finite element analyses. The shape was scaled so that it would have the same cross-sectional area as a 10-gauge wire. Table I gives the cross-sectional area properties that were used for the beam-element analysis. The thickness of the belt was assumed to be 0.500 in. for both the round and flattened spirals. This is the nominal thickness for this type of belt presently being produced.
Both the round and flattened spiral loops were made from 10-gauge wire having a diameter of 0.135 in. A sample of some flattened wire was measured using an optical-comparator equipped with a computer data acquisition system. This data was used in developing a representative flattened cross section (Fig. 4) to be used in the finite element analyses. The shape was scaled so that it would have the same cross-sectional area as a 10-gauge wire. Table I gives the cross-sectional area properties that were used for the beam-element analysis. The thickness of the belt was assumed to be 0.500 in. for both the round and flattened spirals. This is the nominal thickness for this type of belt presently being produced.
Flattened Wire Twist Measurements
In addition, the weaving process causes the flattened spiral to have some degree of twist. The integrity of the interface between the flattened wire spiral and the connecting rod is dependent on this angle of twist. A B-36-10-8-10 belt was used for both the round wire spiral and flattened wire spiral measurements. The actual belt used for the flattened wire twist measurements was slightly different from the modeled belt in that it had flattened seats formed into the connecting wire. Figs. 5a and 5b show the spiral/cross rod configurations in the belt specimens. The measurements acquired from the optical comparator indicated a large variance in twist angle for the flattened wire specimen with a maximum twist angle of 9.75 degrees and a maximum gap of 0.0301 inches between the spiral and connecting rod. The round wire spiral/round connecting rod configuration produces a robust interface that is not sensitive to twist. This specimen had a very uniform appearance with a maximum gap measuring about 0.0206 inches (it should be noted that there were very few gaps in the round wire assembly). The large amount of twist variance for the flattened wire makes it difficult to model this situation. A statistical analysis could be performed to determine the average angle of twist. The finite element analyses were conducted using a 9.75 degree twist angle.
In addition, the weaving process causes the flattened spiral to have some degree of twist. The integrity of the interface between the flattened wire spiral and the connecting rod is dependent on this angle of twist. A B-36-10-8-10 belt was used for both the round wire spiral and flattened wire spiral measurements. The actual belt used for the flattened wire twist measurements was slightly different from the modeled belt in that it had flattened seats formed into the connecting wire. Figs. 5a and 5b show the spiral/cross rod configurations in the belt specimens. The measurements acquired from the optical comparator indicated a large variance in twist angle for the flattened wire specimen with a maximum twist angle of 9.75 degrees and a maximum gap of 0.0301 inches between the spiral and connecting rod. The round wire spiral/round connecting rod configuration produces a robust interface that is not sensitive to twist. This specimen had a very uniform appearance with a maximum gap measuring about 0.0206 inches (it should be noted that there were very few gaps in the round wire assembly). The large amount of twist variance for the flattened wire makes it difficult to model this situation. A statistical analysis could be performed to determine the average angle of twist. The finite element analyses were conducted using a 9.75 degree twist angle.
 |
 |
 |
||
| Fig. 4 Spiral cross section. | Fig. 5 Section belt with (a) round wire spirals and (b) flattened wire spirals. |
Fig. 6 The loaded belt spiral. |
A single spiral loop being loaded by a connecting rod was modeled using ANSYS finite element software. The results were used to determine the relative behavior between the round and flattened spiral wire belts. A step-by-step approach was taken in this analysis, starting with simpler three dimensional (3-D) beam element models with linear material properties to obtain some baseline information and working up to 3-D solid models with nonlinear material properties and contact elements.
Solid models were made first using solid elements with pressure loading and linear material properties, and then were made using contact elements with nonlinear material properties. With each step, the results were compared with the previous step.
A static analysis was used because the conveyor for sintering processes runs at a constant velocity. Also, it was assumed that the loading and unloading of parts did not create any dynamic effects on the tension of the belt. Because the wire mesh furnace belt is constructed from a series of elongated coiled spirals tied together by cross rods, only one loop of the spiral was considered in the model. Also, only one-quarter of the spiral needs to be modeled due to symmetry.
The belt tension in a sintering furnace varies between the entrance and the exit and can vary from furnace-to-furnace. It can be nearly zero pounds at the entrance up to a design load of 240 pounds per linear foot of belt width. This analysis used a design value of 240 pounds per linear foot of belt width divided among the 36 spirals in a foot of belt width resulting in 6.67 pounds of tensile force on each spiral loop (Fig.6).