In February of 1995, a newly designed cast-link belt style furnace was placed in service. This furnace, in the following years, would prove to be the best of this type of equipment yet produced. Specifically, the belt is 54” wide and is supported by hearth rolls on 14” centers and return rolls on 24” centers. One of the unique features of this furnace is that the rolls, along with the drive drum, are synchronized to turn at the same velocity (i.e. feet/min.). This uniform velocity means that the belt floats over the rolls; thus, any stress on the pins and links is at an absolute minimum.
The result of this engineering is that by October of 1998, the furnace had produced in excess of 160 million pounds of heat treated parts, with a bare minimum of maintenance to the belt and rolls. By virtue of its current good condition, it is estimated that this belt will produce over 250 million pounds of parts before it is replaced. When it is removed from service, it will be because of alloy attrition; that is excessive secondary carbide precipitation causing the alloy to become brittle and begin breaking up – not for excessive stretching or loss of pitch.
The use of driven hearth and return rolls in this type of furnace has been around for years; certainly as far back as the 1970’s. Early attempts to employ this technology had generated poor results because the need to accurately synchronize the roll velocities was ignored or dismissed as not essential. In many cases, the roll velocities were intentionally mismatched to drive any belt slack to the charge end of the furnace where the counter-weight or air cylinder take-ups could accommodate such slack.
The result of such a velocity mismatch is that the friction between the belt and rolls, at operating temperature, is enough to cause the stress on the links and pins to exceed the limiting creep stress value of the alloy used for these parts and the belt stretches. Over a period of months, the accumulated stretch is enough that links must be removed from the belt to allow the take-ups to remain effective. The belt is put back in service and the process occurs all over again until the stretch is so bad that the pins cannot be removed; or the belt begins to rip apart. It should be understood that these belts fail not because of the work being processed on them, but because of the poor furnace design.
A second result of mismatched roll speed is the inability, in many cases, to control belt tracking. Excessive speed of return rolls in various furnaces has thwarted all attempts to keep the belt centered and, in several cases, has been severe enough to damage side link guard plates. Retarded speed of the hearth rolls forces the drive drum to pull the fully loaded belt over all of these rolls to make up the mismatched difference; thus adding to the amount of stretch on the belt links and pins.
Only when everything is accurately synchronized, and the belt is floating, can the full value of powered rolls be reached. This accuracy is critical to the point that outside diameters of the rolls should be machined so as to maintain a plus-or-minus ten thousandths of an inch total variation with each other (i.e. hearth roll to hearth roll and return roll to return roll). This is necessary to minimize any out-of-synch condition turn roll to return roll). This is necessary to minimize any out-of-synch condition because when you multiply the diameter by 3.14 (pi) to find the roll circumference, any error will also be multiplied by a factor of three. Thus, if tubes are used “as cast” with a tolerance of one-sixteenth then the roll circumferences through out the hearth could vary by three-sixteenths. Over a great number of furnace cycles, this is enough to substantially reduce belt life.
There are a number of 54-inch wide belt furnaces in operation, of which many have return rolls as part of the motive force in moving the belt through the heat treating cycle. Except for the furnace first described at the beginning of this writing and its sister unit (started in January 1996), none have close tolerance machined rolls that are fully synchronized. Typical production for these non-synchronized furnaces is 80 million to 90 million pounds of heat-treated parts and approximately 40 months of life before excessive stretch warrants replacement of the belt.
In contrast, the first synchronized belt had processed 120 million pounds of parts before the initial eight inches of belt had to be removed and the furnace returned to service (June of 1997). Since that time, the belt take-ups have not shown any appreciable movement; that is, no additional stretch has occurred. It should also be noted that with a slightly different drive system, the belt in the sister furnace has processed 120 million pounds to date (October 1998) with the take-ups having moved a total of one inch over the 2¾ years it has been in service.
The cost to build a new furnace of this type is increased only by the cost of the added hearth and return rolls, and by the additional expense of machining their outside diameters. This added cost will be offset within the first few years of operation by lower maintenance and increased belt life.
Many existing furnaces can be retro-fitted with synchronized rolls for the hearth and, in most cases, return rolls can be added in place of skid tiles or rails. Even if the roll center distance in the hearth cannot be changed and only a few return rolls can be added to the return portion of the furnace, the reduced stress that the belt will be subjected to as discussed above more than warrants the expense to accomplish the re-work.
Since the first two furnaces were place in service, others have been built using the same technology and are as productive and maintenance-free as was anticipated at their inception. One older furnace was re-worked; adding return rolls in place of the original skid tiles, and all of the roll speeds were synchronized. This project has also matched its goal. If any older furnace is to be over-hauled (i.e. re-brick, new burners, replace many rolls, etc.) it is well worth the time to investigate if a synchronized system can be part of that re-work. Generally, the added cost to achieve this advance is small in terms of lower maintenance expense and increased belt life.
If there are specific questions about this technology, please contact Robert Cutshall at Omega Castings, Inc. If he cannot answer your questions, he should be able to suggest other individuals who can.