Twin Screw Citrosuco Report
March 15, 2001 ISSUE #101
Last season a series of citrus feedmill tests were run with the Twin Screw Press prototype. Both limed and unlimed peel were pressed. The report, updated with reference to more recent non-citrus testing, follows:
The test goal was to determine the operating characteristics of the Vincent twin screw design. This was needed in order to establish the design specifications and performance capacities of larger machines.
The performance of the prototype machine met our designers' highest expectations. The areas studied were:
Without any qualification, the way material feeds into the twin screw press is the best ever observed in any screw press. Feeding is normally not a problem with limed peel. However, a great deal of slippage occurs with materials like un-limed (fresh) peel. Normally Vincent de-rates press capacity by 70% with un-limed peel. When raw FMC peel straight from the peel bin was run, it was found that a de-rating of only 25% was necessary. This strong feeding characteristic has been confirmed on raw fish and spent brewers grain, both of which are also slippery materials.
(A consequence of this is that the press Supercharger, so many years in development, has been obsoleted.)
A goal was to measure the capacity of the twin screw press against a known machine. Since the test machine has twin 6" screws, it was compared it to the single screw Model VP-6. Vincent has almost 40 years of experience with the VP-6, and the VP-6 screw configuration was used in the twin screw prototype. It was found that, in seven tests with the Model TSP-6, the capacity averaged 254% of that of the single screw VP-6. (At half speed, 30 Hz, this was 174%.) This allows Vincent to guarantee that the throughput capacity of a twin screw press will be double that of a single screw press with the same screw diameter.
Press Cake Moisture
It was found that the twin screw press has excellent dewatering characteristics. In all moisture tests it was found that the twin screw press removed as much, or a little more, water than the other presses in the feedmill.
The press cake moisture data from four tests follow:
|Test #1||Test #3||Test #2A||Test #4|
|Twin Screw Press||66.5%||64.5%||64.9%||67.0%|
|Gulf Press #2||67.1%||68.5%||68.6%||67.1%|
(with cone withdrawn)
Final press cake moisture is determined by considerations beyond the screw press: the Brix and quantity of molasses added, the amount of waste water present, and the completeness of the lime reaction.
Based on June testing in Mexico on a special VP-22, it has been concluded that it will be best to have seven stages of compression in the press. (The test machine has five.) This will extend the slightly better 30 Hz performance to a 60 Hz machine. It also will give latitude for achieving maximum moisture removal over a wider range of operating conditions (wet peel, underlimed peel, old peel, a worn press, etc.).
It should be noted that the twin screw press is bound by the same laws of chemistry as other presses. A mechanical machine can remove only the free and interstitial water from vegetable material. To remove the hydrogen bound water and the chemically bound water it is necessary to apply heat. This is normally done with combustion energy in a dryer. It also can be done in a screw press by using the drive motor to cause friction heating of material being pressed. The Vincent Twin Screw Press stops short of dewatering by this inefficient use of electrical energy.
It was noted that the twin screw press does not draw as much power as was anticipated. The prototype drew under five horsepower in all citrus testing. This has held true for spent brewers grain, raw fish, and carrot pulp. Only with shrimp shells has there been a need for the full power of the 7-1/2 motor used on the test machine. The lower than expected horsepower requirement is attributed to the slicing action of the overlapping interrupted screw flights.
Susceptibility to Damage from Tramp Iron
During testing and operating four serious incidents of tramp material entering the prototype press have been recorded to date. The items found were a piece of a pump impeller, two valves (one brass, one steel), and a piece of screw conveyor flighting. These were large pieces of metal compared to the diameter and flight thickness of the screw.
The extent of screw and resistor bar damage that occurred was comparable to what is normally experienced in a single screw press, and the damage was very easily repaired in all four cases. It is notable that no appreciable damage to the profile bar screen occurred in any of the four cases.
However it was apparent that a large piece of tramp material will damage more parts inside the twin screw press because of the overlapping screws and the larger number of resistor teeth. For this reason special attention is being given to electrical and mechanical overload conditions.
During the testing observations were made of a number of other areas. Among these were vibration, rigidity, sufficiency of the screen open area, screen deflection and abrasive wear. The prototype design proved quite adequate in all of these.
Overall, Vincent is delighted with the Twin Screw Press. It marks a significant advance in screw press design because the performance is equal or better to anything achieved in the past. In financial terms, it is possible to produce a machine with double the capacity of a single screw press, but at less cost than two single screw presses.