- THE MAGAZINE
- FOOD MASTER
Rapid movement can dramatically reduce cycle times in retorts, but controlling the force created by a heavy basket’s motion required years of additional development.
Accelerated heat transfer has driven many improvements in the conventional retort, or autoclave, over the last century. Container walls are the transfer point, and innovations like the end-over-end retort were developed to stir and rotate the food inside the container to promote even heating and reduce the thermal input needed for sterilization.
Richard Walden of Zinetec Ltd. in the UK spent a decade developing a lab-scale autoclave that relied on rapid back-and-forth reciprocal motion to speed up heat transfer (see “The power of reciprocal motion,” Food Engineering, June 2006). He named his system Shaka and licensed the technology to three OEMs, including Allpax Products, which holds North American rights to the technology. But many challenges remained before a commercial-scale unit could be built. First and foremost was controlling loads weighing up to a ton and rapidly moving back and forth at up to 2 g-force.
One of the three Shaka licensees went out of business before overcoming the challenge, but Allpax, with the support of Pro Mach, its corporate parent, successfully engineered pilot units and high-throughput commercial machines. The first 1,400mm (55-in. wide) retorts went into service in 2011, with three 1,600mm (63-in.) units moving into the field this year.
Overseeing Allpax’s work was Greg Jacob, vice president and general manager. A graduate of the University of New Orleans with a bachelor’s degree in history and political science, Jacob has been affiliated with Allpax since 1985, first as an instrument technician and co-owner of Warriner Controls, a former Allpax subcontractor, and, since 1995, as a shop supervisor and project manager. He also has worked as a welder on Allpax’s shop floor.
FE: When did your involvement with the Shaka system begin?
Jacob: We licensed the technology in 2006. We started by building a 24-in. wide R&D unit. There was a lot of buzz about the technology, and we needed to see if it justified developing a commercial machine.
FE: The R&D machine, which was displayed at a Chicago trade show in 2006, employed springs to control the force of reciprocal motion. Are springs still part of the design?
Jacob: Springs were used to produce side-to-side motion of the basket. We soon abandoned that approach. The size of the springs that would have been required for a commercial retort and questions about their durability after thousands of compressions forced us to explore other solutions.
We experimented with variations in the length of the back-and-forth stroke in the one- to 12-in. range and in the speed, going as fast as 8.3 double strokes per second. Zinetec’s design called for a 6-in. stroke, and we validated that. Three back-and-forth strokes a second now is typical, though the operating speed is specific to the product and is usually in the 140- to180-strokes a minute range.
FE: How many baskets does the retort carry?
Jacob: One. We played with side-to-side shaking with multiple baskets for some time, but the stress on the vessel was significant, and securing the baskets so they didn’t collide with each other, even at 0.5 g-force, was a challenge. We could only imagine what would have happened at 2 g. If something broke, it could be a catastrophic failure. But because processing time is significantly shorter, the throughput with only one basket exceeds that of traditional static retorts with six baskets or rotary machines with four.
FE: Were there other variations on the original design during the development phase?
Jacob: Up-and-down shaking was another approach we tried. That would accommodate loading containers in a vertical orientation, as is done in a conventional retort, but we could not get the air bubble created by the container’s headspace to move all the way to the bottom and create the stirring effect that is critical for rapid heat transfer. The bubble would only descend about three-quarters of the way, leaving an obvious cold spot below. With back-and-forth motion, you get constant stirring, and in most cases, the cold spot is eliminated.
FE: Is air bubble movement similar to the agitation of a rotary retort’s end-over-end movement?
Jacob: Both rely on headspace for agitation, but the activity is much more dramatic with Shaka. There is faster and more frequent displacement from the container’s core to the outside wall. There’s much more control of the air bubble. With end-over-end, you’re limited in finding the rotational-speed sweet spot for a given product, and at some point, as speed increases, the air bubble can stop moving altogether.
Headspace is a critical factor in all agitated retort containers, but maintaining a tight tolerance range is vital with reciprocal motion. The precision of a filler used with existing rotary retorts may be adequate, but an older filler with a static system may not be suitable. X-ray technology is commonly used to monitor headspace.
FE: How does the system manage the stress of a 0.5- to1-ton load rapidly moving back and forth?
Jacob: That was one of the biggest design challenges. While we were brainstorming, someone suggested we use counterweights to balance horizontal and vertical forces, much like a steam engine in a paddlewheel boat. All our engineers are homegrown guys who have been in the engine room of a paddlewheel boat at one time or another, so we started looking at that approach, using the concept of a single-cylinder steam engine with a large flywheel to reduce motion fluctuations and keep the shaft in line. We even dug up an old treatise on steam engines for guidance.
By adding a sufficiently large flywheel, along with crank-arm counterweights to balance forces, there is enough constant inertia to reduce the input power requirements and velocity variation. If the 1,600mm (63-in. diameter) retort relied on brute-force shaking, a 50hp motor would be required. Instead, it only requires a 15hp motor to get going and 5hp to maintain its speed. With this approach, we were able to reduce vessel stress below that of an end-over-end retort, which also helped reduce shaft deflection. Additionally, we tested many shaft seals before settling on a material that would provide the right friction coefficient and sealing capability.
FE: Given the system’s short cycle times, are automated basket loading and unloading necessary to achieve adequate throughput?
Jacob: Unloading is easy, but the loading side can require additional design engineering due to the product and basket clamping requirements. The retorts built so far use trays with cut-outs that retain the containers and rely on traditional loading. We have not had to address glass jars and other challenging containers on a production-scale unit.
With rotary retorts, containers are usually loaded into baskets in a vertical orientation. That still can be done with this system, but then the basket needs to be tipped 90°. Divider sheets are placed between the layers, and the basket is clamped into position.
FE: What types of products lend themselves to heat transfer via reciprocal motion?
Jacob: Viscosity and product type are critical factors. Pet food with the right gravy content would be perfect; cycle time for heating and cooling vegetable soup might be reduced as much as 90 percent. Color, texture and taste improve dramatically, and some products that are impossible to sterilize in a conventional retort are now possible because of the shorter process time. Béchamel sauces, for example, or Mornay sauce, which is Béchamel with cheese, can be processed without any scorching or burning.
After we sent our R&D machine into the field for trials, industry thoroughbreds were very excited by the results but realized the cycle time was so dramatically shorter that they had to reformulate product ingredients. In late 2009, we built a first-article commercial machine that we ran for months while monitoring wear points. That was a 1,300mm unit, with about a 1,000-lb. load. That machine was delivered to a manufacturer in June 2011. Since then, we’ve delivered four 1,600mm retorts with 2,000-lb. capacities to another food company.
FE: Will this technology appeal to a broad number of applications, or is it more of a niche application?
Jacob: It’s not for every type of product, and it’s not for every type of container. But if a processor is interested in the ability to sterilize products that are not viable in today’s retort world, this process puts you about as close as you can get to an aseptic process, at a much lower production cost.