Issue 5/2007


09/02/07

10th Anniversary of MRLs

How it started and where it’s heading

Dr. Harri Hakala
Trends and revolutions
Sometimes technologies have one recognizable driver for development, some aspect of performance that is characteristic to a certain product. A good example of this is the packing density of mass memories used in computer industry. Starting from paper tapes, magnetic tapes, magnetic discs and optical media, they are now approaching a level of performance where an entire movie in HD-format can be packed into a single DV-disc.
Category: Issue 5/2007
Posted by: Editor

The pace of development in the computer industry has been breathtaking. However, sometimes an industry freezes key assumptions for several decades, staying within certain technical paradigms and making only incremental steps in development. This was characteristic of the elevator industry during the early 90’s with respect to both its key technology: hoisting and key driver: space efficiency. The main developments that had just taken place were the computerization of control systems and subsequent rapid improvement in group control performance. On the hoisting side, frequency converters were launched with the introduction of bipolar power transistors and IGBT’s, following the replacement of motor-generator sets by static converters. There had been some attempts towards a more radical hoisting paradigm shift during the years, such as using linear motors in the hoistway to facilitate a multiple-car solution. This would have been particularly interesting in the case of high-rise buildings where required building space for elevators is enormous. However, none of these ideas achieved break-through status. It’s also interesting to note that many of the patents applied for in that area were by construction companies, not elevator companies.

The hoisting technology of a low-rise elevator consisted of two basic alternatives: hydraulic and traction. Space utilization of hydraulic elevators was rather good because the machine room could be in the basement. However, their low speed and restrictions in lifting height limited their use.

The other alternative was a conventional traction elevator. The machine usually consisted of a gearbox and an induction motor. It was also possible to locate the machine in the basement of a building, but then some additional pulleys were needed in the ceiling of the hoistway for the suspension of the counterweight and the cabin. The length of the ropes was double compared with the simple version with the machine in the top of the hoistway. Until the turn of the 90’s, a traction machine with a worm gear was the dominating technology for the mid-rise sector.
For high-rise needs, only direct drive machines were used. Direct drive motors were demanding from a technical point of view, because their efficiency was high enough to require regenerated energy to be fed back to the electrical mains. Big motors were also quite expensive. The introduction of ESW-roping combined with AC-gearless machineries lowered the speed where direct drive was used in the early 90’s.
Each major manufacturer had slightly different versions of the three technologies with minor exceptions. The rather static situation in the market place regarding established hoisting technologies didn’t mean there were no efforts to shift the paradigm towards the hoisting front. Almost all elevator suppliers were examining helical gears to replacement worm gears, with the main motivation being smaller size and better efficiency. Kone even had prototypes ready and implementation plans. The drawback, however, was the required investment in the manufacturing process. There were also technical problems, like the noise level, which was higher in helical gears versus conventional worm gears as well as special solutions, like screw or chain drives, offered by smaller companies or for special applications.
Another and more radical idea was the use of a linear motor as mentioned above. To picture the structure of a linear motor, imagine an ordinary rotating motor that has been sawn from surface to middle point and uncurled as a plane, as in illustration 1. The result is a fl at linear motor. Either the stator or the rotor can be longer, but the same forces that cause the rotation of the motor produce linear movement. Principally, a linear motor is ideal for elevator use because it generates the straightforward movement required in elevators. The author presented one solution for multiple-car linear motor elevators in 1985. Already by the mid 80’s, a report had been prepared on linear motor applications in elevators, but the technology then appeared immature and too expensive.
The prevailing attitudes changed when the “Skylinear”elevator, a linear motorbased product, was introduced in 1990. It was marketed only in Japan, but press releases indicated an intention to cover larger markets later on. Skylinear utilized a hoisting concept that operated with a linear induction motor. In that product, the hoisting machine was located in the counterweight of an elevator. Thus, it was intended for use with ordinary rope suspension and the target market was residential housing, not skyscrapers. Electric components like frequency converters and batteries for the emergency drive were located in a small room next to the shaft. This initiative started a patenting boom in the industry and by the end of 1995 there were about 150 new patents concerning linear motors in elevators.
Development Path of MonoSpaceTM
The challenge of linear motor technology was to be taken seriously. Kone also renewed its study of linear motor applications. A new project was launched late in 1991. The Skylinear solution was based on a tubular linear motor as presented in illustration 1. In this application, the flat linear motor was curled again around the origin resulting in force balance and better efficiency. This solution was quickly patented however, and thus blocked competitive response.
The older alternative, utilizing the fl at linear motor had been patented during the early 70’s, leaving this route open as the patents were outdated. The idea was to use the guide rail of the counterweight as the secondary part and locate the primary part, i.e. the stator winding into the counterweight. The structure studied is presented in illustration 2. The benefit was identifying the possibility to now integrate the secondary part into the ordinary guide rail structure, thus allowing unlimited hoisting height as the secondary part didn’t pass the stator winding as it did in the tubular version. Skylinear, on the other hand, limited the structure height to only 25 m.
But there were still problems:
  • In the motor air gap, there are two kind of forces: one that has an effect along the air gap and causes the movement, and an attraction force working across the air gap that tends to pull the secondary and primary parts against each other (this is not a problem in an ironless rotor, however). This attraction force can be eliminated in perfectly symmetrical motors, but in rotating motors it is far easier to maintain the balance via the bearings.
  • In all linear motors, the asymmetrical three-phase system results in a pulsating torque.
  • The properties of a motor depend on the temperature if its parts. Given that a linear motor was heated only under the stator part, the temperature was uneven and the motor parameters changed continuously when the motor moved.
  • It wasn’t possible to use a slotted secondary part as in standard rotating motors for cost reasons as the conductive part of secondary required massive coating. This restriction also increased the air gap and making the efficiency of the motor very poor.
Many of the major technical problems were solved and patentable solutions filed, but in 1992, the cost level remained the biggest problem. The main issue was the strong interdependence of the building height and the production cost, obviously since the motor was as tall as the hoistway.
As a side effect of this linear motor study, a new idea emerged in late 1992. In essence it was a combination of friction drive and rotating machine but was located in the counterweight – a kind of motorized traction sheave. The idea is presented in illustration 3 (at left). The benefit was mainly the same as those of a linear motor hoisting, namely lack of a machine room, but now at significantly lower cost as the connection between building height and cost was now severed. In the beginning of 1993, we knew that if we were able to build a motor fl at enough to be located in the counterweight, it would be possible to launch an entirely new elevator type. This would then provide the cost level of a traditional friction drive with the benefit of space efficiency offered by a linear motor drive.
As a result, plans to develop the linear motor were cancelled and the whole of 1993 was spent developing the fl at rotating motor type. The point of departure in design was the traditional cylindrical induction motor with outside rotor. However, the thickness of the motor was a problem as the endings of the windings required too much space making the counterweight thicker than in standard elevators, which was not acceptable. An initial condition was that the elevator must fit in the standard shaft. The design was therefore changed to an axial type in which the main air gap fl ux is parallel with the shaft direction. This structure is known as disc motor or pancake motor. In fact the idea isn’t new. Mr. Faraday’s first electric motors in the 1820’s were of this type, the cylinder form appearing later. They weren’t commonly used however, as its design and manufacture was more difficult compared to radial gap motors. Typical applications of this motor have been industrial robots and vehicle wheels with integrated motors.
The structure can be seen as a flat linear motor curled by joining both ends, as presented in illustration 1. The resulting motor consisted of two toroidal parts separated by the air gap. There were several development steps on the path, one of which is presented in illustration 3 (centre). Here the stator part is only a half circle whereas the secondary is a full circle. The idea is that the torque generated in the larger radius is needed in smaller radius. The radius of the traction sheave is smaller than the radius of the stator winding but still located on the same side of the secondary. Although this version was cancelled mainly due to the end effects of the stator, it does highlight the origins of the machine as a “rotating linear motor”.
Another problem was the efficiency of the low-revving induction motor, operating at about 40–70 %. On top of the energy consumption, there was the additional challenge of cooling the heat which generated in the middle of the motor. Therefore, the induction motor was abandoned as principle and we went back to the drawing board. Another attempt focused on reluctance motor, where the secondary part consists only of massive iron rendering rotor losses insignificant. The power factor was only 0.3 in low rpm however, resulting in a high stator current and even lower total efficiency than that of the induction motor.
The viable solution appeared to be a synchronous motor with rare-earth permanent magnets (PMSM). There was a new manufacturer in Finland (the Outokumpu Magnets Corp.) that supplied such magnets. Kone and Outokumpu established a partnership that provided easy access to various magnets for prototype motors. The efficiency of the motor was now excellent as no magnetizing power was needed. The only drawback was the high cost of material. This was compensated by the savings in other elevator parts. Thus the introduction of high-energy permanent magnets not only paved the way for battery-drive portable tools and small HDD’s in laptops but also changed the face of elevator technology as well.
The first running prototype elevator with a flat motor appeared in 1993, fulfilling all requirements for shape, size, performance and cost. This prototype (picture 4) was far from commercial production but was important as evidence of the viability of the concept.
In addition to magnets, another important prerequisite for using a synchronous motor was the frequency converter as it isn’t possible to connect PMSM directly to the mains. The converter had also been developed for induction motors in elevators and Kone had managed to launch the first VVVFs into market in late 80’s. This technology was also easily available.
The next step was to develop all components for a working elevator system. The original idea of locating the motor in the counterweight was abandoned because service work would have been difficult and supplying whole hoisting power to a moving part increased costs. But the basic idea behind Eco- DiscTM was in fact a motorized traction sheave that could be located in many places in the suspension system. Locating it in the counterweight was only one, extreme possibility. The second solution was using the pit of the elevator, as presented in illustation 3 (far right version). The best location was finally identified at the top of the shaft, where the machine was located between the guide rail and the wall of the hoistway. An essential element of the concept was locating the control panel next to the upper most landing. Space must be reserved for the sliding door panel of an automatic door; therefore, no external space was required for the electronics, as it was in competing solutions. Many reasons contributed to this achievement: there was no gearbox, the efficiency of the motor was good and the frequency converter was small. For residential elevators, the motor current was less than 10 Amps, which enabled small components in power electronics and con tactors. The final solution is presented in illustration 5.
The new solution offered many benefits compared to both traditional and linear motor solutions:
  • No reduction gear was needed because the motor was integrated to the traction sheave. Thermal losses therefore dropped significantly. The energy bill was roughly halved as was the required fuse size. The saving effect was multiplied in buildings where emergency energy supply was needed. The simple machine and accurate control also resulted in good riding comfort.
  • In low-rise buildings, space utilization was excellent and the benefit was greatest in the top floor, which is usually the most commercially important.
  • All suspended masses were mounted to guide rails that lead all the forces to the basement of the building. In effect, the new elevator stood on its own feet without the need to design the building for suspended masses.
  • The machine did not require any oil – a clear benefit compared to hydraulic elevators. In the residential low-rise sector, MonoSpaceTM enabled process savings due to the decreased need for case-by-case engineering. Since the machine was in the hoistway, the machine room location didn’t necessitate several layout options. Many elevator components were also cheaper to produce if optimized for the MonoSpaceTM.
The new elevator type was first named “Greenstar”, but later christened MonospaceTM. A typical MonospaceTM for a residential building is presented in picture 6, and is intended for 1 m/s speed and a payload of 630 kg. Several patents were applied for protecting the key features in the new products. The most important claim described “an elevator with a gearless machine located in the hoistway and where the extension of the machine shaft intersected the profile of the elevator car.” The working principle of a motor was not patentable as it was well known, but this application for elevators was new. By the end of 1995, all the most relevant patents were granted, combined with key ideas gathered along the development path that may not have ended in production but were important for making protection complete.
The road to market
By mid 1994, a full scale prototype of the coming MonoSpaceTM-elevator was ready in research center in Hyvinkää, Finland. Internal calculations proved that the product was viable in terms of performance, cost and reliability. Still it was far from commercial success. Main uncertainties to be solved were:
  • Elevator safety codes did not recognize such a product. MonoSpaceTM clearly needed a new path to market.
  • It wasn’t clear that the market would accept a solution that was offered only by one supplier. After choosing Mono- SpaceTM, there was usually no opportunity to return to conventional solutions. If the machine room was left out, no other elevator type was possible without expensive alterations to the building. Many customers might be concerned because MonoSpaceTM required special permission to be officially inspected and certified. It was definitely something new and buyers did not want to be “guinea pigs”.
  • If a machine room was already included in the drawings during the tendering process, then leaving it out meant design change, which either negated the saving, or the construction company absorbed the benefit.
  • Lastly, it was not clear in the beginning if it was possible to utilize the same technology throughout the volume range. A partial solution would have meant just adding to Kone’s existing selection of hoisting technologies, thus resulting in poor cost structure with the need to maintain several technologies at simultaneously.
In the mid 90’s, most countries had their own safety codes, which were usually modified from some general standard, like IEC or DIN. It was also common to have official inspection by authorities before allowing an elevator for public unattended use. This was also the case in Europe during the early 90’s, where regulations and interpretations deviated somewhat from country to country. The importance of the standard varied, too. In some of the countries, standards were only engineering recommendations, while in others they were part of the legislation.
Inside the EU, development was moving towards a common standard. There were already Harmonized European Standards (EN 81-1 and EN 81-2) but they were optional; countries could set their own standards, too. A new directive was under development and it was to be obligatory for EU member states by 31. 12. 1997.
Standards provided quite detailed technical requirements for elevators and they were de facto design rules for elevator manufacturers as inspectors referred to them. For example, paragraph 6.1.2. states:
“The machine and its associated equipment shall be in a special room, comprising solid walls, ceiling and door and/ or trap.”
However, there was a possibility to deviate from the standard. Paragraph (prEN 81-1) 0.0.4 stated:
“When mention is made of a design for the sake of clarity; this should not be considered to be the only possible design; any other solution leading to the same result can be applied if it is equivalent in operation and at least equally safe.”
This option left the suppliers the responsibility to prove equal safety as noted in the code. Therefore it was rarely used and had to be repeated for each country or even every delivered lift. The situation was about to change with the new lift directive 31. 12. 1997. At this point, when some of the “notified bodies” in the EU accepted a certain construction, other EU members had to accept it.
In order to solve the code problem, the designers approached the Dutch authorities by first asking them to act as specialists in risk analysis for the new elevator type. The idea was to utilize the paragraph in the safety code, which allowed other solutions proven to be as safe as specified in the codes. In order to provide such proof, a risk analysis was required.
Leading individual experts and elevator authorities participated in a safety analysis, which was then used when applying for special permission to sell the new elevator. Negotiations started in 1994 and the first elevator in Kone’s premises was accepted for use on 13. 3. 1995. The authorities required certain changes to the original product, but were ready to permit deliveries of MonoSpaceTM in the Netherlands.
The benefits of MonoSpaceTM for customers were obvious: architects gained an increased degree of freedom because the machine room was no longer required, the builder saved in construction costs and users in the energy bill. Still, acceptance in the markets was another key lingering uncertainty. It was clear that some kind of test marketing was needed before Kone could commit solely to Mono- SpaceTM. Test marketing began later in 1995, when the first elevators were delivered to customers as pilots. However, these elevators were not originally sold as MonoSpaceTM but rather as hydraulic elevators. They were later converted to MonoSpaceTM in order to gain experience from real market acceptance and user attitude as soon as possible. Only the necessary new product changes were made, i.e. the new machine and control cabinet were required, but a major part of the other elevator components were available with only minor modifications. Since those standard components were not optimized for MonoSpaceTM, the cost level of such a short- cut product were ultimately higher than the full scale product designed later on. Because Kone did not have its own motor manufacturing facilities, a small scale experimental workshop was established in Hyvinkää.
Its capacity was only a few hundred machines per year, but was sufficient to satisfy demand in the early period.
The Netherlands was selected as the country in which the test was to take place. It was a suitable area because the buildings were rather low and there was a clear need for a machine-room-less (MRL) elevator given that certain types of building, even with only two floors, required an elevator. In such cases, the relative cost of a machine room is significant.
Changing industry standards: from Mono to Mini to Maxi
During test marketing it appeared that nearly all customer groups contacted claimed MonoSpaceTM to be a great innovation and agreed on the benefits. Novelty of the concept remained the main concern and the uncertain reactions of authorities in various countries still to be contacted. Contacting authorities in other countries continued and MonoSpaceTM was released in Europe country by country. The official product launch took place in Brussels in March 1996. At the same time, the R&D began design of the optimized product for real volume scale. All decisions were gradually pointing to a future where MRL would be the dominating technology.
Proof that the same motor technology had clear benefits in bigger elevators also surfaced and the same hoisting principle was possible to apply throughout the range. The same type of machine was equally suited to small residential elevators as for the very biggest elevators. It was now possible to substitute geared traction elevators and hydraulic elevators with only one technology, gearless.
New production facilities for EcoDiscTM were needed. Kone established a new factory for EcoDiscTM in 1997 with a capacity of 5000 units per year, which was then doubled in 1998 and expanded further 1999. Now EcodiscTM-machineries account for over 90 % of Kone’s sales. The license was also sold to Toshiba, who started to sell standard elevators based on the machine manufactured by Kone. Toshiba sold these MonoSpaceTMelevators under the Spacel brand in Japan.
The mid-rise product came onto the market in 1998 under the name MiniSpaceTM . It utilized the energy efficiency and small size of the EcoDiscTM. The idea was to offer an elevator where the machine room was actually an extension of the hoistway. In other words, the area and shape of the machine room were the same as in hoistway.
A high-rise product named AltaTM was launched in 2000. For its development, Kone built a new test facility in Lohja, Finland. It is still the longest test shaft (303 m) for commercial testing of person lifts and it is built in mine shaft called TYTYRI. AltaTM completed the high-end range with the MX100 machine, which has been designed for 17 m/s speed, 5,000 kg payload and 50,000 kg shaft load.
The latest step in the series of elevator products based on EcoDiscTM – technology is Kone’s MaxiSpaceTM. It continues the quest for better space utilization. Maxi SpaceTM is a traction elevator without the counterweight. It therefore offers further space saving that is especially valuable in the modernization business, where it is usually possible to increase the cabin size in a one-step modernization. This has drastic impact on usability as in many cases MaxiSpaceTM can accommodate the use of a wheelchair or baby carriage which may not have been possible in the building’s existing elevator.
On the code front, an elevator without a machine room has become a new industrial standard in many parts of the world. Elevator safety regulations are updated for this very elevator type. All major elevator suppliers have now entered the market with their own MRL-solutions, but the original MonoSpaceTM still competes well.
Elevator engineers will continue their efforts for better space utilization. The unfulfilled dream of a ropeless multiple car elevator still looms ahead, although several possible alternatives have been presented. Perhaps one day we will also see commercially viable solutions of these too but we have a long way to go before we can speak about industry standard. Such elevators must be designed as a part of a building and will therefore present major challenges for our industrial partners in the construction industry.
MonoSpaceTM as a product has earned many rewards, for example the Golden Pyramid in Batimat in Paris 1997 for the most innovative product of the year in the construction industry. Looking back, there is absolutely no doubt that Mono- SpaceTM has changed the elevator business in many ways during its ten-year existence. In summary, perhaps Robert S. Caporale said it best: ”In recent years there has been a lot of discussion about machine-room-less elevators. These technological marvels have really set the elevator industry on its ear. They are without doubt the most innovative designs to come along the elevator industry in a good many years.” (Elevator World, May 2004)
5/2007

^ Top