Conflict Resolution Using TRIZ and Design of Experiment (DOE)

	Bookmark This Page Email This Page Format for Printing

First published in the Proceedings of TRIZCON2001, The Altshuller Institute, March 2001.

Life Fitness, the leader in the exercise equipment industry, manufactures both cardiovascular and weight training equipment. The paper will present a case study of successfully using TRIZ methodology and Design of Experiment (DOE) to solve a longstanding design deficiency in Life Fitness commercial treadmills. Using Algorithm for Conflict Elimination (ACE), a derivation of ARIZ, engineers at Life Fitness dissected a nozzle-clogging problem and came up with a no-cost solution using existing system resources. DOE was used next to optimize the settings of the solution. The resulting feature, called EVERWAX^Ô, was immediately put into production, and it greatly increased customer satisfaction. The solution is currently undergoing the patent application process.

As a research project, TRIZ was introduced to and practiced at Life Fitness by working with a TRIZ consultant, Gregory Yezersky. The project had multiple goals, one of which was using TRIZ to tackle a real life technical problem. The problem the group decided to take on was the nozzle-clogging problem on Life Fitness commercial treadmills.

Figure 1

Most treadmills are mechanically simple with a belt looped around two rollers. One of the rollers is driven by a motor and moves the belt. A deck in between the rollers supports the runners’ weight. Because the belt and the deck have a sliding interface, belt and deck life depend on proper lubrication.

Life Fitness commercial treadmills employ paraffin wax as the lubricant for the belt and deck interface. Initially, solid wax deposited on the deck is the primary source of lubrication, and as the solid wax is worn away with treadmill usage, wax is replenished by dispensing wax emulsion onto the belt.

Figure 2

Wax emulsion consists of paraffin wax, emulsifiers, and water. The emulsifier is used to “suspend” or emulsify wax into a water-based solution. Each time the nozzle sprays wax emulsion, some wax emulsion is left at the nozzle orifice. A tough layer of wax and emulsifier residue forms when the water evaporates. The residue builds up after multiple sprays and begins to clog the nozzle, and eventually the nozzle can be completely clogged.

In the past, several attempts have been made to remedy the nozzle-clogging problem.

1. Different angles of nozzle position have been tried. Although optimized angles of nozzle position alleviated nozzle clogging by changing the drip point location, it only delayed nozzle clogging.
2. Different types of nozzles and different nozzle coatings, such as Teflon, have been tried, but failed because nozzles still clogged.
3. Another attempt was the addition of a heating element to the nozzle. Although the heated nozzle demonstrated some success, added cost, safety concerns, and electrical system limitation prevented the implementation of the heated nozzle.

As a result, Life Fitness recommended changing the nozzle every three weeks. However, customers often complained about the time-consuming maintenance procedure or just ignored it. When wax is not replenished at the belt and deck interface, premature failures occur. Such failures ultimately reflect poor quality on Life Fitness exercise equipment instead of poor maintenance practice on the customers’ part. Ideally, product maintenance should require as little assistance as possible from the customer.

Figure 3

The project team decided to use ACE, a version of ARIZ, to tackle the problem. A flow chart is provided to give a general outline of the ACE algorithm.

The automatic wax dispensing system is designed to dispense a specified volume of wax emulsion uniformly onto the underside of a treadmill belt. The spray pattern is controlled by the nozzle orifice design and pump pressure. The system consists of the wax emulsion, reservoir, peristaltic pump, tubing, spray nozzle, and the spray command program (Figure 2). In order to transfer enough solid paraffin wax onto the belt, wax emulsion was selected. However, this emulsion has a tendency to induce clogging of the spray nozzle, thus disabling the system function.

Component Resources:

Wax emulsion, plastic reservoir, peristaltic pump, plastic tubing, fittings, spray nozzle, spray command program, air, water, wax, emulsifier, and belt.

Energetic Resources:

Electricity, pressure, gravity, velocity, surface tension force, heat, and runner.

ExC-1:

If the percentage of wax in the emulsion is increased, then there is more lubricant sprayed on the belt, but the spray nozzle will be clogged more often.

ExC-2:

If the percentage of wax in the emulsion is decreased, then the spray nozzle will be clogged less often, but there is less lubricant sprayed on the belt.

Goal:

Find a way for the spray nozzle to clog less often and to provide more lubricant sprayed on the belt.

After taking the problem through the extensive algorithm, the project team came up with many potential solutions, three of which are listed below:

1. Move the pump closer to the spray nozzle to reduce pressure loss. Redesign the pump to spray emulsion and air at the end. The air spray will disrupt the surface tension of leftover emulsion, thus reducing emulsions left on the nozzle.
2. Design the nozzle so that the residue can be easily purged on the next spray. The nozzle design should allow a weak bridge to be built between the residue and the nozzle. The weak bridge will break away on the next spray with normal line pressure.
3. Polish the nozzle to reduce adhesion property.

Using 4M (Miniature Men), a brainstorming technique in ACE, both the wax emulsion and wax residue are modeled as little men and analyzed.

Figure 4

First, the function of the emulsifier is identified; it is used to “grab” both water and wax. Closer inspection shows that the wax residue contains only emulsifier and wax. In order to suspend wax in water, water must be introduced again. After looking at available system resources, the idea of “Prespray” emerged. “Prespray” is a short burst of spray prior to the actual wax dispensing spray, and it is used to introduce wax emulsion to the clogged nozzle. The idea is to get the nozzle wet and let it “soak” for a period of time. By wetting the nozzle and letting it “soak,” the tough residue of wax and emulsifier will soften. During the actual wax dispensing spray, the nozzle clears itself by pushing away the softened residue.

The mechanism at work is the affinity of emulsifier to both water and wax. Emulsion leaves a tough residue of wax and emulsifier around the orifice of the nozzle after it dries. When the dried residue comes in contact with a new supply of emulsion, two things happen as the residue is “soaking” for a time period. First, the emulsifier in the emulsion grabs onto wax residue and emulsifies the wax residue. Second, and more importantly, the dried residue of emulsifier and wax, still chemically attached, “suspends” or emulsifies in the presence of water. The combination of two chemical interactions leads to the softening of the residue. This cycle continues with each spray and the lubrication delivery can then be maintained.

The Prespray solution uses only system resources and did not require significant design change. The only modification was done through software. The cost of this solution was the cost of making the software change, which was minimal.

Some preliminary tests were performed using the concept of Prespray and the results were very positive as all of the nozzles cleared themselves. Through these preliminary tests, it was found that long soaking time and large Prespray amounts contributed to better performance.

For field implementation, both the Prespray amount and soaking time must be minimized. The Prespray amount must be minimized because excess wax can have anti-lubricious property. For the system to spray wax, the belt must be moving for the wax emulsion to be evenly distributed and prevent “puddling” and leaking out. The soaking time must be minimized because the probability of Prespray and the wax dispensing spray occurring during the same workout greatly increases. A simple two-factor, three-level DOE was used, and three repetitions were run for each of the nine settings.

Table 1

The output of the experiment is a numerical representation of how clogged the nozzle is; the lower the number, the worse the nozzle clogging is. An output of 160 represents a no-clogging condition, where an output of 0 represents a total-clogging condition. The data is then fed into DOE KISS, an analysis software, and a solution space is generated.

Figure 5

Based on the solution space, a setting is picked. The selection criteria was the lowest setting that produced a no-clogging condition. Confirmation trials were run and the results conferred with the prediction of the solution space. After successful field testing, the solution was put into production and named EVERWAXÔ.

Figure 6

Using ACE (Algorithm for Conflict Elimination), a methodology derived from ARIZ, in combination with DOE (Design of Experiment), Life Fitness Research Engineering solved a long- standing design deficiency in the Commercial Treadmill product line. The innovative feature, Everwaxä, had virtually zero implementation cost because changes were only made in software. Treadmill uptime is greatly increased because the maintenance procedure of the lubrication system was reduced from once every three weeks to an inspection every six months. Other analysis methodologies, such as DOE, complement TRIZ and will play an important role in the advancement of TRIZ practice and development.

Born in Taiwan, Republic of China, and immigrated to United States in 1985.

Graduated from University of Illinois-Urbana/Champaign with Bachelor of Science in Mechanical Engineering in 1997. Currently working in Research Engineering at Brunswick-Life Fitness, Franklin Park, Illinois.

Contact Information:
10601 West Belmont Ave.
Franklin Park, IL 60131
Phone: (847) 288-3454
Fax: (847) 288-3709
Email: jhsing@lifefitness.com

• Altshuller, G., “The Innovation Algorithm”, 1999, Technical Innovation Center, Inc.,Worcester, MA.
• Launsby, R., Schmidt, S., “Understanding Industrial Designed Experiments”, 1998, Air Academy Press & Associates, Colorado Springs, CO.
• Yezersky, G., Systemology, 1998, Systemology Research Center, Farmington, MI.
• Yezersky, G., ACE-2000, 2000, Systemology Research Center, Farmington, MI.